Weathering, Soils & Paleosols - Martini & Chesworth (1992) PDF

Developments in Earth Surface Processes
1. PALEOKARST: A SYSTEMATIC AND REGIONAL REVIEW

P. BOSAK, D. FORD, J. GLAZEK and I. HORÂCEK (Editors)
Developments in Earth Surface Processes 2
WEATHERING,
SOILS & PALEOSOLS
Edited by
LP. MARTINI and W. CHESWORTH

Department of Land Resource Science, University ofGuelph,
Guelph, Ont. NIG 2W1, Canada
^ ELSEVIER
Amsterdam — London — New York — Tokyo 1992
ELSEVIER SCIENCE PUBLISHERS B.V.
Sara Burgerhartstraat 25
P.O. Box 211, 1000 A E Amsterdam, The Netherlands
Library of Congress Cataloging-in-Publication Data
Weathering, soils & paleosols / edited by I.P. Martini and W.

Chesworth.
p. cm. — (Developments in earth surface processes ; 2)
Includes bibliographical references and index.
ISBN 0-444-89198-6
1. Weathering. 2. Soils. 3. Pa 1eopedo1ogy. I. Martini, I. P.
(Ireneo P e t e r ) , 1935- . II. Chesworth, Ward. III. Title:
Weathering, soils, and paleosols. IV. Series.
QE570.W43 1991
551.3Ό5—dc20 91-36393
CIP
ISBN 0-444-89198-6
© 1992 Elsevier Science Publishers B.V., All rights reserved.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form
or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior
written permission of the publisher, Elsevier Science Publishers B.V., Permissions Department,
P.O. Box 521, 1000 AM Amsterdam, The Netherlands.
Special regulations for readers in the USA - This publication has been registered with the Copyright
Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC
about conditions under which photocopies of parts of this publication may be made in the USA. All
other copyright questions, including photocopying outside of the USA, should be referred to the
publisher.
No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as
a matter of products liability, negligence or otherwise, or from any use or operation of any methods,
products, instructions or ideas contained in the material herein.
This book is printed on acid-free paper.
Printed in The Netherlands

V
Preface and Acknowledgements
This book is an attempt to connect present happenings on the planetary land

surface with ancient ones, and to determine similarities and differences in weath
ering and pedogenic conditions. A single volume cannot possibly cover all facets
of the complex problems involved. A complete book could be written solely about
the terminological differences between Soil Science and Geology alone. We have
chosen therefore to be selective and to examine some of the fundamental concepts
to be dealt with in depth. Each chapter is written by an expert or experts in their
respective fields, and each includes a short review, some new information or ap
proach, and an indication of the direction future research should take to achieve
some significant advancement in understanding. Most of the people we have con
tacted responded favorably and their efforts are included here. A few researchers
suggested that perhaps some topics were sufficiently dealt with in other books or in
journals and did not need to be covered again. For instance, the rates of weathering
and of other Earth-surface processes have received recent coverage in two excellent
volumes edited by Colman and Dethier (1986) and Lerman and Meybeck (1988),
referred to in Chapter 2. Consequently, although such topics are important to the
themes of this book, they are treated only indirectly, in terms of their incidental im
pact on our major topics. Again, some soil types of great interest to the soil scientist
as well as the agriculturalist (Chernozem for example) have not received coverage
because their fossil representatives are unknown.
The subjects included reflect the following preoccupations. First, there was a
need to define our present understanding of the process of weathering, of the
nature of weathering products, and how they change during diagenesis. Geologists
and geochemists have a long history of involvement in this area. They have not,
however, shown the same depth of interest in the second of our preoccupations,
which concerns the nature of soils and the ways in which soils reflect the physics,
chemistry and biology of their environment. There is a great need for geologists,
geochemists, soil scientists and others to bring their various talents to bear on this
topic cooperatively, and we have tried to present the subject in such a way that an
earth scientist from one field has access to the complementary research in other
fields. The point, ultimately, has been to use what insight we have gained from
VI Preface and acknowledgements
the study of the present land surface, or the surface of other planets, to interpret
the past. Thus our final preoccupation has been to examine paleosols to see what
they tell us about ancient Earth surfaces. In so doing we find ourselves pushing the
principle of uniformitarianism to its limits.
We thank all the contributors to this book, for their excellent work and for the
grace with which they accepted the comments of referees and editors. One editorial
decision we were forced to take because of space limitations, was to remove indi
vidual acknowledgements from each chapter. Each author, including us, received
help from colleagues, students, relatives and organizations, to such an extent that
the cumulative number grew to the size of a small army. We are grateful to them all:
without their effort this book would not have been written. However, one individual
must be singled out as contributing as much as any of us, to the task of bringing this
book to print. We are deeply grateful to Mrs. Marilyn Metcalf for her profession
alism and skill in correcting and retyping parts of all contributions, and for dealing
with reference lists that might have come from the Tower of Babel.
We hope the reader will find some stimulation in these pages and we look for
ward to discovering the results of this stimulation in future published research.
LP. MARTINI and W. CHESWORTH (Editors)

Vil
List of Contributors
P.W. BIRKELAND Department of Geological Sciences, University of Colorado at Boulder,

Boulder, CO 80309-0250, USA.
B. BOULANGE Laboratoire de Géosciences de l'Environnement, URA 132 CNRS

et ORSTOM, Faculté de Sciences de St. Jérôme, Université d'Aix-
Marseille III, 13397 Marseille Cedex 13, France
LB. CAMPELL New Zealand Soil Bureau, Department of Scientific and Industrial
Research, Prìvate Bag Lower Hutt, New Zealand
W. CHESWORTH Department of Land Resource Science, University of Guelph, Guelph,

Ont. NIG 2W1, Canada
F. COLIN Laboratoire de Géosciences de l'Environnement, URA 132 CNRS

Marseille III, 13397 Marseille Cedex 13, France
G.G.C. CLARIDGE New Zealand Soil Bureau, Department of Scientific and Industrial
Research, Prìvate Bag Lower Hutt, New Zealand
PJ. DeMARIS Illinois Geological Survey, Natural Resources Building 615 East Pea-
body Road, Champlain, IL 61820, U.S.A.
L.J. EVANS Department of Land Resource Science, University of Guelph, Guelph,

Ont.NlG2Wl, Canada
R.E. HUGHES Illinois Geological Survey, Natural Resources Building 615 East Pea-
body Road, Champlain, IL 61820, U.S.A.
M.J. KRAUS Dept. of Geological Sciences, University of Colorado at Boulder, Cam

pus Box 250, Boulder, CO 80309-0250, U.S.A.
J.P. LEGROS INRS-ENSA, Science du Sol, 9 Place Pierre Viala, 34060 Montpellier
Cedex, France
J.F. LINDSAY Bureau of Mineral Resources, Geology and Geophysics, Onshore Sedi
mentary and Petroleum Program Bureau of Mineral Resources (BMR),
GP.O. BOX378, Canberra City, A.C.T. 2601, Australia
vi i i List of contributors
F. MACIAS Catedratico de Edafologia y Geologia, Facultad de Biologia, Universi-

dad de Santiago, Santiago de Compostela, Spain
LP. MARTINI Department of Land Resource Science, University of Guelph, Guelph,

A.R. MILNES CSIRO, Division of Soils, Private Bag No. 2, Glen Osmond, S.A. 5064,
Australia
D.B. NAHON Laboratoire de Géosciences de VEnvironnement, URA 132 CNRS

Marseille III, 13397Marseille Cedex 13, France
H.W. NESBITT Department of Geology, University of Western Ontario, London, Ont.

N6A 3K7, Canada
G.J. RETALLACK Department of Geology, University of Oregon, Eugene, OR 97403Ί272,

USA.
M. ROBERT C.N.RA., Station de Science du Sol, Rue de St. Cyr, 78000 Versailles,
France
C. ROQUIN Departement de Géologie, Université de Louis Pasteur, 1 Rue Blessig

67084 Strasbourg Cedex, France
W SHOTYK Geological Institute, University of Berne, Baltzerstraße 1, Berne CH-

3012, Switzerland
G. SPIERS Department of Land Resource Science, University of Guelph, Guelph,

Y. TARDY Departement de Géologie, Université de Louis Pasteur, 1 Rue Blessig

67084 Strasbourg Cedex, France
D. TESSIER C.N.RA., Station de Science du Sol, Rue de St. Cyr, 78000 Versailles,
France
M. THIRY Ecole Nationale Supérieure de Mines de Paris, Centre de Géologie

Generale et Miniere, 77305 Fontainebleau, France
A. WATSON 9 Hartley Terrace, Boston, MA 02134, U.SA.
WA. WHITE Illinois Geological Survey, Natural Resources Building 615 East Pea-
body Road, Champlain, IL 61820, USA.
VP. WRIGHT P.R.I.S., University of Reading Reading Berkshire, RG6 2AB, U.K.
3
Chapter 1
Reflections on soils and paleosols

LP. ΜΑΚΠΝΙ and W. CHESWORTH
Introduction
The soil constitutes a superb, albeit complex record of the land surface environ
ment. Since it is axiomatic amongst geologists that the present is the key to the past,
it is a little surprising that, until recently, the soil has largely been ignored in the
environmental interpretation of ancient rock sequences. It is the purpose of this
book to reinforce this geological aspect of soil science.
We study paleosols for the information they provide about their former envi
ronmental setting (Holland, 1984). As with any other stratigraphic-sedimentological
study where the principle of uniformitarianism is utilized, younger and/or less mod
ified deposits are easier to interpret than their opposites. In particular it may be
difficult to interpret paleoclimate from a paleosol because a profile constantly ad
justs in response to water percolation and physico-chemical changes (Nahon et al.,
this book). For instance, lateritic soils presumably formed under a regime of high
temperature, high rainfall and rain forest vegetation, now exist as vetusols (Cre-
maschi, 1986) in arid conditions in Australia (Nahon et al., this book). In effect
changing conditions have a polygenetic effect on the weathering mantle and soils
form a palimpsest of all the changes they have been subject to (Ruhe, 1965).
Other difficulties in interpreting paleosols are related to the fact that they "are
often only fragments of what was [once a temporal] continuum of soil" (Lavkulich,
1989, p. 35), in part modified by burial and diagenesis. Also our knowledge of
soil-forming processes and rates on Earth, when land-organisms and atmospheric
oxygen were absent or negligible, is deficient.
Difficulties in interpreting paleosols relate also to problems we encounter in un
derstanding certain fundamental features in present soils. For instance, microfabric
analysis of soils (micromorphology) is a potentially powerful tool in the study of
paleosols, but the micromorphological features in recent soils "may be as much af
fected by the chemical, physical and biological properties of the matrix as by the
nature of the environmental influence" (Dumanski, 1969, p. 45). Presence of relict
4 LE Martini and W. Chesworth
features of earlier pedogeneses, occurrence of allochthonous materials and effect of

subsequent processes, may further complicate the picture so that great care needs
to be taken in interpreting micromorphological observations.
The chemistry of weathering and pedogenesis is relatively well understood. How
ever, the interpretation of chemical and mineralogical analyses of paleosols can
present problems because the Earth's surface and near surface acts as an open sys
tem, and soils and paleosols may change materially during weathering, burial and
exhumation. Nevertheless, Nesbitt (this book) indicates that chemical variations in
relatively simple mineralogical systems, such as those associated with granitic parent
rocks, enable weathering and diagenetic trends in paleosol to be distinguished. Fur
thermore, paleosols can provide valuable information about diagenetic processes
and trends, as their texture, mineralogy and bulk chemical composition change dur
ing burial.
Definition of soils
In this book the soil spectrum is treated principally in terms of its well defined
extremes, from the initial stages of weathering and Leptosols to Ferralsols and duri-
crusts. The relationship of soil to bedrock is summarized by Macias and Chesworth
(this book) for humid climates and by Watson (this book) for arid ones.
At the outset is the problem of definition. The purposes of this book require a
definition that can be applied to other times than the present and other planets
than Earth. Furthermore, it must distinguish soils from sediments. The definition of
Retallack et al. (1984, p. 7) serves this purpose. Soil is "material at the surface of a
planet or similar body, altered in place by physical chemical or biological agencies,
or by a combination of them.,, This definition is sufficiently general to encompass
the soil scientists' definition of soil on Earth: " a three dimensional natural body
of unconsolidated matter on the immediate surface of the Earth that has been
subject to, and influenced by, genetic and environmental factors of parent material,
climate, biotic influence and topography, all acting over a period of time to produce
a product that differs in physical, chemical, biological and morphological properties
and characteristics from the material (parent material) from which it was derived
and is capable of supporting the growth of land plants" (Lavkulich, 1969, p. 26). Yet
it allows for the absence of land plants on other planetary bodies, or even on the
early Earth.
The engineering definition of soil being "all unconsolidated material above
bedrock" (Bates and Jackson, 1987) is also encompassed, but whereas the engi
neer does not distinguish between soil and unconsolidated sediment, Retallack et
al. (1984) stress that a soil forms in place, whereas sediment is transported material.
This does not mean that the definition is without difficulty. Since sediments are usu
ally transported soils in one form or another, a gradation between soil and sediment
Reflections on soils andpaleosoh 5
is possible. Soliflucted soils and turbated soil are still recognizably soils, while having
been moved like a sediment.
Soil and energy
Energy is unevenly distributed at the surface of the Earth. This is the primary
cause of change, and weathering and soil formation are changes that move land
surface systems in directions aimed at an even redistribution of energy. If an even
distribution of energy is ever achieved, all change will stop. This is a statement of the
second law of thermodynamics and such a state is said to have reached maximum
entropy. Another way of saying the same thing is that spontaneous change leads to
an increase in entropy, which is often equated to an increase in randomness, with
complete chaos being the ultimate state of maximum entropy.
So what is to be made of soils in these terms? The alteration of parent material by
the spontaneous change of weathering, leads to an increase in entropy. But the or
ganization of a resulting soil into horizons appears to run counter to the second law
of thermodynamics. Johnson and Watson-Stegner (1987) recognized this dichotomy
in referring to the randomising weathering stage (haploidization, homogenization)
as "regressive pedogenesis" and the increase in order resulting from horizonation as
"progressive pedogenesis". The paradox is however only apparent and arises from
the fact that we are considering a system independently of its thermodynamic sur
roundings. In "progressive pedogenesis" solar energy enters the soil profile mostly
by virtue of moving fluids (Johnson and Watson-Stegner, 1987; Runge, 1973). Any
decrease in entropy within the soil, is ultimately achieved by a greater increase in
entropy by the sun.
Provided that the external energy source continues to produce and that fluids
continue to migrate within surficial Earth materials, changes will continue to modify
the soil. If conditions are uniform, a steady haploidization occurs. Resulting soils
will evolve in the direction of decreasing available energy (increasing entropy),
along an energy scale extending from the high potential state of igneous rocks, to
lows represented by the generation of sediments or in situ regolith or saprolite,
or to the quasi-complete homogenization in parts of lateritic soils. Lateritic (Tkrdy
and Roquin, this book) and ferrallitic (Macias and Chesworth, this book) soils
may approximate a quasi-equilibrium with environmental conditions, over a short
time span because of rapid chemical reaction rates, or over longer time spans if
conditions remain sufficiently stable for slow reactions to proceed to completion.
Some scientists go further than to claim a quasi-equilibrium as a possible state for
soils. Nikiforofï (1969, p. 31) for example speaks directly of "dynamic equilibrium".
Lavkulich (1969, p. 31) rightly objects to this, preferring the term "steady state".
Yet even the term steady state must be qualified by the prefix "quasi" since the soil
environment is constantly changing in space and time, and soil changes with it in a
6 LP. Martini and W. Chesworth
non-steady fashion. Without doubt, relatively time-independent states are achieved

in soils, but to use the terms equilibrium (from the thermodynamics of reversible
processes) and steady state (from the thermodynamics of irreversible processes) is
to risk abusing the terms badly. Perhaps a better concept is that of pedogenic inertia
which relates to the fact that certain soils, once formed are difficult to dislodge from
the state they have reached (Bryan and Tfeakle, 1949).
None of this discussion should be taken to mean that equilibrium or steady
state models are useless in soil science. As in many other systems, such models are
necessary to constrain our hypotheses and speculations about the real world.
Zonality of soils
Another aspect of the soil-energy relationship relates to the global distribution,

or zonality, of soils. An uneven distribution of energy at the surface of Earth leads
to unequal rates of chemical and biochemical activities and of physical processes
in weathering systems. Since it is these activities and processes that produce soil, it
should be possible to relate soil variation to energy distribution. Volobuyev (1984)
attempts this (Table 1.1), the resulting picture showing an obvious relationship to
the climatic scheme presented by Chesworth (this book). In fact, climate as a first
order derivative of the energy balance can be looked on as a major driving force
of the weathering system. In other words, there should be a close correlation be
tween soil zones and climatic zones, and on those continents where there has been
a relatively stable geological environment in the recent past (Africa and Australia
for example) soil zonation does indeed parallel climatic zonation (Table 1.1). On
no other continent is this unambiguously the case, though the pre-eminence of cli
mate (Duchaufour, 1982) was first recognized in the concept of climatic zonation
of soils in Eurasia by Dokuchayev (1899). According to him soils are distributed in
essentially latitudinal zones since climate, when not disturbed by large mountains
and bodies of water, shows a latitudinal zonation. He was aware that other factors
could potentially disrupt this latitudinal zonation, though two of the principle ones
(parent material and topography) were also zoned latitudinally in the transect he
studied, and a third (vegetation) closely follows climate, so that no significant devi
ation from a latitudinal distribution of soil zones was found. However, when other
soil-forming factors do not vary sympathetically with climate, the climatic zonation
of soils follows no simple pattern. The point though, is that the zonation is still
an expression of energy zones at the surface of the Earth, and at their simplest
they correlate well with climatic zones (Volubuyev, 1984), and hence such climatic
variables as temperature and precipitation.
Soil zonality shows up in other ways. In mountainous regions climate is variable
on a vertical rather than horizontal scale. Consequently soils tend to be zoned
in a vertical sense too, as Legros (this book) clearly demonstrates. The nature
Reflections on soils andpaleosols
TABLE 1.1
Ecosystem and energetic conditions at the Earth's surface

100
extreme semidesert dry savanna equatorial

desert tropical savanna forest
80
60 parkland- moist
\ steppe subtropical
dry
40 H desert forest- meadow-
desert semidesert steppe forest
meadow- forest-
20 thin forest tundra
tundra
0.2 0.4 0.6 0.8

a—
Vertical axis: R = radiation balance of the Earth's surface, kcal cm~ 2 per year ( = thermal series);
horizontal scale: a = index of radiant energy utilization in biogeocenoses which in part is correlative
to the arid-humid gradient utilized by Birkeland in his analysis of the Quaternary soil chronosequences
(= hydrologie series). From Volobuyev (1984).
of the proton donor in soils also varies with climate. In the boreal zone, where
organic matter accumulates, its decomposition products are the important proton
donors. Further north, production of organic matter is not so prolific and H2CO3
becomes the dominant donor of protons (Ugolini, 1986). Further south, where
microbial breakdown of organic matter prevents the litter accumulation, H2CO3
again dominates the acidity characteristics of the system. Exactly the same zonation
is seen related to altitude in mountains (Legros, this book).
The complicating effect of the water balance is seen in the limits imposed on
ferralsolic and lateritic areas. These are bounded by belts of Vertisols and smectitic
soils, where semi-arid or arid conditions prevail (Tardy and Roquin, this book).
In lateritic areas, as in other zones, clay mineral distribution is strongly affected
by drainage conditions: gibbsite especially, and kaolinite to a lesser degree, are
indicative of good drainage, while smectites may indicate poor drainage (Tardy and
Roquin, this book).
Clay distribution is also commonly related to climatic zones as Evans (this book)
shows, with 1:1 sheet silicates dominant in hot wet climates and 2:1 sheet silicates
and chlorites being more important in cooler climates. However, the production
of a given mass of kaolinite from feldspar say, requires an equivalent amount of
energy no matter where the reaction occurs. In colder climates, it takes longer to
accumulate that energy than in hotter ones. So the reaction may proceed but the
effect will be less obvious over the short term in higher than in lower latitudes. The
Fig. 1.1. Soils map of the world (source: Soil Conservation Service, U.S. Department of Agriculture, 1971, Provisional map). For identification of principal
suborders and great groups, see original map or Miller and Donahue (1990).
Reflections on soils andpaleosols 9
type of clay mineral cannot, of itself, be a climatic indicator, but the intensity of
its production might be. Degree of alteration needs to be distinguished from type
or alteration (Pedro et al., 1975). A good example is the rubification of weathered
materials. It develops in thick sequences of red beds in tropical and subtropical
regions (Tardy, this book), which Pedro places in his zone of rubification (Pedro,
1968). However, some reddening occurs in arctic deserts, and in sub-arctic, boreal
and temperate zones, though compared with reddening in the tropics, it is much less
intensely developed. Yet there is one aspect of it which does show a clear climatic
zonation: hematite is associated with the phenomenon in lower latitudes and is
absent in higher ones (Schwertmann, 1988).
It is clear from all this that temperature (the intensity factor of the thermal
energy state of a system) is one of the most important but not the only variable
involved in establishing soil zonalities. Other environmental factors such as atmo
spheric circulation (hence precipitation, evaporation and snow cover), and ground
elevation (hence insulation), affect soil processes and distributions. Superimposed
on these soil gradients there is the parent material factor particularly important
in recent orogens (actively developing mountains) or otherwise recently exposed
landscapes, where numerous types of rocks outcrop, from freshly exhumed igneous
plutons and volcanic rocks to highly dissected sedimentary and metamorphic rocks.
A second point to state is that whereas the whole soil profile needs to be studied
and understood in recent settings, the features and mineralogies that form under
certain energetic conditions in deeper soil horizons (horizons that are more readily
preserved than surficial ones) are of great importance in the study of paleosols. A
corollary is that few soil features, if any, may be unique to a particular zone on
Earth. In spite of this, their parageneses and frequency of occurrence are the keys
to use in discriminating between paleo-landscapes.
Mineral concentrations in soils
Geology and soil science overlap in the fact that concentrations of minerals of
economic or environmental importance may occur in soils. These concentrations
develop either during weathering or are fostered by the presence of a soil pro
file during burial (the diagenetic stage). Nahon et al. (this book) demonstrate the
importance both of parent material and of dissolution and re-precipitation during
weathering, in the formation of concentrations of gold and aluminum in latentes. In
such a process, weathering is an efficient natural agent for refining ores such as Au,
Al, Mn, Ni, Cu (Nahon et al., this book).
Concentrations of metals of environmental rather than economic significance are
also possible in organic soils (peat), with Fé, Cu, U, Zn, S being examples. It is
well known, for instance, that dissolution of mafic minerals under acid anaerobic
conditions in some peatlands leads to a much faster rate of release of Fe than in
oxygenated conditions. Release of iron and its transport and re-precipitation in ox
idized zones, leads to the formation of the "bog ores" of Fe and Mn. Shotyk (this
book), inter alia, analyzes ways in which dissolved organic acids in peatland waters
promote mineral dissolution through hydrolysis and complexation. An interesting
byproduct of this is that any volcanic ash blown on peatland is readily modified into
well-crystallized kaolinite forming the "tonstein" layers common in Paleozoic coal
measures (Hughes et al., this book). Another interesting consideration is that recent
(and ancient) wetlands and peatlands are indeed powerful regulators of surficial
waters and the Earth's atmosphere. They constitute a sink for carbon and other
biophile elements, and they are a source of methane, one of the so-called green
house gasses. Also iron may be re-precipitated locally in peatland environments in
the form of sideritic nodules, for instance. This regulation of the carbon cycle by
organic soils on a global scale is currently under intensive study. Such study is a
necessary preliminary to any interpretation of comparable paloenvironments such
as existed in Permo-Carboniferous times, when enormous peatlands formed on the
Earth's surface and regulated the evolution of the atmosphere and the hydrosphere,
and hence of living species.
Sedimentary facies and soil catenas
One of the parallels between geology and pedology is provided by the similar
concepts of sedimentary facies and soil catenas. In sedimentary geology the term
facies is used to label sediments or rocks which are characterized by a unique set of
properties related to lithology, texture, structure and organic remains. Confusingly,
the term has been used to describe a rock unit, to refer to the product of a process,
or to establish environments of sedimentation. When used in relation to environ
ments, a hierarchical system is usually established, where (1) assemblages of single
characters of the deposits (composition and texture of particles) form "structures"
(laminations) which indicate processes, (2) assemblages of structures and other fea
tures, such as organic remains, determine the "facies" and provide information on
some aspects of the environment, and (3) assemblages of facies give information on
the overall environment.
Tb be useful in paleo-reconstructions, the sedimentary system must be aggrada-
tional so that at least parts of the facies sequence are buried and preserved. Under
such circumstances sediments and sedimentary rocks are most readily studied in
vertical sections. Walther (1893-1894) stated that in the absence of major uncon
formities (which represent missing record) the vertical sequence of facies observed
at one locality is similar to the lateral sequence of facies developed on a land
scape, representing a specific time surface. This "law" is extremely useful because
it allows lateral extrapolation of environmental interpretations to be made on the
basis of vertical profiles (exposures, or cores) at one or a few localities, leading to
Reflections on soih and paleosols 11
a reconstruction of paleogeography at any recognizable (datable) geological time

(Middleton, 1973). It is important to note that time lines do not follow lithological
boundaries but on the contrary, they cut across lithologies. For example at the same
time sand is deposited on a beach, silt is formed on a shelf and clay farther offshore.
Some of these concepts are applicable to the study of soils and paleosols, while
others are not. Recognizable units such as skeletal particles, plasma, organic mate
rials and so on, are organized into horizons, which are grouped into pedons. The
pedons grade into each other as the soil blanket on a slope, to make up a catena,
and an ensemble of catenas makes up a soil landscape. Each soil group on the
landscape is equivalent to a sedimentary facies as being an indicator of a particular
environment, with different, but contemporaneous soil groups indicating local and
regional changes in the environment or landscape.
Unlike a sedimentary facies however, a soil system is generally not formed in an
aggradational situation, so vertical changes of soil groups are not generally found.
Thus Walther's law of facies does not apply. Exceptions occur locally, of course, for
instance in the terrestrialization processes (Shotyk, this book) of certain peatlands,
whereby aquatic organic matter grades laterally and vertically into partly decom
posed fen peat deposits of graminaceous matter, which are laterally and vertically
transitional into fibrous, Sphagnum bog peat deposits. Additionally, buried soil pro
files can be treated as traditional sedimentary facies. For instance, the degree of
development of soil profiles or their frequent alternation with flood deposits al
lows, under certain circumstances, a reconstruction of the "architecture" of fluvial
deposits, to reveal the evolving paleogeography of floodplains (Kraus, this book).
The idea of facies can be applied to systems ranging from the macro to the
microscopic, with each level of analysis providing valuable information for the un
derstanding of the system as a whole. Microfacies analysis is a routine procedure
in sedimentary geology in the study of carbonate rocks (Carozzi, 1960). In soils,
Robert and Tfessier (this book) show that a full understanding of weathering and its
products requires a similarly detailed study of microsites and microsystems, with for
instance, the tip of the root system providing a very distinct, aggressive environment.
Furthermore, the study of facies and microfacies has lead to a better understanding
of silcretes (Milnes and Thiry, this book): diagnostic characters have been estab
lished to distinguish pedogenic from groundwater precipitated silcretes. The former
require a longer period of land surface stability to develop than the latter.
Paleosols and time
In spite of Einstein, geologists follow Aristotle in thinking of time as a one way,

evenlyflowingstream. Apart from putting a date to events, time is the variable with
respect to which duration and rate of change are measured. As far as the duration
of specific weathering or sedimentary events are concerned, they will in general be
12 LP. Martini and W. Chesworth
expected to persist for a relatively longer geological time in a cratonic area (tecton
ically stable parts of continents), than in tectonically active zones affected by earth
quakes, faulting and a more intense erosional regime. The information imprinted in
soil profiles varies within the different tectonic realms as well. In tectonically active
areas, along the San Andreas fault for instance, the presence of soils repeatedly
buried under alluvium or colluvium along fault scarps, can be used to estimate the
duration of quiet periods between faulting events (Birkeland, this book). For very
different reasons, similar repetitive changes in the occurrence and degree of devel
opment of soil profiles, occur in cratonic alluvial environments. The changes in this
case depend primarily on the frequency offloodingand siltation events, which may
be independent of local tectonic settings. Thus thefloodingof certain cratonic areas
may result from the climatic conditions of distant mountain belts. This recurring
rejuvenation of the soil profile is used to advantage by Kraus (this book) to ana
lyze in detail the formation and significance of paleosols in aggrading sedimentary
environments, that is, in environments where short lived exposure of the flooded
land surface to weathering can occur. In some cases weathering is sufficiently rapid
to leave paleoecological markers of each exposure of short duration. Kraus (this
book) views the "composite" soil sequences of alluvial deposits formed in this way,
and the more mature soils formed on unconformities, in terms of two processes.
These are intrabasinal processes, such as the shifting of position of crevasse splays,
and extrabasinal processes, such as tectonic activity and eustatic sea level changes.
Similar concepts have been used by Hughes et al. (this book) to dispel the long held
idea that underclays form in situ in a homogeneous, single event. They interpret the
underclays of the Illinois Basin to be composite soils which were developed mostly
during pre-peatland periods, through several episodes of sedimentary deposition
and pedogenesis. Possibilities of this kind, as well as the effect of changing environ
mental conditions on a single soil profile [concepts of the relict soil of Ruhe (1965)
and of the vetusol of Cremaschi (1986)], produce polygenetic soils and paleosols
that are difficult to interpret (Birkeland, this book; Wright, this book).
Ancient stable continental surfaces which have experienced little or no, or low
rates of, deposition and erosion, may be lined by discontinuous but locally thick
paleosols (Wright, this book). In a stratigraphie sequence such surfaces represent a
period for which the geological record is missing, that is, they are unconformities
or diastems of variable duration. Most Precambrian paleosols, for example, line
major unconformities of a duration far greater than that required for soil formation
(Retallak et al., 1984). Most mature paleosols, therefore, do not record a single set
of environmental conditions, but contain evidence of events that predate the regime
that existed just before burial.
Further changes are possible after burial. Unconformities commonly act as con
duits for mineralisingfluids.Where the unconformity is lined by a paleosol, metaso-
matic changes are possible. Although the paleosol may still be recognizable, infor
mation (especially of a paleoecological kind) will be lost.
Reflections on soils andpaleosols 13
Paleosols and life forms
Ultimately, the analysis of paleosols yields information on the biosphere. Retal-

lack (this book) addresses this fundamental question and traces the evolution of
types of soils in relation to evolving life forms. The ancient soils indicate that vascu
lar plants and advanced life forms such as millipedes existed at the surface of Earth
as early as Ordovician times. This is much earlier than commonly thought. Strong
evidence of colonization by vascular land plants during Silurian times is provided by
globally widespread and diverse fossil spores, sporangia and stems. Evidence for the
development of woodlands and forests in Devonian times, is provided by well devel
oped root-traces in paleosols. Similar evidence indicates early freshwater marshes
possibly in late Silurian, and certainly in Devonian times. Finally, peats developed
in late Devonian, and proliferated during the Carboniferous.
Retallack (this book) also indicates that soil groups can be recognized through
out geological time. Some extend back to the Precambrian (Entisols, Inceptisols,
Aridisols, Vertisols), while others have been recognized from Paleozoic times (His-
tosols, Alfisols). It may be possible, as Lindsay (this book) suggests, to use evidence
from the moon or other planets to provide insight into the Earth during the early
Archean, a period for which the simple application of uniformitarianism based on
modern terrestrial soils in not possible.
Classification
The process of classifying is complex and never-endingly controversial. Ideally

a classification should be able to carry basic information about a set of items in
a simple fashion that makes it understandable and useful to researchers. The best
classifications for scientific use have heuristic content, that is they contain the funda
mental property of prediction. From such general classifications, derivative ones can
be constructed to deal with local needs or interests. In this book some of the basic
concepts used in classifying Earth materials are put forward by Chesworth et al. (this
book); however, no single classification has been imposed on the various authors.
Consequently several soil classifications have been used in the different chapters,
and to ensure some common understanding, each term is provided with synonyms
from the FAO-UNESCO or American Soil Tkxonomy Classifications. Watson (this
book), for instance, deals in some detail with the classification of desert soils of pri
marily warm regions, and compares and contrasts various procedures. Campbell and
Claridge (this book) deal with the soils of cold climate regions. Such classifications
are based primarily on morphological and mineralogical criteria, and although the
names of taxonomic units may have genetic connotations, the terms are used here
in purely descriptive ways.
Retallak (this book) and other authors try to recognize paleosols and hence clas
sify them, utilizing modern soil classifications. This is a classic use of the principle
of uniformitarianism and, where applicable, it transfers whatever environmental in

formation modern soil terms carry, to the interpretation of the paleosols. But the
principle of uniformitarianism cannot be applied indiscriminately to soils. Caution
and more research are necessary. As Retallack (this book) mentions, "extinct" soils
may exist in the Precambrian record. Furthermore, Chernozems as typical grassland
soils could not develop on Earth before the Mesozoic. A second major drawback of
this transfer of conceptual models, is that recent soil groups and classes represent
"soil facies" which can develop under various environmental conditions (Lavkulich,
1969). It is the associations and the sequences of these facies which unambigu
ously establish the environment of formation. Except for some areas, the paleosol
record is not extensive enough to establish time calibrated, regional and strati-
graphic trends. Hence, interpretations must be based on one or few profiles, rather
than on associations or sequences. Furthermore, paleosols are incomplete. When
for instance a paleosol is truncated, as it often is, and when it is modified by diagen-
esis, much of the information needed to place them in recent soil classifications has
been lost.
If modern soil classifications are to be used for paleosols, then new groups and
criteria must be developed in defining the classes. Variables such as CEC or the
various forms of extractable Al or Fe which are used in classifying modern soils, are
not retrievable from paleosols. Perhaps it is better to analyze paleosols for their own
sake, that is as unique stratigraphie entities. Only after that is done, through the
fundamental axiom of the principle of uniformitarianism can we try to tie them to
quasi-equivalent, recent environments of soil formation.
Future research trends
Soils are very detailed and sensitive indicators of the environment and of en
vironmental changes. Because of their importance to life on Earth, and especially
because of the crucial role they play in sustaining human civilization, there are many
good, practical reasons for studying them. To a geologist, however, their interest lies
in the degree to which they can help in interpreting the past and, to further this
ideal, more interaction between geologists and soil scientists is needed. Interdis
ciplinary cooperation invariably leads to a valuable cross-fertilization of ideas and
techniques. Specific research areas that would benefit from this cooperation can be
singled out.
(a) Before paleosols can be used to provide useful information about past en
vironments, pedological changes need to be disentangled from diagenetic ones.
Unfortunately pedogenesis and diagenesis have a gradational relationship, an over
lap called pedodiagenesis by Klappa (1983). A distinction on the basis of water
table, with pedogenic processes above (in the unsaturated, vadose zone) and di
agenesis below (in the phreatic zone) does not remove the ambiguity. Gleying,
Reflections on soils andpaleosoh 15
a well recognized pedological process occurs below the water table. Ideally the
distinction could be made in the same way that metamorphic petrologists distin
guish between different mineral facies, that is in terms of mineralogical reactions
in a pressure-temperature framework. The stability fields of clay mineral polytypes
might be useful in this regard.
(b) The most comprehensive area that paleopedologists have and can contribute
to, is the study of the evolution of the biosphere. Ancient soils are a repository of
important clues to changes in the biosphere, atmosphere and hydrosphere. Inter
preting the clues is beset with difficulties, but increasingly sophisticated techniques
of analysis provide powerful means of circumventing at least some of them. Since
the Precambrian, the weathering land surface has evolved with life and with an in
creasingly oxygenated atmosphere. Retallack (this book), says "diversification of soil
types [occurred] hand in hand with the evolution of new kinds of vegetation" and we
need to develop models and hypotheses for this symbiotic relationship.
(c) Parallel to the development of the soil-plant weathering system on land, there
were concomitant changes in the marine environment. Much more detail needs to
be obtained in order to tie the pedological and the marine environments together.
Pioneering studies exist (Erhart, 1967; Bryan and Tbakle, 1949), but much remains
to be done.
(d) New technologies need to be pursued. Many geological hypotheses in the past
could not be tested. Some at least can now be modelled using supercomputers.
One final remark: soil scientists have always been aware of the importance of
geology to their science. Many geologists on the other hand have seldom seen soil as
anything other than an inconvenient blanket, obscuring more important materials.
Now that an era has begun in which the geologists attitude is changing, many new
and exciting scientific advances can be expected. We hope that this book points the
way to some of these advances.
References
Bates, R.L. and Jackson, J.A. (Editors), 1987. Glossary of Geology. American Geological Institute,
Alexandria, Va., 788 pp.
Brunn, J.H. and Deprairies, A., 1965. Étude sédimentologique préliminaire de formations à caractère
flysch et molasse. Rev. Géol. Phys. Géol. Dyn., 7: 339-354.
Bryan, W.H. and Teakle, L.J.H., 1949. Pedogenic inertia; a concept in soil science. Nature, 164: 969
Carozzi, A.V., 1960. Microscope Sedimentary Petrology. John Wiley, New York, N.Y., 483 pp.
Cremaschi, M., 1986. Paleosols and vetusols in the central Po plain (Northern Italy). A study in Quater
nary Geology and soil development. UNICOPLI, Ed. Milan, 306 pp.
Dokuchayev, V.V., 1899. Kucheniyu o zonakh prirody. GorizontaPnyye i vertikal'nyye pochvennyye zony
(Concerning the doctrine of natural zones. Horizontal and vertical soil zones). St. Petersburg (as
reported by Volobuyev, 1984).
Dumanski, J., 1969. Micromorphology as a tool in Quaternary research. In: S. Pawluk (Editor), Pedology
and Quaternary Research, National Research Council of Canada, Proceedings of Symposium held
at Edmonton, Alberta, May 13-14, 1969, University of Alberta, pp. 39-52.
Erhart, H., 1967. La genèse des sols en tant que phénomène géologique. Masson, Paris, 177 pp.
Holland, H.D., 1984. The Chemical Evolution of the Atmosphere and Oceans. Princeton University
Press, Princeton, N.J., 582 pp.
Johnson, D.L. and Watson-Stegner, D., 1987. Evolution model of Pedogenesis. Soil Sci., 143: 349-366.
Klappa, CF., 1983. A process response model for the formation of pedogenic calcretes. In: R.C.L.
Wilson (Editor), Residual Deposits: Surface Related Weathering Processes and Materials, Geol.
Soc. Spec. Pubi., Blackwell Scientific Publications, Oxford, pp. 211-220.
Lavkulich, L.M., 1969. Soil dynamics in the interpretation of paleosols. In: S. Pawluk (Editor), Pedology
and Quaternary research, National Research Council of Canada, Proceedings of Symposium held at
Edmonton, Alberta, May 13-14 1969, University of Alberta, pp. 25-37.
Middleton, G.V., 1973. Johannes Walther's law of correlation of facies. Bull. Geol. Soc. Am., 84: 979-
988.
Miller, R.W and Donahue, R.L., 1990. Soils. Prentice-Hall, Englewood Cliffs, N.J., 768 pp.
Nikiforoff, C.C., 1959. Reappraisal of the soil. Science, 129: 186-196.
Pedro, G., 1968. Distriubtion des principaux types d'altération chimiques à la surface du globe. Présen
tation d'une Esquisse Géographique. Rev. Géol. Phys. Géol. Dyn., 10: 457-470.
Pedro, G., Delmas, A.B. and Seddoh, F.K., 1975. Sur la nécessité et l'importance d'une distinction
fondamentale entre type et degré d'altération. Application au problème de la definition de la fer-
ralitisation. C R. Acad. Sci., Sér. D, 280: 825-828.
Retallack, G., Grandstaff, D. and Kimberley, M. 1984. The promise and problems of Precambrian
paleosols. Episodes, 7: 8-12.
Ruhe, R.V., 1965. Quaternary paleopedology. In: H.E. Wright, Jr, and D. Frey (Editors), The Quaternary
of the United States, Princeton University Press, Princeton, N.J., pp. 755-764.
Runge, E.C.A., 1973. Soil development sequences and energy models. Soil Sci., 115: 183-193.
Schwertmann, U., 1988. Occurrence and formation of iron oxides in various pedoenvironments. In:
J.W. Stucki, B.A. Goodman and U. Schwertmann (Editors), Iron in Soils and Clay Minerals. Reidel,
Dordrecht, pp. 267-306.
Ugolini, F.C, 1986. Pedogenic zonation in the well drained soils of the Arctic regions. Quat. Res., 26:
100-200.
Volobuyev, V.R., 1984. Two key solutions of the energetics of soil formation. Soviet Soil Sci., 16(4): 1-8.
Walther, J., 1893-1894. Einleitung in die Geologie als Historische Wissenschaft. Fischer, Jena, 3 volumes
(statement of Walther's law is in Vol. 3, Chapter 27).
19
Chapter 2
Weathering systems
W. CHESWORTH
Conceptual view of weathering systems
This chapter is concerned with chemical weathering, which is described in terms

of the factors affecting the process, and the trends that develop. The latter are con
sidered in terms of Eh (pe) and pH as master variables. Geographical patterns and
the connection between soils and other sediments are also examined. Yatsu (1988)
provides a review of the definitions of weathering to be found in the literature. The
following is a composite view, owing much to the authors he cites: weathering is the
alteration of materials at, or near, the earth's surface, as a response to the condi
tions that prevail there. It results from a complex set of interactions, usually far from
equilibrium, between the lithosphère, atmosphere, hydrosphere, and biosphere. The
process takes place in a zone that extends from the surface of the earth down to the
zone of diagenesis. Since the transition is gradual, any boundary between the two
zones will be arbitrary.
Conventionally, weathering is divided into physical and chemical types. In reality
both occur together, with physical weathering, whether by physical disaggregation
of rocks in place, or by dynamic disintegration of particles during erosion and
transport, playing the important role of exposing fresh surfaces to chemical attack.
Dokuchaev introduced a "factorial" approach to weathering systems, when he
considered soil to be a product of the combined activity of the following factors: liv
ing and dead organisms, parent rock, climate and relief (Strzemski, 1975). Outside
the Russian sphere, Jenny's (1961) version of this approach has had the most influ
ence. Figure 2.1 is an attempt to display weathering factors within the lithosphere-
atmosphere-hydrosphere-biosphere framework of the definition of weathering pre
sented above. Unlike Jenny's approach, which stresses the independent action of his
factors, Figure 2.1 clearly emphasises interaction and feedback.
As the term "weathering" implies, a case can be made for considering climate to
be the principal driving force and hence for climatic factors being the fundamen
tal ones. For example Tardy (1986) shows how weathering is completely dependent
20 W Chesworth
EXTERNAL INPUTS
Solar energy.
New material
Atmo w/ from mantle
\£> Meteorites
Type ana density of x::::::: ATMOSPHERE

living species, dead xjxj:. . . . . . . . . . . . . . . . . . . . . . ...^...r+
and decaying oiganic;:::;:;:;:;:::;:;:;:;:;:;:^^
matter, human ^^ψχ^χ|χψχ^:|χ·:|χ|: : :'ρέΓΐίουΐ3ΐβ8, climate,
activity:::::v::::?fc£:£
• x ^ x WEATHERING
::SYSTEMv:
Water aval lability J ^ x x x ï l

residence time, drainage, Χ χ Χ ; Χ : \
position and fluctuation of::::::::"
water takle:::::X:\x::::::
HYDROSPHERE ::oèòmòrph'ology,
^ x j x c h e m i s t r y , mineralogy
Jtexture, structure,
Sxgeological history
LITHOSPHERE
Fig. 2.1. Factors affecting the weathering system: a schematic representation of equations such as Jenny's
(1969), but emphasizing interrelationships and feedback.
on the water cycle with the latter being driven by climate. Prima facie evidence
is provided by the distribution of soils in areas of the earth's surface which have
been geologically quiescent in the recent past. Thus, in regions such as the African
continent, soil zones are broadly coincident with climatic zones. However, where
geological changes have been recently active, for example in areas where Ceno-
zoic orogenies and Quaternary glaciations have taken place, climatic factors alone
are insufficient in deciphering the modifications of the earth's surface caused by
weathering. In such cases, a large geological and geomorphological effect is seen.
The qualitative view of the weathering system shown in Figure 2.1 contains fac
tors that are difficult, if not impossible, to quantify. They are not therefore, con
ducive to a rigorous or even a semi-quantitative, analysis of the chemistry of weath
ering. For this latter purpose, a better approach is to simplify the process into the
interaction of inorganic solids with an aqueous phase. This may seem to concentrate
on the lithosphère and the hydrosphere to the exclusion of all else, but atmospheric
and biospheric factors can always be considered as acting through the medium of
water, the virtual pre-requisite phase in chemical weathering. In effect the aqueous
phase constitutes the environment to which the solids are responding by weathering,
Weatherìng systems 21
and through which the physico-chemical factors that operate at the surface of the
earth are channelled.
The inorganic solids
The land surface has a lithospheric base, and the lithosphère is generated by the
internal geological cycle (Figure 2.2), which is driven by radiogenic heat. Overall,
the composition of the surface (Table 2.1) is close to andésite (Taylor and McClen-
nan, 1985), though in a given locality the lithospheric materials that are available
to the weathering system will depend on the previous geological history of the sys
tem being considered. The various tectonic environments impose distinct chemical,
mineralogical and petrological signatures on the land surface. An important fac
tor in determining weatherability is whether the land surface is recent, ancient or
renewed by erosional processes, such as glaciation (Chesworth, in press). The inor
ganic solids that make up the lithosphère are conventionally divided into primary
(inherited) and secondary (authigenic) phases, the implication being that the pri
mary minerals in a weathering system are produced by high temperature and/or
high pressure igneous and metamorphic processes, and constitute the weathering
parts of the system. The secondary minerals, in this usage, are complimentary to the
primary and are taken to be neoformations produced at the expense of pre-existing
phases.
Fore Fore 0ld

Ocean Ocean Accretionary Magmatic
arc arc
ridge basin wedge arc continental
ridge basin crust
Fig. 2.2. Relationship of the endogenic and exogenic cycles in the context of the theory of plate tectonics.
The subduction zone model is based on dimensions from modern Sumatra, and is derived from Hamilton
(1989).
22 W Chesworth
TABLE 2.1
Composition of the surface of the continental crust—the average parent material of the weathering
system
Chemical composition Mineralogical composition

(weight %) (volume %)
Si0 2 66.0 Quartz 12

Ti0 2 0.5 K-feldspar 12
AI 2 O 3 15.2 Plagioclase 39
FeO 4.5 Mica 5
MgO 2.2 Pyroxene 11
CaO 4.2 Olivine 3
Na 2 0 3.9 Clay minerals (and chlorite) 4.6
K20 3.4 Calcite (and aragonite) 1.5
99.0 Dolomite 0.5
Magnetite (and titanomagnetite) 1.5
Others 4.9
Source: Taylor and McClennan, 1985; Ronov and Yaroshevsky, 1969.
As phases inherited from parent materials, most earth-surface systems contain

inorganic solids that are predominantly silicate (including aluminosilicate) in na
ture. A second important group is the carbonates, particularly calcite and dolomite.
Sulphides may be of more local significance than either silicates or carbonates, and
under special circumstances, various soluble salts, particularly sulphates and chlo
rides, may be present.
The principal secondary phases are the aluminosilicate clay minerals, oxides and
hydroxides, with carbonate precipitation common in some environments. Less com
monly the precipitation of sulphides, sulphates and chlorides may significantly in
fluence the weathering path and the concomitant composition of natural waters.
Initially formed phases will tend to be metastable and age to more stable phases
with time. Neoformations will also tend to increase in grain size. If weathering
conditions change, neoformations under an older weathering regime may become
weathering phases in the new one. Thus, the classical division between primary
and secondary minerals is not absolute. The most soluble of the weathering phases
present will determine the initial chemical composition of any aqueous phase, and
the initial weathering trend. Important features are grain size and crystallinity, with
decreases in both parameters, increasing the rate of chemical attack on unstable
solids.
The aqueous environment
The functions of water, an absolute pre-requisite if chemical weathering is to

take place to any significant degree, in the weathering zone are summarised in Table
Weathering systems 23
2.2. This is sometimes described as a catalytic role, though it is not a case of true
catalysis in the sense that at the completion of the process the water is regenerated,
unchanged.
The aqueous phase, in the simplest case, enters the weathering system as atmo
spheric precipitation, atmospheric (including climatic) factors determining chem
istry and abundance. The oxidation state of the incoming water is equivalent to the
Po2 °f the atmosphere (0.21 bars) while the pH of about 5.7 is the pH in equilibrium
with atmospheric PQO2 at 10~3 5 bars. It may be more acid than this because of the
presence of airborne contaminants from industry, automobile exhausts, and so on.
Atmospheric precipitation also has a similar compositional ratio of dissolved salts
(picked up as aerosols) as sea water, even when distant from the sea (Gibbs, 1970).
Within the weathering system, the aqueous phase is rapidly modified by interaction
with solids, including the biomass. Dissolution of the most soluble minerals present
in the system as weathering phases, and precipitation of the least soluble minerals
forming as new phases will tend to control the inorganic composition of the aqueous
phase. Residence time of the water will be another important factor in this context,
so that a complex set of feedback loops connects the aqueous phase with other
weathering factors.
The hydrodynamic situation is also of importance since theflowof water through
a weathering system is one of the major determining factors in the evolution of
TABLE 2.2
Functions of water in the weathering zone
Physical functions
a. Agent of physical transport.
Gravitational and evapotranspirational potentials are determining factors on a macro-scale.
Capillary forces are important on a micro-scale.
b. Medium through which reactants diffuse at reaction sites,
such as at the solid-liquid interface.
c. Exerts a partial pressure.
Directly re la table to chemical potential and activity.
d. May help physical disintegration of solids,
such as the freeze-thaw mechanism.
Chemical functions
a. A solvent.
Its anomalously high dielectric constant accounts for its ability to dissolve more solids than
any other solvent in the natural world.
b. Necessary component of all the typical reactions of the weathering zone,
such as hydration/dehydration, acid/base, solution/precipitation, ion-exchange.
c. Constituent of the principal new phases formed during weathering,
such as hydroxides, clays, amorphous substances.
d. Acts as a chemical buffer.
In humid climates the aqueous throughput of a system generally ensures that natural waters
belong to only a few types, each with a fairly uniform chemistry.
24 W Chesworth
the system. This factor depends upon the volume of water available, texture of
the weathering materials, landscape position and position of the water table. The
importance of landscape position is obvious in the case of a soil catena. Members
of the catena downslope, receive drainage waters that may have already reacted
with materials upslope. This topographical influence is best illustrated in the clay
minerals commonly found in catenary sequences of the humid tropics.
On the highest point of a well drained surface gibbsite is common, with kaolin-
ite downslope, while in the valley bottom smectite occurs. Such a sequence can be
explained by the increasing concentrations of H4S1O4 and metal cations in ground-
waters travelling laterally downslope.
The effect of the microenvironment on the aqueous phase must also be con
sidered. Microenvironemt is important in two ways. The size and interconnectivity
of pores controls hydrodynamics at the local scale and the solids surrounding the
aqueous phase will vary on a micro-scale and produce local variations in aqueous
geochemistry. Such local variations will yield secondary reaction products which
reflect these variations, (Meunier and Velde, 1979). At this micro-scale equilib
rium thermodynamics may possibly be applied (Korzhinskii, 1957; Thompson, 1959;
Chesworth and Dejou, 1980), with the total macrosystem being treated as an ensem
ble or mosaic of equilibrium microsystems.
A further point concerns the thermodynamic activity of water (ΛΗ2Ο) in the
aqueous phase. Although it is normally assumed to be 1 (the value for pure water)
or close to it, there are weathering environments (saline ones) where ΛΗ2Ο may
be much less than one. In such conditions, a number of important dehydration
reactions become possible (Chesworth, 1972; Tkrdy and Nahon, 1985; Mattigod and
Kittrick, 1980), such as gibbsite to boehmite, and goethite to hematite.
A major influence on the chemistry and behaviour of water in the weathering
system is the biomass. Organic compounds are clearly important in soils, although
decomposition by bacterial action and attenuation with depth, determine that the
major effect is generally concentrated in the upper metre or usually less of the soil
surface. Consequently, Jackson and Sherman (1953) suggest that the soil compart
ment weathering zone is qualitatively different from the rest of the weathering zone.
In other words pedochemical weathering may be distinguished from a more general
geochemical weathering.
The biomass is important in that it produces breakdown products which serve
two functions. First, many of the products are acid and can directly attack a min
eral surface to release ions into solution. The insoluble representatives of these
materials are called humic acids, and the soluble fulvic acids, two operational terms
for organic materials formed on the land surface from biological debris. They con
tain a number of functional groups such as carboxyl, phenol, alcohol, carbonyl and
methoxyl. Thefirsttwo functional groups are potentially important in providing H +
to the weathering system, with simple carboxylic acids, such as oxalic and citric,
being the principal ones concerned in mineral breakdown. The simple acids do not
persist to any great depth and have no complexing role. Acidity in the soil compart
ment is initially set by carboxylic H + and may be lowered to the range 3.5 to 4.5.
With attenuation of the organic factor in the upper solum, dissolved CO2 becomes
the principal determinant of pH.
Certain organic compounds are important in soils as complexants. In the ful-
vic fraction the complexants are capable of holding in solution, cations that would
otherwise precipitate (Robert et al., 1980). In the humic fraction they act as com
plexants on mineral surfaces polarising metal-oxygen bonds and initiating mineral
breakdown, as well as serving as a temporary stockpile of cations (Stumm, 1986).
Microbial activity, especially in the soil, is one of the most important, though
poorly understood, driving mechanisms of the weathering system. Microbial redox
reactions in particular are important mechanisms for mobilizing metals in natural
environments (Beveridge, 1989).
Major weathering trends in a pe-pH framework
The foregoing discussion and Figure 2.1 indicate something of the complexity of
weathering systems. However, simplifying the environment of the system to an inor
ganic solid phase/aqueous phase interaction provides a convenient way of classifying
weathering trends. In essence the boundary of the system is drawn around the solid
phases, an approach introduced by Korzhinskii (1957) in the study of metasoma
tism, and thus the aqueous phase constitutes the thermodynamic surroundings or
environment.
Considering the prevalence of redox and hydrolysis reactions in the weathering
zone, the activities of the electron and of the proton have a special significance.
Thus, the most generally useful master variables of the weathering environment will
be pe (or Eh) and pH. The spread of Eh-pH conditions that can be expected in
the weathering environment is approximately as shown (within the stability field
of water) in Figure 2.3. If a number of rather rare environments at the surface
of the earth (acid sulphate soils and weathering vanadium deposits for example)
are ignored, the normally expected conditions can be seen to cover an area shown
in Figure 2.3. This figure shows three salients, which correspond to three lines of
evolution in soils and other weathering materials.
The trends are directly related to the behaviour of water. The acid and alkaline
trends are in the oxidising or water- unsaturated zone of weathering. The acid trend
requires a net leaching or downward movement of water to the water table. The
alkaline trend is found in dryer environments where a net deficit of water because
of high évapotranspiration potentials, persists. The reduced trend is found in water-
saturated conditions in a weathering profile. In this chapter, soils which have not
reached the extremes of the three trends will be grouped together as "non-extreme"
soils, or weathering systems.
26 W Chesworth
Fig. 2.3. The Eh-pH framework of alteration under Earth-Surface conditions. Extremely acid environ
ments (weathering sulphide deposits, acid sulphate soils for example) are ignored. The irregular outline
within the stability field of water is meant to represent most weathering environments. It comprises
three salients, each representing a specific weathering trend: A the acid trend, S the alkaline trend and
R the reduced trend.
The acid trend
The acid trend is the common one in weathering and is found in regions of humid
climate on materials with a generally unimpeded drainage. A combination of hy
drolysis between water and solid phases, and residence time of water in the system,
controls the composition of the aqueous phase. The movement of water is pre
dominantly downwards, hydrolysis is the principal reaction and leaching produces
a change in the products of weathering towards compositions made up chiefly of
Goldschmidt's hydrolysate, resistate and oxidate elements (Figure 2.4). A progres
sive acidification of the weathering materials and a trend towards the acid salient of
Figure 2.3 is produced.
The chief characteristic of acidification of the weathering environment is the pro
gressive replacement of cations on exposed mineral surfaces by protons. At its most
fundamental, it can be looked upon as the titration of quenched igneous rock (as the
primary parent material) against acids derivable ultimately from volcanic sources or
from the biomass. As such, this results in the land surface being "inevitably over
titrated, acting as a sink for protons" (Edmond et al., 1979). Alternatively, the pro
cess can be considered to be a form of low temperature hydrogen metasomatism,
analogous to the higher temperature form recognized in hydrothermal systems by
Hemley and Wones (1964).
Weathering systems
B. Mobilised as hydroxy complexes in acid trend, or as low
A. Long range mobility as simple cations in all but the extreme oxidation state ions (Fe, Mn) in reduced trend. Immobilised in
alkaline trend, where they precipitate in semi-arid and arid alkaline trend by precipitation as carbonates. May be immobi
environments as carbonates, sulphates or chlorides. May be lised in oxidised systems by adsorption on Fe and Mn oxides
s
immobilised by adsorption on clay sized aluminosilicate and hydroxides.
phase.
1.6 12.Ba
8-Ba
C . Immobile in acid (except podsolisation) and alkaline trends.
- 1.2k Precipitate as hydroxy (Fe, Mn) as crystalline and amorphous
Rare > ^ ^ aluminosilicate phases. Mobilised in podzolic systems as
Earths metal-organic complexes (Al, Fe), and in reduced systems as
8«Zr
^ low oxidation state ions (Fe, Mn).
6 »Cu U
Mg<.'
Zn/6f.Ga
6
-J
# {·Μη ^ J D. Mobilised as hydroxy acids in all trends. Si incorporated in
4( Ga 4^]^0 aluminosilicate clays, allophanes, and as biogenic opal.
<·ΑΙ z / r . l 6
4-S
Ionic charge (z) E. Long range mobility as oxyanions, except in alkaline trend.
Precipitate with alkali metals and alkaline earths under arid
conditions.
Fig. 2.4. Mobility/immobility characteristics of chemical elements in terms of ionic potential. This is a highly modified version of the Goldschmidt-Mason
diagram (Mason and Moore, 1982, p. 162) and the diagram of Pedro and Delmas (1970). Sections C and D coincide with Goldschmidt's hydrolysate
elements. The numerals beside each point on the graph indicate coordination number.
s
28 W. Chesworth
In Jackson and Sherman's (1953) division of the weathering zone into two parts,
the general geochemical compartment of the weathering zone, has the system H 2 0 -
CO2 as its principal source of protons and is dominated by bicarbonate waters. The
acidity of such a system is initially determined by the atmospheric H2O-CO2 equi
librium, which results in a pH of 5.7. Within the soil, the respiring cells of roots
produce more C0 2 (Epstein, 1977) and, with organic decay, this can raise pC0 2
some two orders of magnitude higher than the atmospheric value, to produce an
equilibrium pH around 4.5. In other words, the pedochemical compartment is likely
to be more acidic than the geochemical one, so that a more vigorous hydrolysis will
tend to take place there. In addition, in soils organo-metallic complexes may be sig
nificant, so that Pedro and Sieffermann (1979) recommend the term complexolysis
for the resulting combination of hydrolysis and complexation.
In detail, hydrolysis, mediated by H + , can be envisioned as a two stage process
(Nicholls, 1963). First, a mineral surface can be considered as a break in the ideal,
electrostatically balanced, mineral structure. The broken bonds at the surface are
ultimately satisfied by whatever ions the local environment is capable of providing.
The commonest are OH", which attaches to positive bonds, and H + (or Η3θ + ),
which attaches to negative bonds (normally to oxygen). The net effect is to hydrox-
ylate the surface. This surface hydroxylation can be considered the first stage of
weathering.
The second stage involves the breaking of internal bonds between cations and an-
ions, a process encouraged by the polarisation of near surface metal-oxygen bonds.
The resulting effect is to release hydroxylated or hydrated cations into solution and
to leave behind a refractory residuum. The solvated cations are transported in pro
portion to their solubilities (Figure 2.4) in the aqueous medium which is commonly
oxidising and acid. The residual products provide the raw materials of the clastic
sediments.
If igneous rocks are considered as the initial solid reactants of the weathering
system (the pristine, or primitive, unweathered system), the compositional changes
attendant upon the evolution of the system towards the acid salient are shown di-
agrammatically in Figure 2.5, and the weathering environment can be classified in
terms of its geochemical and mineralogical maturity as shown. Of course, weather
ing does not need to start with primitive materials, but can begin with sedimentary
products or their metamorphic equivalents, which have already progressed some
way along the maturity scale. Progression towards a mature endpoint may therefore
involve more than one cycle of weathering. This progression is inherent in the con
cept of a rock cycle, which, in the plate tectonic model, is more of a conveyor belt
than a cycle with the ocean floor moving from spreading centre to subduction zone,
and acting as a conveyor of rock from mantle back to mantle.
Weathering systems
50% CaO + MgO 50% CaO + MgO
20% Na. Al 2 0 3 + Fe 2 0 3 20% Na20 27 Si0 2 + Al ? 0 3 -e re ? 0 3

Oxides, hydroxides and 1:1 clays dominate in clay fraction.
Low in plant nutrients, high in hydrolysate elements.
Fig. 2.5. Three gradational weathering stages for igneous rock in humid climates, based on Goldschmidt's (Mason and Moore, 1982, p. 174) division of
the major components of the earth's surface into résistâtes, hydroiysates and oxidates (the Si0 2 + A1203 4- Fe 2 0 3 apex), carbonates (the CaO + MgO
apex) and evaporates (the Na 2 0 + K 2 0 apex).
8
30 W Chesworth
The alkaline trend
The alkaline trend occurs in relatively arid regions with a net annual water deficit.
Net movement of water is upwards in the weathering zone, although a seasonal dis
tribution of rainfall may determine that, for part of the year, there is enough water
to effect a solution and downward leaching of soil components. However, capillary
rise and evaporation of soil water are the distinctive features of pedogenesis, with
evaporation and precipitation of minimal phases being the results. Geochemically,
the result is that a number of electrolytes, especially Na + , Mg2+, Ca2+, HCOJ,
CO3-, SO4", Cl"and COJ, may reach high concentrations in the system. These
solutions migrate in the profile and salts may accumulate in certain horizons (Spiers
et al., 1984). Mineral phases, such as alkaline carbonates and zeolites, precipitate in
such systems in order of increasing solubility product, with each precipitating salt
acting as a chemical barrier separating pairs of evolutionary trends (Eugster and
Hardy, 1970). Salts precipitated in soils appear to be controlled by similar chemical
barriers as salts precipitated in evaporite systems in the East African rift system, or
in the Permo-THassic of western Europe.
In pedology, alkalinisation is also known as solonization (Buoi et al., 1980). It
may lead to the accumulation of toxic amounts of certain elements, such as B and
Se for example. The related process of dealkalinisation (solodization) occurs when
water becomes available, perhaps as a consequence of climatic change or of land
irrigation, toflushthe salts out of the soil system to the groundwater table.
The reduced trend
The fundamental characteristic of the reduced trend of weathering is the pres

ence of an excess of water so that the weathering profile is completely or partially
submerged. Generally, this saturation with water leads to a lowering of the partial
pressure of oxygen in the system to the degree that anaerobic conditions develop.
Subsequently elements such as Fe and Mn are present in a reduced state, a state
in which they are readily mobilised. This mobilisation is in direct contrast to their
behaviour in oxidising environments. Close to the water table,fluctuationsin level
produce alternating redox conditions, with a resulting alternation of mobilisation
andfixationof Fe.
The major visible changes are those associated with iron, and the most signifi
cant feature is the production of ferrous iron in solution. In this form the iron can
leave the soil profile in solution. The latter therefore loses the dark colours asso
ciated with the presence of iron-bearing solids and acquires lighter colours tending
towards grey. The process is referred to as gleying. What happens next depends
upon other features of the chemical environment. For example if the environment
contains sulphate ion and sulphate reducing bacteria are present, pyrite may form.
If carbonate ions dominate, siderite may result. If sulphide and carbonate ions are
scarce, but sufficient silica is present in solution, authigenic silicates may appear.
Of the latter, only glauconite is known to form at the present day, but in past sed
imentary environments, and particularly in the Precambrian, iron formations were
produced with several authigenic iron silicates. Brookins (1988, p. 73-81) illustrates
a number of possibilities.
A distinctive pedogenic process takes place in reduced environments and has
been called ferrolysis (Brinkman, 1970). Ferrolysis can be considered to be a two
stage process. The first stage, in which reduction replaces a relatively high Ca and
K on exchange sites by Fell, is followed by oxidation. Oxidation might result from
a lowering of the water table and it causes Fell to be replaced by H + to produce
acid clays and Fe (and Mn) hydroxides as solid phases. The latter produce mottles,
concretions and extensive indurated layers (hardpan).
A comparable two stage process may have given rise to the Precambrian iron
formations (Drever, 1974).
A pedogenic grid
Redox-pH diagrams have been used in the examination of weathering systems

since the pioneering studies of Krumbein and Garrels (1952) and Baas-Becking et
al. (1960). This is true in spite of a number of difficulties (Berner, 1970), espe
cially in choosing what minerals to show on them, whether to consider metastable
phases, whether to consider solid solution in condensed phases, and what to do
about bacterial reductions. Above all there is the major difficulty of obtaining good
measurements of Eh in the field. Even so, a consensus exists that the diagrams are
useful in indicating general conditions and general tendencies in a system even if
they cannot be relied upon for more precise information.
In considering the whole weathering zone, the Eh-pH diagram of Krumbein
and Garrels (1952) is too restrictive since it does not cover the whole pH range
shown on Figure 2.3. Consequently their diagram has been expanded (Figure 2.6)
by the inclusion of fences based on Al3+ equilibria with gibbsite and kaolinite, and
on the breakdown of Ca-saturated smectites, the latter equilibria coming closest to
justifying the Krumbein and Garrels (1952) neutral fence. In addition the organic
fence of Krumbein and Garrels (1952) is replaced by one more in keeping with the
conditions found in organic-rich horizons in soil.
The term pedogenic grid is suggested for Figure 2.6 since it can be used to ex
amine weathering under the broadest range of conditions found in nature, that is,
in soils. For example, nodes where fences cross are analogous to invariant points
in pressure-temperature diagrams. Pedogenic inertia (Bryan and Tbakle, 1949),
wherein a soil persists after conditions that formed it have changed, can be ex
plained by the pseudo-invariance of reactions taking place at a node. Furthermore,
the end points of the three general weathering trends can be explained in terms of
nodes representing the extremes of all but a few terrestrial systems.
32 W Chesworth
Fig. 2.6. The pedogenic grid, an extension of the diagram of Krumbein and Garrels (1952). Fence A is
defined by equilibria in the system SÌO2-AI2O3-H2O; fenced by reaction of a Ca-saturated smectite;
fence C is the calcium carbonate fence (Krumbein and Garrels, 1952); fence D is defined by equilibria
between alkali carbonates or zeolites and aqueous solutions; fence E is the organic matter fence of
Krumbein and Garrels (1952) modified to relate more closely to conditions in the A horizon of soils; F
and G are sulphate-sulphide (Krumbein and Garrels, 1952) and goethite fences, respectively.
A further analogy with the PT diagram or at least with the PT interpretation of

mineral fades can be suggested. This is that the divariant pe-pH fields bounded
by essentially univariant (or pseudo-univariant) fences represent phase assemblages
that constitute mineral facies of the weathering zone.
Weathering rates
The problem of determining weathering rates has been attacked in the field and
in the laboratory. Field studies depend upon finding a datable surface that has
undergone weathering for a measurable time. Estimates are dependent upon all the
factors of Figure 2.1 and are therefore site specific and without general significance.
Laboratory studies of weathering rate throw light on mechanism. The rate limiting
step appears to be surface controlled reactions in most cases (Grandstaff, 1986),
though the weathering of glass may be diffusion controlled.
Different minerals weather at different rates, but, as Eggleton (1986) points out,
no adequate explanation exists. The susceptibility of a mineral to weathering may
possibly be related to its position in Bowen's reaction series (Goldich, 1938) or to its
free energy of formation (Keller, 1954). Ultimately, "weathering rate will depend on
the mechanism whereby the weathering agents break the bonds between the atoms
of the crystal" (Eggleton, 1986).
Colman and Dethier (1986) and Lerman and Meybeck (1988) provide two recent
collections of papers containing much information about weathering rates. Again
they confirm the view that rate is site specific with generalisation being difficult.
Judged in terms of soil production, rates range generally from 1 to about 500 mm
per thousand years, with most soils forming at rates less than 100 mm per thousand
years. It seems reasonable to state that in regions with a temperate, humid climate
the rate of soil formation on aluminosilicate parent materials will be less than 50
mm per thousand years, with soils on carbonate parent materials forming at rates
between 50 and 100 mm per thousand years (Reynolds et al., 1987).
Geographical distribution
Whether we view Jackson and Sherman's (1953) two compartments of the weath
ering zone as distinctly separate or as being gradational, one into the other, weath
ering at depth will be largely determined by the ability of the system H2O-CO2
to produce protons. Deep weathering will therefore be relatively uniform, rate dif
ferences apart, throughout the world. On the land surface, however, with organic
acids and complexants significantly present, regional variations in weathering chem
istry occur. These variations reflect geographical changes in the coupled factors of
climate and vegetation, and the products are the different soils of the earth's sur
face. How many distinctly different weathering environments in the soil zone this
amounts to, is a debatable point. Pedro and Sieffermann (1979) for example recog
nise four, to which Brinkmann (1970) has added a fifth. Gaucher (1977) on the
other hand itemises ten different types of soil-chemical environment. Macias and
Chesworth (this book) discuss the matter further.
Regardless of how many separate types are recognized the acid end points of
weathering differ with latitude. Thus the podzol is typical of humid temperate cli
mates, and the ferralsol of humid tropical ones. The former is the prototypal system
for demonstrating the effect of organic constituents on weathering, while the latter,
since organic debris breaks down rapidly in the tropics, is not basically different
from a weathering regime dominated by the H2O-CO2 system. Stated differently,
34 W. Chesworth
the pedochemical compartment of the weathering zone becomes indistinguishable

from the geochemical compartment in the humid tropics. Despite this chemical
convergence, physical differences remain.
The role of climate is worth stressing. As stated earlier, there is a sense in which
climate is the driving force and not simply one of the factors of weathering. An old,
but still useful way of looking at climate, is incorporated into Figure 2.7, where it
is considered simply in terms of temperature (hot/cold) and availability of water
(wet/dry). Vegetation patterns and soils, at least to a first approximation follow
the general climatic scheme. The two oxidising weathering trends, the acid and
alkaline trends respectively, relate to the same diagram, though the wet/dry factor
is more important than the hot/cold factor in determining their relative distribution.
Temperature, however, is of obvious importance in its influence on the kinetics of
reactions in the weathering system.
The relationship of the third general pedogenic trend, the reducing trend to Fig
ure 2.7 is more complicated. This trend can be expected anywhere below the water
table. Yet, if the examination of weathering is confined to surface, or immediate,
A. CLIMATE
Cold
S N O W A N D ICE
C. SOILS
S N O W A N D ICE
TUNDRA
TAIGA TUNDRA SOILS
PODZOLS
Dry Û Q Wet
cc 2 Û
co PODZOLS
1-
cc < 2 LLj
N
< 2 m X £
1X1
CO
D
CO o LUVISOLS
z
cc
KAOLISOLS
Hot
B. VEGETATION D. CHEMISTRY
SNOW AND ICE S N O W A N D ICE
TUNDRA ACID TRENDS Γ -~7\ Fe

A l , Fe preferentially f l u s h e d to lower L
TAIGA preferentially
levels; Si a c c u m u l a t e d at higher levels
flushed from
CO ALKALINE TRENDS system.
uu
coco Q i
I
com z CO co I C a salts ppt.
<z>
tree Û.
S 1-
co
LU
ce Na I in w e a t h e r i n g
Ü I co LU
o chlorides i c o l u m n . Calcite
,_co
CEO —
(
CO <
CO CC
o
Li.
Z
and principally Si preferent
ially r e m o v e d . Al|
LLIZ
co<
cr
O La < |carbonates|
ppt. I
with g y p s u m
a s aridity |Fe a c c u m u l a t e d
LU er
Q I increases near surface
_l
Fig. 2.7. Schematic relationship between climate, vegetation, soil and chemistry in the weathering zone.
The reduced trend will be found anywhere below the water table. In surface and near-surface envi
ronments it will be commonest as shown on the wet side of the climate spectrum, in poorly drained
environments (modified from U.S. Department of Agriculture, 1938).
Fig. 2.8. Soil weathering zones on a global scale. Acid trend: 1 — ferra Hi tisation; 2 = podzol isation; 3
= andosolisation. Alkaline trend: 4 = calcic soils; 5 = sodic soils. Reduced trend: 6 = gleysolic and
histosolic materials; these may be found anywhere under humid climates — in the large area shown on
the map they are intermittently distributed in the zone of permafrost. Non extreme soils: 7 = these are
soils mostly with 2:1 sheet silicates, with or without calcite, and well within the limits of the Eh-pH
field of common weathering environments shown in Figure 2.3.
subsurface conditions, the trend can be considered to be confined to the relatively

wet section of the diagram. This section is essentially coincident with the area af
fected by the acid trend. However, other factors are important, particularly texture,
with its effect on drainage rates, and the presence of an impermeable substratum
(for example a clay or shale layer, or permafrosted materials), to support a high
water table.
Considerations such as those incorporated in Figure 2.7 have been used to devise
maps that show a distribution of weathering types (Pedro, 1968). Figure 2.8 is a
highly generalised map showing the distribution of the weathering systems of Figure
2.7 on a global scale.
Conclusions: the soil-sediment connection
The semi-serious statement that "soil is sediment on its way to the sea" embodies
the truism that virtually all sediment starts its life as soil. Sediment might therefore
be expected to reflect this heritage.
36 W. Chesworth
There are a number of general ways in which pedogenesis could be expected to

impact on the nature of sediment derived from soil:
(a) the chemical (including mineralogical) composition of the sediment might
correspond to a particular soil, or a particular soil horizon;
(b) the soil organic matter might leave its signature on sediment; and
(c) morphological features of a soil might be inherited by the sediment.
All three impacts could be achieved by in situ burial to produce a paleosol.
Erosion and transport, however, will inevitably destroy many, if not all, pedogenic
features. The question is: what survives?
Brunn and Desprairies (1965) construct a hypothesis illustrative of the survival
of compositional features, to explain cyclic sediments offlyschand molasse type in
Greece. Regular alternations of red clays with sandy marls richer in smectite, are
taken to be formed by the stripping of soil profiles whereby the B horizon yields the
red clay and the C horizon the marl.
For the chemical and mineralogical features of a pedological environment to
leave their imprints on derived sediment, the soil would need to be well differ
entiated, its horizons would have to be eroded sequentially, and the process of
derivation would have to avoid as much as possible any sorting. Unfortunately, to
some degree these criteria are mutually exclusive. For example the dismembering
of a profile, horizon by horizon, implies a relatively careful sedimentary process,
whereas a lack of sorting imphes a rapid and indiscriminate one. The most probable
expectation would be that the sedimentary process would homogenise differences
between soil and parent material rather than preserve them. In other words a hy
pothesis such as Brunn and Desprairie's (1965) would be inherently unlikely.
The more probable relationship is seen between greywackes and their source
areas in active continental margins. Here is a region where sorting is minimised and
where distinctive soil-forming processes take place, andosolisation on intermediate
volcanics, and podzolisation on intermediate plutonics. However, the composition
of greywacke from such a locality reflects the parent materials rather than any
distinctive horizonation resulting from pedogenic processes. No selective stripping
of soil horizons is revealed in the derived sediment, any pedogenic differentiation
would seem to have been homogenized out of existence, with a mixing of horizons
to give a composition similar to that of the parent material.
This does not rule out a more attenuated connection between sediment and
soil. Consider the clay mineralogy of modern oceanic sediments. Kaolinite (and less
commonly gibbsite) are characteristic of, though not restricted to, the pedogenic
process of ferrallitisation, a process occurring at present in the tropics and near
tropics. Consequently, the rivers and winds of such regions, feed kaolinite into ad
jacent seas (Hathaway, 1979). On the other hand, chlorite is generally concentrated
in the oceans at high latitudes. The latter might reflect the fact that temperate zone
pedogenesis is less rigorous than in the tropics, or it might be due to the fact that
recent glaciations dictate that temperate zone soils are younger than tropical zone
soils.
Ferrallitisation also leads to the concentration of Fe in surface and near-surface
horizons of soil. Derived sediments may therefore be Fe-rich and have the distinc
tive colours of Fe in oxidised form. The connection with continental red beds is
widely accepted.
In a series of publications Erhart (1963,1964,1967) has taken these ideas further.
He recognizes two contrasted states in the weathering regime of continents. A
state of biostasis holds when a continental region is stable, with a well established
vegetation within the restrictions imposed by climate, and with intense pedogenetic
alteration. Ferrallitic soils predominate, with thicknesses of 100 m or more. The
elements K, Na, Ca, Mg and Si are leached out and proceed towards the ocean.
In the sea, the subsequent development of organisms that utilise Ca and Mg in
constructing exoskeletons, eventually produces carbonate sediments, and detrital
sediments are less important.
The contrasted state of rhexistasis is one in which vegetation cover is poor,
and intense erosion removes the ferrallitic soils produced in a previous period of
biostasis. Detrital sediments predominate over carbonates. Rhexistasis alternating
with biostasis might explain some types of cyclic sedimentation.
Erhart (1963, 1966) also believes that the biological cycle in soil is important in
the genesis of some of the minerals common in sediments. For example Si and Al
can concentrate in plant tissues as phytoliths and on the death of the plants may
concentrate in the upper horizon of soil. Opal phytoliths are very common, and
phytolithic material with almost a 1:1 ratio of Al to Si is known according to Erhart
(1963). The opal phytoliths can have a profound effect on the geochemistry of Si,
since they are perhaps the most readily soluble form of S1O2 at the surface of the
earth (Wilding et al., 1977). Erhart (1963) suggests that great radiolarian episodes
in the geological past could indicate reworking of soil opal. He further suggests
(Erhart, 1966) that Al-rich phytoliths may be precursors of a significant proportion
of the kaolinite of the sedimentary record.
Much of this, of course, is highly speculative, but the ideas involved deserve
further investigation. Some of the more obvious connections, and some of the sedi
mentary environments with which comparisons may be instructive, are summarized
in Figure 2.9.
38 W. Chesworth
ACID TREND \-+- Most siliciclastic sediments

PROCESSES
1. Podzolisation
Si accumulates NON-EXTREME
in upper horizon SOILS
Active Fé, Al flushed to Bisiallitisation
margin "* lower ALKALINE
Si, Al, Fe
greywacke ^— 2. Andosolisation accumulation TRENDS
Particular/ in Ca, Mg, saturated PROCESSES
weathering of glassy exchange complex 1. Precipitation of
volcanics. Si, Al calcium salts:
form amorphous
T—
1. calcite in less
phase arid, sulphate in
3. Ferrallitisation. more arid areas.
Continental Fé, Al accumulate
red Beds * "1 near surface at
""^^ 2. Precipitation of
sodium salts: Evaporitic and
Sediments expense of Si. Ü.
carbonates and zeolitic sediments
with 1:1 4. Formation of chlorides.
sheet Fe and Al 0. 3. Silicification
silicates duricrusts
and/or mobilisation of
gibbsite u. Si at high
alkalinites produces
£- 0.2 cherts
Archaean
banded Fe
formations
Fig. 2.9. Soil chemical processes in an Eh-pH framework, with speculative relationships to derived
sediments shown. A, S and R represent acid, alkaline (sodic) and reduced soils, respectively, and a, s
and r are generalised acid, alkaline and reduced trends. The term non-extreme soil is used to indicate a
soil that is removed from the outlined boundary of most weathering environments.
References
Baas-Becking, LG.M., Kaplan, I.R. and Moore, O., 1960. Limits of the natural environment in terms of
pH and oxidation potentials. J. Geol., 68: 243-284.
Berner, R.A., 1970. Behaviour of iron during weathering, sedimentation and diagenesis. In: Handbook
of Geochemistry, Section 26G. Springer, Berlin, pp. 1-8.
Beveridge, TJ., 1989. Metal Ions and Bacteria. Wiley, New York, N.Y., 461 pp.
Brinkman, R., 1970. Ferrolysis, a hydromorphic soil-forming process. Geoderma, 3: 199-206.
Brookins, D.G., 1988. Eh-pH Diagrams for Geochemistry. Springer, Berlin, 176 pp.
Brunn, J.H. and Desprairies, A., 1965. Etude sédimentologique préliminaire de formations à caractère
flysch et molasse. Rév. Géol. Phys. Géol. Dyn., 7: 339-354.
Bryan, WH. and Teakle, LJ.H., 1949. Pedogenic inertia; a concept in soil science. Nature, 164: 969.
Buoi, S.W., Hole, ED. and McCracken, R.J., 1980. Soil Genesis and Classification. Iowa State Univ.
Press, Ames, Iowa, 404 pp.
Chesworth, W, 1972. The stability of gibbsite and boehmite at the surface of the earth. Clays Clay
Miner., 20: 369-374.
Chesworth, W, 1973. The parent rock effect in the genesis of soil. Geoderma, 10: 215-225.
Chesworth, W, 1975. The system S i 0 2 - A l 2 0 3 - F e 2 0 3 - H 2 0 and the kaolinitic stage of the goethite
facies. Clays Clay Miner., 23: 389-392.
Chesworth, W, in press. Geochemistry of micronutrients. In: Micronutrients in Agriculture. Soil Sci.

Soc. Am., Book Ser. No. 4.
Chesworth, W. and Dejou, J., 1980. Are considerations of mineralogical equilibrium relevant to pedol
ogy? Evidence from a weathered granite in central France. Soil Sci., 130: 290-292.
Coincon, R., Tardy, Y. and Godard, A., 1975. Les enseignements d'ordre morphogénique at paléo
climatique apportés par l'étude de bassins de l'ouest de la Margeride. Rev. Geomorph. Dyn., 25:
81-91.
Colman, S.M. and Dethier, D.P. (Editors), 1986. Rate of Chemical Weathering of Rocks and Minerals.
Academic Press, London, 603 pp.
Drever, J.I., 1974. Geochemical model for the origin of Precambrian banded iron formations. Geol. Soc.
Am. Bull., 85: 1099-1106.
Eggleton, R.A., 1986. The relation between crystal structure and silicate weathering rates. In: S.M.
Colman and D.P. Dethier (Editors), Rates of Chemical Weathering of Rocks and Minerals. Academic
Press, London, pp. 21-40.
Edmond, J., Corliss, J.B. and Gordon, L.I., 1979. Ridge crest hydrothermal metamorphism in the
Galapagos spreading center and reverse weathering. Maurice Ewing Ser. 2, Am. Geophys. Union,
Washington, D.C., pp. 383-390.
Epstein, E., 1977. The role of roots in the chemical economy of life on earth. BioScience, 27: 783-787.
Erhart, H., 1963. Sur le cycle de la silice hydratée dans la biosphère. C.R. Acad. Sci., 256: 3731-3734.
Erhart, H., 1964. Conséquences logiques de la bio-rhexistasie sur le chimisme des mers anciennes et sur
les sédiments qu'elles ont engendrés. C.R. Acad. Sci., 259: 2669-2671.
Erhart, H., 1967. La genèse des sols en tant que phénomène géologique. Masson, Paris. 177 pp.
Eugster, H.P. and Hardie, L.A., 1970. Saline lakes. In: A. Lerman (Editor), Lakes: Chemistry, Geology,
Physics. Springer, New York, N.Y., 363 pp.
Friedman, G.M. and Krumbein, W.E., 1985. Hypersaline Ecosystems. Springer, Berlin, 484 pp.
Garrels, R.M. and Christ, C , 1965. Solutions, minerals and equilibria. Harper and Row, New York,
N.Y., 450 pp.
Gaucher, G., 1977. Vers une classification pédologique naturelle basée sur la geochimie de la pédoge
nèse. Catena, 4:1-27.
Gibbs, R.J., 1970. Mechanisms controlling world water chemistry. Science, 170:1088-1090.
Goldich, S.S., 1938. A study in rock weathering. J. Geol., 46: 17-58.
GrandstafF, D.E., 1986. The dissolution rate of forsteritic olivine from Hawaiian beach sand. In: S.M.
Hamilton, W.B., 1989. Crustal geologic processes of the United States. Geol. Soc. Am. Mem., 172:
743-781.
Hathaway, J.C., 1979. Clay minerals. Chapter 5 in Marine Minerals. Am. Min. Soc. Rev. Mineral., 6:
123-173.
Hemley, J.J. and Wones, W.R. 1964. Chemical aspects of hydrothermal alteration with emphasis on
hydrogen metasomatism. Econ. Geol., 59: 538-569.
Jackson, M.L. and Sherman, G.D., 1953. Chemical weathering of minerals in soils. Adv. Agron., 5:
219-318.
Jenny, H. 1961. Derivation of state factor equations of soils and ecosystems. Soil Sci Soc. Am. Proc., 25:
385-388.
Keller, W.D., 1954. Bonding energies of some silicate minerals. Am. Min., 39: 783-793.
Korzhinskii, D.S. 1959. Physicochemical Basis of the Analysis of the Paragenesis of Minerals. Consul
tants Bureau, New York, N.Y., 142 pp.
Krumbein, WC. and Garrels, R.M., 1952. Origin and classification of chemical sediments in terms of
pH and oxidation-reduction potentials. J. Geol., 60:1-33.
40 W Chesworth
Lelong, F., Tardy, Y., Grandin, G., Trescases, J.J. and Boulangé, B. 1976. Pedogenesis, chemical weath
ering and processes of formation of some supergene ore deposits. In: K.L. Wolf (Editor), Handbook
of Strata-bound and Stratiform Ore Deposits. Elsevier, Amsterdam, 3: 93-173.
Lerman, A. and Meybeck, M., 1988. Physical and chemical weathering in geochemical cycles. NATO
Adv. Study Inst. Ser. C, 251: 375.
Mason, B. and Moore, C.B. 1982. Principles of Geochemistry. Wiley, New York, N. Y., 344 pp.
Mattigod, S.V. and Kittrick, J.A., 1980. Temperature and water activity as variables in soils mineral
activity diagrams. Soil Sci. Soc. Am. J., 44: 149-154.
Meunier, A. and Velde, B. 1976. Mineral reactions at grain contacts in early stage of granite weathering.
Clay Miner., 11: 235-240.
Nicholls, G.D., 1963. Environmental studies in sedimentary geochemistry. Sci. Prog., 51:12-31.
Pedro, G., 1968. Distribution des principaux types d'altération chimique à la surface du globe. Présen
tation d'une esquisse géographique. Rév. Géogr. Phys. Géol. Dyn., 10: 457-470.
Pedro, G. and Delmas, A.B., 1970. Les principes geochimiques de la distribution des éléments-traces
dans les sols. Ann. Agron., 21: 483-518.
Pedro, G. and Sieffermann, G., 1979. Weathering of rocks and formation of soils. Earth Sci., 16: 39-55.
Reynolds, B., Hornung, M. and Stevens, P.A., 1987. Solute budgets and denudation rate estimates for a
mid-Wales catchment. Catena, 14: 13-23.
Robert, M., Veneau, G. and Berrier, J., 1980. Solubilisation comparée des silicates, carbonates et
hydroxydes en fonction des conditions du milieu. Bull. Minéral., 103: 324-329.
Ronov, A.B. and Yaroshevsky, A.A., 1969. Chemical composition of the earth's crust. In: PJ. Hart (Edi
tor), The Earth's Crust and Upper Mantle. Geophys. Monogr. 13, Am. Geophys. Union, Washington,
D.C., pp. 37-57.
Spiers, G.A., Pawluk, S. and Dudas, M.J., 1984. Authigenic mineral formation by solodization. Can. J.
Soil Sci., 64: 515-532.
Strzemski, M., 1975. Ideas Underlying Soil Systematics. U.S. Dept. of Commerce, National Technical
Information Service, Va., 542 pp.
Stumm, W, 1986. Coordinative interactions between soil solids and water — an aquatic chemist's point
of view. Geoderma, 38: 19-30.
Tardy, Y. and Nahon, D., 1985. Geochemistry of latérites, stability of Al-goethite, Al-hematite, and
Fe 3+ -kaolinite in bauxites and ferricretes: an approach to the mechanism of concentration forma
tion. Am. J. Sci., 285: 865-903.
Taylor, S.R. and McClennan, S.M., 1985. The Continental Crust. Its Composition and Evolution. Black-
well, Oxford, 312 pp.
Thompson, J.B., 1959. Local equilibrium in metasomatic processes. In: P.H. Abelson (Editor), Re
searches in Geochemistry. Wiley, New York, N.Y., pp. 427-457.
U.S. Department of Agriculture, 1938. Soils and Men. U.S. Government Printing Office, Washington,
D.C., 1232 pp.
van Breemen and Brinkman, R., 1976. Chemical equilibria and soil formation. In: G.H. Bolt and M.G.
M. Bruggenwert (Editors), Soil Chemistry: A. Basic Elements. Elsevier, Amsterdam, 271 pp.
Velde, B., 1985. Clay Minerals: A Physico-Chemical Exploration of their occurrence. Elsevier, Amster
dam, 427 pp.
Wilding, L.P., Smeck, N.E. and Drees, L.R., 1977. Silica in soils: quartz, cristobalite, tridymite and opal.
In: J.B. Dixon and S.B. Weed (Editors), Minerals in Soil Environments. Soil Sci. Soc. Am., Madison,
Wise, 948 pp.
Yatsu, E., 1988. The Nature of Weathering. Sozosha, Tokyo, 624 pp.
41
Chapter 3
Extraterrestrial soils — the lunar experience

J.F. LINDSAY
Introduction
Soils form on planetary surfaces, that is, at the interface between the planet and
its interplanetary environment, in response to the release of internal planetary (en-
dogenic) energy and thefluxof extraplanetary (exogenic) energy (Figure 3.1). It may
seem simplistic to make such a statement but it is essential to divest ourselves of an
earthbound conceptual framework if we are to place terrestrial soils into the larger
framework of planetary processes. On planets such as the earth the atmosphere
and hydrosphere act as intermediaries, redistributing the available energy flux in a
complex way across the whole planetary surface. As a consequence, extraterrestrial
energy fluxes, for the most part solar energy, are redistributed by mechanisms such
as wind, waves or stream flow. On smaller planetary bodies with proportionately
smaller gravitational fields, a planet's ability to hold an atmosphere is reduced such
that on a body the size of the moon, for all practical purposes there is no atmo
sphere or hydrosphere. This absence is critical in determining the path that soil
formation takes. Without air and water many of the things we take for granted in
the terrestrial environment simply disappear; there is no rainfall, no runoff and no
stream erosion and consequently there are no clays, no hydrated minerals, no water
table and so on. We literally find ourselves in another world.
The surface of the moon and other smaller planetary bodies is exposed directly
to the interplanetary environment. Without fluid intermediaries solar energy, the
dominant energy source in the terrestrial setting, becomes largely ineffective as
a source of erosive soil-forming energy because it is simply re-radiated back into
space from the planet's airless surface. Conversely, the meteorite flux, a feeble
exogenic energy source in the terrestrial setting, becomes, by default, the domi
nant source of erosional and transportational energy available at the surface of
the planet. The only other soil forming processes that may affect soil development
are mass wasting, which aids erosion on steeper slopes (Swann et al., 1972), and
electrostatic transport, which occurs along the planet's terminator (Criswell, 1972).
42 J.E Lindsay
ENDOGENIC EXOGENIC
PROCESSES LUNAR EVOLUTION PROCESSES
PLANETARY
FORMATION
ί ACCRETION
ACCRETIONARY
I
SEGREGATION EARLY HEAVY
HEATING AND OF METEORITE
DIFFERENTIATION CRUST (HIGHLANDS) BOMBARDMENT
PARTIAL MELTING FORMATION FINAL

AND ERUPTION OF MARIA LARGE IMPACTS
OF BASALT
CONTINUED METEORITE
EVOLUTION BOMBARDMENT
OF SOIL AT CONTEMPORARY
LEVELS
Fig. 3.1. Simplified schematic relationship between planetary processes and soil formation on the moon
(after Veverka et al., 1986). The soil is a product of the composition of the moon's surface and its
geologic history which is a product of both endogenic and exogenic processes.
Electrostatic transport may be important in transporting finer material less than 10

/xm in diameter.
The moon
When viewed from the earth the moon's surface is obviously divided, using
albedo, into maria and highlands (terra; Figure 3.2). The highland areas are consis
tently rougher and more intensely cratered than the maria suggesting that there is a
considerable difference in age. These simple observations, first made by Galileo in
1610, provided the first direct information that the moon, like the earth and other
planets, has had a complex evolution. The moon was probably formed 4.65 x 109
years ago. The large-scale chemical differentiation of the moon began almost imme
diately and was complete shortly after 3.85 x 109 years ago. By this time the lunar
crust, which is of most interest to us here, had formed and become cold and rigid;
the moon, to all intents, had ceased to evolve and had become a dead planet (Figure
Extraterrestrial soils — the lunar experience 43
Fig. 3.2. A view of the intensely cratered lunar farside and part of the nearside including Mare Crisium.
Note the lack of large craters on the younger mare surface indicating a considerable age difference
(NASA AS-16-3028).
3.2). The two-fold subdivision of the lunar surface thus reflects its evolution. The
highlands, which on average are 3-4 km higher than the maria, are underlain by
much altered lunar crust, whilst the maria were formed by lava flows in the final
stage of the planet's evolution. Highland lithologies are, in general, anorthositic in
composition whilst the lavas of the maria are basaltic. This compositional differ
ence determines, in large part, the composition of the soil formed on the bedrock
surfaces.
The meteorite flux

The marked difference in crater density between the maria and highland surfaces
long suggested that thefluxof meteorites had changed through time and many early
writers had postulated an early intense bombardment of the moon (Urey, 1952;
Kuiper, 1954; Kuiper et al. 1966; Hartmann, 1966). Whilst their arguments seemed
reasonable it was not until the Apollo landings in the late 1960s and early 1970s
that it was possible to date the surfaces independently and establish some estimates
of the intensity changes. Dating revealed that initially thefluxwas intense and that
it had declined rapidly, at first with a half life of approximately 108 year, until
approximately 3.3 x 109 years ago, after which it had remained relatively constant
or had perhaps increased slightly (Figure 3.3). The flux of meteorites at the lunar
surface evolved in three stages (Hartmann, 1970).
44 J.E Lindsay
Pre--Imbrian (Terrae + Basins)

Imbrian (Mare Fili)
Eratosthenian
MMMBM
Copernican
256 -
?
128 - i
1
64 - 1
1
1
1
o> 32 -
\
OD \
ώ 16
- \
a»
>
\
1 8 - \
α>
OC
\
4 \\
" 1
\
2 r- 1 \
\
\
1 I- \
ne _J | 1 1 1
'5 4 3 2 1 0
Age (by)
Fig. 3.3. Relative cratering rate (present day flux = 1) of the lunar surface as a function of time (from
Hinners, 1971).
(a) An intense bombardment of low velocity (1.7-2 km s - 1 ) circumterrestrial

particles left over after the formation of the moon.
(b) An intense bombardment of the last, planet-forming planetesimals swept up
from low-eccentricity solar orbits with collision half-lives of the order of 108 year.
These were medium velocity (2-10 km s"1) collisions which probably formed the
large circular maria.
(c) The present phase of high velocity (8-40 kms" 1 ) sporadic meteorites and
cometary particles. The modern flux produces approximately 2 to 8 x IO3 erg cm" 2
per year: a very modest energy flux compared to alternative sources, such as wind or
water erosion, available at the earth's surface.
Thus, we might expect that soils formed on highland areas differ significantly
from those formed on maria. We would expect them to be different compositionally
because of the underlying bedrock differences and we could expect that highland
soils would be thicker and more highly evolved because of the longer exposure to
the meteorite flux and because they were exposed to a more intense flux early in
their history (Figure 3.1).
Extraterrestrìal soils — the lunar experience 45
Lunar soil
The surface of the moon is blanketed by a thin layer of weakly-cohesive, low-
velocity, detrital material that is generally referred to as "soil" or "regolith" (Figures
3.4 and 3.5; Lindsay, 1976). The use of the term "soil" in a setting where biologic ac-
Fig. 3.4. A clearly defined footprint in the lunar soil at the Apollo 11 site (NASA AS 11-40-5878).
Fig. 3.5. Sampling the lunar soil at the Apollo 12 landing site with a drive tube (NASA AS 12-49-7286).
46 LE Lindsay
tivity is absent is perhaps an extension of the normal usage. However, the earth and
its biosphere appear to be unique, at least in the solar system, and there is a need
to broaden the use of the term in the extraterrestrial environment. The lunar soil
is a continually evolving blanket produced in large part by hypervelocity meteoroid
impact (Figure 3.6), although locally a small pyroclastic volcanic contribution may
be present. The morphology of small craters penetrating the soil showed that me
dian soil thickness varied from 3.3 to 16 m (Oberbeck and Quaide, 1968). Further,
the median thickness of the regolith correlates directly with the density of impact
craters and with the age of the underlying substrate (Table 3.1).
Fig. 3.6. A closeup view of a small crater in the lunar soil surrounded by a thin éjecta blanket. The crater
has not penetrated the soil layer and has therefore not added new material to the soil but reworked
the existing soil to produced a new stratigraphie unit which may or may not survive future reworking
(NASA AS15-85-11466).
TABLE 3.1
Soil parameters at the Apollo landing sites
Apollo site Soil thickness Mean grain size Age of substrate

(m) (Mm) (b.y.)
11 4.4 15.7 3.51
12 3.7 54.4 3.16
14 8.5 37.2 3.95
15 4.4 28.4 3.26
16 12.2 - -
17 4.0, 8.0 - 3.74
Soil structure
The lunar soil is not homogeneous but is complexly layered (Lindsay, 1975,1976).
The thickness-frequency distribution of the units is bimodal with the strongest
mode at 1.0 to 1.5 cm and a secondary mode at 4.5 to 5.0 cm. The units are
distinguished on the basis of colour and texture, particularly variations in grain size.
The contacts of the units are generally sharp, although at times difficult to observe
because the textural differences are subtle. Some units have transitional contacts
and structures similar to flame structures were observed at the base of other units.
Both normal and reverse graded beds have been described (Lindsay et al., 1971;
Duke and Nagel, 1975).
Physical properties
The mean density of the paniculate material forming the lunar soil ranges from
2.90 to 3.24 g cm""3 depending upon the nature of source materials. Soils formed
on the basaltic maria tend to be denser than soils formed on the more anorthositic
highland, that is, the soils in large part reflect the composition of their substrate.
The porosity of lunar soils may range from 41 to 70% with the result that soil
density is variable (Carrier et al., 1974). The main reason for the range in porosities
is probably connected with the abundance of extremely irregularly-shaped glass
particles (agglutinates) which are discussed in a following section. Soil density also
increases in a logarithmic manner with depth (Carrier et al, 1974). The density
profile can be approximated by:
p=Po + 0.121 ln(Z + l)
where p0 and p are the density at the surface and at some depth Z (cm). p0 is
approximately 1.38 g cm"3. The continual reworking of the lunar surface apparently
keeps the surface layers of the soil loose but at depth the vibration due to the
passage of numerous shock waves causes the soil to increase in density.
Composition
It has been estimated that 95% of a soil sample at any given point on the moon
is derived from within 100 km (Shoemaker et al, 1970). The proportion of exotic
components decreases exponentially with distance. Thus, the chemistry and miner
alogy of the lunar soils, for the most part, reflect the composition of the underlying
bedrock (Table 3.2), so that soils from maria areas have an overall basaltic composi
tion with a high Fe content, while the highland soils tend to be more anorthositic in
composition and have high Al and Ca values (Table 3.3). The boundaries between
the two are sharp at the available resolution (Adler et al., 1972a, b). Lateral move
ment of detrital material thus cannot be rapid despite continued reworking for long
time periods by the meteoroid flux.
48 J.E Lindsay
TABLE 3.2
Major element chemistry of average Apollo 15 soils and basalts
Basalt Average soil Olivine basalt Quartz

Si02 46.61 44.20 48.80
Ti02 1.36 2.26 1.46
AI2O3 17.18 8.48 9.30
FeO 11.62 22.50 18.60
MgO 10.46 11.20 9.46
CaO 11.64 9.45 10.80
Na20 0.46 0.24 0.26
K2O 0.20 0.03 0.03
P2O5 0.19 0.06 0.03
MnO 0.16 0.29 0.27
Cr 2 0 3 0.25 0.70 0.66
Total 100.00 99.41 99^7
Note: Apollo landed on the lunar maria but at a site close to the highlands (after Taylor, 1975). The
higher AI2O3 values in the soil reflect the addition of highland materials to the soil.
TABLE 3.3
Average major element chemistry for highland and maria soils
Element Percent of atoms Weight percent oxides

Mare Highland Average Mare Highland Average
O 60.3 61.1 60.9 - - -
Na 0.4 0.4 0.4 0.6 0.6 0.6
Mg 5.1 4.0 42 9.2 7.5 7.8
Al 6.5 10.1 9.4 14.9 24.0 22.2
Si 16.9 16.3 16.4 45.4 45.5 45.5
Ca 4.7 6.1 5.8 11.8 15.9 15.0
Ti 1.1 0.2 0.3 3.9 0.6 1.3
Fé 4.4 1.8 2.3 14.1 5.9 7.5
Source: Tlirkevich (1973).
Petrography
The lunar soil consists of three basic components: (a) rock fragments, (b) mineral
grains, and (c) glass particles. The composition of these components varies consider
ably from one locality to another, depending upon the nature of the bedrock. They
also vary in abundance in a more local sense (laterally and with depth) in response
to the addition of more distant exotic components, the degree to which the soil has
been reworked (its age), and in response to bedrock inhomogeneities.
Rockfragmentsare the dominant soil component in particle size ranges larger
than 1 mm (60 to 70% of the 1-2 mm size range). Below 1 mm their importance
decreases rapidly as they are disaggregated to form mineral and glass particles. The
coarser the mean grain size of the soil the greater, in general, the abundance of rock
fragments.
Three categories of clastic fragments should be considered: igneous rock frag
ments, crystalline breccias and soil breccias.
Igneous rock clasts. Basalts are by far the most common igneous rock fragment
in maria soils whilst in the highlands the igneous fragments come from grabbroic-
noritic-anorthositic sources or, less frequently, from a highland basalt source. Be
cause the grain size of the minerals forming the highland lithologies is large com
pared to the size of the soils, positive identification of many rock types is difficult.
Crystalline brecciafragments.These fragments dominate the lithic clasts in all of
the highland soils. They are bedrock fragments derived from éjecta units resulting
from large meteorite impacts. Fragments of all of the matrix types are represented
but again, because of the size of the clasts in relationship to the grain size of the
detrital materials forming the breccias, it is not possible to evaluate the relative
abundance of different lithologie types.
Soil breccias. The breccia fragments are poorly sorted agglomerations of rock,
mineral and glass fragments which to a large extent preserve the texture of the
parent rock. They are characterized by a very open discontinuous framework and
are estimated to have a void space of about 35%. The breccias have densities
of around 2 g cm"3 which is only slightly greater than unconsolidated soil (Wa
ters et al., 1971; McKay et al., 1970). The spaces between the larger clasts of the
framework are filled with glass-rich clastic materials with an average grain size
of about 50 /xm (close to the mean grain size of typical lunar soils). The glass
forming this matrix tends to be well sorted, closely fitted and plastically moulded
against the larger clastic rock fragments. In a general way the texture of these rocks
resembles terrestrial ignimbrites (Waters et al., 1971). Glass fragments exhibit a
wide variety of devitrification features, ranging from incipient to complete devitri
fication.
There are two prominent textural features of the soil breccias which provide in
sights into their depositional history: layering and accretionary lapilli. Very weakly-
defined layering or bedding is seen in many of the soil breccias, generally in samples
where accretionary lapilli are abundant (Waters et al., 1971; Lindsay, 1972a). Accre
tionary lapilli appear to be an intrinsic characteristic of the lithology (McKay et al.,
1970, 1971; Lindsay, 1972a, b). They are similar to terrestrial volcanic accretionary
lapilli (Moore and Peck, 1962) and range in size from 50 μτα to 4 mm. The lapilli
generally have a core of one or more larger detrital grains surrounded by alternating
layers of darkfine-grainedglassy material and of larger detrital grains. Soil breccias
are thus formed directly from the lunar soil during meteorite impact and are im
portant in the evolution of the soil because they consolidate finer grained materials
forming larger particles.
Mineral grains are the dominant detrital particles in the intermediate grain sizes,
particularly between 30 and 60 μηι, on the coarse side of the mean grain size of the
50 J.E Lindsay
TABLE 3.4
Modal composition of a mature and an immature highland soil from the Apollo 14 site
Detrital Immature soil Mature soil

component
150-200 90-150 60-75 20-30 150-250 90-150 60-75 20-30
(Mm) (μτη)
Agglutinates 5.3 5.2 6.5 12.5 54.5 60.3 56.5 43.5
Breccias 64.4 54.0 53.0 20.7 23.6 23.5 17.5 7.5
Angular glass 2.9 8.8 5.5 12.0 9.4 8.3 11.5 10.0
Rotational glass 0.8 2.0 0.5 2.0 2.4 0.6 2.0 7.0
Mineral grains 11.4 17.0 31.0 50.0 8.8 5.9 8.5 34.0
Rock fragments 15.1 12.4 3.5 2.5 1.0 1.3 1.0 0.0
Note: The mature soil is finer grained and contains an abundance of agglutinates (from McKay et al.,
1972).
bulk soil (Tkble 3.4). This distribution probably reflects the grain size of the source
rocks to a large extent, although the physical properties of the minerals themselves
are also important. Some minerals, notably feldspar, tend to comminute faster than
others.
As is the case for lithic fragments, the detrital mineral grains present in a soil
reflect, for the most part, the nature of the underlying bedrock. The grains are gen
erally angular except where rounded grains have been inherited from other source
materials. Most mineral grains, but particularly the plagioclase, show much evidence
of shock modification as a consequence of the hypervelocity impact environment.
Many plagioclase crystals have been disrupted by shock to such an extent that they
have been converted to diaplectic glasses.
Plagioclase is the ubiquitous mineral in the lunar soils. The anorthite content
is generally high in all soils although the range of values in maria soils is gener
ally greater than in highland soils (Apollo Soil Survey, 1974). A small number of
potassic-feldspar grains have been encountered in the lunar soils. Most are small
and attached to, or intergrown with, plagioclase.
Pyroxenes are present in most soils and like the plagioclase frequently show
signs of shock modification. In maria soils they are almost exclusively clinopyroxene
(augite) derived from the maria basalts. Highland soils contain a relatively large
proportion (over a third) of orthopyroxene as well as clinopyroxene. The pyroxenes
in the highland soils probably were mainly derived from norite and highland basalt
clasts in the crystalline breccia basement rock, although pyroxenes are common as
mineral clasts in many highland breccias.
The pyroxene/plagioclase ratio of most texturally mature soils consistently in
creases with decreasing grain size (Tkble 3.4). This may simply reflect the differing
mechanical properties of the two minerals (Lindsay, 1976), or it may be due to
fine grain sizes being transported over greater distances (Finkelman, 1973). How-
ever, the net result is an increase in mafic elements in finer grain sizes which has
considerable bearing on the chemistry of some impact generated glass particles.
Olivine is present in most soils but is very variable in proportion. Only some
maria basalts contain a large percentage of olivine so there is a tendency for it to
vary from sample to sample at any one mare site. Olivine is generally only present in
small amounts (7%) in highland soils, where it is probably derived from pre-existing
troctolite fragments and perhaps from a dunite source. In general, the compositional
range of olivines is larger in highland soils than in mare soils.
Most soils contain some opaque minerals, mainly ilmenite. The ilmenite is prob
ably largely basaltic in origin at both mare and highland sites. It is very variable in
amount and can be quite abundant in some mare soils. A variety of other minerals,
such as spinels, is present in the soils but in minor amounts only.
Metallic grains are present in very small numbers in all of the lunar soils. The
particles are generally small, few being larger than 100 μπι, and consist largely of
kamacite and taenite and in some cases Schreibersite and troilite. Some particles are
free standing metal whereas others are associated with a silicate assemblage. Bulk
chemistry of the particles indicates that some come from igneous rocks, especially
basalts while others are from meteoritic materials.
Glass particles are abundant in the lunar soils and provide some of the clearest
insights into its provenance and to some extent its evolution. Compositionally and
morphologically the glasses are extremely complex but they can conveniently be
divided into two broad categories: glasses which are essentially homogeneous, and
agglutinates which are extremely inhomogeneous.
Homogeneous glasses. The homogeneous glasses are morphologically diverse.
However, most are angular jagged fragments obviously derived from larger glass
fragments. A smaller number of glass particles have rotational forms (dumbbells,
spheres, spheroids (Figure 3.7) and teardrop shapes). Some particles have detrital
rock or mineral fragments as cores. Others have a small number of mineral frag
ments dispersed through otherwise homogeneous glass. Some glasses are vesicular
with vesicles forming as much as 30% of the particle volume. In the extreme some
spheres are actually hollow bubbles up to 2 cm in diameter. Variations of the mor
phology of rotational glass forms may be explained in terms of the surface tension of
the molten glass, and the angular velocity of the spinning glass mass as it is ejected
from the impact crater (Fulchignoni et al, 1971). The regular form of many homo
geneous glasses suggests that they are impact melts sprayed onto the lunar surface
during crater excavation.
Homogeneous glasses come in a wide variety of colors (colorless, white, yellow,
green, orange, red, brown and black) although most particles have darker colors,
generally brown or black. In general terms, the darker glasses contain more Fe
and Ti whereas the lighter glasses tend to be more aluminous. The lighter colored
glasses tend to be more anorthositic or "highland" in composition, whereas the
darker glasses tend to be basaltic or "mare" in composition. The angular and rota-
52 J.E Lindsay
tional glass forms have the same range of colors and refractive indices, suggesting a
common origin.
The chemistry of the glass particles generally reflects the composition of local
bedrock (Table 3.5). Since the Apollo landing sites were chosen to give insights
into the origin of various major features visible on the lunar surface, especially the
distinction between maria and highlands, it is instructive to look at chemistry of
soils collected at two of these sites, Apollo 14 and Apollo 15. The Apollo 15 site
is essentially a mare site, and the Apollo 14 site is situated in the lunar highlands.
The Apollo 15 data are particularly instructive in that the site is on the mare surface
but a short distance from the highlands (Apennine Front). Both highland and maria
rock compositions are, therefore, well represented. Tb add to the complexity of the
Apollo 15 site the soils also contain a modest proportion of green-glass particles of
basaltic composition which are believed to be pyroclastic in origin.
A total of eleven compositional types occur among the homogeneous glasses
from the Apollo 15 site (Reid et al., 1972), of which the green glass particles are
the most common. The green glasses, which are mostly spheres containing a few
olivine needles, appear to be pyroclastic materials formed in fire fountains during
the flooding of the lunar maria. Similar orange glasses occur at the Apollo 17 site.
There is considerable variation in the pyroclastic contribution in surface samples
and rotational pyroclastic glass forms occur vertically through the soil blanket.
The soils from the Apollo 14 site contain a similar range of compositional types
to the Apollo 15 samples, but the relative proportions of the glasses change in such
a way that the amount of highland-derived Fra Mauro basalts increases whereas
Fig. 3.7. a. A broken homogeneous glass particle with a rotational shape from the lunar soil. The
particle is approximately 220 μιη in length (NASA S-72-52308). b. An almost perfect glass sphere 390
μπι in diameter formed as a response to surface tension in the airless lunar environment. The particle
is a basaltic green glass that is probably volcanic in origin but its surface has been spalled by later
micrometeorite impacts at the surface of the lunar soil (NASA S-72-53599).
TABLE 3.5
Average composition of the most common homogeneous glass types found at the Apollo 14 site
Mare type Fra Mauro type Anorthositic Gabbroic Granitic Low-silica

basaltic basaltic gabbroic anorthositic glass glass
Si02 45.48 48.01 45.23 47.37 71.54 37.97
Ti02 2.77 2.02 0.36 0.14 0.39 0.23
AI2O3 10.86 17.12 25.59 31.32 14.15 34.54
FeO 18.14 10.56 5.59 2.98 1.79 1.19
MgO 11.21 8.72 7.84 2.18 0.70 5.57
CaO 9.56 10.77 14.79 14.78 1.97 20.39
Na20 0.39 0.71 0.25 0.95 0.93 0.00
κ2ο 0.32 0.55 0.12 0.22 6.53 0.00
Total 98.73 98.46 99.77 99.94 98.00 99.89
Source: Apollo Soil Survey (1974).
the maria basalts are of less importance (Table 3.5). A large number of glasses at
the Apollo 14 and 15 sites are characterized by a high Fe and low Al content and
resemble the composition of maria basalts. Brown ropy basaltic glasses analogous to
terrestrial tholeiitic basalts with a higher potassium, rare-earth element and phos
phorus content (Meyer et al., 1971) have been found in soils from several Apollo
sites. Glasses characterized by a high A1203 content have been called highland
basalt glasses (anorthositic gabbro). This is the anorthositic component generally
held to be characteristic of the highlands.
Agglutinates. Agglutinates are one of the most distinctive particle types found
in lunar soils and are very important in understanding the origin and evolution of
the soil (Lindsay, 1971,1976). They are intimate mixtures of inhomogeneous dark-
brown to black glass and mineral grains, many of which are partially vitrified (Figure
3.8). Compositionally an agglutinate consists of approximately 50% mineral grains.
A few agglutinates are vesicular (particularly larger particles, Figure 3.8), most are
massive and overall less vesicular than the homogeneous glasses.
Agglutinate surfaces have a coating offinedetrital fragments which gives them a
dull saccharoidal texture (Figure 3.8). Most particles are extremely complex in shape
but have a general rounded form suggesting that their final shape was determined
by the viscosity and surface tension of thefluidglass. Some consist of dendritic glass
projections radiating from a central mineral grain while others are bowl shaped or
take the form of rings or donuts, suggesting that they are "pools" of melt formed in
the bottom of microcraters infine-grainedlunar soil (Lindsay, 1971,1972c, 1975).
Agglutinates are thus constructional particles made up fromfine-graineddetrital
materials. The included particles are generallyfinerthan 125 μχα with a median size
of close to 38 μπι (Lindsay, 1972c) and a size distribution similar to the fine tail of
the bulk soil.
54 J.E Lindsay
Fig. 3.8. a. A scanning electron microscope image of an agglutinate particle from a mature lunar soil
(NASA SP-53160). b. The same particle close up (NASA S-72-53161). c. A very delicate agglutinate
consisting of a mineral grain acting as a core with thin projections of brown glass. The maximum length
of the particle is 230 /im (NASA S-71-51084). d. A thin section of a large agglutinate (approximately
1.5 mm in length) consisting of fine mineral grains bonded by dark glass (NASA S-71-38420).
Single agglutinates may be as large as 1.5 mm, although rarely, and comminuted
agglutinate fragments occur in abundance in particle sizes as small as 16 μτα. How
ever, most unbroken agglutinates occur in a narrow size range between about 250
and 178 μπι. The proportion of agglutinates in any one sample varies considerably
as a function of exposure age (McKay et al, 1974). Their abundance also varies
with depth. Typically the size distribution of unbroken agglutinates has a mean of
184 /im and is moderately to moderately-well sorted. The distribution is generally
fine skewed, the coarse end of the distribution being truncated. Agglutination thus
removes the fine end of the bulk soil grain size distribution and shifts it to the coarse
side of the mean grain size of the bulk soil.
The chemical composition of individual agglutinates broadly reflects the compo
sition of the underlying bedrock. Thus, agglutinates from the maria areas have a
TABLE 3.6
Chemical analyses of three agglutinate particles from soils collected at the Apollo 12 site
Sample 1 Sample 2 Sample 3
Si02 46.7 42.2 44.8
Ti02 2.32 3.34 2.60
AI2O3 15.4 9.8 15.6
FeO 12.4 18.3 13.7
MgO 7.9 12.0 7.8
CaO 10.7 8.9 11.6
Na20 0.56 0.32 0.54
κ2ο 0.92 0.21 0.23
P2O5 0.62 0.19 0.17
MnO 0.10 0.12 0.12
Cr 2 O s 0.21 0.35 0.29
NiO 0.00 0.04 0.04
Total 98 96 97
Source: Chao et al. (1970).
general basaltic composition (Table 3.6). Agglutinate glasses may be very homoge
neous but in general are very variable in chemistry probably as a result of incom
plete mixing (Papike, 1981; des Marais et al., 1973). Because of the inhomogeneity
of these glasses more is to be gained from a study of their bulk chemistry than from
the chemistry of individual particles. There are significant differences in chemistry
between the agglutinate and non-agglutinate fractions of the soils, and neither frac
tion is comparable chemically to a major homogeneous glass group. In particular
it has been found that the composition of agglutinate glasses is very similar to the
finest (<20 /xm) fraction of the soil (Laul and Papike, 1980). Comminution by im
pact produces fractionated fine materials because the different mineral components
comminute at different rates (Lindsay, 1976; Papike et al., 1982). Fines are then
selectively fused by the agglutination process (Lindsay, 1976, Papike et al., 1982) Ex
perimental studies have generally supported these observations (Horz et al., 1984;
Simon et al, 1985,1986).
Texture of the lunar soil
Grain size
Typically the lunar soils are fine grained and poorly sorted. They are moderately
coarse skewed and near log normally distributed (Figure 3.9). There is, however, a
considerable variability in the grain size parameters of samples from any one site
(Lindsay, 1976).
The mean grain size of the soils ranges from a coarse extreme of 380 μιη to
as fine as 32 /xm, although most soils have means close to 62.5 μπι. The coars
est soils are believed to be primary detrital materials excavated from the bedrock
56 J.E Lindsay
90
80
70
60
^50
#40
α>
•S 30
J2
J 20
10
0.5 0.25 0.125 0.062 0.031

Grain Size (mm)
Fig. 3.9. Cumulative grain size distributions for typical lunar soil from the Apollo 11 site (after Lindsay,
1971).
beneath the soil whereas the finest soils appears to be the product of gaseous
sorting, perhaps in a base surge. Variations in mean grain size are not random
but form part of a time sequence relating to the amount of reworking each soil
layer has undergone. When plotted stratigraphically the mean grain size of the
soil decreases upwards in a regular manner, with minor erratic excursions which
are probably due to the introduction of either older coarse soils or freshly exca
vated bedrock material (Lindsay, 1973). Further, the mean grain size of the soil
is strongly related to the content of agglutinates on the coarse side of the mean
(Figure 3.10; McKay et al., 1974). As the soil blanket evolves the accumulation rate
decreases and the soil is subjected to longer periods of reworking by micromete-
oroids resulting in an increased agglutinate content and a finer mean grain size.
It is also apparent from Figure 3.10 that the scatter about the regression line in
creases as the grain size of the soil decreases. This scatter probably relates to
random destruction of large numbers of agglutinates by larger layer-forming impact
events.
The sorting of the lunar soils likewise varies. Like the mean grain size, the sorting
of the lunar soils conforms to a time sequence in which older soils are better sorted
(Lindsay, 1973). There is also a strong relationship between mean grain size and
Extraterrestrial soils — the lunar experience
250μίτι 125ΜΠΙ 62.5μπι 31.3ΜΠΙ

Mean grain size
Fig. 3.10. Mean grain size of lunar soils as a function of agglutinate content in the 90 to 150 μηι fraction
of the soil (after McKay et al., 1974).
sorting which shows that finer soils are better sorted (Lindsay, 1976). Similarly, the
better sorted soils contain a greater abundance of agglutinates on the coarse side of
the mean.
Agglutinates are extremely delicate, fragile particles that can be destroyed by
comminution much more readily than rock or mineral fragments of equivalent size.
The relationship between agglutinate content and exposure age indicates that under
normal conditions micrometeoroid reworking produces more agglutinates than it
destroys by comminution (Lindsay, 1972c, 1976). The destruction of agglutinates
reduces the mean grain size of the soil and makes the grain size distribution more
symmetrical (that is, less coarse skewed). The larger the agglutinate content of the
soil the more dramatic the effect. Once excavated to the surface micrometeoroid
reworking again takes effect and the grain size parameters begin to converge on
the ideal evolutionary path. We thus see evidence of a series of random excursions
during which agglutinates arefirstformed at the expense of the fine tail of the grain
size distribution by micrometeoroid reworking, and are then crushed and shifted
back to the fine tail again. This effect has been referred to as cycling and occurs in
the most mature soils.
Shape
The shape of soil particles varies considerably, from the smooth spherical form
of glass droplets to the extremely irregular and complex shapes of agglutinates.
58 LE Lindsay
Between these two extremes are the blocky angular comminuted rock, mineral and
glass fragments. The overall mean sphericity of the soil particles is 0.78 (a value of
1.0 being a perfect sphere). However, perhaps predictably, the sphericity of particles
varies markedly with grain size and depth in the lunar soil (Lindsay, 1972c, 1974,
1975).
If we look at the sphericity of detrital particles forming a primitive soil (con
sisting largely of comminuted rock and mineral fragments) we find that sphericity
changes in a regular linear manner with grain size. Smaller particles become increas
ingly more spherical. If we then investigate a texturally mature soil we find that a
sinusoidal distribution is superimposed on the general linear trend. T\vo zones of de
pressed sphericity develop; one at between 125 μτη and 1 mm, the other at between
8 and 31 /im. The midpoint of these two zones lies close to one standard deviation
either side of the mean grain size of the bulk soil. The zones of reduced sphericity
coincide with whole agglutinates on the one hand and comminuted agglutinates on
the other. The relationship between particle shape and size thus relates entirely to
the textural evolution of the soil and in particular to micrometeoroid reworking at
the lunar surface.
Evolution of the lunar soil
In spite of the tenuous nature of the energy source and the large inefficiencies
involved, the meteoroidfluxhas, over several aeons, produced an extensive body of
soil over most of the lunar surface. This soil body probably contains stratigraphie
information covering most of the life time of the solar system and perhaps a record
of events of galactic magnitude. Clearly the soil body is dynamic and the processes
forming it are ongoing which raises questions about its evolution and how informa
tion concerning its origins is preserved in the soil stratigraphy.
Energy partitioning and thefluxof detrital materials

Lunar soils are the product of hypervelocity impacts that occur when meteorites
traveling at orbital velocities strike the planetary surface. The energy released by
hypervelocity impact at the lunar surface is partitioned in a complex way. The ki
netic energy of an individual impact event is released through heating and either
fusing or vaporizing both target and projectile, comminuting the substrate and fi
nally ejecting the heated and comminuted materials. The kinetic energy of the
meteorite flux is also partitioned in different ways according to the mass spectrum
of that flux. Different portions of the mass spectrum contribute to, or modify, the
soils in different ways (Figure 3.11). It is as yet difficult to evaluate these variables
but some order of magnitude estimates have been made (Lindsay, 1975). The post-
mare meteoroid flux at the lunar surface produces a primary sediment flux of the
order of 2.78 x 10"7 g cm"2 per year, which represents an erosional efficiency of
approximately 1.6% (Gault et al., 1972; Lindsay, 1975). The sediment flux per unit
Log (mg)
Fig. 3.11. The size distribution of the meteoroidflux,a primitive lunar soil, a mature lunar soil and the
agglutinates from a mature lunar soil all expressed on the same scale as weight percent per decade of
mass to allow direct comparison (from Lindsay, 1975,1976).
area entering the earth's oceans is 175 times larger than the lunar sediment flux.
Bedrock material appears to be excavated most efficiently by meteoroids larger than
103 to 104 g. In contrast, agglutination is caused by micrometeoroids in the mass
range of 10"7 to 10~4 g or by about 68% of the flux mass. Layer-forming events
60 J.E Lindsay
appear to be the product of meteoroids larger than 7 g or less than 1% of the flux
mass.
Soil accumulation is a self-damping process such that the average accumulation
rate decreases with time (Lindsay, 1976; Quaide and Oberbeck, 1975). If the mete-
oroid flux had remained constant over time its effectiveness as an agent of erosion
would gradually be reduced as the soil blanket grew in thickness (Figure 3.12). For
new material to be excavated from the bedrock beneath the soil an impact must be
energetic enough to penetrate the pre-existing soil layer. As the soil blanket grows,
more and more energetic events are required to accomplish the same result. How
ever, it is evident that the numberfluxof particles decreases rapidly with increasing
particle size and with it the available erosional energy must also decrease. The en
ergy actually available for erosion of bedrock is probably considerably less than 1%
of the total meteoritic energy incident on the moon (Figure 3.11; Lindsay, 1975).
Whatever the history of bombardment, there should be rapid initial accumulation of
soil followed by gradually decreasing growth rates.
3 4 5
Log (mg)
Fig. 3.12. Soil thickness in relation to the minimum meteoroid mass capable of penetrating the soil to
excavate bedrock (after Lindsay, 1975).
The variations in median soil-thickness thus reflect differences in elapsed time

since the production of the new rock surface upon which the soil is evolving. The
soil begins as a thin deposit of nearly uniform thickness and gradually changes to a
thicker deposit with a greater spread of thickness values. That is, the standard devi
ation of the soil thickness frequency distribution increases with time. The increasing
spread of thickness values highlights the stochastic nature of the impact process
— soil accumulation is a discontinuous process dependent upon random (in both
space and time) hypervelocity impacts.
Modelling of such complex processes requires more knowledge than is currently
available about the meteoroidflux,and therefore some simplifying assumptions are
necessary. Quaide and Oberbeck (1975) have used a Monte Carlo approach to study
soil development on maria surfaces. If the meteoroid flux has been constant since
the flooding of the maria the thickness (Th) of the soil blanket is related to its age
<A)br.
Th = 2.0&40 M
where Th and ,4 are in meters and aeons, respectively. The accumulation rate (dThl
6A) decreases exponentially with time. This observation has considerable bearing
on the understanding of the texture and mineralogy of the soil. As the accumulation
rate declines more and more of the kinetic energy of the meteoroidfluxis redirected
into mixing and reworking the soil. This extra energy is particularly important in
modifying the grain size parameters of the soil which in turn affects the nature
of the impact-produced glass particles and the glass content of the soil. Because
smaller volumes of fresh bedrock material are added with time, any materials that
are cumulative will increase in concentration; this is true of impact glasses (Lindsay,
1971,1972c) and should be true of the meteoritic materials themselves.
Textural evolution
The texture of the lunar soil evolves in a regular manner in a direct response to
continued reworking by the meteoroid flux (Lindsay, 1976). The two main dynamic
processes responsible for this developmental sequence are comminution and agglu
tination. The textural maturity of the lunar soil is determined almost entirely by
the balance between these two opposing processes, one destructive the other con
structive. As a result of this complex interaction between the soil and the meteoroid
flux the soil passes through three transitional evolutionary stages (Lindsay, 1975).
For convenience the stages are designated: the comminution dominated stage, the
agglutination dominated stage and the steady state or cycling stage (Figure 3.13).
The comminution dominated stage

Because of the extreme age of the lunar surface few soils have survived that
could be assigned to this evolutionary stage. Energetic impact events striking lunar
62 J.E Lindsay
Standard deviation 0
Fig. 3.13. Stages in the evolution of the lunar soil shown by changes in mean grain size and sorting. The
solid line indicates the evolutionary path of an idealized soil. Contours show the density distribution of
analyzed soil samples. Note that most soils are finer than the mean agglutinate size and they concentrate
at a mean of 58 μνα (4.1 phi), the stable point between the two glass modes. Grain size is expressed in
phi units where grain size in mm = 2~ p h l .
bedrock produce coarse grained and poorly sorted éjecta. When such loose par-
ticulate materials are subjected to repetitive fracture the grain size distribution is
gradually modified in such a way that it asymptotically approaches a log-normal
form (Kolmogoroff, 1941; Halmos, 1944; Epstein, 1947). The mean grain size of
the paniculate materials is gradually reduced and they become more poorly sorted.
These primitive soils are coarse grained with means in excess of 1 mm. During this
stage of soil development comminution is the only effective process as most mete-
oroids interact with single detrital particles. The particles of soil are so large that
most micrometeoroid impacts are either involved in catastrophic rupture of grains
or they simply form pits on the larger grains.
Experimental studies suggest that cratering occurs in loose paniculate material
only when the particle mass is less than an order of magnitude larger than the
impacting projectile. Thus comminution will dominate soil dynamics until the soil
particles are reduced to masses of the order of 10"5 g, that is, an order of magnitude
larger than the mean mass of the meteoroidflux(10"~6 g). This suggests that the soils
would have a mean grain size of the order of 95 //m. At the very least it is unlikely
that agglutination can become effective until the mean grain size of the soil is less
than 125 μπι, the mean grain size of the agglutinate distribution.
The agglutination dominated stage

Most of the soils presently at the lunar surface appear to be passing through
this stage of development. Once the mean grain size of the soil is small enough
for agglutination to be effective, it begins to balance the effects of comminution
(Lindsay, 1976). This results in a continuing decrease in the mean grain size as larger
lithic and mineral grains are comminuted; the standard deviation of the grain size
distribution is reduced causing the ideal evolutionary path in Figure 3.13 to reverse
its slope. During this stage there is a strong correlation between mean grain size and
standard deviation, and the distribution becomes skewed towards the coarser grain
sizes. The glass particles produced by the two opposing processes (agglutination and
comminution) are formed in two narrow size-ranges. Agglutinates concentrate at
approximately 177 /xm while comminuted agglutinates accumulate at approximately
16 /xm. The mean grain size of most lunar soils lies between these two glass modes
(Lindsay, 1972c). Ultimately, as the coarse rock and mineral fragments are broken
down by comminution, the mean grain size is forced to lie between the two glass
modes at close to 63 μτη and the standard deviation of the soils is roughly coinci
dent with the glass modes. At this point the soils are approaching a steady state.
However, agglutination still dominates in that more agglutinates are formed than
are destroyed by comminution.
The grain size distribution of the agglutinates appears to be controlled by a
complex of variables, but to a large extent it is a direct function of the meteoroid
mass-frequency distribution. At the root-mean-square velocity of 20 km s"1 a me
teoroid fuses or vaporizes about 7.5 times its own mass of detrital materials (Gault
et al., 1972; Lindsay, 1975). Experimental work shows that most agglutinates are
formed at shock induced pressures of between 381 and 514 kbar (Gibbons et al.,
1975). Assuming that the detrital material enclosed in the agglutinate forms part
of the solid-liquid-mixture transition, an agglutinate should be about 5 times the
mass of the impacting meteoroid. The size distribution of agglutinates on the fine
side of 217 μτηfitsthe model well. However, on the coarse side of 212 μτη the size
64 J.E Lindsay
distribution is depleted by comparison with the micrometeoroid flux. This implies

that agglutination is inhibited by a second variable, which relates to scale (Lindsay,
1975). The most likely variables are viscosity and surface tension of the melt. Ag
glutinate formation relies on the ability of the melt to maintain its integrity during
crater excavation. As event magnitude increases the probability that an agglutinate
will form decreases, because the glass is dissipated as small droplets. Consequently,
the agglutinate grain size distribution isfineskewed and by comparison with the me-
teoroid flux it is excessively deficient in coarser particles. The standard deviation of
the agglutinate distribution cannot, therefore, be larger than the standard deviation
of the meteoroidflux;it is in fact about half of the equivalentfluxvalue.
The mean agglutinate size distribution suggests that most agglutinates are formed
by micrometeoroids in the mass range 5.5 x 10"5 to 7.0 x 10"8 g. This mass range
includes approximately 68% of the mass and hence of the kinetic energy of the
meteoroid flux. Obviously, agglutinate formation is one of the major processes
active in the reworking of mature lunar soils.
The steady state stage

As the evolving soil moves closer to the point of stabilization, where the mean
lies between the glass modes, the grain size parameters become more and more
dependent on the presence of the agglutinate mode.
At this point a soil may enter a cycling mode. Agglutinates are fragile and readily
crushed so that when a soil is redistributed by a larger layer-forming impact event a
large proportion of the agglutinate population is destroyed. Removal of the aggluti
nates causes the mean and standard deviation to decrease and become more coarse
skewed. That is, the soil tends to take on the grain size parameters of the second
glass-mode. The soil is then exposed at the lunar surface to the micrometeoroid flux
and agglutination once again becomes effective, shifting the grain size parameters
back to the steady state (Lindsay, 1975).
Probabilistic models for mixing and turnover rates due to hypervelocity impact
suggest that the upper millimeter of the lunar soil is the primary mixing zone.
This mixing zone is probably coincident with the zone in which agglutinates form.
Below this mixing zone the rate of turnover and mixing decreases very rapidly with
increasing depth which explains the well-preserved soil stratigraphy. Surface soil
layers (Figure 3.6) appear to survive from about 1 to 50 m.y. before burial (Gault et
al., 1972).
The presence of normal and reverse graded bedding in the lunar soil suggests
that a single set of depositional processes were operative during the formation of at
least some stratigraphie units. Grain size data from individual units suggests that two
processes may be active in the transport of soil materials following a hypervelocity
impact: base surge and grainflow(see Lindsay, 1976).
This model represents an ideal situation where processes are assumed to be op
erating in a closed system. In reality the soils are continually reworked and soils
from two or more stages are continually being mixed. Soils from earlier erosional
episodes may be fossilized by burial and later re-excavated to mix with more mature
soils at the lunar surface (Lindsay, 1976). At the same time larger impact events may
penetrate the soil blanket and excavate bedrock material. The result is a complex in-
terbedded sequence of mixed soils that gradually converge to a steady state upward
through the stratigraphy (Figure 3.13).
Soil maturity
As the lunar soils evolve they change in texture and composition, that is, they
mature. The term "maturity", when applied to detrital material in a terrestrial
setting, is generally used in the sense of a measure of chemical and physical progress
along some predetermined evolutionary path. If we consider sand accumulating on
a beach we see several processes that operate to modify the texture and composition
of the detrital materials. Wave action abrades, rounds and sorts the detrital particles
while less stable minerals deteriorate chemically in the aqueous environment and
are removed. The end product is a well-sorted sand consisting of well-rounded
quartz grains. In the case of beach sand the proportions of detrital quartz can
be used as a measure of maturity. Thus in Pettijohn's (1957) concept of maturity
monomineralic quartz sand is "the ultimate end product" to which beach sand is
"driven by the formative processes that operate upon it".
Tb apply Pettijohn's (1957) concept to the waterless depositional environment
of the lunar soil requires a deeper insight into what we are actually attempting to
measure with a maturity index. The answer is simply the total energy expended
in the formation of the sediment (Lindsay, 1971, 1976). The formative processes
operating on the lunar soil in large part relate to the kinetic energy of the meteoroid
flux at the lunar surface. Each impact on the lunar surface releases a significant
proportion of its kinetic energy in the form of heat energy (Gault and Heitowit,
1963; Braslau, 1970). In turn a proportion of this heat energy is expended in fusing
some of the detrital materials excavated by the impact. Since the soil is continually
reworked by the meteoroid flux the glass content of the soil increases with time.
Thus the total glass-content of the soil is a measure of mineralogie maturity (see
Lindsay, 1976).
However, the maturity index is not a simple linear function with time. While
it is obvious that the total glass-content of the soil must increase with time the
rate of increase in the glass content will decline exponentially as the accumulation
rate decreases and more energy is expended in remelting pre-existing glass. The
agglutinate content of the soils has also been suggested as a measure of maturity.
This measure offers a more sensitive index of textural maturity during the later
stages of soil development and reflects more the recent exposure age than the
overall soil maturity (Lindsay, 1976). However, very young or very old soils do not
conform to the agglutinate model.
66 J.F. Lindsay
Other worlds
In spite of the obvious differences in mineralogy, chemistry and texture between

lunar and terrestrial soils from the view point of soil mechanics and engineering they
share many similarities. Because they were formed in a vacuum in an environment
free of the effects of oxygen and water the fine soil particles have "clean" surface
that allow electrostatic bonding and clumping to occur readily with the result that
the lunar soils behave in much the same manner as might be expected of a granular,
slightly cohesive (perhaps damp) terrestrial soil. An astronaut stepping onto the
surface of the moon for the first time walked on soil with a familiar physical feel, if
not a familiar environment. The soil is firm and it retained clear impressions of the
astronaut's boots (Figure 3.4).
Soils occur on most if not all planetary surfaces. They are the product of geologic
processes and reflect the planet's history (Figure 3.14). Through lack of knowledge
it is difficult to know what we might expect on the Jovian planets but on the smaller
planets such as Mercury or the Asteroids we can confidently predict a soil very much
like that existing on the moon. Similarly, on planets intermediate in size between
the earth and the moon such as Mars we see what might best be described as hybrid
soils where the meteorite flux has been involved in generating clastic materials that
have then been modified and redistributed by the effects of a tenuous but significant
atmosphere. Venus likewise seems to share many features in common with the
Martian setting for although its atmosphere is dense it does not have a hydrosphere
and the resulting surface features suggest a combination of meteorite impact and
eolian processes.
At the lower end of the size range, rocky asteroids of modest size (300 km) could
be expected to have regoliths as much as 3.5 km thick. Asteroids smaller than 300
km (and small satellites such as those of Mars) will have thin regoliths because
of éjecta loss resulting from their small gravitational fields. Unlike the moon the
ENDOGENIC
ATMOSPHERIC
1.
PLANETARY DYNAMICS METEORITIC
OCEANIC
WIND, WAVES, HYPERVELOCITY

GLACIATION, ETC. SOLAR WIND IMPACT
Fig. 3.14. Simplified schematic outlining processes affecting soil formation on planetary surfaces.
regoliths on these small bodies will not show evidence of extensive reworking be
cause the rate at which éjecta blankets form exceeds the excavation rate (Bunch
and Rajan, 1988). Evidence of the importance of meteorite impact on the surfaces
of the smaller planets and satellites is abundant in the form of intense cratering.
All satellite surfaces so far studied, with the exception of Io and to a lesser extent
Europa, are cratered, as are Mars and Mercury. In the case of Mercury the moon is
probably a very effective analog, whilst on Mars and Venus the atmosphere plays an
important role in soil formation, and subdues the role of hypervelocity impact. The
spectral characteristics of Mercury suggest that it is covered with a mature lunar-like
soil. Io is exceptional, for in spite of its modest size gravitational interaction with
the main planet results in active volcanism; soil formation is therefore dominated
by endogenic processes. Europa is sheathed in ice and has more in common with
the other outer planet satellites which consist of mixtures of water frost and dark
silicate or carbonaceous materials. All of the icy satellites are expected to have
impact generated regoliths. Evidence of impacts is clear and there are significant
variations in albedo that imply reworking of the surface. Endogenic processes were
also important on many of these larger satellites. However, "water vulcanism", for
example, is little understood. Latitude dependent variations in albedo in some cases
suggest other active surface processes, perhaps relating to temperature, but as yet
not understood (Veverka et al., 1986). On other less well understood satellites such
as Titan and TViton, which have significant atmospheres, other processes come into
play; the soils there would be the product of chemical weathering and other atmo
spheric processes, much as on earth but modified by their distance from the sun and
reduced solar radiation. The formation of soil is thus a complex process dependent
upon many variables: the size of the planet, the composition of its surface, its dis
tance from the sun and, in the case of satellites, its orbital relationship to the main
planetary body (Figure 3.14).
Given all of these complexities, understanding the lunar soil is all the more im
portant because it provides a point of reference for any planetary study. While
only a superficial blanket on a planetary surface it is the soil that provides the first
point of contact for any scientific study, whether the study be carried out remotely
from earthbound telescopes or orbital platforms, or by direct observations using
manned or unmanned spacecraft on the planetary surface. Clearly, much remains to
be learnt about soils or regoliths on other planetary surfaces. Manned landings on
Mars planned for the early 21st century are likely to extend our knowledge of sur
face processes significantly, but the main progress in understanding extraterrestrial
soils is likely to come from unmanned planetary missions. Surface landings by un
manned spacecraft offer the greatest potential, but, in the case of the smaller planets
and asteroids lacking an atmosphere, orbital experiments using spectral character
istics and remote methods such as passive X-rayfluorescencewill probably provide
the great bulk of data for the foreseeable future.
68 J.E Lindsay
References
Adler, I., Trombka, J., Gerard, J., Lowman, P., Schmadebeck, R., Blodgeth, H., Eller, E., Yin, L.,
Lamothe, R., Gorenstien, P. and Bjorkholm, P., 1972a. Apollo 15 geochemical X-ray fluorescence
experiments. Preliminary report. Science, 175: 436-440.
Adler, I., Trombka, J., Gerard, J., Lowman, R., Schmadebeck, R., Blodgeth, H., Eller, E., Yin, L.,
Lamothe, R., Gorenstien, P., Bjorkholm, P., Gursky, H., Harris, B., Golub, L. and Harnden, R. E,
1972b. X-ray fluorescence experiment. In: Apollo 16 Preliminary Science Report, NASA Sp-315,
19-1-19-14.
Apollo Soil Survey, 1974. Phase chemistry of Apollo 14 soil sample 14259. Modern Geol., 5: 1-13.
Braslau, D., 1970. Partitioning of energy in hypervelocity impact against loose sand targets. J. Geophys.
Res., 75: 3987-3999.
Bunch, T.E. and Rajan, R.S., 1988. Meteorite regolithic breccias. In: J.E Kerridge and M.S. Matthews
(Editors), Meteorites and the Early Solar System. Univ. of Arizona Press, Tucson, Ariz., pp. 144-
164.
Carrier, W.D., Mitchell, J.K. and Mahmood, A., 1974. Lunar soil density and porosity. Proc. 5th Lunar
Sci. Conf., Suppl. 5, Geochim. Cosmochim. Acta, 3: 2361-2364.
Chao, E.T.C., Boreman, J.A., Minkin, J.A., James, O.B. and Desborough, G.A., 1970. Lunar glasses of
impact origin: physical and chemical characteristics and geologic implications. J. Geophys. Res., 75:
7445-7479.
Criswell, D.R., 1972. Lunar dust motion. Proc. 3rd Lunar Science Conf., Suppl., 3, Geochim. Cos
mochim. Acta, 3: 2671-2680.
des Marais, D.J., Hayes, J.M. and Meinschein, W.G., 1973. Accumulation of carbon in lunar soils.
Nature (London), 246: 65-68.
Duke, M.B. and Nagel, J.S., 1975. Stratification in the lunar regolith — a preliminary view. The Moon,
13: 143-158.
Epstein, B., 1947. The mathematical description of certain breakage mechanisms leading to the loga
rithmic-normal distribution. J. Franklin Inst., 44: 471-477.
Finkelman, R.B., 1973. Analysis of the ultrafine fraction of the Apollo 14 regolith. Proc. 4th Lunar Sci.
Conf., Suppl., 4, Geochim. Comsochim. Acta, 1:179-189.
Fulchignoni, M., Funiciello, R., Taddeucci, A. and Triglia, R., 1971. Glassy spheroids in lunar fines from
Apollo 12 samples 12070.37,12001.73, and 12057.60. Proc. 2nd Lunar Sci. Conf., Suppl., 2, Geochim.
Cosmochim. Acta, 1: 937-948.
Gault, D.E. and Heitowit, E.D., 1963. The partitioning of energy for hypervelocity impact craters formed
in rock. Proc. 6th Hypervelocity Impact Symp., Cleveland, Ohio, 2: 419-456.
Gault, D.E., Horz, F., Brownlee, D.E. and Härtung, J.B., 1974. Mixing of the lunar regolith. Proc. Fifth
Lunar Sci. Conf. Suppl. 5, Geochim. Cosmochim. Acta, 3: 2365-2386.
Gault, D.E., Horz, F. and Härtung, J.B., 1972. Effects of microcratering on the lunar surface. Proc. 3rd
Lunar Science Conf., Suppl., 3, Geochim. Cosmochim. Acta, 3: 2713-2734.
Gibbons, R.V., Morris, R.V., Horz, F. and Thompson, T.D., 1975. Pétrographie and ferromagnetic res
onance studies of experimentally shocked regolith analogs. Proc. 6th Lunar Sci. Conf., Geochim.
Halmos, P.R., 1944. Random alms. Ann. Math. Stat., 15: 182-189.
Hartmann, W.K., 1966. Early lunar cratering. Icarus, 5: 406^18.
Hartmann, W.K., 1970. Preliminary note on lunar cratering rates and absolute time-scales. Icarus, 12:
131-133.
Hinners, N.W., 1971. The new moon; a view. Rev. Geophys. Space Phys., 9: 447-552.
Horz, F., Cintala, M.J., See, TH., Cardenas, F. and Thompson, T.D., 1984. Grain size evolution and
fractionation trends in an experimental regolith. Proc. 15th Lunar and Planet. Sci. Conf., Part 1, J.
Geophys. Res., Suppl., 89: C183-C196.

Kolmogoroff, A.N., 1941. Über das logarithmisch normale Verteilungsgesetz der Dimensionen der
Teilchen bei Zerstückelung. Dokl., Akad. Nauk SSR, 31: 99-101.
Kuiper, G.P., 1954. On the origin of the lunar surface features. Proc. Nati. Acad. Sci. U.S., 40: 1096
Kuiper, G.P., Strom, R.G. and LePoole, R.S., 1966. Interpretation of the Ranger records. J.P.L. Tech.
Rep. 32-800, pp. 35-248.
Laul, J.C. and Papike, JJ., 1980. The Apollo 17 drill core: Chemistry of size fractions and the nature of
the fused soil component. Proc. 11th Lunar and Planet. Sci. Conf., pp. 1395-1413.
Lindsay, J.F., 1971. Sedimentology of Apollo 11 and 12 lunar soil. J. Sediment. Petrol., 41: 780-797.
Lindsay, J.F., 1972a. Sedimentology of clastic rocks from the Fra Mauro region of the moon. J. Sediment.
Petrol., 42: 19-32.
Lindsay, J.F., 1972b. Sedimentology of clastic rocks returned from the moon by Apollo 15. Geol. Soc.
Am. Bull. 83: 2957-2970.
Lindsay, J.F., 1972c. Development of soil on the lunar surface. J. Sediment. Petrol., 42: 876-888.
Lindsay, J.F., 1973. Evolution of lunar grain-size and shape parameters. Proc. 4th Lunar Sci. Conf.,
Suppl., 4, Geochim. Cosmochim. Acta, 1: 215-224.
Lindsay, J.F., 1974. A general model for the textural evolution of lunar soil: Proc. 5th Lunar Sci. Conf.
Suppl. 5, Geochim. Cosmochim. Acta, 1: 861-878.
Lindsay, J.F., 1975. A steady state model for the lunar soil. Geol. Soc. Am. Bull., 86: 1661-1670.
Lindsay, J.F., 1976. Lunar Stratigraphy and Sedimentology. Elsevier, Amsterdam, 302 pp.
Lindsay, J.F., Heiken, G.H. and Fryxell, R., 1971. Description of core samples returned by Apollo 12.
NASA TM X-58066,19 pp.
McKay, D.S., Fruland, R.M. and Heiken, G.H., 1974. Grain size and the evolution of lunar soils. Proc.
5th Lunar Sci. Conf., Suppl., 5, Geochim. Cosmochim. Acta, 1: 887-906.
McKay, D.S., Greenwood, W.R. and Morrison, D.A., 1970. Origin of small lunar particles and breccia
from the Apollo 11 site. Proc. Apollo 11 Lunar Sci. Conf., Suppl., 1, Geochim. Cosmochim. Acta, 1:
673-693.
McKay, D.S., Heiken, G.H., Taylor, R.M., Clanton, U.S. and Morrison, D.A., 1972. Apollo 14 soils: Size
distribution and particle types. Proc. 3rd Lunar Sci. Conf. Suppl. 3, Geochim. Cosmochim. Acta, 1:
983.994.
McKay, D.S., Morrison, D.A., Clanton, U.S., Ladle, G.H. and Lindsay, J.F., 1971. Apollo 12 soil and
breccia. Proc. 2nd. Lunar Sci. Conf., Houston, Texas, Suppl. 2, Geochim. Cosmochim. Acta, 1: 755-
773.
Meyer, C , Brett, R., Hubbard, R., Morrison, D., McKay, D., Aitken, F, Takeda, H. and Schonfeld,
E., 1971. The mineralogy chemistry and origin of the KREEP component in soil samples from the
Ocean of Storms. Proc. 2nd Lunar Sci. Conf., Suppl. 2, Geochim. Cosmochim. Acta, 1: 393-411.
Moore, J.G. and Peck, D.L., 1962. Accretionary lapilli in volcanic rocks of the western continental
United States. J. Geol., 70: 182-193.
Oberbeck, V.R. and Quaide, W.L., 1968. Genetic implications of lunar regolith thickness variations.
Icarus, 9: 446-465.
Papike, J.J., 1981. The lunar regolith: Significance of the chemistry of the <10 μ fraction (F 3 ). Abstracts,
Lunar Planet. Sci., 12: 805-807.
Papike, J.J., Simon, S.B. and Laul, J.C, 1982. The lunar regolith: Chemistry, mineralogy, and petrology.
Rev. Geophys. Space Phys., 20: 761-826.
Pettijohn, F.J., 1957. Sedimentary Rocks. Harper and Row, New York, N.Y., 718 pp.
Quaide, W.L. and Oberbeck, V.R., 1975. Development of mare regolith: some model implications. The
Moon, 13: 27-55.
Reid, A.M., Warner, J., Ridley, W.I. and Brown, R.W., 1972. Major element composition of glasses in
three Apollo 15 soils. Meteoritics, 7: 395-415.
70 J.E Lindsay
Shoemaker, E.M., Hait, M.H., Swann, G.A., Schleicter, D.L., Schaber, G.G., Sutton, R.L., Dahlem,
D.H., Goddard, E.N. and Waters, A.C., 1970. Origin of the lunar regolith at Tranquility Base. Proc.
Apollo 11 Lunar Sci. Conf., Suppl., 1, Geochim. Cosmochim. Acta, 3: 2399-2412.
Simon, S.B., Papike, JJ., Horz, F and See, TH., 1985. An experimental investigation of agglutinate
melting mechanisms; shocked mixtures of sodium and potassium feldspars. Proc. 16th Lunar and
Planet. Sci. Conf., Part 1, J. Geophys. Res., 90: D103-D115.
Simon, S.B., Papike, JJ., Horz, F and See TH., 1986. An experimental investigation of agglutinate
melting mechanisms: shocked mixtures of Apollo 11 and 16 soils. Proc. 17th Lunar and Planet. Sci.
Conf., Part 1, J. Geophys. Res., 91: E64-E74.
Swann, G.A. and many others, 1972. Preliminary geologic investigations of the Apollo 15 landing site.
In: Apollo 15 Preliminary Science Report. NASA SP-289, 5-1-5-112.
Taylor, S.R., 1975. Lunar Science: A Post-Apollo View. Pergamon Press, New York, N.Y., 372 pp.
Turkevich, A.L., 1973. Average chemical composition of the lunar surface. The Moon, 8: 365-367.
Urey, H.C., 1952. The Planets; Their Origin and Development. Yale, New Haven, Conn., 245 pp.
Veverka, J., Thomas, P., Johnson, TV, Matson, D. and Housen, K., 1986. The physical characteristics
of satellite surfaces. In: J.A. Burns and M.S. Matthews (Editors), Satellites. Univ. of Arizona Press,
Tucson, Ariz., pp. 342-402.
Waters, A.C., Fisher, R.V., Garrison, R.E., and Wax, D., 1971. Matrix characteristics and origin of lunar
breccia samples no. 12034 and 12037. Proc. 2nd Lunar Sci. Conf., Suppl., 2, Geochim. Cosmochim.
Acta, 1: 893-907.
71
Chapter 4
Incipient weathering: some new concepts on

weathering, clay formation and organization
M. ROBERT and D. TESSIER
Introduction
In the last twenty years, both new methodologies and new concepts have been
developed which allow a better understanding of weathering, clay formation and
organization. For instance, scanning microscopy has improved our knowledge of
primary mineral weathering and pointed out the importance of microsites or the
role played by the biological factor in weathering.
High resolution transmission electron microscopy and associated techniques (mi
croprobe and microdiffraction) on ultrathin sections of small-size secondary miner
als have also permitted the weathering at the unit cell scale to be more accurately
followed, for example in phyllosilicate sheets or layers. Thus, more attention must
be paid to the importance of clay texture and its variations according to the main
factors of soil formation.
Weathering processes in relation to mineral structure
Hydrolysis is the general phenomenon which governs weathering (Pedro and

Sieffermann, 1979). If intense leaching by water occurs, all the silicates behave
similarly and this may lead to dissolution of the structure, desilicification and pre
cipitation of new secondary minerals (generally 1:1 clay minerals or oxides and
hydroxides). This process is the main source of clays in tropical areas. An illustra
tion is given in Figure 4.1 where feldspars and micas experimentally weathered by
water show pits in their structure in addition to formation of hydroxides of Al or
Fe. The same phenomenon was found for feldspars (Berner and Holdren, 1979) in
extreme natural weathering conditions (ferrallitisation in the tropics, and podzoliza-
tion in the temperate zone).
In the temperate zone where weathering intensity is relatively lower, reactions
such as cation exchange, diffusion, oxidation-reduction (Mn, Fe) can become dom
inant and solid state transformations occur in some minerals, depending on their
72 M. Robert and D. Tessier
Fig. 4.1. a and b. Feldspar and mica weathered by water, c. Goethite individualization on weathered
biotite. d. Evolution of mica to smectite (Kounestron et al., 1977).
structure and composition. The most common example concerns 2:1 phyllosilicates
and especially micas, which, amongst the rock-forming minerals, constitute the main
source of clay minerals (Dejou et al, 1977).
Figure 4.2a presents a general scheme for the experimental evolution of mica
in relation to solution concentration: alkaline or alkaline-earth elements and pH.
Under neutral or alkaline conditions, mica weathers to vermiculite by K exchange
(Barshad, 1948; Robert 1971) or even to a low-charge vermiculite having smectitic
swelling behavior (Robert and Barshad, 1972).
At pH less than 5, Al becomes mobile and is released from all silicate structures,
but can also be reprecipita ted in 2:1 interlayer space (Jackson, 1963). In this way
"intergrade" or Al hydroxy vermiculite may form. These minerals are characteristic
Incipient weathering: some new concepts on weathering clay formation and organization 73
TRANSFORMATION
Trioctahedral mica
N/1CH:v:%%Γπή
ftOX — ► vermiculite
WxCOv
£ N/100
WPï Maintenance
o of the layer structure
Ü
N/1000-R$8$
ÈÎ^^^^ÔPTHË'STRUCTURË^^^^
12 14
pH
Different acids
15
<
Œ Oxalic
LU
Z c
DESTRUCTION BY
10 À iS COMPLEXATION Citric
CM'8
LL W
ΟΌ Tartaric
zP
y co _
l~ O) I
ACIDITY PROMOTES THE Lactic
°£ I TRANSFORMATION OF fumarie
2:1 MINERALS
PS i AL HYDROXY INTERLAYER
Hydroxybenzoic
c/) ' FORMATION
LU I HCI
O I
250
pH
Fig. 4.2. a. General scheme for trioctahedral mica experimental evolution in function of pH and solution
concentration (from Robert, 1971). b. Changes in trioctahedral micas as a function of the biospheric
conditions (pH vs \ogKc where Kc is the stability constant of Al complexes with the organic acids) (from
Robert and Berthelin, 1986).
of acidic soils such as acid brown earths (Rich, 1968; Righi et al., 1988). But they
become unstable under very acidic conditions on account of the dissolution of both
interlayer and octahedral layers (pH < 3).
Mild acidic conditions (pH 6 to 5) may be associated with presence of higher
amounts of CO2 in soil solutions than in surface water. More dissociated acids are
necessary to have pH values lower than about 5. The presence of inorganic acids in
soil has certainly been underestimated; their formation is mainly due to oxidation
mechanisms. For instance H2SO4 can be formed by oxidation of sulfur compounds
(S, H2S, FeS2) present in the rock itself or in acid rain. The resulting incipient
weathering can be very strong (van Oort and Robert, 1988). A more usual source
of acidity is NH4+ oxidation by bacteria giving rise to both nitrate and H + (Simon-
Sylvestre et al, 1987). Seasonal nitrification can give rise to a flush of Al in soil
water.
Reduction and oxidation of iron can also be a source of protons in soils that
are seasonally submitted to hydromorphic conditions. Fe2+ ions can migrate to
exchangeable positions on clays and upon oxidation and Fe3+ hydroxylation, release
protons to give a corresponding release of Al and the formation of uintergrade,,
minerals (Brinkman, 1970).
If such oxidation occurs in the octahedral sheets of 2:1 minerals, it releases
iron, an increase in the dioctahedral character of the clays (with a corresponding
increase in the stability to weathering) and a decrease in charge, which can explain
the formation of swelling minerals. Such oxidation was also cited (Farmer et al,
1971) as an explanation for the formation of regular interstratified mica-vermiculite.
For iron-rich chlorite also, it represents the main mechanism that leads to the solid
state transformation to vermiculite or smectite (Mackumbi and Herbillon, 1972).
The main sources of acidity in soil are organic acids, especially aliphatic and
aromatic, coming either from organic matter decomposition or more directly from
secretions of living organisms (Stevenson, 1967; Bruckert, 1970).
For these acids, it is possible to define a gradation in complexing properties
(Figure 4.2b; Huang and Keller, 1972; Razzaghe-Karimi and Robert, 1975). In the
presence of 2:1 phyllosilicates a competition for Al exists between the hydroxy-
interlayer and Al complexed by organic matter in solution. Thus, a stability constant
higher than 5 is necessary to remove Al from clays or to inhibit Al intergrade
formation. It can serve to differentiate two systems — acid, or acid and complexing
— both for the release, transport and deposition of elements. Thefirstleads to acid
brown soil formation, the second to podzolization (Robert et al, 1987).
Field experiments conducted with vermiculite as a "test mineral" permitted us
to verify that such weathering processes occur very quickly in acid soils and that
they often have a seasonal variation (Ranger and Robert, 1985, Ranger et al., in
press). Thus, at the subalpine level (in the Alps), podzolization (accompanied by
removal of Al) occurs during the cold season and acidification (with an Al interlayer
accumulation) during summer (Dambrine et al., 1988). In the Amazonian forests,
podzolization processes are linked with the wettest season (Righi et al., in press).
In podzol E horizons, mica can be transformed into vermiculite or even smectite
(Gjems, 1963) and thus be considered as a relict 2:1 phyllosilicate, the Al of which
is being constantly removed.
Between the two great domains of weathering and clay formation which cor
respond to tropical and temperate regions, transition zones exist. For example, in
Ultisols and Alfisols, water hydrolysis, which is less intense than in Oxisols, gives
rise to 2:1 phyllosilicate which become unsaturated with the formation of Al inter-
Incipient weathering: some new concepts on weathering, clay formation and organization 75
grade vermiculite which can represent an intermediate stage to kaolinite formation

(Wada and Kakuto, 1983). A similar interstratification of 2:1 and 1:1 clay minerals
has been found in volcanic soils (Delvaux et al, 1990). In Vertisols, or when soil
solutions are rich in silica and divalent cations like Ca-Mg, all the tecto- or phyl-
losilicates lead to the same paragenesis which is rich in smectites. In the case of
mica however, smectite grows at the border and in the continuum of the mica layers
(Figure 4.1; Jackson and Shridar, 1974; Kounestron et al., 1977).
All these examples of solid state transformation are specific to 2:1 phyllosilicates
and, under certain conditions, to chlorites which are now classified in the same
family. Such a transformation is related to the existence of an interlayer space
where specific reactions occur. Large differences in mineral stability, which can be
of an order of magnitude between di- and trioctahedral minerals, are also explained
by the octahedral layer occupancy (Robert and Veneau, 1978).
Amongst tectosilicates the feldspars have a structure that is too dense to be
transformed and K, Na or Ca cannot be exchanged without severe destruction of
the structure. This is illustrated later at the nanoscale. In feldspathoids, tunnels in
the structure permit exchange and diffusion reactions so that leucite —► analcite
transformations occur at low temperatures (Robert and Veneau, 1974). The struc
ture of natural and synthetic glasses may be similar to those of feldspathoids and
cation-depleted layers can be formed upon weathering (Robert et al., in press).
Microsystems and microsites
Weathering systems can also be considered as a succession of macroscopic weath

ering levels in which rock minerals gave rise to specific clay minerals (Ildefonse
et al, 1979; Meunier and Velde, 1979). Each macroscopic weathering level can be
considered as an assemblage of small-size weathering microsystems. It is possible to
distinguish three geochemical microsystems (Figure 4.3), where:
(a) the reaction is dominated by the rock type and occurs in the coherent rock at
the grain boundaries; neogenesis is host mineral-dependent (contact microsystems);
(b) rock and solution interact to give plasmic microsystems in more advanced
weathering stages (saprolite zone); and
(c) the fissured zones facilitate the circulation of the solution and recrystallization
is induced.
The contact microsystem may be characterized by the growth of sericite-like
illites. The fissurai zone where high water leaching occurs leads to another para
genesis, that of oxides and hydroxides associated with kaolinite and beidellite. The
last system can give different clay mineral parageneses which are dependent on the
composition of the rock, especially for basic and ultrabasic rocks.
Biochemical or biological weathering may give rise to more specific microsystems
which can be found either close to the forest litter horizon rich in organic acids,
76 M Robert and D. Tessier
Fig. 4.3. Weathering microsystems. Principal steps in the weathering of granite during weathering, a.
Unweathered rock: B = biotite; Q = quartz; P = plagioclase; O = orthoclase; M = muscovite; S =
sericite; 1 = widening of the joints; 2 = microfissuration, b. Development of the contact microsystem:
1 = reactional micas; V = simultaneous destabilization; i " = diffusion (1-V-l" = contact reactions);
2 = inert contact; 3 - beginning of the internal destabilization of primary minerals, c. Development of
the plasmic system: 1 = internal destabilization; 2 = plasma; 3 = argilan; 4 = fissure, d. Development
of the fissurai system: 1 = plasma; 2 = fissures; 3 = clay concentration; 4 = recent clay deposit; 5 =
subvertical rusty band. (From Meunier and Velde, 1979.)
or at the interface between constituents and living organisms. Direct observations

of the earliest steps of colonization of rocks and minerals in mountain or arctic
areas make it possible to study incipient weathering (Figure 4.4). Large amounts of
bacteria, algae, fungi and lichens occur at the rock surface and weathering features
occur at the interface (Robert et Berthelin, 1986). Such microorganisms excrete H + ,
Fig. 4.4. Incipient weathering by biological factor (subalpine level on granite), a. Algal colony on
plagioclase. b and c. Hyphae at the rock surface, d. Penetration of lichen hyphae between packs of mica.
(Robert and Berthelin, 1986.)
acids and complexing agents. Dissolution features on different silicates are shown
in Figure 4.5 where even quartz is dissolved. For calcium or magnesium carbonate
rocks, endolithic lichens give galleries inside the rocks. Most often, precipitation of
dissolved products, Si-Al, Si-Fe or Fe and Ca occur very close to the weathering
sites or even around the organism itself (Figure 4.5).
Oxalic acid is frequently secreted by fungi and lichens and is the most active
agent of dissolution, however Ca, Mg or iron oxalates often precipitate (Wilson and
Russell., 1980). The microbial microsystem thus appears to be a very specific one,
with dissolution-precipitation phenomena occurring over a short distance. Iron or
manganese compounds, most often protoferrihydrite, are precipitated in the sheath
around Fe2+ oxidizing bacteria such as Galionella or Leptothrix.
Fig. 4.5. Incipient weathering by the biological factor (Robert and Berthelin, 1986). a. Quartz weathering
by lichen Acarospora hospitaris. b. Mica weathering, c. Si-Al gel formation at the surface of granite, d.
Calcium oxalate precipitation in relation to lichen, e. Microcrystalline lepidocrocite formed in Euglena
sp. (unicellular alga): intracellular and in the cell membrane (Mann et al., 1987). f. Calcite crystals
formed in roots (Jaillard, 1983).
Microorganisms can also concentrate mineral compounds inside their cells (Fig
ure 4.5). Algae, such as Euglena, precipitate iron crystalline compounds (lepidocro
cite) intracellularly or in their cell walls (Mann et al., 1987). But calcium carbonate
is the most common crystalline compound formed either by bacteria, fungi, algae or
roots (Callot et al., 1984), and it precipitates most generally inside the cells (Jaillard
1983).
Many authors have shown that the rhizosphere is certainly the most complex
microsystem. Such a microsystem includes several different zones (Figure 4.6), from
the apex to the main root, where polysaccharides (polygalacturonates), low weight
organic acids, electrons (giving reduction), and protons, are successively excreted.
The rhizosphere shows also an increase in the number of bacteria and the presence
of ecto- and endo-mycorrhizae and it is indeed the most efficient microsystem in
weathering, where minerals can be either transformed (mica —► vermiculite) or
dissolved (Mortland et al., 1956; Spyridakis et al., 1967; Leyval and Berthelin, 1986;
Robert and Berthelin, 1986; TMbuth et al., 1987; Hinsinger et al, 1988). Biological
microsites or microsystems are very important in incipient weathering, and also in
i l Η20 Ν0 3 " Ca 2+
Κ+ Ρ
Vi
ti
Distance from root tip
Schematic flux variations along a root

Fig. 4.6. The rhizosphere as a weathering system (from Jaillard, 1983). Schematic view of root functions.
Some elements are absorbed and others excreted.
the rock cycle (Westbroeck, 1983). For example, Jaillard (1983) showed that in some
calcic soils more than 25% of carbonates pass through the roots and thus appear as
former root cells with a regular 100 μτη size (Figure 4.5).
Because of acid secretions and organic matter decomposition, indirect biochemi
cal weathering is important to all the soil upper horizons. Its intensity decreases with
increased temperature because of organic matter mineralization, and with depth in
the profile.
The most specific system is the complexing system which results in the release
and mobilization of Al and Fe in chelate-like compounds. It also inhibits crystalliza
tion, leading to the formation of ferrihydrite and allophane-like minerals (Huang
and Violente, 1986; Robert and Berthelin, 1986). Acidity, even when the rate of dis
solution is increased, is less specific when compared with normal hydrolysis. When
pH is less than 5.5, aluminum is released in an uncomplexed form such as Al 3+ , or
even as Si-Al protoimogolite (Farmer, 1979) or hydroxy polymer form. Colloidal
transport is also important in such a system. Neogenesis is not disturbed and the
crystallization of oxides and hydroxides and the formation of Al intergrades can
occur (Robert et al., 1987).
The nanoscale: characterization of weathering primary stages
Development of a methodology
Soon after the discovery of X-ray diffraction (XRD) it was applied to the investi
gation of the main phyllosilicate structures (Mauguin, 1928; Pauling, 1930).
Transmission electron microscopy techniques were developed in 1940 and ap

plied with success in biology, especially to the understanding of virus structures. It
was not immediately applied to clays and wasfirstused on orientated clay particles,
where only the shape in the 001 plane can be characterized (Beutelspacher and van
der Marel, 1968; Eberhart, 1976). It took more than twenty years to overcome the
main technical difficulties so that direct information on clay organization in ultra-
thin sections (less than 15 nm) perpendicular to the 001 plane could be observed.
These thin sections can be obtained either by ion milling, which permits the study
of primary mineral weathering (feldspars, micas, chlorites), or by embedding in a
resin and making an ultrathin section with an ultramicrotome. Embedding was first
performed with metaacrylate or araldite under vacuum (Eberhart and THki, 1972;
Tthoubar et al, 1973; Brown and Jackson, 1973; Lee et al., 1975), which permitted
the differentiation of 1.0 and 0.7 nm layers, evidence for interstratified smectite/
kaolinite and the presence of mica layers in kaolinite crystals.
Another difficulty which had to be overcome was to keep intact the clay organiza
tion, and especially the open interlayer space in order to differentiate the main 2:1
clay minerals. If the swelling clay is embedded under vacuum and subsequently sub
mitted to high vacuum in the microscope, the 001 distance is close to 1 nm. These
difficulties were overcome by Spurr in 1969 for biomaterials and applied to clays by
Tfessier (1978). To do so, the sample isfirstbrought to a controlled hydration state
under gas pressure, water is replaced by alcohol, then by propylene oxide before
embedding with Spurr's resin.
Main studies on prìmary mineral weatherìng
Non phyllosilicates
For the last twenty years high resolution transmission electron microscopy
(HRTEM) has given information at the unit cell scale or the structure, polytypism
and weathering of the main minerals (Brown and Rich, 1968; Buseck and Iigima,
1974). Channels can be imaged in the structure of ring and chain silicates and lay
ers in the phyllosilicates. Veblen and Busek (1980) showed the existence of various
intergrowth microstructures of pyroxene —► amphibole, pyroxene —» sheet silicate,
and postulated several different mechanisms of hydration.
Weathering of olivine to iddingsite or goethite was described by Nahon et al.
(1982); Eggleton (1984) showed the role of defects and dislocation and the exis
tence of different weathering stages. The first stages of weathering, (Veblen and
Busek, 1980; Eggleton and Boland, 1982), very often involve solid state recrystal-
lization of phyllosilicates or hydroxides with dimensions controlled by the host min
erals. These authors have shown that structural continuity is possible between all
the primary minerals which contain edge-linked octahedra and between secondary
phyllosilicates (Eggleton, 1975, 1986). An example is the topotaxial alteration of
pyroxene to a single crystal of nontronite over domains which can be greater than
2 //m (Figure 4.7). Similarly in a soil from Southern Portugal, Abreu and Vairinho
(1990) showed such a solid state transformation of amphibole into trioctahedral
vermiculite. Feldspars have no octahedrally coordinated cations and there is no
inheritance either for the octahedral or even for the tetrahedral layers (Eggleton
and Busek, 1980). So destruction of the structure and rebuilding of new minerals is
necessary. Etch pits develop at the twin planes between mono- and triclinic domains
and are the sites for clay mineral formation (Figure 4.8). Recently Tkzaki and Fyfe
(1987) described intermediate states between feldspars and clays which had circular
structures 15 to 20 nm in diameter with 1.4-2 nm lattice fringes or long fiber forms
with various lattice spacings. These structures are precursors of halloysites. Eggle
ton (1987) found similar packed hollow spheres close to 20 nm in non crystalline
Fe-Si-Al oxyhydroxides which could develop in 1:1 or 2:1 clays minerals. Tkzaki et
al.(1989) have shown that these structures can also exist in glasses.
However, all these studies have not been placed in the context of soil genesis.
More recently, Romero et al. (in press) studied the first stages of clay genesis on
different crystalline rocks in NW Spain where climatic conditions are fairly similar
to subtropical conditions (high precipitation, but relatively low mean temperature
15°C). Ultrathin sections of small-size feldspars showed that the first stage is an
amorphization without any visible reorganization, with chemical release of K, Na,
Ca and decrease of Si/Al ratio. At other stages, exsolution and mass transformation
gave paracrystalline compounds such as allophane or imogolite which are possible
sources of crystalline phyllosilicates (Figure 4.8).
Phyllosilicates
The weathering of mica macrocrystals is well documented because HRTEM im
ages are easier to obtain. During hydrothermal processes (300°C), chloritization of
mica occurs through brucitization of a trioctahedral mica. This process has given
rise to two hypotheses (Figure 4.9). Olives Banos (1985) and Olives Banos et al.
(1983) suggest that brucitization occurs in the interlayer space of biotite (potassium
planes) where partial slip or cleavage has occurred. This leads to single chlorite
layers which are interstratified in the mica structure (Figure 4.9a). Such a mecha
nism of transformation results in an increase in volume of about 30%. Eggleton and
Banfield (1985) suggest that another mechanism is involved which transforms two
layers of biotite into one layer of chlorite. There is a 35% loss in volume, and only
the tetrahedral layer is inherited intact by the chlorite, whereas the octahedral sheet
undergoes important elemental redistribution. These two different mechanisms can
occur under different conditions, and we may also explain hydroxyaluminum vermi
culite formation in temperate regions.
Banfield and Eggleton (1988) studied the weathering of biotite into vermiculite
and showed that the transformation mainly occurs by replacement of K by hydrated
cations. Sometimes however two biotite sheets gave rise to only one layer of ver
miculite. Further evolution towards kaolinite and goethite involves dissolution and
84
00
• Mg at 1/2
OLIVINE GOETHITE O Mg at zero
AT 7 HALF OCCUPIED Ä ^ ^ FULLY OCCUPIED SAPONITE

/ W / OCTAHEDRAL Ä ^ Y OCTAHEDRAL Si-TETRAHEDRAL
/ V COLUMNS ÊÊKm COLUMNS
PYROXENE NONTRONITE
f ^ ^ ^ W A A f f f f ^ f f f Si TETRAHEDRA
F· OCTAHEDRA
i/T\A/TTTTT
o o
M. Robert and D. Tessier

lo o
' 1 ^ . ▲ Jr^^l·.
o o
A. A A A A A A A
I
a.
S3
Fig. 4.7. Evolution of ring and chain silicates into secondary minerals (hydroxides or phyllosilicates). a. Relations between olivine and goethite (viewed
down Z or normal to the oxygen close packed plane), b. Relation between pyroxene and nontronite (viewed along the chain axis). (From Eggleton, 1975,
1984.)
Incipient weatherìng: some new concepts on weatherìng clay formation and organization 83
Fig. 4.8. Weathering of non phyllosilicate primary minerals, a. Etch pits on feldspar (SEM), b. High
resolution transmission electron micrograph of primitive clays on K feldspar showing 7 Â lattice on
sphere walls and curving layer structure having 12.5 Â domain structure (Tazaki and al., 1989). c.
Granite feldspar giving allophane (Romero et al., in press).
reprecipitation, but crystallographic orientation is maintained between clay, hydrox

ides and the octahedral chains of the biotite. During kaolinitization of biotite, two
types of kaolinite occur as 30 nm packs of layers interstratified in biotite, or as
two layer units (1.4 nm) irregularly interlayered in biotite (Ahn and Peacor, 1987).
Some TEM pictures suggest the reaction of one biotite layer to give two kaolinite
layers. Such a mechanism would imply complete biotite dissolution and kaolinite
MECHANISM ®
ONE
TWO BIOTITE
CHLORITE
MECHANISM
mimmnMMiniimmm
Mg-saturated and air-dried
T o o o o WWWW T
| v\zvvyv^^
4 A 14 Â
1 /\/\/^/\/\/\/\/\^^^ΑΉΠΠΧ. 1
O Me Kaolin
Verini c u i i t e Kaol i n
K-saturated and air-dried
T
T
14Â
10 A
1 1
• K
Fig. 4.9. a. Schematic diagram of the two mechanisms considered for the development of chlorite
from biotite, following Eggleton and Banfield (1985). In mechanism 7, a bruci te-like sheet replaces
the K-interlayer of biotite; in mechanism 2, the loss of tetrahedra from one biotite layer reduces it to
a brucite-like layer, so that two biotites become one chlorite (from Eggleton and Banfield, 1985). b.
Schematic drawing of an interlayer vermiculite-kaolin mineral (from Wada and Kakuto, 1983).
crystallization at linear boundaries.

A study by Romero et al. (in press) on clay formation from crystalline rocks in
NW Spain showed the different stages of mica weathering in subtropical conditions
(Figure 4.10). There isfirstan interlayer opening along 001 planes which, combined
with exfoliation, results in microdivision of the mica crystal. Such a microdivision
can progress until monolayer individualization, which represents the last metastable
unit. The next stage is gel formation and individualization of different generations
of halloysite; with time, kaolinite develops.
Under such conditions where hydrolysis is intense due to leaching, but where
the temperature is relatively low, it is therefore possible to demonstrate the limit
between solid state transformation and neogenesis. Both on crystalline and volcanic
glassy rocks, halloysites are the metastable minerals which are first formed, and
irregularly interstratified 2:1 and 1:1 phases are intermediate stages (Wada and
Kakuto, 1983; Delvaux et al., 1990). Such intergrade 2:1 1:1 minerals formed in
Ultisols or Alfisols are different from Al vermiculite found in brown acid soils
(Figure 4.9b).
The transformation of illite or mica to vermiculite or smectite in sediments dur
ing diagenesis, is relatively well documented: Amouric and Parron (1985) clearly
demonstrated that the formation of glauconite involves growth at the expense of a
non-crystalline mineral. In the case of the smectite-illite sequence, the two possible
mechanisms, namely direct transformation by closing of the layers or dissolution-
precipitation reaction, are sources of constant discussion.
The second type of reaction certainly occurs under hydrothermal conditions to
form a 2 M illite, and recent results of Andreoli (1989) showed that interstratified
illite-smectite (70%), which is characteristic of sediments, can be formed by trans
formation of smectite by Kfixationand through alternate wetting and drying cycles
(Figure 4.11a). This reaction can be reversed quite easily (Andreoli et al, in press)
during weathering. It occurs by exfoliation and layer separation giving mono- or
bilayers with a swelling behavior (Figures 4.10 and 4.11b).
Discussion
The nanoscale studies of primary mineral weathering in relation to clay formation

(Malia and Komarneni, 1989) complement the general statements on processes and
our views of microsystems in Henin et al. (1968). The parent rock mineral structure
is of great importance and many cases exist between solid state transformation and
neoformation in solution. Inheritance plays the greatest role in the case of phyllosil-
icates where the layers are stable even under the form of bilayers or monolayers and
can be the last stages of transformation. For other silicates inheritance is possible
in part, either for tetrahedral (ortho- and ring silicates) or for octahedral sheets or
oxyhydroxide formation (chain silicates). Feldspar weathering gives by exsolution a
non-crystalline compound which will evolve in different directions either by desilici-
^::5iki^^l?li
Fig. 4.10. Weathering of phyllosilicate primary minerals (mica), a. Exfoliation phenomenon, b. Micro-
division. c. Solid phase diffusion leading to monolayers and gel. d. Solid phase leading to halloysite.
(Romero et al., in press.)
fication (allophane-halloysite, then kaolinite) or without great variation of the Si/Al

ratio, giving smectitic clays.
The first stages of transformation, can be readily studied in the temperate zone
where hydrolysis is relatively mild; the end stages appear in subtropical regions
where hydrolysis is more active. In tropical areas where weathering is intense all
cases give directly kaolinite + hydroxides. However, even in the last named case,
phases with short range order occur and must be taken into account as transitional
stages.
Fundamental IIIite' Particles
20 A
30 A
40 A
50 A
Fig. 4.11. Smectite-illite evolutions, a. Evolution of smectite to I/S interstratified and illite (bentonites
of Silesia), b. Evolution of illite to swelling interstratified (in "underclays" sediments (from Andreoli,
1989). c. Diagramatic representation of fundamental particles (from Nadeau et al., 1984).
The nanoscale: characterization of clay texture and its variability
Small-angle scattering (SAS) gives information about clay organization (Ben

Rhaïem et al, 1987), but HRTEM is a more suitable technique because the lay
ers and their association in particles or crystals and in aggregates can be displayed
from samples prepared in different ways. This technique will bring new data on the
organization of both reference clays (from quarries) or natural soil clays, taking into
account both the clay structure (layer organization) and clay texture (particle size,
shape and arrangement).
Reference clays
Because of obvious differences in ion exchange, hydration, and shrink-swell, we

shall distinguish between smectites and other clay minerals (illites, kaolinites and
even vermiculites).
Smectites
Most of the studies carried out on clay hydration have been made using minerals
from clay deposits. Their main characteristic is a low-charge layer (lower than about
0.45 negative charge per half cell), and nearly 100% compensated by exchangeable
cations (Tessier and Pedro, 1987), that is no interlayer closed by potassium. Thus
the surface accessible to a molecule as ethylene glycol monoethyl is close to the the
oretical surface of the clay layers (800 m2 g" 1 ). X-ray diffraction patterns also show
that unit layer spacings contain hydrated cations (Mooney et al., 1952; Glaeser and
Mering, 1968). Layer spacings differ depending on the nature of the compensating
cations and the value of the water vapor tension in equilibrium with the clay mate
rial. In addition, basal spacings correspond to the thickness of one to three water
molecules.
Electron diffraction studies show that the stacking of the layers in the ab direction
is generally fully disordered if the layer charge is low and localized in octahedral po
sitions (Mering and Oberlin, 1971). Evidence of systems with intermediate ordering
has been made with both increasing total and tetrahedral layer charges.
In the last decade, another approach to clay mineralogy has been to study the
microstructure of clays in relation with their behavior. Many attempts have been
made to describe clay-water system fabric using scanning electron microscope
(SEM) (Gillot, 1969; Tessier and Berrier, 1979) and transmission electron micro
scope (TEM) (Shomer and Mingelgrin, 1978; Tfessier and Pedro, 1982). For SEM,
the microscope can be specially equipped to observe small wet samples at low tem
perature (-80°C) which can be rapidly frozen in freon cooled with liquid nitrogen
(Delage et al., 1982). For TEM studies, to avoid altering the structural organization
of the sample, thin sections can be prepared by successive replacements of the clay
solution by solvents andfinallyby epoxy resin (Spurr, 1969; Tbssier, 1984).
Figure 4.12 shows changes in structural organization for a Greek montmorillonite

during drying at three values of the swelling pressure: 0.032 bar, 1 bar and 10
bar. The scanning electron microscope does not make it possible to distinguish
layers and crystals forming the clay. Only a three dimensional network of particles
is visible. Results also show that, at low pressures, particle arrangement changes
without any modification of the interlayer distances. It is necessary to reach about 50
bars pressure to observe the change from 1.86 to 1.56 nm for a Ca-montmorillonite
(Ben Rhaïem et al., 1987).
0.032 bar
w = 3.69
1 + e = 9.8
1 bar
w = 1.14
1 + e = 4.0
10 bars
w = 0.82
1 + e = 3.2
Fig. 4.12. Volume and organization changes of a Greek montmorillonite at three pressures; SEM pic
tures of a freeze sample (Tessier, 1984). a. 0.032 bar; b. 1 bar; c. 10 bars, w = water content; e = void
ratio (Kv/K$)
Clay size variability in relation to chemical composition

Relationships between layer charge values and charge location were observed
by Tburet et al. (1990). Low charge smectites (<0.35) have the lowest tetrahedral
charge and the largest development of the layers in the ab direction. This is espe
cially true for montmorillonites which have no iron in octahedral positions. In this
case clay layers can be extended up to about 1 μπι (Figure 4.13). For a Wyoming
montmorillonite having 0.18 Fe3+ per half cell and a tetrahedral charge of 0.4, layer
lateral extension is smaller (<0.1 μτα) (Tfessier, 1984).
Clay layer extension of the Cheto montmorillonite or nontronite is still smaller
as shown in Figure 4.13 (Tburet et al., 1990). These authors concluded that layer
extension plays a considerable role in pore size distribution and the Theological
properties of smectites.
Textural changes
TEM and low-angle X-ray scattering (LA-XRS) are suitable methods for observ
ing clay textural changes. The passage from discrete layer spacings (from 12.6 to
18.6 Â) to "illimited" intracrystalline swelling is observed. In fact, LA-XRS shows
interlayer distances of 3.5 to about 10 nm which correspond to the formation of a
diffuse double layer (Norrish, 1954). At the same time, tactoids (stacks of 3 to 8
layers) can be observed for monovalent cations such as Li, Na or K.
For divalent cations (Mg and Ca) the diffuse double layer is not expanded even at
low salt concentrations (Quirk, 1968). The number of layers making up the network
observed with SEM is in this case thicker (Figure 4.12) since stackings of 50 layers
can be observed (Ben Rhaïem et al., 1987; Figure 4.13).
Increasing the salt concentration of the interstitial solution can have the same ef
fect as the change from monovalent to divalent cations. The threshold concentration
for the collapse of the diffuse layer is about 0.3 M NaCl (Tfessier, 1984).
Similar changes in texture and layer spacing can be observed depending on the
intensity and number of wetting-drying cycles. Particles are thicker and layer stack
ings become ordered and even closed (Gaulthier and Mamy, 1979; Andreoli, 1989).
Finally, it is not possible to define a single particle characteristic for smectites.
They have particles of variable size depending on environmental conditions. The
evolution of clay smectite organization actually reflects what may occur at three
levels: arrangement (orientation-disorientation), particle size (following the a, b
and c planes) and internal particle structure (variable in interlayer spacing and order
in ab direction). These changes can explain clay smectite formation, weathering in
soils and smectite transformation to more illitic clay systems in sediments.
Kaolinites, halloysites, illites and vermiculites
For these clays, crystallites are the smallest particles that can be isolated. In
kaolinite minerals, only one organizational level corresponding to crystallites has
Incipient weathering: some new concepts on weatherìng clay formation and organization 91
Fig. 4.13. Variability of smectite organization, a. Na Greek montmorillonite. b. Na nontronite. c. Ca

Greek montmorillonite.
been demonstrated (Figure 4.14). TEM examination of glauconite or illite shows

that the particles that can be extracted are very often crystallite aggregates, domains,
generally 0.1 to 1 μτα in size. However EGME on BET surface areas reach about
100 to 150 m2 g"\ which means that crystallites are made of few layers (5 to 7
maximum). The shape of halloysite particles is different with spheres and cylinder
like objects (Figure 4.14).
These characteristics induce particular swelling-shrinking properties which show
little variation with the environmental parameters.
Soil clays
Only a few recent results are available on soil clay organization. It has been seen
that the main processes of soil formation lead to primary mineral microdivision.
So the coarse clay fraction (>0.2 μπι) is mainly composed of inherited primary
minerals (tecto- or phyllosilicates) and most typical soil clays have a very small size
(<C0.2 /im). When compared to reference clays, they display crystallites which are
both thinner with only a few layers (along the c axis) and shorter (small extension
along a and b directions).
However, there is a large size heterogeneity especially for the number of layers
which can range from monolayer to about 10 layers. In the case of 2:1 clay minerals
in temperate regions, we can imagine a continuous solid solution from one illitic
end member with closed 1.0 nm layers to a smectitic end member with all the
intermediate stages present.
All the clays of this sequence can be placed in increasing order of CEC and
total surface area values (Figure 4.15), which represent good criteria for the study
of soil clay mixtures. Several illustrations of the main clay facies are presented in
Figure 4.16 (Robert et al., in press). Ttoo kinds of illite must be distinguished: (a)
one is a micromica with 10% K2O and more than 10 layers (10 Â) in each crys
tallite or particle, and (b) the second has only 5 layers and 7.5% K2O content on
average. Such an illite which seems more specific to soils, does not correspond to
the end member of Srodon and Eberl (1984) nor to Jackson's definition, but it
is close to the "Le Puy 111116" model. Pure smectites are very rare because closed
10 Â layers are often present. They have a beidellite composition (Dixon, 1982;
Wilson, 1987; Robert et al, in press) and quasi-crystals are formed by the super
position of small and thin particles. Such smectite morphology can be compared to
the Cheto-type montmorillonite or to nontronite. The most abundant soil clay in
temperate regions is formed by interstratified minerals. Interstratification can occur
inside the same type of particles (interstratification sensu stricto or structural), but it
mainly results from the superposition of very thin particles with one to five layers.
We propose to call such interstratification a "textural interstratification,, (Robert,
1989).
100 nm
Fig. 4.14. Texture of different reference clay minerals, a. Kaolinite. b. Halloysite. c. Halloysite. d.
Venniculite (<2 μπι Santa Oliala).
The small particles that we find in soils are not strictly analogous to the fun
damental particles of Nadeau et al.(1984), but interparticle swelling is indeed an
important phenomenon to be considered (Figure 4.11c).
Vermiculites have to be studied separately from this illite-smectite sequence
which is mainly valid for soils developed on sedimentary rocks. Particles and layers
can be thin, but have a larger lateral extension (Malia et al, 1989), and keep a
certain rigidity which is inherited from the structural order of mica. Mica layers can
always be present in the core of large crystals (Vali and Koster, 1986) and the three
dimensional order seems to be a remenant memory even under extreme conditions.
For all the 2:1 clay minerals found in temperate regions, transformation and
microdivision phenomena are dominant; the stable species even for vermiculite,
have a dioctahedral composition and a considerable layer charge heterogeneity as
shown using alkylammonium treatment (Malia et al., in press).
For subtropical or tropical soils, textural evolution with a decrease in particle size
can also be an important factor even for kaolinite. In Guyana, crystallites have a
10 μτα size in the saprolite and less than 0.01 μτη in the B horizon (Robain et al.,
1990).
103m2/Kg
| T I i 1
14*
*5
SMECTITES
• 9
10*
• 13
1
L
4*
6* INTERSTRATIFIED
15* MINERALS I
3*
8*
_ 2* J
ILLITES
7 *X
12 v
l
KAOLINITE S '
I 1 1
0 .2 .4 .6 .8 1.0
eq/Kg
CEC Ba
Fig. 4.15. Relation between total surface area (EGME) and cation exchange capacity for sedimentary
soils clays.
Incipient weathering: some new concepts on weathenng clay formation and organization 95
40 nm
Fig. 4.16. Examples of soil clay organization, a. Micro mica showing some layer opening, b. Illite
particles, c. Interstratified clay (textural interstratification), d. Soil clay smectite. (Robert et al., 1990.)
Another feature revealed by TEM studies is the widespread occurrence of hal-

loysites, generally found in soils developed on volcanic material, but also in young
soils developed on crystalline rocks.
Clay associations
All the methods used to characterize soil clay fractions begin with different treat
ments in order to disperse and purify clays (Mehra and Jackson, 1960). These
methods, which do not greatly change the layer unit for XRD studies, introduce
many modifications in the layer organization.
We have seen that TEM studies make it possible to characterize different orga
nization levels in clays: layers and particles which represent the elementary level,
domains which are quite usual for illites.
If no drastic treatment has been introduced, clay particles are commonly asso
ciated in small nano- or micro aggregates (Figure 4.17). Electrostatic forces (for
domains), iron, aluminum or organic matter are the most usual factors involved.
In synthetic associations prepared in the laboratory, it is possible to localize a
coating on the clay surface which represents the cement between particles (Figure
4.17); the most active coatings are polycations or low-range crystalline compounds
(Saleh and Jones, 1984; Robert et al., 1987). In natural conditions, crystalline iron
oxides or hydroxides are generally of very small size (3 to 5 nm large). Different
examples are presented in Figure 4.17 in soils developed in temperate or tropical
climates. Most commonly, clay-oxide associations are very diffuse.
In Oxisols developed from schists in French Guyana, weathering of the parental
material allows the formation of large kaolinite crystals with goethite sandwiched
in between (Tandy, et al., 1990). Gilke and Suddhiprakan (1979) suggest that such
an orientation of goethite parallel to biotite crystals is evidence that there are no
epitaxic relationships. However, for some authors such epitaxial development of
goethite on kaolinite can occur as demonstrated by Boudeille and Müller (1988).
There are only few studies of what is called the clay-humic complex at the
nanoscale level. It is illustrated in Figure 4.17 for a Vertisol and shows very dif
fuse relationships between a "cloud" of small humic particles and smectite particles.
Organic matter is mainly located between small particles and not between layers.
Clay behavior
In order to establish close relationships between the type of clay and chemical
and physical properties, both surface characteristics (CEC, total surface area) and
a better knowledge of clay organization are necessary; organization of clays at a
different scale (microaggregates) is also very important as is the presence of Al and
Fe compounds which can change both the physical and chemical properties (Robert
and Tferce, 1989).
Fig. 4.17. Clay associations, a. Microaggregate (fersiallitic soil from Portugal) with an iron cement b.
Fe coating on kaolinite basal surface (Robert et al., 1987). c. Goethite crystals sandwiched between
kaolinite particles French Guyana (Tandy et al., 1990). d. Illustration of the clay-humic complex from a
Vertisol (Guadeloupe West Indies).
Ue
□ Air
I l Solution
P(bor) 0.010
Na-Wyoming montmorillonite IO M NaCI
He -iT
15-1-
\ e*tf
P(bar) 0.010
Ca-Wyoming montmorillonite I 0 M CaC^
.2T
s#i7
P(bar) 0,010 1.0
Le Puy illite
we -tr
5
1 4-
o
P(bar) 0.010 1.0 io
St Austell Kaolinite
Fig. 4.18. Evolution of water and void ratios for different clay minerals.
As an illustration of clay behavior, Figure 4.18 shows the parallel evolution of the
water ratio and void ratio for four clayey materials: Ca- and Na-smectite, illite and
kaolinite (Tessier, 1984). Na-smectites prepared with a diluted solution are the most
hydrated clays. At low pressures the water content decreases in the following order:
Na and K smectites > Ca-Mg-smectites > illites > kaolinites

Incipient weatherìng: some new concepts on weatherìng clay formation and organization 99
At high pressures the void volume of the samples is smaller as the clay is initially
more hydrated. On the other hand, the clay matrix is saturated with water at low
pressures. At higher pressures the stress with which air penetrates into the samples
is lower as the clay is initially less hydrated. Furthermore, the minimum values of the
porosity (expressed in void ratio) are significantly different ranging from e = 1.1 for
a coarse kaolinite to about 0.35 for a smectite. Both air entry point and shrinkage
aptitude, and CEC were found by Tessier et al. (1990) to be strongly correlated with
its mineralogical nature.
Conclusion
Among the new methodologies developed in the last few years electron mi
croscopy has permitted the most important progress through a better knowledge
of incipient weathering processes and mechanisms either at a microscale or at a
nanoscale.
Inheritance of secondary minerals from primary minerals has been better defined
and new mechanisms such as epitaxy, epigenesis can be studied. A specific sequence
from illite to smectite has been isolated in which the very important textural changes
(microdivision) lead to very small particles or even monolayers. Such data have to be
explored not only for incipient weathering, but also for other soil forming processes
such as clay illuviation, planosolisation or even podzolization.
It is well known that short range ordered compounds (allophanes-imogolites-
ferrihydrites) are very important for soils developed from volcanic rocks with a
glassy structure. The first stages of crystalline rock weathering show that such com
pounds are also abundantly formed, even inside primary minerals such as feldspars.
So thermodynamic data and theories of clay formation which take into account only
soluble species and crystalline compounds, must be reviewed.
These studies applied to clay minerals also throw new light on clay organization
in relation to water content, and its changes with the physical or chemical environ
ment.
New concepts have also been developed. The ideas of Nadeau et al. (1984) de
fined fundamental particles in sediments. Our studies show that soil clay particles
are very small, and that interparticle space is a more important consideration than
interlayer space either for swelling, localization of potassium, organic matter and
other properties. At that scale, also relationships with chemical and physical proper
ties are easy to establish.
Future nanoscale studies with HRTEM will investigate weathering in different
climatic environments, and at the interface between organisms and minerals. But it
is certain that other methods will be also applied in the future to soil studies. One
can cite spectroscopic methods, especially EXAFS, which have already been used
with success to study the genesis and evolution of iron compounds (Combes et al.,
1986) or liquid and solid NMR (Si, Al), which makes it possible to follow aluminous
or Si-Al compounds through a sequence from cation -> polycation —> crystalline
compounds (Bottero et al., 1980; Mackenzie et al., 1989).
References
Abreu, M. and Vairinho, M., 1990. Amphibole alteration to vermiculite in a weathering profile of
gabbro-diorite in soil micromorphology. In: L.A. Douglas (Editor), Soil Micromorphology, A. Basic
and Applied Science. Developments in Soil Science, 19. Elsevier, Amsterdam, pp. 493-500.
Ahn, J.H. and Peacor, D.R., 1987. Kaolinitization of biotite: TEM data and implications for an alteration
mechanism. Am. Miner., 72: 353-356.
Amouric, M. and Parron, C , 1985. Structure and growth mechanism of glauconite as seen by high
resolution transmission electron microscopy. Clays Clay Miner., 33: 473-482.
Andreoli, C, 1989. Evolution des phyllosilicates 2:1 en fonction de la dynamique du potassium. Thèse
Univ. Paris VI, 233 pp.
Andreoli C , Robert M. and Elsass F, in press. Origin and evolution of interstratified 2:1 clay minerals.
An HRTEM study of sediments under coal. Clays Clay Miner.
Banfield, J.F. and Eggleton, R.A., 1988. Transmission electron microscope study of biotite weathering.
Clays Clay Miner., 36: 47-60.
Barshad, I., 1948. Vermiculite and its relation to biotite. Am. Miner., 33: 655-678.
Ben Rhaïem, H., Pons, C.H. and Tessier, D., 1987. Factors affecting the microstructure of smectites.
Role of cation and history of applied stresses. In: Schultz, van Olphen and Mumpton (Editors),
Proc. Int. Clay Conf. Denver, Colo. The Clay Minerals Society, pp. 292-297.
Berner, R.A. and Holdren, G.R., 1979. Mechanisms of feldspar weathering, II. Observations of feldspars
from soils. Geochim. Cosmochim. Acta, 43: 1173-1186.
Beutelspacher, H. and van der Marel, H.W., 1968. Atlas of Electron Microscopy. Elsevier, Amsterdam,
333 pp.
Bottero, J.Y., Casas, J.M., Fiessinger, F. and Poirier, J.E., 1980. Studies of hydrolyzed chloride solutions.
Nature of aluminum species and composition of aqueous solutions. J. Phys. Chem., 84: 2933-2939.
Boudeille, M. and Muller, J.P., 1988. Structural characteristics of hematite and goethite and their
relationships with kaolinite in a laterite from Cameroon. A TEM study. Bull. Mineral., I l l : 149-
166.
Brinkman, R., 1970. Ferrolysis a hydromorphic-soil forming process. Geoderma, 3:199-206.
Brown, J.L. and Jackson, M.L., 1973. Chlorite examination by ultramicrotomy and high resolution
electron microscopy. Clays Clay Miner., 21: 1-7.
Brown, J.L. and Rich, C.I., 1968. High-resolution electron microscopy of muscovite. Science, 161: 1135-
1137.
Bruckert, S., 1970. Influence des composés organiques solubles sur la pédogenèse en milieu acide. Ann.
Agron., 21: 421-452.
Buseck, PR. and Iijima, S., 1974. High resolution electron microscopy of silicates. Am. Miner., 59:1-21.
Callot, G., Guyon, A., Mousain, D. and Plassard, C , 1984. Rôle des hyphes mycéliens dans la formation
des sols et des roches. Agronomie, 5: 143-150.
Combes, J.M., Manceau, A. and Calas, G., 1986. Study of the local structure in poorly-ordered precur
sors of iron oxyhydroxides. J. Phys., Colloq. C8, 47: 697-701.
Dambrine, E., Robert, M. and Ranger, J., 1989. La podzolisation à l'étage subalpin: un processus
typiquement hivernal. C.R. Acad. Sci., Paris, II, 308: 1797-1802.
Dejou, J., Guyot, J. and Robert, M., 1977. Evolution superficielle des roches cristallines et cristallophyl-
liennes dans les régions tempérées. INRA, Paris, 464 pp.
Delage, P., Tessier, D. and Mareil Audiguier, M., 1982. Use of cryoscan 20 apparatus for observation of
freeze fractured planes of a sensitive Québec clay in SEM. Can Geotech. J., 19: 111-114.
Delvaux, B., Herbillon, A.J., Viehaye, L. and Mestdagh, M.M., 1990. Surface properties and clay min
eralogy of hydrated halloysitic soil clays. II. Evidence for the presence of halloysite smectite mixed
layers. Clays Clay Miner., 141-160.
Dixon, J.B., 1982. Mineralogy of vertisols. In: Vertisols and Rice Soils of the Tropics. Symposia Papers
II, 12th ICSS, New Delhi, pp. 48-59.
Duplay, J., 1988. Géochimie des argiles et géothermométrie des populations monominérales de parti
cules. Thèse Doctorat Univ. Strasbourg, 222 pp.
Eberhart, J.P., 1976. Méthodes physiques d'étude des minéraux et des matériaux solides. Doin, Paris,
507 pp.
Eberhart, J.P. and Triki, R., 1972. Description d'une technique permettant d'obtenir des coupes minces
de minéraux argileux par ultramicrotomie. J. Microsc, 15: 111-120.
Eggleton, R.A., 1975. Nontronite topotaxial after hedenbergite. Am. Miner., 60: 1063-1068.
Eggleton, R.A., 1984. Formation of iddingsite rims on olivine. A transmission electron microscope study.
Eggleton, R.A., 1986. The relation between crystal structure and silicate weathering rates. In: S.H.
Eggleton, R.A., 1987. Non crystalline Fe-Si-Al-oxyhydroxides. Clays Clay Miner., 35: 29-37.
Eggleton, R.A. and Banfield, J.F., 1985. The alteration of granitic biotite to chlorite. Am. Miner., 70:
902-910.
Eggleton, R.A. and Boland, J.N., 1982. The weathering of enstatite to talc through a series of transitional
phases. Clays Clay Miner., 30: 11-20.
Eggleton, R.A. and Buseck, P.R., 1980. High resolution electron microscopy of feldspar weathering.
Farmer, V.C., Russell, J.D., McHardy, W.J., Newman, A.C.D., Ahlricks, J.L. and Rimsaite, J.Y.H., 1971.
Evidence for loss of protons and octahedral iron from oxidized biotites and vermiculites. Mineral.
Mag., 38: 121-137.
Farmer, V.C., 1979. Possible roles of a mobile hydroxyaluminum orthosilicate complex (proto-imogolite
and other hydroxyaluminum and hydroxy-iron species) in podzolization. Coll. Int., CNRS No. 303,
pp. 275-279.
Gaultier, J.P. and Mamy, J., 1979. Evolution of exchange properties and crystallographic characteristics
of bionic K-Ca montmorillonite submitted to alternate wetting and drying. Trans. Int. Clay Conf.,
Oxford, pp. 167-175.
Gilkes, R. and Suddhiprakan, A., 1979. Biotite in deeply weathered granite I. Morphologic, mineralog-
ical and chemical properties. II. The oriented growth of secondary minerals. Clays Clay Miner., 27:
349-367.
Gillott, J.E., 1969. Study of the fabric of fine grained sediments with the scanning electron microscope.
J. Sediment. Petrol., 39: 90-105.
Gjems, O., 1963. A swelling dioctahedral clay mineral of vermiculite smectite type in the weathering
horizons of podzols. Clay Miner. Bull. 5: 183-193.
Glaeser, R. and Mering, J., 1968. Domaines d'hydratation homogène des smectites. C.R. Acad. Sci.,
Paris, Sér. D, 267: 463-466.
Henin, S., Pedro, G., Robert, M., 1968. Considérations sur les notions de stabilité et d'instabilité des
minéraux en fonction des conditions du milieu; essai de classification des systèmes d'agression. Trans.
9th Int. Cong, of Soil Sci., Adelaide, III, pp. 79-90.
Hinsinger, Ph., Dufery, J.E. and Jaillard, B., 1988. Biological weathering 21 of micas in the rhizosphere
as related to potassium absorption by plant roots. C.R. Int. Symp. Root Res. Soc, Uppsala.
Huang, P.M. and Violente, A., 1986. Influence of organic acids on crystallization and surface properties
of precipitated products of aluminum. In: Interactions of soil minerals with natural organics and
microbes. SSSA Spec. Pubi., pp. 159-221.
Huang, W.H. and Keller, W.D., 1972. Organic acids as agents of chemical weathering of silicate minerals.
Nature Phys. Sci., 293: 149-151.
Ildefonse, Ph., Proust, D., Meunier, A. and Velde, B., 1979. Rôle de la structure dans l'altération
des roches cristallines au sein des microsystèmes. Science du Sol, Séminaire Altération des Roches
Cristallines, pp. 239-257.
Jackson, M.L., 1963. Aluminum bonding in soils: A unifying principle in soil science. Soil Sci. Soc. Am.
Proc., 27:1-10.
Jackson, M.L. and Shridar, K., 1974. Scanning electron microscopic and X-ray diffraction study of natu
ral weathering of phlogopite through vermiculite to saponite. Soil Sci. Soc. Am. Proc, 38: 843-
847.
Jaillard, B., 1983. Mise en évidence de la calcitisation des cellules corticales de racines de graminées en
milieu carbonate. C.R. Acad. Sci., 297: 293-295.
Kounestron, O., Robert, M. and Berner, J., 1977. Nouvel aspect de la formation des smectites dans les
vertisols. CR. Acad. Sci., 284: 733-735.
Leyval, C. and Berthelin, J., 1986. Comparison between the utilization of phosphorus from insoluble
phosphates by ectomycorrhizal fungi and rhizobacterisa in physiological and genetical aspects of
mycorrhiza. In: V. Gianinazzi-Pearson and S. Gianinazzi (Editors), INRA, Paris, pp. 340-345.
Lee, S.Y., Jackson, M.L. and Brown, J.L., 1975. Micaceous occlusions in kaolinite observed by ultrami-
crotomy and high resolution electron microscopy. Clays Clay Miner., 23: 125-129.
Mackenzie, K.J.D., Bowden, M.E., Brown, I.W.M. and Meinhold, R.H., 1989. Structure and thermal
transformation of imogolite studied by 29Si and 27A1 high resolution solid state nuclear magnetic
resonance. Clays Clay Miner., 37: 317-324.
Mackumbi, L., and Herbillon, A.J., 1972. Vermiculitisation expérimentale d'une chlorite. Bull. Groupe
Fr. Argiles, 25: 37-41.
Malia, PB. and Komarneni, S., 1989. High resolution transmission electron microscopy (HRTEM) in
the study of clays and soils. Adv. Soil Sci. 12: 159-186.
Malia, PB., Douglas, L.A. and Robert, M., 1989. Weathering and layer change dynamics in 2:1 layer
silicate clays. In: Weathering, I, Processes. Theophrastus Publications, pp. 231-256.
Malia, PB., Robert, M., Douglas, L.A. and Tessier, D., in press. Charge heterogeneity and microstruc
ture of 2:1 layer silicates: A HRTEM study. Am. Miner.
Mann, H., Tazaki, K., Fyfe, WS., Beveridge, T.J. and Humphrey R., 1987. Cellular lepidocrocite precip
itation and heavy metal sorption in Euglena Sp. (Unicellular Alga): implications for biomineraliza-
tion. Chem. Geol., 63: 39-43.
Mauguin, C, 1928. Étude des micas au moyen des rayons X. Bull. Soc. Fr. Minéral. Cristallogr., 51: 285.
Mehra, O.P. and Jackson, M.L., 1960. Iron oxide removal from soils and clays by a dithionite-citrate
system buffered with sodium bicarbonate. Clays Clay Miner., 7: 317-327.
Mering, J. and Oberlin, T, 1971. The smectites. In: The Election-Optical Investigation of Clays, Min
eral. Soc., London, pp. 135-145.
Meunier, A. and Velde, B., 1979. Weathering mineral facies in altered granites: the importance of local
small scale equilibria. Mineral. Mag., 43: 261-268.
Mooney, R.W, Keenan, A.C. and Wood, L.A., 1952. Adsorption of water vapor by montmorillonite.
II. Effect of exchangeable ions and lattice swelling as measured by X-ray diffraction. J. Am. Chem.
Soc., 74: 1371-1374.
Mortland, M.M., Lawton, K. and Uehara, G., 1956. Alteration of biotite to vermiculite by plant growth.
Soil Sci., 82: 477-481.
Nadeau, PH., Wilson, M.J., McHardy, W.J. and Tait, J.M., 1984. Interstratified clays as fundamental
particles. Science, 225: 923-925.

Nahon, D., Colin, E and Tardy, Y., 1982. Formation and distribution of Mg, Fé, Mn smectites in the
first stage of laterite weathering of forsterite and tephroite. Clay Miner., 17: 339-348.
Norrish, K., 1954. The swelling of montmorillonite. Disc. Faraday Soc, 18: 120-134.
Olives Banos, J., 1985. Biotites and chlorites as interlayered biotite-chlorite crystals. Bull. Mineral.,
108: 635-641.
Olives Banos, J., Amouric, M., Fouquet, C D . and Baronnet, A., 1983. Interlayering and interlayer slip
in biotite as seen by HRTEM. Am. Miner., 68: 754-758.
Pauling, L. 1930. The structure of micas and related minerals. Proc. Nati. Acad. Sci., 16: 123-129.
Pedro, G. and Sieffermann, G., 1979. Weathering of rocks and formation of soils. In: FR. Siegel (Editor),
Review in Modern Problems of Geochemistry. UNESCO, pp. 39-55.
Quirk, J.P., 1968. Particle interaction and soil swelling. Isr. J. Chem., 6: 213-234.
Ranger, J., Dambrine, E., Robert, M., Righi, D. and Felix, C, 1990. Current forming processes charac
terized by an in situ experimental method using a mineral (vermiculite) and resin bags. Geoderma.
Ranger, J. and Robert, M., 1985. Étude des conditions physicochimiques des milieux forestiers par une
approche expérimentale in situ utilisant l'altération des minéraux primaires. Agronomie, 2: 157-168.
Razzaghe-Karimi, M.H. and Robert, M., 1975. Altération des micas et géochimie de l'aluminum: rôle
de la configuration de la molécule organique sur l'aptitude à la complexation. C.R. Acad. Sci., Sér.
D, 280: 2645-2648.
Rich, CL, 1968. Hydroxy interlayers in expansible layer silicates. Clays Clay Miner., 16: 15-30.
Righi, D., Ranger, J. and Robert, M., 1988. Clay minerals as indicators of some soil forming processes
in the temperate zone. Bull. Minéral., I l l : 625-632.
Righi, D., Bravard, S., Chauvel, A., Ranger, J. and Robert, M., 1989. In situ study of podzolization in
an oxisol-spodosol sequence of Amazonia (Brazil). Soil Sci.
Robain, H., Tessier, D., Grimaldi, M. and Elsass, F, 1990. Importance de la texture des kaolinites
dans la caractérisation des couvertures ferrallitiques. Conséquences sur l'organisation et la stabilité
physique des sols. C.R. Acad. Sci., Paris, II, 311: 239-246.
Robert, M., 1971. Les aspects du processus de vermiculitisation. Ann. Agron., 22: 43-93.
Robert, M., 1989. In: Les minéraux argileux. I. Structure, propriétés et applications. Les interstratifiés
II, 6, 288-304. Soc. Fr. Minéral. Cristallogr., Paris, 586 pp.
Robert, M. and Barshad, L, 1972. Sur les propriétés et la détermination des minéraux argileux 2:1
expansibles (vermiculites smectites). C.R. Acad. Sci., 275: 1463-1465.
Robert, M. and Berthelin, J., 1986. Role of biological and biochemical factors in soil mineral weathering.
In: Interactions of soil minerals with natural and microbes. SSSA, Spec. Pubi., 17, pp. 453-495.
Robert, M, Hardy, M. and Elsass, F, in press. Crystallochemistry, properties and organization of soil
clays derived from major sedimentary rocks in France. Clay Miner.
Robert, M., Hardy, M., Elsass, F and Righi, D., 1990. Genesis, crystallochemistry and organization of
soil clays derived from different parent materials in temperate regions (France). 14th Int. Conf. of
Soil Science, Kyoto, 1990, VII: 42-47.
Robert, M., Magonthier, M.C and Razzaghe-Karimi, M.H., in press. Experimental weathering of glassy
rocks in acid medium. Application to natural conditions and high level waste storage. Applied
Geochem.
Robert, M., Razzaghe-Karimi, M.H. and Ranger, J., 1987. Rôle du facteur biochimique dans la pod-
zolisation. In: D. Righi and A. Chauvel (Editors), Podzols et Podzolisation, AFES-INRA, pp. 207-
223.
Robert, M. and Tercé, M., 1989. Effect of gels and coating on clay mineral chemical properties. In:
Inorganic Contaminants in the Vadose Zone. Ecological Studies, 74. Springer, Berlin, 17: 57-71.
Robert, M. and Veneau, G., 1974. An experimental evaluation of the effects of pH and concentration of
salt solutions on the alteration of leucite at low temperature. Geoderma, 11: 209-219.
104 M Robert and D. Tessier
Robert, M. and Veneau, G., 1978. Stabilité des minéraux phylliteux 2:1 en conditions acides. Rôle de
la composition octahédrique. Proc. 6th Int. Clay Conf., Oxford, pp. 385-394.
Robert, M., Veneau, G. and Hervio, M., 1983. Influence des polycations du fer et de l'aluminum sur les
propriétés des argiles. Sci. Sol, 34: 235-251.
Robert, M., Veneau, G. and Abreu, M., 1987. Études microscopiques d'associations aluminum-argiles
ou fer-argiles. Proc. of the VII Int. Working Meeting on Soil Micromorphology, Paris, 1985, pp.
467-474.
Romero, R., Robert, M., Elsass, F. and Garcia, C, in press. HRTEM studies of weathering microsystems
developped on different crystalline rocks (N.W.Spain). Clay Miner.
Saleh, A.M. and Jones, B.F., 1984. The crystallinity and surface characteristics of synthetic ferrihydrite
and its relationship to kaolinite surfaces. Clay Miner., 19: 745-755.
Shomer, I. and Mingelgrin, U., 1978. A direct procedure for determining the number of plates in
tactoïdes of smectites: the Na/Ca-montmorillonite case. Clays Clay Miner., 26:135-138.
Simon-Sylvestre, G., Robert, M., Veneau, G. and Beaumont, A., in press. Experimental nitrification
related to acidification and silicate weathering. In: Proc. 8th Int. Symp. on Environmental Biochem
istry, Nancy, 1987. Marcel Dekker, New York, N.Y.
Spurr, A.R., 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultra-
struct. Res., 26: 31-43.
Spyridakis, D.E., Chesters, G. and Wilde, S.A., 1967. Kaolinization of biotite as a result of coniferous
and deciduous seedling growth. Soil Sci. Soc. Am. Proc, 31: 203-210.
Srodon, J., Andreoli, C, Elsass, F. and Robert, M., 1990. Direct high resolution transmission electron
microscopic measurement of expandability of mixed-layer illite/smectite in bentonite rock. Clays
Clay Miner., 38: 373-379.
Srodon, J. and Eberl, D.D., 1984. Illite. In: S.W. Bailey (Editor), Micas. Mineralogical Society of Amer
ica, Washington, D.C., 24: 495-544.
Stevenson, F.J., 1967. Organic acids in soil. In: A.D. McLaren and G.H. Peterson (Editors), Soil Bio
chemistry, Vol. 1. Marcel Dekker, New York, N.Y., pp. 119-146.
Tandy, J.C., Grimaldi, M., Grimaldi, C. and Tessier, D., 1990. Mineralogical and textural changes in
French Guyana oxisoils and their relation with microagregation. In: L. A. Douglas, (Editor), Soil
Micromorphology, A Proc. Int. Micromorpholohy meeting, San Antonio, Texas, Elsevier, pp. 191-
198.
Tazaki, K. and Fyfe, WS., 1987. Primitive clay precursors formed on feldspar. Can J. Earth Sc, 24:
506-527.
Tazaki, K., Fyfe, WS. and van der Gaast, S.J., 1989. Growth of clay minerals in natural and synthetic
glasses. Clays Clay Miner., 36: 47-60.
Tchoubar, C, Rautureau, M., Clinard, C. and Ragot, J.P., 1973. Technique d'inclusion appliquée à
l'étude des silicates lamellaires et fibreux. J. Microsc., 18:147-151.
Tessier, D., 1978. Étude de l'organisation des argiles cal ci que s. Évolution au cours de la dessiccation.
Ann. Agron., 29: 319-355.
Tessier, D., 1984. Étude expérimentale de l'organisation des matériaux argiles. Hydratation, gonflement
et structuration au cours de la dessiccation et de la réhumectation. Thèse Univ. Paris VII, INRA
Versailles Pub., 360 pp.
Tessier, D., and Berrier, J., 1979. Utilisation de la microscopie électronique à balayage dans l'étude des
sols. Observation de sols humides soumis à différents pF. Sci. Sol, 1: 67-82.
Tessier, D., Beaumont, A. and Pedro, G., 1990. Influence of clay mineralogy and rewetting rate on clay
microstructure. In: L.A Douglas (Editor), Soil Micromorphology, A. Basic and Applied Science.
Developments in Soil Science, 19. Elsevier, Amsterdam, pp. 115-121.
Tessier, D., Bruand, A. and Beaumont, A., in press. Relationship between clay mineralogy and soil
behaviour in Paris Basin clayey soils. Proc. Int. Clay Conf. Strasbourg, 1989.
Tessier, D. and Pedro, G., 1982. Electron microscopy of Na smectite. Role of layer charge, salt concen
tration and suction parameters developments in sedimentology. Int. Clay Conf., Elsevier, Amster
dam, pp. 165-176.
Tessier, D. and Pedro, G., 1987. Mineralogical characterization of 2:1 clays in soils. Importance of the
clay texture. Proc. Int. Clay Conf., Denver, pp. 78-84.
Touret, O., Pons, C.H., Tessier, D. and Tardy, Y., 1990. Etude de la répartition de l'eau dans les argiles
saturées Mg 2 + aux fortes teneurs en eau. Clay Miner., pp. 217-233.
Tributh, H., Boguslawski, E., Lieres, A., Steffens, D. and Mengeilk, G., 1987. Effect of potassium
removal by crops on transformation of illitic clay minerals. Soil Sci., 143: 404-409.
Vali, H. and Koster, H.M., 1986. Expanding behavior, structural disorder, regular and random irregular
interstratification of 2:1 layer-silicates studied by high-resolution images of transmission electron
microscopy. Clay Miner., 21: 827-859.
van Oort, F. and Robert, M., 1988. Rôle de l'oxydation du soufre dans les premiers stades de l'altération
superficielle des calcschistes (Beaufortain, Alpes du Nord). Conséquences pédologiques sur l'évolu
tion pédologique. C.R. Acad. Sci., Paris, II, pp. 819-825.
Veblen, D.R. and Buseck, PR., 1980. Microstructure and reaction mechanisms in biopyriboles. Am.
Miner., 65: 599-623.
Wada, K. and Kakuto, Y., 1983. Intergradient vermiculite-kaolin mineral in a Korean ultisol. Clays Clay
Miner., 31: 183-190.
Westbroek, P., 1983. Biological metal accumulation and biomineralization in a geological perspective.
In: P. Westbroek and E.W. De Jong (Editors), Biomineralization and Biological Metal Accumulation.
Reidel, Dordrecht, pp. 1-11.
Wilson, M.J. and Russell, J.D., 1980. Glushinskite, a naturally occurring magnesium oxalate. Miner.
Mag., 43: 837-840.
Wilson, M.J., 1987. Soil smectites and related interstratified 2:1 minerals: recent developments. Proc.
Int. Clay Conf., Denver, 1985, pp. 167-173.
107
Chapter 5
Alteration products at the earth's surface — the

clay minerals
LJ. EVANS
Introduction
The minerals that comprise most soils and sediments were formed under condi
tions vastly different from those found at the earth's surface today. These primary
minerals in rocks originated at high temperatures and pressures and are generally
unstable in terrestrial environments. They are altered in the presence of water,
carbonic acid, organic acids and oxygen to yield more stable constituents — the
secondary minerals. These secondary minerals are the major products of weathering
of minerals in soils; of diagenetic alteration of transported material in sediments;
and of hydrothermal alteration of rocks in terrestrial and oceanic environments.
The weathering and diagenetic alteration of minerals in soils and sediments is an
important part of the geological cycling of elements.
Pre-eminent amongst the products of mineral alteration at the earth's surface are
the phyllosilicate minerals, although various oxides, hydroxides and oxyhydroxides
may assume importance in some weathering environments. These secondary prod
ucts vary in their crystallinity and are found predominantly in the clay-sized fraction
(<2μπι) of soils and sediments. Although the term clay mineral is often restricted
to the phyllosilicate clays, in this chapter the term will be used to describe the major
secondary minerals found in clay fractions and, as defined, the clay minerals are
comprised of three structurally distinct groups of minerals: the phyllosilicate clays,
the poorly ordered aluminosihcates, and various oxides and oxyhydroxides of Fe, Al
and Mn (Table 5.1).
A number of mechanisms have been recognized to account for the presence of
clay minerals in soils, sediments and sedimentary rocks. Firstly, they may have been
inherited, that is, they have a detrital (allogenic) origin; or, secondly, they may have
been formed in situ, that is, they have an authigenic origin. Authigenesis may have
occurred either by alteration of inherited clays through transformation processes,
or, more rarely, by direct precipitation from solution through the weathering of
both primary and secondary minerals (Millot, 1970; Eberl, 1984; Curtis 1987).
108 LJ. Evans
TABLE 5.1
Classification of the clay minerals
A. Crystalline
Phyllosilicate minerals Non-silicate minerals
1:1 layer Kaolinite, halloysite Hematite c*-Fe2C>3
2:1 layer Clay mica (illite) Pyrolusite β-ΜηΟι
Smectite, vermiculite Gibbsite Al(OH) 3
Chlorite Goethite a-FeOOH
2:1 ribbon Sepiolite Lepidocrocite 7-FeOOH
Palygorskite
B. Poorly and non-crystalline

Alwninosilicate minerals Non-silicate minerals
Allophane Ferrihydrite Fe5HOs.4H20
Imogolite
The structures of clay minerals
Phyllosilicate clays
The phyllosilicate clays are composed of alternating planes of oxygens, hydroxyls
or oxygens/hydroxyls. These planes arise from the joining together of sheets of O,
OH, or O/OH's with tetrahedral and octahedral coordination into 1:1 or 2:1 layers.
Within the spaces created by the various packing arrangements of these anions are
smaller cations, such as Si, Al, Mg or Fe. These layers may be separated from one
another by various interlayer materials, such as cations, hydrated cations, organic
molecules, and groups and sheets of hydroxide octahedra. The combination of a
layer and any interlayer material is referred to as a unit structure. The following
discussion on the structures of the clay minerals is largely taken from reviews by
Brown et al. (1978), Bailey (1980,1988a) and Newman and Brown (1987).
The classification of the phyllosilicate minerals is based on the arrangement of
the tetrahedral and octahedral sheets, the net layer charge on the unit structure,
and the nature of the interlayer material between the 1:1 or 2:1 layers. The struc
tures of the phyllosilicate clay minerals are comprised of two fundamental building
blocks: sheets of silica tetrahedra, [SÌ2O5]2", and sheets of aluminum or magnesium
hydroxide octahedra, Al2(OH)6 or Mg3(OH)6. The plate-like morphology of most
phyllosilicate clay minerals reflects this underlying atomic arrangement.
Ideally, the tetrahedral sheet is a continuous two-dimensional arrangement of
hexagonally open packed O's, with the fourth corner of the tetrahedron — the
apical O, directed normal to the sheet. The tetrahedra are normally occupied by
Si 4+ , but may contain Al3+ to create a negative charge within the sheet.
In the octahedral sheet, each hydroxyl carries one negative charge and, because
the octahedra are joined together by sharing edges, each hydroxyl is shared between
Alteration products at the earth 's surface — the clay minerals 109
three adjacent octahedron. The total negative charge ascribable to each octahedra
is two. Tb maintain electrical neutrality within the sheet, two out of three octahedra
can be filled with trivalent cations (dioctahedral sheets) or all three can be filled
with divalent cations (trioctahedral sheets).
In the phyllosilicate minerals the silica tetrahedral sheet is joined to the octa
hedral sheet by replacing two hydroxide anions from each octahedra with apical
oxygens from the tetrahedral sheet. In this way the two negative charges on the
[SÌ2O5]2" sheet are neutralized by the charges created in the octahedral sheet. By
joining together either one or two silica tetrahedral sheets with one octahedral sheet
two groups of phyllosilicate minerals are formed: (a) the 1:1 clay minerals, such as
kaolinite and halloysite, consisting of one tetrahedral sheet joined to one octahedral
sheet, and (b) the 2:1 clay minerals, such as the smectites, vermiculites, clay micas
(illites) and chlorite, consisting of two tetrahedral sheets joined to one octahedral
sheet.
1:1 clay minerals

Kaolinites comprise the major group of clay minerals with 1:1 structures. They
are dioctahedral minerals with the octahedral positions occupied by Al3+ cations.
Structurally they consist of one [Si205]2"sheet joined to one [Al2(OH)6] sheet (Fig
ure 5.1a). This structure has a half unit cell of chemical composition Si2Al205(OH)4.
There is little or no isomorphous substitution in either the tetrahedral or octahedral
sheets and consequently the mineral has a very low permanent layer charge.
As the silica tetrahedral sheet is slightly larger than the alumina octahedral sheet,
the silica tetrahedra in kaolinite rotate slightly into a ditrigonal arrangement to
accommodate the smaller octahedral sheet. In the octahedral sheet, the vacant site
is surrounded by six occupied octahedra in such a way that the edges shared between
adjacent Al cations are shortened, whereas the unshared edges are lengthened
(Brown et al., 1978). Successive kaolinite structures are joined together through
hydrogen bonding between basal O's on silica tetrahedral sheets and basal OH's on
octahedral sheets. This stacking leads to three polymorphs — kaolinite, dickite and
nacrite.
In the 1:1 dioctahedral mineral halloysite, rotation of the silica tetrahedra does
not occur and the mineral generally has a tubular morphology with the basal O's of
the silica tetrahedra positioned on the outer surface of the rolled-up layers (New
man and Brown, 1987). The morphology of most halloysites is fibrous, although
some spherical forms have recently been identified by scanning electron microscopy
(Brown et al. 1978). Halloysite also differs from kaolinite in that it contains ad
ditional interlayer water located between the 1:1 layers. This water is driven off
irreversibly at 110°C.
110 L.J. Evans
2:1 clay minerals

The 2:1 clay minerals are structurally related to pyrophyllite and talc, but differ
in that they have considerable isomorphous substitution in both tetrahedral and/or
octahedral sheets. Isomorphous substitution generally involves the substitution of
elements of higher charge with those of lower charge, and the 2:1 clay minerals
consequently have permanent negative structural charges.
The clay micas are 2:1 phyllosilicates in which the interlayer space is occupied
by K+ ions (Figure 5.1b). The ionic radius of K+ is of such a size that it will fit
into the hole created by two basal planes of O's in the ditrigonally open packed
configuration of adjacent silica tetrahedral sheets. The micaceous mineral found in
clay fractions is usually termed illite, although many minerals formerly identified
as illites have since been shown by X-ray diffraction to contain expandable layers
(Nadeau and Bain, 1986). Hence the exact status of this name is still not resolved
(Bailey, 1980b) and the term clay mica is preferred in some reviews (Brown et al.
a Kaolinite (1:1 structure) b· Smectite or vermiculite (2:1 structure)
Fig. 5.1. Structures of the phyllosilicate clays.

Alteration products at the earth's surface — the clay minerals 111
1978), whereas the term illite is preferred in others (Newman and Brown, 1987).
Because ilhte is still commonly used in the geological hterature, the term illite will
be used in this chapter. Illites generally have lower K+ contents than macroscopic
micas, ranging in content from 7 to 10% K2O for clay micas (illites) as compared to
12-13% for macroscopic micas. The status of the term sericite, also often used to
describe microscopic micas, is also unresolved.
The common macroscopic micas of igneous and metamorphic rocks are musco-
vite, phlogopite and biotite. Muscovite is a dioctahedral mica in which the octahe
dral positions are occupied by Al 3+ , whereas phlogopite and biotite are trioctahe-
dral micas with Mg2+ and Mg2+ plus Fe2+, respectively, occupying their octahedral
positions. Tb accommodate the positive charge contributed by the addition of the
K+ ion to the 2:1 structure, a negative charge is generated in the tetrahedral
sheets by replacement of a Si4+ ion by an Al3+ ion. Ideal chemical compositions of
muscovite and phlogopite are SÌ3Al3O10(OH)2K and Si3AlMg3Oi0(OH)2K, respec
tively.
By contrast, celadonite and glauconite are Fe3+ -rich dioctahedral micas in which
the negative charge is generated principally in the octahedral sheet by replacement
of a Fe3+ ion with a Mg2+ ion. Although the amount of substitution of Si4+ by Al3+
in tetrahedral sheets is small, this substitution is always greater in glauconite than in
celadonite (Buckley et al., 1978).
The smectites are clay minerals with 2:1 structures whose unit structures are
separated by interlayer hydra ted cations, generally Ca2+, or to a lesser extent Mg2+
or Na+ (Figure 5.1c). Smectites can be either dioctahedral or trioctahedral and
their negative charges generated in either the tetrahedral or octahedral sheets.
These charges can be generated by either isomorphous substitution or by non-
ideal octahedral occupancy. The classification of the smectite minerals is shown in
Tkble 5.2.
Vermiculites have similar structures to smectites (Figure 5.1c), with these two
groups of phyllosilicate minerals differing only in the magnitude of the charges in
their unit layers. The minerals with layer charges up to 1.2 per C>2o(OH)4 formula
unit are classified as smectites, whereas those with charges between 1.2 and 1.9 are
classified as vermiculites. Vermiculites also differ from smectites in that vermiculites
can often occur in macroscopic forms, chiefly as a result of the alteration of biotite
mica by supergene solutions.
The chlorites are also 2:1 minerals, but differ from the minerals discussed above
in that they contain an additional octahedral sheet within the interlayer space be
tween the 2:1 layers (Figure 5.Id). The octahedral sheets in the chlorite structure
can be either di- or trioctahedral and there is consequently a wider range of chem
ical composition in chlorites than for the other clay minerals (Weaver and Pollard,
1973). These additional octahedral sheets are however generally positively charged
and are bonded in part by electrostatic attraction to the negatively charged 2:1
layers. In addition to this electrostatic attraction, weak hydrogen bonds are formed
112 L.J. Evans
TABLE 5.2
Classification of the smectite minerals

Source of Nature of Octahedral Mineral
layer charge octahedral layer cation(s)
Tetrahedral Dioctahedral Al 3 + Beidellite
Dioctahedral Fe 3 + Nontronite
Trioctahedral Mg2+,R3+ Saponite
Octahedral Dioctahedral Al 3 + , Mg 2 + Montmorillonite
Trioctahedral Mg2+,Li+ Hectorite
Trioctahedral Mg 2 + Stevensite
between the basal O's on the tetrahedral sheets and the basal OH's on the octahe
dral sheets so that macroscopically they are larger than smectites or vermiculites.
Sepiolite and palygorskite arefibroustrioctahedral Mg-rich 2:1 minerals. They
differ in structure from talc however in that the 2:1 layers do not form a continuous
sheet but are arranged into ribbons with finite widths. In sepiolite and palygorskite
the tetrahedral silica sheet inverts itself after eight and five octahedral positions
respectively so that the basal plane of O's at the bottom of one 2:1 layer becomes
the basal plane of O's at the top of the next 2:1 layer. This arrangement is repeated
throughout the structure to give a very open framework in which the channels
are relatively wide and are filled with exchangeable cations. Of the five octahedral
positions in palygorskite, only four are filled — two with Mg and two with Al. In
many soils and sediments the clay suite often contains clay minerals with properties
and X-ray diffraction characteristics intermediate between those discussed above.
In these phyllosilicate structures two or more kinds of layers occur in a vertical
stacking sequence and are referred to an interstratified minerals. Interstratified
smectite/illite has been commonly reported in sedimentary rocks.
The crystalline non-silicate minerals

The most important group of non-silicate minerals found in soils and sediments
are the oxides, hydroxides and oxyhydroxides. A listing of some of the more com
mon crystalline non-silicate minerals found in soils and sediments is shown in Ta
ble 5.1.
Hematite, a-Fe203, consists of O atoms arranged in hexagonal open packing
with Fe3+ ions in octahedral coordination, whereas in maghemite, 7-Fe 2 0 3 , one
third of the Fe3+ atoms are tetrahedrally coordinated. Mn0 2 exists in a variety of
crystalline forms with a diverse and complex chemistry. Amongst the more common
forms found in soils and sediments are birnessite, £-Mn02, pyrolusite, /?-Mn02,
and hollandite, α-Μη02. Structurally they are composed of various arrangements
of Mn0 6 octahedra in which the negative charges on the octahedra are balanced by
incorporation of K, Ca, Ba, Al and Li into the structures. The oxides of Fe and Mn
can form readily in soils and sediments if environmental conditions are suitable.
Gibbsite, Al(OH)3, is the only commonly occurring crystalline hydroxide and it
consists of Al2(OH)6 octahedra joined together by sharing edges. T\vo oxyhydrox-
ides of Al, boehmite, 7-AIOOH, and diaspore, α-ΑΙΟΟΗ, have structures similar
to gibbsite, but differs in that the octahedra are composed of alternate planes of O's
and OH's held together through hydrogen bonding.
TWo oxyhydroxides of Fe commonly occur — goethite, ß-FeOOH, with a struc
ture similar to boehmite, except for the replacement of Al3+ by Fe3+ in the oc
tahedral positions, and lepidocrocite, 7-FeOOH, a polymorph of goethite. A third
polymorph, akageneite, /?-FeOOH, has rarely been reported.
The non-crystalline and poorly crystalline minerals

The non-crystalline minerals do not have a well defined unit cell and are charac
terized by having order only over a short range. They are amorphous to X-rays and
are therefore difficult to identify in soils and sediments. Poorly crystalline minerals
differ from non-crystalline minerals in that they give a weak but discernable X-ray
diffraction pattern.
Amorphous silica, S1O2, is common in the surface horizons of soils and is believed
to be derived from plant material in the form of opal phytoliths. These phytoliths
form by uptake of Si by plants and the subsequent deposition of S1O2 around the
cell walls of the plant.
Allophanes are a series of naturally occurring hydrous aluminosilicates which are
essentially X-ray amorphous. Allophanes contain varying amounts of Si0 2 , A1203
and H2O, with the ratio of Si to Al ranging from 1 to 2. A typical composition of
an allophane is 2SiO2.Al2O3.3H2O. Because of the range in chemical composition
of allophane, there is a corresponding range in structures (Okada et al, 1975). Alu
minium can occur in both tetrahedral and octahedral coordination, with the amount
of Al* increasing as the Al content increases. The presence of A1IV within the
structure imparts to the mineral a permanent negative structural charge; whereas
the presence of hydroxyls in surface aluminol groups imparts a pH-dependent net
proton charge that may be either positive, neutral or negative.
Imogolite is a poorly crystallinefibrousmineral originally identified in weathered
volcanic glasses. It has a chemical composition of SiO2.Al2O3.H2O in which all
the Al is in octahedral co-ordination (Farmer et al, 1983). Transmission electron
microscopy shows it to have afibrousmorphology, with the fibres approximately 2
nm in diameter and several microns long. The inside of the tube consists of apical
hydroxyls from silica tetrahedra, whereas the outside of the tube consists of basal
hydroxyls from the alumina octahedra. Hydrous aluminosilicates with characteristics
intermediate between the two groups described above have been termed proto-
imogolite and even proto-imogolite allophane.
114 LJ. Evans
Ferrihydrite, formally called "amorphous" iron hydroxide, is a poorly ordered

oxyhydroxide of Fe that commonly occurs in soils and sediments. It has a stoichiom-
etry of FeOOH, modified by small crystal size and a high surface water content to a
composition between Fe4(OH)i2 and Fes03(OH)9. Its structure is based on double
hexagonal close packed O and OH with two adjacent layers of Fe™ followed by two
layers of Fe* (Eggleton and Fitzpatrick, 1988).
Occurrence of clay minerals
Soils
Soils result from the action of weathering processes at the earth's surface and
the intensity of these weathering processes dictates to a large extent the type of clay
mineral formed. There is thus a distinct zonal distribution in the occurrence of clay
minerals within the world's soils (Gradusov, 1974; Chamley, 1989).
Kaolinite, and to a lesser extent halloysite, for example, are common minerals
in soils developed in tropical and sub-tropical areas, particularly on acidic rocks in
well-drained environments, whereas smectites most commonly occur in soils from
warm temperate regions, particularly in non-glaciated areas. Smectites are also com
mon in poorly drained soils developed from volcanic materials and in the E horizons
of podzolic soils.
Chlorites, particularly those containing trioctahedral sheets, tend to be relatively
unstable in soils with a neutral to acidic pH and generally weather to give vermi-
culite. In soils of high latitudes or low precipitation chlorite is generally preserved.
Illites have a similar distribution to chlorite being relatively unstable in intense
weathering regimes.
Sepiolite and palygorskite are largely restricted to soils of arid environments,
particularly those containing basic rocks, such as basalt. They may however occur in
other soils if those soils have developed in material derived from sedimentary rocks
containing either or both of the two minerals.
The oxides and oxyhydroxides of Fe, Al and Mn also have a marked world-wide
pedological zonal distribution, with both hematite and maghemite commonly occur
ring in highly leached tropical soils, such as ferralsols, particularly those developed
on basic soil parent materials, such as basalts and gabbros. There is also increasing
evidence to suggest that the presence of maghemite in some, if not all soils, is the
result of previous burning which causes the conversion of other Fe oxides, oxyhy
droxides and hydroxides to maghemite (Schwertmann, 1988). Akaganeite has rarely
been reported in soils, except for some New Zealand soils (Logan et al., 1976). The
various oxides of Mn are not generally as common in soils as those of Fe but have
been found in manganese nodules in imperfectly and poorly drained soils.
Gibbsite commonly occurs in association with kaolinite in tropical soils and is
believed to be the major end-product of weathering of aluminosilicate minerals. It,
together with lesser amounts of boehmite, is a common constituent of bauxitic de

posits. Diaspore and corundum, AI2O3, are rare constituents of soils and sediments
but have been reported in diagenetically altered bauxites.
Goethite is the most abundant Fe-containing oxyhydroxide in soils, being particu
larly common in temperate soils. It is formed by the hydrolysis of Fe3+ ions released
during the oxidation and weathering of Fe-containing minerals. Lepidocrocite is
believed to form through the oxidation of Fe2+ ions, particularly in poorly drained
soils. Variations in pH, partial pressure of C0 2 and the rate of oxidation of Fe2+ de
termine to a large extent which of the oxyhydroxides of Fe will form in a particular
soil environment (Schwertmann, 1988).
Allophane and imogolite were at first thought to be largely confined to soils
developed from volcanic parent materials in tropical and sub-tropical environments,
but now are believed to occur in small amounts in many other soils and sediments.
Ferrihydrite is a poorly crystalline form of Fe in soils and recent evidence suggests
that this mineral is more common than previously appreciated. It may be the major
Fe-containing secondary mineral in many Fe-concretions of imperfectly to poorly
drained soils and in the Fe accumulations of podzolic B horizons. Ferrihydrite
is however metastable and will transform to hematite in warm climates, and to
goethite in cool, temperate climates. In tropical soils it converts to hematite through
a process of dehydration and recrystallization.
Sediments
Lacustrine and estuarine sediments

The clay detritus carried by rivers to the sea consists of both material developed
within soil profiles and unaltered material inherited from the bedrock. Results from
a large number of studies on freshwater lacustrine sediments indicate that the clay
minerals present primarily reflect source lithologies and that there is no definitive
evidence for the authigenesis of clay minerals in freshwater lacustrine environments
(Jones and Bowser, 1978; Chamley, 1989).
For example, Court et al., (1972) found no evidence for diagenetic changes in the
sediments of Lake Tàhoe through time and attributed the chlorite and vermiculite
present to the weathering of granitic rocks and the smectite to the weathering of vol
canic rocks on adjacent land surfaces. Similarly the progressive westward increase
in the ratio of smectite to illite in sediments of Lake Constance, northern Europe,
have been ascribed to a progressive enrichment in detrital material of Cenozoic
origin in contrast to the upstream crystalline detritus which is richer in chlorite and
illite (Müller and Quakernaat, 1969). Other close relationships between the clay
mineralogical composition of freshwater lacustrine sediments and the mineralog-
ical assemblage of associated soils and sediments have been reported by Moore
(1961), Dean and Gorham (1976), Englund et al. (1976) and Jones and Bowser
(1978).
116 L.J. Evans
By contrast, some studies on saline lakes associated with volcanic terrain, such as
Lake Chad in north Africa, have provided definitive evidence for the formation of
Mg- and Fe-rich smectites, such as saponite, stevensite and nontronite (Carmouze et
al. 1977; Gac et al, 1977). Illitization of detritai smectites has also been reported in
saline lakes (Singer and Stoffers, 1980; Jones and Weir, 1983), as has the alteration
of diatoms to smectites in Bolivia (Badaut and Risacher, 1983). The formation of
phyllosilicate clays in saline lakes is, however, the exception rather than the rule
(Chamley, 1989).
The transport and deposition of sediments in estuaries involves various inter
acting factors that include rate of sedimentation and different source areas for the
clays. Sediment source may not only be from soils associated with adjacent water
sheds but also from bottom sediment transported into the estuary by tidal action
(Sawhney and Frink, 1979). It has been commonly reported that kaolinite, illite
and smectite are selectively distributed in estuarine sediments because of differing
settling rates due to their varying aggregate sizes and densities in suspension (Bar-
baroux and Gallenne, 1973; Edzwald and O'Melia, 1975; Gibbs, 1977; Manickam
et al., 1985). The presence of certain clay minerals in various stratigraphie sections
from the Pennsylvanian allowed Brown et al. (1977) to infer cyclic changes in marine
to non-marine transgressions in shallow water sediments.
Although it was thought from early studies on estuarine sediments that chemical
alteration played an important role in controlling the composition of the sediments,
further work and a réévaluation of the original data suggests that such alterations
are insignificant (Chamley, 1989).
Oceanic sediments
As oceanic sediments are derived largely from terrestrial sources, there is a zonal
distribution of clay minerals within the sediments of the world's oceans correspond
ing in large part to the clay mineralogy of adjacent land masses. The occurrence of
clay minerals in oceanic sediments has been described by Biscayne (1965), Griffin
et al. (1968), Rateev et al. (1969), Windom (1976) and Chamley (1989) and the
following review is taken from those sources.
The highest contents of chlorite occur in the oceans of high latitudes, that is,
near the North and South Poles. In northern latitudes large areas of Canada and
northern Europe are covered by chlorite-containing low-grade metamorphic rocks
of the Precambrian Shield or sedimentary rocks of Paleozoic age. As these rocks
have been subjected to minimal chemical weathering, chlorite is preserved in eroded
sediments.
Vast areas of the Indian, Pacific and Atlantic Oceans near the tropics contain
little or no chlorite — except for small areas near volcanic islands, such as Hawaii,
where alteration of volcanic ash has occurred on the oceanfloor.The occurrence of
chlorite in sediments of the Gulf of Arabia and Pacific sediments east of Australia
have been ascribed to desert conditions encouraging the preservation of chlorite.
The distribution of illite in oceanic sediments is similar to that of chlorite and is

similarly believed to be largely detrital in origin. In the Atlantic Ocean, illite con
tents are generally highest near northern or southern latitudes and as the equator is
approached illite gives way to smectite and kaolinite.
Kaolinite is formed by intensive weathering of minerals in soils and is thus a ma
jor soil mineral in tropical environments. It is therefore commonly found in oceanic
sediments in tropical regions, with concomitantly lower contents of kaolinite occur
ring in northern and southern oceans. Further evidence that kaolinite is derived
from terrestrial sources is its common association with other end products of weath
ering, such as rutile, anatase and zircon. In the Atlantic ocean, maximum contents
occur around the coast of equatorial west Africa drained by the Niger and Congo
rivers and off the coast of South America drained by the Sao Francisco river. In
the Pacific Ocean, lower contents of kaolinite are found in sediments south of the
equator, probably because there are no large terrestrial source areas for kaolinite
within the region. High amounts of kaolinite occur in the oceanic sediments around
Australia, India, Madagascar and Java as a result of tropical weathering.
Smectite has its highest distribution near equatorial regions — often associated
with gibbsite and kaolinite. Unlike the other clay minerals however, the distribution
of smectite in oceanic sediments is believed to be the result of both neoformation in
oceanic sediments and inheritance from terrestrial sources. Three principal modes
of neoformation have been recognized — alteration of volcanic rocks and glasses,
low temperature combination of biogenic silica and Fe-oxyhydroxides, and direct
precipitation from hydrothermalfluids(Cole and Shaw, 1983).
On the fringes of the Pacific ocean, for example New Zealand, Indonesia and
Japan, and in sediments associated with mid-Oceanic ridges, there is considerable
tectonic activity and hence the sediments contain large amounts of volcanic ash.
There is more smectite in southern than northern oceans probably because of the
greater occurrence of volcanic activity and the subsequent conversion of volcanic
ash to smectite.
On the other hand, sediments in the Caribbean and the Amazon basin contain
high contents of smectite, up to 26-33%, derived largely from terrestrial sources.
Similarly, sediments of the Yellow River and the Mississippi contain abundant smec
tite because of the weathering of loess, which provides a detrital source of smectite.
Azonal occurrences of some other clay minerals also occur, such as the presence
of illite in tropical sediments near the north African coast and the Indian Ocean
south of Sri Lanka. These occurrences are believed to be due to eolian origin from
Europe and India respectively. Other azonal occurrences in the Indian Ocean are
the illite-rich sediments of the Bays of Bengal and Arabia. These sediments are
derived from rivers draining the arid areas of India and Pakistan whose soils contain
large amounts of illite.
118 LJ. Evans
Sedimentary, igneous and metamorphic rocks

In clastic sedimentary rocks, such as sandstones, clay minerals are generally
found only in small amounts, with kaolinite, illite and chlorite being the most com
monly reported phyllosilicate minerals (Millot, 1970; Chamley, 1989). The minerals
are believed to be largely inherited but there is considerable evidence that kaolin
ite (Hurst and Irwin, 1982; Scotchman et al., 1989), illite (Ehrenberg and Nadeau,
1989; Warren and Curtis, 1989) and chlorite (Humphries et al., 1989) can be ne-
oformed by alteration of feldspars after deposition. Clay minerals are much more
abundant in shales, with rocks older in age than the Mississippian dominated by
illite and chlorite (Weaver, 1959). Minerals in these older rocks are believed to be
largely diagenetic (neoformed) whereas in shales younger than the Mississippian
there is a much more variable mineralogy, reflecting a mixture of both inheritance
and neoformation.
Coal measure underclays are commonly comprised of kaolinite, illite and illite/
smectite. The origin of these clays is still disputed. A pedological origin in which the
clays are believed to result from the alteration of various rocks by acidic, organic-
rich waters is favoured by some authors (Rimmer and Eberl, 1982), while others be
lieve that their composition reflects the clay mineralogy of the source area (Schultz,
1958; Alonso and Brime, 1990). Bentonite beds are dominated by smectite, formed
as a result of alteration of volcanic ashes in a marine environment (Grim and
Güven, 1978), although illitic and kaolinitic bentonites are also known.
Although illite and kaolinite have been reported in some limestones, and chlorite
in some dolostones, chemical precipitates, such as evaporites, generally contain very
little in the way of clay minerals with sepiolite and palygorskite being identified in
some chalks and evaporites.
Apart from chlorites and the macroscopic micas, other clay minerals are not
abundant in igneous or metamorphic rocks. Chlorite can result from low-grade
metamorphic alteration of igneous rocks containing hornblende or biotite and also
by hydrothermal alteration of pyroxenes, amphiboles and biotites. Similarly, kaolin
ite can form from the hydrothermal alteration of feldspars in granitic rocks; Mg-rich
smectites from the alteration of volcanic rocks and vermiculite from the alteration
of biotites (Chamley, 1989).
Formation of clay minerals
Weatherìng
The weathering of primary minerals to form secondary clay minerals is an im
portant pedological process, the extent of which depends on soil climate, age and
the geological nature of the parent material. A vast body of information on the
secondary products of weathering has accumulated over the years and only some
typical results will be presented below.
Gilkes et al. (1973) studied deeply weathered regoliths in Australia and from the
results of mineralogical analyses suggested the following reaction pathways for the
weathering of granite in a tropical environment:
feldspar kaolinite gibbsite boehmite

biotite goethite hematite
Similar studies by Stoch and Sikora (1976) have suggested:
muscovite muscovite/smectite smectite kaolinite

biotite kaolinite
Studies on the weathering of basalt have shown that the nature of the principal
weathered products — kaolinite and/or smectite — depends on a number of differ
ent factors, such as rainfall, soil drainage, age of the soil and soil pH, with kaolinite
being more abundant in wetter (Sieffermann and Millot, 1966; Singer, 1966), poorly
drained (Sherman and Uehara, 1966), older (Briner and Jackson, 1970) or low pH
environments (Swindale, 1966). The relationship between rainfall and the clay min
eral composition of soils from different rock types (Figure 5.2) has been illustrated
by Barshad (1966).
Some possible reaction pathways to produce clay minerals from weathering re
actions in soils are shown in Figure 5.3. The ultimate products of these reactions
depend largely on climatic and pedological factors, although there is with time a ten
dency in freely draining environments to approach a residua system of weathering
as defined by Chesworth (1973).
80
Kaolinite 80 r
\ Smectite and halloysite
c 60
o 60
k_
o
Q_
O 40
- ΙΛ_ Vermiculite 40
k_
Q. _ / λ lllite\y
ω a. 20 h
Gibbsite^/
è 20
(D —·—i ^ ^ ι^ - — ι ι
50 150 200 50 100 150 200
Mean annual rainfall (cm)
Fig. 5.2. Effect of precipitation on the distribution of clay minerals from a. acidic and b. basic rocks in
California (Barshad, 1966).
120 L.J. Evans
k
1 mite 1 1 Chlorite | 1 Feldspars Gibbsite |
k
\
V / -Al 3 +
(Al-rich)
Allophane
(Halloysite)
-H4S1O4 η
-H4S1C
Vermiculite **| Smectite Kaolinite |

(Mg, Fe-rich)
A \
+AI 3+ >v -H4S O4
'
Hydroxy- Ferromagnesium Hematite |
interlayered minerals
verm cunie
Fig. 5.3. Some possible weathering pathways of minerals.
Hydrothermal alteration
The clay minerals resulting from the reactions between hydrothermal fluids and
various rock types depend not only on the nature of the geological material but
also on the chemical composition of the hydrothermal fluids and the pressure-
temperature conditions under which the reactions are occurring. These reactions
can occur at temperatures from 60 up to 400°C, with the reaction products of low
temperature hydrothermal alteration often resembling those found in weathering
regimes. At higher temperatures and pressures the resultant products resemble
more closely those from higher grade metamorphic processes (Table 5.3).
Mineralogical reactions at high temperatures are not only more rapid but because
of the temperature dependence of reaction equilibrium constants different phases
TABLE 5.3
Some examples of hydrothermal formation of clay minerals
Andesitic and rhyolitic lava flows3 Plutonic and metavolcanic rocks b

60-190°C Smectite 30°C Smectite and vermiculite
110-210°C Kaolinite 100°C Smectite/vermiculite
100-150°C Illite 160°C Smectite/chlorite
150-310°C Chlorite 200°C Illite, chlorite, smectite/chlorite
Volcanic rocks c Sandstones and shales d

120-160°C Smectite 190°C Illite/smectite
160-230°C Smectite/illite 190-325°C Illite
>230°C Illite >326°C Biotite, vermiculite
a
Cathelineau et al. (1985).
b
Steiner (1968).
c
Meunier et al. (1987).
d
McDowell and Elders (1980).
Alteration products at the earth's surface — the clay minerals 121
-lb
-10
Muscovite
pK -5
0 Kaolinite
1 1 1 1
0 100 200
Temperature (°C)
Fig. 5.4. Temperature dependence of the equilibrium constant, K, for kaolinite <-*■ muscovite.
may form for a given ionic composition of the aqueous phase as the temperature
is raised. Figure 5.4 illustrates the changes in the equilibrium constant for the
incongruent reaction between kaolinite and muscovite as a function of temperature
as calculated by the van't Hoff equation:
2Si3Al3Oio(OH)2K + 3H 2 0 + 2H+ <-► 3Si2Al205(OH)4 + 2K+
Diagenesis
There is general agreement that shale diagenesis involves, to a large extent, the
conversion of smectite to illite/smectite and to illite, and also possibly to chlorite.
With increasing depth of sediment, there is a progressive decrease in the content
of smectite and an increase in the amount of interstratified illite/smectite, as well
as decreases in the amounts of calcite and K-feldspar (Hower et al., 1976). The
conversion of smectite to illite generally ceases at depths below 3,700 m. Conversion
may however occur at much shallower depths if there is an increased geothermal
gradient induced by igneous intrusions (Andrews, 1987).
Analyses of illite/smectite from burial diagenetic sequences indicate that the
main compositional changes with increasing amounts of illite layers are a gain in
interlayer K, increasing substitution of Al for Si in tetrahedral layers, coupled with
reduction of octahedral Fe3+ to Fe2+ and loss of octahedral Mg and Fe (Hower,
1981).
Experiments by Eberl and Hower (1976) suggest that the time span for the
conversion of smectite to low expandable interstratified phyllosilicates is of the
order of 106 years at 60°C and Hower et al. (1976) regard the overall reaction as:
K+ + Al3+ + smectite -> illite + Si

122 L.J. Evans
with the K+ and Al3+ ions being supplied by the dissolution of K-feldspars. The illite
formed from such reactions has an average K content of 0.75 atoms per Oio(OH)2
unit cell (Srodon et al, 1986).
This simple diagenetic transformation of smectite to illite by the gain of K+ ions
has however been questioned and it has been suggested that neoformation of smec
tite may be more important than previously suspected (Jeans, 1989). Nadeau and
Bain (1986), for instance, have shown that the chemical composition of smectites
in Cretaceous illitic bentonites precludes their having served as transformation pre
cursors for most Cretaceous illites. These latter authors favour an origin involving
neoformation.
Authigenic kaolinite, illite and chlorite have all been reported in sandstones, with
temperatures of formation being in the range 70-160°C (Velde, 1985). Jahren and
Aagaard (1989) have shown that, while illite compositions do not vary substantially
with depth of burial in some Jurassic sandstones, there is a compositional change
in chlorites with A1IV increasing and octahedral occupancy decreasing with increas
ing temperature. They suggest that the clay minerals are continually recrystallizing
under equilibrium conditions in response to increasing depth of burial.
Future research
Tb date much of our knowledge of the occurrence of clays has been derived from
the use of X-ray diffraction analysis. As a result there has been much emphasis
on the crystalline components of clay fractions, with the poorly ordered minerals,
such as the allophanes, having received less attention. The increased use of selec
tive extradants to characterize these poorly ordered minerals should improved our
understanding on the occurrence and distribution of these minerals, particularly in
sediments and sedimentary rocks.
Through the increased use of techniques such as high resolution scanning and
transmission electron microscopy, some of the established modes of origin of clay
minerals have been questioned. These techniques, coupled with chemical character
ization of individual clays by energy dispersive microchemical analysis, will be more
widely used than previously in clay studies as these techniques gain wider exposure.
References
Alonso, O.E. and Brime, C, 1990. Mineralogy, geochemistry and origin of the underclays of the central
coal basin, Asturias, Spain. Clays Clay Miner., 38: 265-276.
Andrews, J.E., 1987. Jurassic clay mineral assemblages and their post-depositional alteration: upper
Great Estuarine Group, Scotland. Geol. Mag., 124: 261-271.
Bailey, S.W., 1980a. Structure of layer silicates. In: G.W. Brindley and G. Brown (Editors), Crystal
Structures of Clay Minerals and their X-ray Identification. Miner. Soc, London, pp. 1-123.
Bailey, S.W., 1980b. Summary of recommendations of AIPEA Nomenclature Committee. Clay Miner.,
15: 85-93.
Bailey, S.W. (Editor), 1988. Hydrous Phyllosilicates. Reviews in Mineralogy, No. 19, Miner. Soc. Am.,
725 pp.
Barbaroux, L. and Gallenne, B., 1973. Répartition des minéraux argileux dans les sédiments récent de
l'estuaire de la Loire et du plateau continental voisin. C.R. Acad. Sci., 277: 1609-1612.
Barshad, I., 1966. The effects of variations in precipitation on the nature of clay mineral formation in
soils from acidic and basic igneous rocks. Int. Clay Conf., Jerusalem, 1: 167-173.
Badaut, D. and Risacher, F., 1983. Authigenic smectite on diatom frustiles in Bolivian saline lakes.
Geochim. Cosmochim. Acta, 47: 363-375.
Biscayne, P.E., 1965. Mineralogy and sedimentation of Recent deep sea clay in the Atlantic Ocean and
adjacent seas and oceans. Geol. Soc. Am. Bull., 76: 803-832.
Briner, G.P. and Jackson, M.L., 1970. Mineralogical analysis of clays developed from basalt in Australia.
Isr. J. Chem., 8: 487-500.
Brown, L.F., Bailey, S.W., Cline, L.M. and Lister, J.S., 1977. Clay mineralogy in relation to deltaic
sedimentation patterns of Desmoinesian cyclotherms in Iowa, Missouri. Clays Clay Miner., 25: 171-
186.
Brown, G., Newman, A.C.D., Rayner, J.H. and Weir, A.H., 1978. The structure and chemistry of soil
clay minerals. In: DJ. Greenland and M.H.B. Hayes (Editors), The Chemistry of Soil Constituents.
Wiley, New York, N.Y., pp. 29-178.
Buckley, H.A, Bevan, J.C., Brown, K.M., Johnson, L.R. and Farmer, V.C., 1978. Glauconite and celado-
nite: two separate mineral species. Miner. Mag., 42: 373-382.
Carmouze, J.-R, Pedro, G. and Benier, J., 1977. Sur la nature des smectites de néoformation du lac
Tchad et leur distribution spatiale en fonction des conditions hydrogéochimiques. C.R. Acad. Sci.,
284: 615.618.
Cathelineau, M., Oliver, R., Nieva, D. and Garfias, A., 1985. Mineralogy and distribution of hydrother
mal mineral zones in Los Azufres (Mexico) geothermal field. Geothermics, 14: 49-57.
Chamley, H., 1989. Clay Sedimentology. Springer, Berlin, 623 pp.
Chesworth, W, 1973. The residua system of chemical weathering. J. Soil Sci., 24: 69-81.
Cole, T.G. and Shaw, H.F., 1983. The nature and origin of authigenic smectite in some recent marine
sediments. Clay Miner., 18: 239-252.
Court, J.E., Goldman, C.R. and Hyme, N.J., 1972. Surface sediments in Lake Tahoe, California, Nevada.
Curtis, CD., 1987. Clay diagenesis. In: R. Rodriguez-Clemente and Y. Tardy (Editors), Geochemistry
and Mineral Formation in the Earth Surface. Centre Nat. Rech. Sci., Madrid, pp. 585-596.
Dean, WE. and Gorham, E., 1976. Major chemical and mineralogical components of profundal surface
sediments in Minnesota lakes. Limnol. Oceanogr., 21: 259-284.
Eberl, D.D., 1984. Clay mineral formation and transformation in rocks and soils. Philos. Trans. R. Soc.
London, Ser. A, 311: 241-257.
Eberl, D.D. and Hower, J.H., 1976. Kinetics of illite formation. Geol. Soc. Am. Bull., 87: 1326-1330.
Edzwald, J.K. and O'Melia, C.R., 1975. Clay distributions in Recent estuarine sediments. Clays Clay
Miner., 23: 39-44.
Eggleton, R.A. and Fitzpatrick, R.W, 1988. New data and a revised structural model for ferrihydrite.
Ehrenberg, S.N. and Nadeau, PH., 1989. Formation of diagenetic illite in sandstones of the Garn
Formation, Haltenbanken area, mid-Norwegian continental shelf. Clay Miner., 24: 233-253.
Englund, J.O., Jorgensen, P., Roaldset, E. and Aagaard, P., 1976. Composition of water and sediments
in Lake Mjosa, south Norway, in relation to weathering processes. In: H.L. Golterman (Editor),
Interactions Between Sediments and Fresh Water. Proc. Int. Symp., Amsterdam, pp. 125.132.
124 LJ. Evans
Farmer, V.C, Adams, M.J., Fraser, A.R. and Palmieri, F., 1983. Synthetic imogolite: properties, synthesis
and possible applications. Clay Miner., 18: 459-472.
Gac, J.Y., Droubi, A., Fritz, B. and Tardy, Y., 1977. Geochemical behaviour of silica and magnesium
during evaporation of waters in Chad. Chem. Geol., 19: 215-228.
Gibbs, R.J., 1977. Clay mineral segregation in the marine environment. J. Sediment. Petrol., 47: 237-
243.
Gilkes, R.J., Scholz, G. and Dimmock, G.M., 1973. Lateritic deep weathering of granite. J. Soil Sci., 24:
523-536.
Gradusov, B.P., 1974. A tentative study of clay mineral distribution in soils of the world. Geoderma, 12:
49-55.
Griffin, J.J., Windom, H. and Goldberg, E.D., 1968. The distribution of clay minerals in the world's
oceans. Deep Sea Res., 15: 433-459.
Grim, R.E. and Giiven, N., 1978. Bentonites. Elsevier, Amsterdam, 256 pp.
Hower, J., Eslinger, E.V., Hower, M.E. and Perry, E.A., 1976. Mechanisms of burial metamorphism of
argillaceous sediments. 1. Mineralogical and chemical evidence. Geol. Soc. Am. Bull., 87: 725-737.
Hower, J., 1981. Shale diagenesis. In: F.J. Longstaffe (Editor), Clays and the Resource Geologist.
Mineral. Assoc. Can., pp. 60-80.
Humphreys, B.A, Smith, S.A. and Strong, G.E., 1989. Authigenic chlorite formation in late Thassic
sandstones from the Central Graben, North Sea. Clay Miner., 24: 427-444.
Hurst, A. and Irwin, H. 1982. Geological modelling of clay diagenesis in sandstones. Clay Miner., 17:
5-22.
Jahren, J.S. and Aagaard, P., 1989. Compositional variations in diagenetic chlorites and illites, and
relationships with formation-water chemistry. Clay Miner., 24: 157-170.
Jeans, C.V., 1989. Clay diagenesis in sandstone and shale: an introduction. Clay Miner., 24: 127-136.
Jones, B.F. and Bowser, CJ., 1978. The mineralogy and related chemistry of lake sediments. In: A.
Lerman (Editor), Lakes: Chemistry, Geology and Physics. Springer, New York, N.Y., pp. 179-235.
Jones, B.F. and Weir, A.H., 1983. Clay minerals of Lake Abert, an alkaline, saline lake. Clays Clay
Miner., 31: 161-172.
Logan, N.E., Johnson, J.H. and Childs, C.W., 1976. Mossbauer spectroscopic evidence for akaganeite
(/?-FeOOH) in New Zealand sols. Austr. J. Soil Res., 14: 217-224.
Manickam, S., Barbaroux, L. and Ottman, F, 1985. Composition and mineralogy of suspended sediment
in the fluvio-estuari ne zone of the Loire river, France. Sedimentology, 32: 721-741.
McDowell, S.D. and Elders, W.A., 1980. Authigenic layer silicate minerals in Borehole Elmore 1, Salton
Sea Geothermal Field, California, U.S.A. Contrib. Miner. Petrol., 74: 293-310.
Meunier, A., Velde, B., Beaufort, D. and Parneix, J.-C, 1987. Dépots minéraux et altérations liés aux
microfracturations des roches: un moyen pour caractériser les circulations hydrothermales. Bull.
Soc. Géol. France, Ser. 8: 971-979.
Millot, G., 1970. Geology of Clays. Chapman and Hall, London, 429 pp.
Moore, J.E., 1961. Petrography of northeastern Lake Michigan bottom sediments. J. Sediment. Petrol.,
3: 402-436.
Müller, G. and Quakernaat, J., 1969. Diffractometric clay mineral analysis of recent sediments of Lake
Constance (Central Europe). Contrib. Miner. Petrol., 22: 268-275.
Nadeau, P.H. and Bain, D.C., 1986. Composition of some smectites and diagenetic illitic clays and
implications for their origin. Clays Clay Miner., 34: 455-464.
Newman, A.CD. and Brown, G., 1987. The chemical constitution of clays. In: A.CD. Newman (Editor),
Chemistry of Clays and Clay Minerals. Miner. Soc, London, pp. 1-128.
Okada, K., Morikawa, H., Iwai, S., Ohira, Y. and Ossaka, J., 1975. A structural model for allophane.
Clay Sci., 4: 291-303.
Rateev, M.A., Gorbunova, Z.N., Lisitzyn, A.R and Nosov, G.L., 1969. The distribution of clay minerals
in the oceans. Sedimentology, 13: 21-178.
Rimmer, S.M. and Eberl, D.D., 1982. Origin of an underclay as revealed by vertical variations in
mineralogy and chemistry. Clays Clay Miner., 30: 422-430.
Sawhney, B.L. and Frink, C.R., 1978. Clay minerals as indicators of sediment source in tidal estuaries
of Long Island Sound. Clays Clay Miner., 26: 227-230.
Schultz, L.G., 1958. Petrology of underclays. Geol. Soc. Am. Bull., 69: 363-402.
Schwertmann, U., 1988. Some properties of soil and synthetic iron oxides. In: J.W. Sucki, B.A. Goodman
and U. Schwertmann (Editors), Iron in Soils and Clay Minerals. Reidel, Dordrecht, pp. 203-250.
Scotchman, I.C., Johnes, L.H. and Miller, R.S., 1989. Clay diagenesis and oil migration in Brent Group
sandstones of NW Hutton Field, UK, North Sea. Clay Miner., 24: 339-374.
Sherman, C D . and Uehara, G., 1956. The weathering of olivine basalt in Hawaii and its pedological
significance. Soil Sei. Am. Proa, 20: 337-340.
Sieffermann, G. and Millot, G., 1969. Equatorial and tropical weathering of recent basalts from Ca
meroon: allophanes, halloysite, metahalloysite, kaolinite and gibbsite. Int. Clay Conf., Tokyo, pp.
417-430.
Singer, A., 1966. The mineralogy of the clay fraction from basaltic soils in the Galilee, Israel. J. Soil
Sci., 17: 136-147.
Singer, A. and Stoffers, P., 1980. Clay mineral diagenesis in two East African lake sediments. Clay
Miner., 15: 291-307.
Srodon, J., Morgan, D.J., Eslinger, E.V., Eberl, D.D. and Karlinger, M.R., 1986. Chemistry of illite/
smectite and end-member illite. Clays Clay Miner., 34: 368-375.
Steiner, A., 1968. Clay minerals in hydrothermally altered rocks at Wairakei, New Zealand. Clays Clay
Miner., 16: 193-213.
Stoch, L. and Sikora, W., 1976. Transformations of micas in the presence of kaolinization of granites
and gneisses. Clays Clay Miner., 24: 145-162.
Swindale, L.D., 1966. A mineralogical study of soils derived from basalt and ultrabasic rocks in New
Zealand. N. Z. J. Sci., 9: 484-506.
Velde, B., 1984. Clay Minerals: A Physico-chemical Explanation of their Occurrence. Elsevier, Amster
dam, 427 pp.
Warren, E.A. and Curtis, CD., 1989. The chemical composition of authigenic illite within two sandstone
reservoirs as analysed by ATEM. Clay Miner., 24: 137-156.
Weaver, CE., 1959. The clay petrology of sediments. Clays Clay Miner., 6: 154-187.
Weaver, C E . and Pollard, L.D., 1975. The Chemistry of Clay Minerals. Elsevier, Amsterdam, 213 pp.
Whitney, G., 1983. Hydrothermal reactivity of saponite. Clays Clay Miner., 31: 1-8.
Windom, H.L., 1976. Lithogenous material in marine sediments. In: J.P. Riley and R. Chester (Editors),
Chemical Oceanography, 5. Academic Press, Toronto, Ont., pp. 103-135.
127
Chapter 6
Diagenesis and metasomatism of weathering

profiles, with emphasis on Precambrian paleosols
H.W. NESBITT
Introduction
There is interest in ancient weathering profiles (paleosols) because they may yield
information about the composition of paleo-atmospheres (Holland, 1984; Gay and
Grandstaff, 1980). The data are of particular use to geochemists and geophysicists
attempting to unravel the early evolution of the planet. Paleosols are preserved
by burial under variable thicknesses of sediments. Those now exposed have been
uplifted and invariably subjected to recent weathering regimes. All paleosols, there
fore, have been affected by diagenetic processes during and after burial and, un
fortunately, the older the paleosol, the more likely it is to be modified. Diagenetic
reactions may alter textures, mineralogy and bulk compositions of profiles, and suffi
ciently deep burial may result in metamorphism, with destruction of original textural
and mineralogical properties possible and likely. Modifications may be so exten
sive that recognition of a textures or original mineralogy is difficult or impossible;
recognition then is dependent primarily upon geological setting. Unconformities
immediately above the paleosol (regolith), and inclusion of parental material and
partially weathered soils in the overlying sediments are some of the least ambiguous
geological evidence for paleosols. Where the geological conditions are met, other
properties of the profiles, mineralogy, textures and bulk compositions, must be used
to confirm the existence of a paleosol.
Even where ancient weathering profiles are recognized, interpretations may be
suspect where effects of diagenesis and metamorphism are incompletely docu
mented. These effects must be clearly recognized and separated from effects related
to the original weathering processes if the genesis of the profile is to be understood.
There is, as a result, impetus to study the changes to weathering profiles resulting
from diagenesis and metamorphism; without the studies, interpretation of ancient
weathering processes and atmospheric conditions will remain uncertain. Compari
son of recent weathering profiles with diagenetically altered paleosols may provide
new insights into diagenetic reactions. The information is valuable, therefore, to
128 H.W.Nesbitt
sedimentary petrologists, petroleum and economic geologists concerned with metal

transport and secondary porosity in sedimentary basins.
Comprehensive studies of effects of diagenesis on weathering profiles are scarce
(Wahlstrom, 1948). Few studies, for example, have compared bulk compositional
properties of recent and ancient weathering profiles, yet bulk compositional prop
erties exert an important control on the mineralogy of paleosols. Here, bulk com
positional trends developed in recent profiles are compared to, and contrasted with,
trends found in paleosols. Tïends in both are compared with compositional trends
predicted from mass balance and kinetic considerations, and insights gained are
applied to some Precambrian paleosols.
Methodology
Introduction
Discussion of recent weathering profiles is here restricted to mature, well-devel
oped profiles produced from homogeneous plutonic (granitic) and volcanic rocks.
These rock types have been chosen to minimize ambiguities in interpretation re
lated to inhomogeneity of parental (source) rocks to the profiles. Compositional
trends in immature soils may be poorly developed or difficult to interpret, whereas
trends of mature profiles generally are well developed (Feakes and Retallack, 1988).
The study is therefore restricted to mature profiles to avoid ambiguities related to
incipient soil formation (Craig and Loughnan, 1964; Grant, 1963; Lovering, 1959).
Bulk compositions of soils or rocks (paleosols) yield weight percent or parts per
million (ppm) of oxides or elements, thus only proportions of these constituents
are provided. The gains or losses of constituents therefore must be documented
by comparing the weight percent (or ppm) of an element to either the original
sample volume or the weight percent of another element or oxide. THangular plots,
which are graphical comparisons of the percentages of three oxides or elements,
are useful in this regard (Figure 6.1). Compositional trends produced by weathering
or diagenesis are readily displayed, but importantly, trends are simple to interpret
because only changes to one or more of the plotted constituents will give rise to
compositional trends on the triangles. As well, the effects of diagenetic reactions on
the bulk compositions of weathering profiles and paleosols can be analyzed using
simple geometric analogs to vector analytical methods.
TheA-CN-K (feldspar) diagram

Feldspars constitute 70-80% of minerals of the upper continental crust (Wede-
pohl, 1969; Nesbitt and Young, 1984). Compositional trends of profiles developed
on granitic rocks are strongly influenced and probably controlled by the weath
ering of feldspars; consequently triangular diagrams reflecting the weathering of
Diagenesis and metasomatism of weathering profiles
A1 2 0 3
1.0,
A1 2 0 3
1 0
- AKa.,Ha.,lm.,AI.,Gi.
PI
>TG"Ar-
A Ga
Ba
Hb.v
PRIMARY
C
P·, MINERALS
0.0
Ca0+Na 2 0
0.0
Ca0 + Na20
Fig. 6.1. a. Compositional (molar proportions) relationships among plagioclase, K-feldspar, other com
mon primary minerals and average bulk compositions of some common unweathered igneous rocks.
b. Relations among the two feldspars, common secondary weathering products, average and average
Mississippi Delta muds. Mineral compositions are shown by diamonds. PL = plagioclase; K-sp. = K-
feldspar; BL = biotite; Hb. - hornblende; Cp. = clinopyroxene; Ka. = kaolinite group minerals; Ha. =
halloysite; Im. = imogolite; AL = allophane; GL = gibbsite; Gr = av. alkali granite; A = av. adamellite;
G = av. granodiorite; T — av. tonalité; D — av. diorite; Ga = av. gabbro; Ba = av. continental basalt;
Mis. Delta = Mississippi Delta muds. Average igneous rock compositions are from Nockolds (1954) and
average shale and Mississippi Delta muds is taken from Pettijohn (1975).
the feldspar minerals, plagioclase solid solution (NaAlSÌ308-CaAl2Si208) and K-

feldspar (KAISÌ3O8), are required to illustrate the compositional changes accom
panying weathering of feldspathic rocks. Figure 6.1a has AI2O3 as one apex, the
second apex is CaO + Na 2 0 and the third is Κ2Ο. The diagram is referred to as
the A-CN-K or feldspar weathering diagram. Molar proportions are used, hence
plagioclase solid solution plots on the left-hand boundary of the triangle at 50%
AI2O3 and K-feldspar plots on the right-hand boundary, again at 50% AI2O3. The
line connecting the two feldspars is called the "feldspar join". Other common, but
less abundant primary minerals (minerals of the granitic or volcanic rocks) are plot
ted on Figure 6.1a. The primary minerals plot either close to or below the feldspar
join. Biotite, one of the most common mafic minerals of granitic rocks, plots close
to the right-hand boundary between approximately 50 and 55% AI2O3.
Compositions of some average granitic rocks and basalt (Nockolds, 1954) are
plotted on Figure 6.1a. Most plot close to the feldspar join, indicating that the
feldspar minerals are most abundant in these rocks. Gabbro and basalt plot well
below the feldspar join and very close to the left-hand boundary. The positions
indicate that they contain appreciable „ nounts of mafic minerals (clinopyroxene
and/or hornblende) relative to the feldspars, and that plagioclase greatly dominates
130 H.W.Nesbitt
K-feldspar (if present). Averages for the volcanic rocks (Nockolds, 1954) plot over
the same range as the averages for granitic rocks. The mineralogical proportions
of volcanic rocks cannot be evaluated because they contain glassy materials the
composition of which is unconstrained. These data now are compared with the
compositions of secondary minerals (weathering products), and rocks containing
secondary minerals.
Common secondary minerals (minerals produced in the weathering or diagenetic
environment) are plotted on Figure 6.1b. Gibbsite, the kaolinite-group minerals
(kaolinite, halloysite, dickite, nacrite), imogolite and allophane plot at or very close
to the A1203 apex. Smectites (montmorillonites and beidellites) plot along the left-
hand boundary between 70 and 85% A1203 and illites plot along the right-hand
boundary between 70 and 85% AI2O3. The position of vermiculite is variable and
is determined by the extent to which it is weathered. Its derivation from biotite
involves rapid release of K from the mineral (Karathanasis, 1988) and uptake of
Mg and/or Al onto exchange and other sites. Vermiculite, therefore plots well above
biotite, and once all K 2 0 is removed and Al taken up on exchange sites, it plots
close to the AI2O3 apex.
The secondary minerals are formed from primary minerals either within weather
ing profiles or diagenetically within sedimentary basins. Sediments and rocks com
posed of these minerals reflect the compositions of the included secondary minerals.
Average shale and average Mississippi Delta surface mud (Pettijohn, 1975) plot well
above the feldspar join (Figure 6.1b). The average shale (Clarke, 1924) plots near
the right-hand boundary at 73% AI2O3, indicating that it is composed of secondary
minerals primarily, of which illite probably is most abundant. The aluminous na
ture of the secondary mineral groups certainly is reflected in muds and shales. The
Mississippi delta mud is representative of recent muds deposited from the Missis
sippi River. The sediments include glacial debris, loess and other glacially derived
materials, transported from the upper reaches of the drainage basin. The debris,
derived partly from the Canadian Shield, includes feldspars and probably other
primary minerals. Compared with average shale, the delta mud plots closer to the
feldspar join, indicating that the percentage of A1203 is low in the recent deltaic
muds compared with average shale.
It is apparent from comparison of the data plotted on Figure 6.1a and b that the
diagram is useful for studying the effects of weathering. The fields of the primary
minerals and unweathered rocks plot close to or below the feldspar join, whereas
the secondary products of weathering plot high on the diagram (Figure 6.1b). As
primary minerals of a parent rock are altered during weathering, and the propor
tions of secondary minerals increase relative to the primary minerals, weathering
trends will be defined on Figure 6.1, with the trends directed from the field of the
unweathered rocks (near the feldspar join) up the diagram, eventually towards the
AI2O3 apex.
Diagenesis and metasomatism of weathering profiles 131
TheA-CNK-FM (mafics) diagram
Although the feldspar minerals are the most abundant labile minerals of igneous
rocks, the mafic minerals are important in that they contain the bulk of the tran
sition elements. The A-CNK-FM (mafics) diagram (Figure 6.2) retains AI2O3 as
one apex, groups the constituents CaO, Na 2 0 and K2O at the left-hand apex (CNK
apex) and total iron (as FeOr) and MgO as the third apex (FM apex). Both plagio-
clase and K-feldspar plot on the left-hand boundary at 50% AI2O3 (Figure 6.2). The
mafic and secondary minerals are also plotted on the diagram. The primary miner
als plot close to, and below, the line joining feldspar composition with the FM apex
(feldspar-FM join). Average igneous rocks plot close to, and below the join, with
the rocks containing the greatest proportion of feldspars (granites) plotting closest
to the feldspar point.
The secondary minerals plot well above the feldspar-FM join as does the average
shale. As shown subsequently, separation of the fields of the primary and secondary
phases results in development of distinctive compositional trends as fresh rock is
progressively weathered. The trends commence at the fresh rock composition (near
the feldspar-FM join) and are directed upwards and to the right hand side of the
diagram as Ca, Na, K, Mg and possibly Fe are removed in solution in preference
to Al.
A1 2 0 3
Fig. 6.2. Compositional relations among primary and secondary minerals, including the mafic minerals,
some common igneous rock types and av. shale. The diagram emphasizes the proportion of mafic
minerals relative to feldspars. Abbreviations are given in the caption to Figure 6.1.
132 H.WNesbitt
Predicted trends using mass balances
Shaw et al., (1967), Wedepohl (1969) and Taylor and McLennan (1981) estimate
the average bulk composition of the upper continental crust (UCC) to be of gran-
odioritic composition. The three averages are similar (Nesbitt and Young, 1984) and
the average of Shaw et al., (1967) is plotted on Figure 6.3 (point 5).
Wedepohl (1969) estimates the mineralogical composition of the average granitic
rock of the upper continental crust (UCC) to be 41% plagioclase, 21% each of
quartz and K-feldspar, 6% amphibole, 4% each of biotite and pyroxene, 2% oxides
and 0.6% olivine. Quartz and the oxides are résistâtes, thus the feldspars represent
80% of the labile minerals (minerals readily susceptible to weathering). Calculation
of the mineralogy of the upper continental crust using the estimates of Shaw et al.
(1967) and Taylor and McLennan (1981) indicate that feldspars constitute 70 to 80%
of the labile minerals (Nesbitt and Young, 1984). Feldspar weathering is, therefore,
the focus of subsequent study.
Congruent dissolution offeldspars

If plagioclase dissolves congruently (dissolves stoichiometrically and without a
residue) from UCC (upper continental crust, Figure 6.3a, point 5), the effect on the
bulk composition is indicated by the vector labelled Viy which emanates from S and
is directed away from the plagioclase composition, terminating at point p (Figure
6.3a). Its magnitude (length) is determined by the amount of plagioclase in UCC.
If K-feldspar dissolves congruently, the effect on the original bulk composition
(Figure 6.3a, point S) is indicated by the vector labelled V2, which terminates at
k. If both feldspars dissolve congruently and completely, the resultant vector is the
vector sum of V\ and V2; thus the resultant is directed away from the AI2O3 apex.
However, the bulk composition does not necessarily evolve along the resultant as
the feldspars dissolve. The evolutionary path of S is controlled by the rate at which
each feldspar dissolves (kinetic constraints). If initially, plagioclase dissolves more
rapidly than K-feldspar, then the trend will be convex towards p (Figure 6.3a),
and where K-feldspar dissolves more rapidly (initially), the compositional trend is
convex towards k.
Incongruent dissolution ofplagioclase

Al immobile
As an extreme condition, consider all A1203 released during the weathering
of plagioclase to be included in a secondary phase containing, as the essential
constituents, Al, Si and H 2 0, and little or no alkalis or alkaline earths. Whether the
secondary phase is amorphous, or crystalline, is immaterial to the vector analysis
shown in Figure 6.3a. Si and H 2 0 are excluded from Figure 6.3a, consequently
Fig. 6.3. Effect of weathering of feldspars on the bulk composition of average upper continental crust
(UCC). S = average bulk composition of the upper continental crust; V\ a vector representing congruent
(stoichiometric) dissolution of plagioclase;/? = final composition of 5 after weathering of all plagioclase;
V2 a vector representing congruent dissolution of K-feldspar; k = final composition of S after weathering
of all K-feldspar; K3 a vector representing formation of an Al-silicate (devoid of alkalis or Ca), such
as Ka., Ha. or Im.; V4 a vector representing formation of illite. Vp = resultant vector of V\ and K3,
thus it represents weathering of plagioclase in S coupled with production of an secondary Al-silicate. Kk
= resultant vector of Vi and V4, and represents the weathering of K-feldspar to produce illite. Figure
6.3b illustrates the mass balance relations as a result of simultaneous weathering of plagioclase and
K-feldspar. Vp and Kk are vectors representing the incongruent weathering (residue formed) of each
feldspar to produce Al-silicate and illite. Vr is the vector sum of Vp and Kk or alternatively, the vector
resulting from removal of Ca + Na2 and K2 in the proportion indicated by A. Points is the composition
of the weathering solution derived from average upper continental crust (5) according to reaction (4).
their contents in the secondary phases do not at all affect the following analysis.
There is substantial evidence that phases containing primarily Al, Si and H 2 0,
form. Kaolinite and halloysite have been observed as secondary products of many
profiles (Markovics, 1977; Grant, 1963). Al-silicates of variable Al : Si ratio (possibly
amorphous), allophane and imogolite have been observed (Farmer, 1982; Eggleton,
1987; Tkylor, 1988; Nesbitt and Muir; 1988).
Consider, for convenience, kaolinite to form as the secondary product after pla
gioclase. Halloysite, imogolite or allophane can be used in its place and the argu
ments that follow are valid provided the phase contains little Na, Ca or K, and
effectively all Al released from plagioclase is incorporated into the secondary phase.
Incongruent dissolution (dissolution with formation of residual solid products) of
plagioclase (here albite is used) to kaolinite is then given by:
NaAlSi3Og + H+ + 4.5 H 2 0 = |Al 2 Si 2 0 5 (OH) 4 + Na+ + 2H 4 Si0 2 (aq)
albite kaolinite
134 KWNesbitt
The reaction may be viewed as the combination of two reactions, congruent

dissolution of plagioclase (albite) and formation of the secondary phase from so
lution. The amount of secondary phase precipitated is determined entirely by the
amount of Al released from the plagioclase. The vector representing the first re
action, V\y is constructed assuming that all plagioclase dissolves. Formation of the
secondary phase is represented by vector K3; Vp is the resultant vector (Figure 6.3a).
The resultant vector can be obtained by vector summation of V\ and V$. There is
an alternative, more direct construction. The resultant Vp emanates from the bulk
composition (5) and is directed away from the CaO + Na 2 0 apex because, by re
action (1), the bulk composition S loses Na+ to solution. With the direction of the
resultant, Vp known, the magnitude or length of V$ is obtained by drawing a vector
from the end of Vx (Figure 6.3a, point/?) towards the secondary mineral that forms
(kaolinite in this case), to intersect the resultant vector Vp. The point of intersection
is the composition of the residue after plagioclase has been completely removed
from the sample. The evolutionary bulk compositional trend follows the resultant
Vr Where applicable, the premise upon which Vv was calculated (Al immobile)
removes the need for much kinetic data.
Al mobilized
The vector V„ provides a good approximation to real weathering trends of plagio
clase where only a small proportion of the total Al in feldspar is released to solution.
In fact the limit to the amount of Al that may be released without modifiying the
vector properties is dependent upon the precision with which AI, Na, Ca and K
are analyzed in bulk compositions. For X-rayfluorescenceanalysis, the error in Al
is about 1-2% of quoted values. One to two percent of Al in altered plagioclase
therefore can be dissolved and removed from the profile without affecting predicted
or observed weathering trends.
The effects of release of significant amounts of Al to solution can be evaluated
by stipulating the proportion of Na/Al released to solution during dissolution of
plagioclase. Reactions analogous to (1) are written and mass balances obtained. The
resultant vector is constructed byfirstplotting on Figure 6.3a, the Na/Al value dis
solved from plagioclase (ratio is plotted on the left-hand boundary). The resultant
vector then is constructed; it emanates from S and is directed away from the Na/
Al value (on the boundary) representative of plagioclase dissolution. The extreme
case, where all Al released from plagioclase is taken into solution, is represented
by vector Vx. The resultant vector for real weathering trends falls between these
extremes with V? representative of trends produced in mildly acidic environments
and V\ more representative of trends produced in extremely acidic settings. The rate
of plagioclase dissolution relative to the rate of "precipitation" of secondary phases
determines the trajectory of the resultant vector; the trajectory is curved where rates
of these reactions vary as a function of solution composition and other weathering
conditions.
Incongruent dissolution of K-feldspar

The alteration products of K-feldspar are uncertain but illite is a common sec
ondary mineral of weathering profiles (Grant, 1963; Weaver, 1967; Nesbitt et al.,
1980). The source of most potassium incorporated into illite of profiles is derived
ultimately from K-feldspar (Nesbitt and Young, 1989). Although illite probably is
not the initial weathering product of K-feldspar, it is the final, major K-bearing
secondary product. Intermediate secondary products may form but they are incon
sequential to the mass balances (Figure 6.3a) where Al is conserved in the profiles.
The reaction is:
2.7KAlSi308 + 2H + + 9.6 H 2 0 =
K-feldspar
Ko.7Al27SÌ3 30,o(OH)2 + 2 K + + 4.8H 4 Si0 2 (aq)

illite
Illite composition is chosen arbitrarily to reflect the K deficiency and Si en

richment commonly observed for illites. Other compositions (including the proxy,
muscovite) yield the same qualitative results, hence the arguments and conclusions
that follow are unaffected by the choice of composition.
Conservation of Al in secondary illite requires the bulk of K+ to be released to
solution [equation (2)]. Reaction (2) can be viewed as the combination of two re
actions, congruent dissolution of K-feldspar (Figure 6.3a, vector V2) and formation
of illite from solution (Figure 6.3a, vector K4). The resultant is shown as Vk (Fig
ure 6.3a). Al-conservation in the secondary phases requires the bulk composition to
evolve along the resultant Fk as weathering of K-feldspar proceeds.
Some Al may be released to solution during weathering. K-feldspar weather
ing vectors can be determined where the ratio, K/Al, released to solution during
leaching, is known or can be stipulated; thus the evolutionary trend of the bulk
composition (S) can be predicted even when Al is not conserved. If all Al released
from K-feldspar is taken into solution and removed from the profile, the resultant
vector and the evolutionary trend of the bulk composition are identical to the path
for congruent dissolution of K-feldspar (Figure 6.3a, V2).
Incongruent dissolution of both feldspars
The average mineralogical composition of the UCC (upper continental crust)

contains plagioclase and K-feldspar in the molar ratio 3:1, and a normative cal
culation indicates that the plagioclase composition is 2/3 albite and 1/3 anorthite
(CaAl2Si208; Nesbitt and Young, 1984, appendix, table 2). The weathering reac
tion for anorthite component of plagioclase to kaolinite (or other Al-silicate) is (Al
conserved in the secondary phases):
136 H.WNesbitt
CaAl2Si208 + 2 H+ + H 2 0 = Al2Si205(OH)4 + Ca2+

anorthite kaolinite
The proportions of feldspar components in the UCC are, from the above data,
two parts albite and one part each of anorthite and K-feldspar (molar basis). Con
servation of Al in the secondary phases permits calculation of the "weathering
vectors" for the average UCC. Reactions (l)-(3) are combined in the proportions
2:1:1 (see above) to yield the overall reaction:
2Ab + lAn + lKsp + 4.74H+ + 13.56 H 2 0 =
2Kaol + 0.37111 + 2Na + + lCa 2+ + 0.74K+ + 5.78H4Si04(aq) (4)
The reaction may be viewed as three congruent dissolution reactions plus two
reactions by which kaolinite and illite are formed. The direction of the resultant
mass balance vector is illustrated on Figure 6.3b (Vr). Its direction is determined by
calculating the composition of the weathering solution from reaction (4). The ratio:
Na2(aq) + Ca(aq)
Na2(aq) + Ca(aq) + K2(aq)
is required to plot the weathering solutions on the feldspar diagram. Concentrations
of the monovalent ions are halved to conform to plotting parameters of the dia
grams. From reaction (4) the ratio is (2+1)/(2+1+0.37) = 0.89 (Figure 6.3b, point
Ay base of triangle). It represents the proportions of constituents extracted from
the bulk composition (5) by the weathering solution. The residues of weathering
necessarily fall along a vector emanating from point S (Figure 6.3b), and directed
away from points, as shown on Figure 6.3b (vector Vr). Thefinalbulk composition,
after complete weathering of the feldspars, is represented by the star at the end of
vector VT (Figure 6.3b).
The evolutionary path followed by the bulk composition is not restricted to the
resultant vector VT, but is dependent upon the relative rates at which plagioclase and
K-feldspar are weathered. Where plagioclase weathers more rapidly than K-feldspar
the evolutionary trend is convex towards vector Vp (Figure 6.3b). Where K-feldspar
weathers more rapidly, the bulk compositional evolutionary trend is convex towards
vector Vk. Only where release rates of constituents from the two feldspars are
similar will evolutionary weathering trends follow the vector VTy because only then
are the leach rates controlled by feldspar abundances in the rock (mass balances
alone).
The evolutionary path of 5 is, in either case constrained within the compositional
space bounded by vectors Vp and Vk provided Al is essentially conserved within the
profile.
Predicted trends using kinetic data
Kinetic data are available to predict the initial weathering trend of feldspathic
rocks in the compositional space of Figure 6.1. Bulk compositional evolutionary
trends have been predicted (Nesbitt and Young, 1984) using the kinetic data of
Busenberg and Clemency (1976) and the compositions of various types of granitic
rocks. Busenberg and Clemency (1976) demonstrated that longterm leach rates of
Na, Ca and K from feldspars follow linear rate laws:
Q = KiT + b (5)
where C/ is the concentration of constituent /, Ki is the release rate constant, T is
time and b is a constant equal to the concentration of / in solution before leaching.
The instantaneous release rate is:
—- = Ki (6)
and this holds for all constituents obeying the linear rate law. Equation (6) pro
vides the instantaneous rate of release of Ca, Na and K to solution, normalized to
unit mass of feldspar from which each element is released; consequently the bulk
instantaneous release rate (BIRR) is obtained by multiplying the mass of feldspar
containing the element i by the instantaneous release rate (Ki):
bulk instantaneous release rate of / = Ms * Ki (7)
where * is the operator, multiplication. There is, for each element released to
solution, an equation equivalent to equation (7). The ratio of the BIRR of Na, Ca
and K released to solution from the average UCC is, therefore (Nesbitt and Young,
1984):
MAb * ATNa : MAn * KQ* : MRS * KK (8)
where MAO, M\n and MKS are the total moles of albite, anorthite (in plagioclase)
and K-feldspar, respectively. The proportions of Na, Ca and K removed from UCC
during the initial stage of weathering can be calculated from equation (8) knowing
the composition of the rock (Shaw et al., 1967; Wedepohl, 1969; Nesbitt and Young,
1984) and the appropriate rate constants (Busenberg and Clemency, 1976). The cal
culations are performed as indicated by Nesbitt and Young (1984). The release rate
constants for Na+ and Ca2+ from plagioclase composed of 2/3 albite and 1/3 anor
thite, are 10" 158 and 10"16 °, respectively (Busenberg and Clemency, 1976; Nesbitt
and Young, 1984) and the release rate of K+ from microcline is 10"15 8. The three
release rate constants are similar; consequently, and as is apparent from equation
(8), the proportions of Na, Ca and K released from average upper continental crust
(granodiorite) is controlled largely by the proportion of the two feldspars rather
than by the release rate constants. Since mass balances, not kinetic constraints, are
140 H.W.Nesbitt
Fig. 6.4. Calculated (a) and observed (b) weathering trends for some crystalline rocks. Small dots are
idealized mineral compositions. The arrows illustrate the weathering trend of some average igneous
rocks, calculated from kinetic and mass balance constraints [equation (8)]. Symbols of a: large dot =
av. granite; filled square = av. adamellite; filled triangle = av. granodiorite; inverted triangle (filled)
= av. tonalité; filled diamond = av. gabbro. Open symbols represent calculated compositions of solu
tions derived from the rock with the same (filled) symbol. Symbols of b: filled squares, open squares
and dotes are fresh and weathered samples for the Stone Mountain Granite (Grant, 1963), Mazanini
Granite (Lovering, 1959) and the Toorongo Granodiorite (Markovics, 1977), respectively. Calculated
proportions of Ca, Na and K in weathering solutions from these rocks are plotted on the baseline of b,
using the same symbols. The arrows represent the calculated initial weathering trends for these rocks.
Figure is modified after Nesbitt and Young, 1984.
the major control on the release of Na, Ca and K during weathering, its evolutionary
trend is close to the resultant vector Vr (Figure 6.3b). The plagioclase composition
used above (2/3 albite component) is typical for granodioritic rocks; consequently
the weathering trends followed by granodiorites of similar composition will be close
to or along the resultant vector (Kr). Granodiorites are among the most common of
granitic rocks; consequently vector VT is representative of weathering trends devel
oped on a large number of exposed plutonic rocks. Tb emphasize this aspect note
that the weathering trend for average granodiorite (Figure 6.4a) is similar to the
resultant vector Vr of Figure 6.3b, which is calculated from mass balance constraints
alone. Vector VT is also the "typical" trend for the upper crust because the average
upper continental crust is of granodioritic composition.
The weathering of granites contrasts markedly with the weathering of granodior
ites. Calc-alkali granites (Nockolds, 1954) contain subequal amounts of plagioclase
and K-feldspar and the composition of plagioclase typically is 85-90 mole percent
albite and 10-15 percent anorthite. The release rate constants of these albitic pla-
gioclases are 10"15 ° to 10"15 4 for Na+ and 10"16 ° to 10"16-1 for Ca 2+ . The release
rate constant of K from microcline is approximately 10" 158 . The release rate con
stants for Na and K differ greatly whereas the proportions of the two feldspars are
similar. The controls on release of the cations from granites therefore reflect both
kinetic and mass balance constraints. The high release rate of Na from the albitic
plagioclase of granites implies that plagioclase decomposes more rapidly than K-
feldspar. The release rates also imply that weathering solutions draining granitic
terrain should contain more Na than K and that secondary weathering products
after plagioclase should be found associated with primary K-feldspar.
Weathering trends for average gabbro, diorite, granodiorite, adamellite and gran
ite (all compositional data from Nockolds, 1954) are predicted using kinetic con
straints [equation (8)] and are plotted on Figure 6.4a (Nesbitt and Young, 1984 for
details of calculations).
The most important aspect of the mass balance and kinetic calculations is that,
regardless of the major control on weathering, the predicted and observed trends
are sub-parallel to the (CaO+Na20)-Al2C>3 boundary.
Predicted and observed trends in A-CN-K diagrams
Evolutionary compositional trends produced by weathering of rocks cannot be

sampled directly because they form over long periods. Weathering profiles devel
oped from parental source rocks, however, can be sampled. On the assumption that
the compositional variations observed within a profile (that is, spatial variations)
mimic compositional variations through time, the predicted evolutionary trends and
compositional trends within profiles can be compared and contrasted. The com
parison may be misleading for some situations because the above assumption is
required to facilitate comparison. The compositional trends from active profiles of a
granodiorite, Tborongo Granodiorite (Markovics, 1977), and two granites, Mazaruni
Granite (Lovering, 1959) and Stone Mountain Granite (Grant, 1963), are plotted on
Figure 6.4b. The observed trends are sub-parallel to the left-hand boundary and are
similar to the predicted trends of Figures 6.3b and 6.4a. It is apparent that these sim
ple thermodynamic, mass balance and kinetic concepts are sufficient to explain the
general compositional trends observed in recent profiles (Nesbitt and Young, 1984;
1989). As well, the resultant Vx of Figure 6.3b is similar to the predicted granodior
ite trend of Figure 6.4a and to the observed granodiorite weathering trend of Figure
6.4b. It is also apparent that Al of the feldspars has been largely conserved within
secondary phases of the active profile. Furthermore, the plagioclase composition of
the Tborongo Granodiorite is close to 2/3 albite (approx. 65%); consequently, the
release rate constants of Ca, Na and K from the feldspars should be similar and
140 H.W.Nesbitt
the weathering trend should be controlled by the proportions of feldspars in the

granodiorite. In fact the observed trend (Figure 6.4b) is similar to Vx (Figure 6.3b)
confirming the generalizations made for weathering of granodiorites; mass balances
are the major control on the weathering trend of granodiorites in the feldspar di
agram. Confirmation of the predictions provides justification to proceed with the
interpretation of paleosols.
Profiles developed on volcanic rocks can be (and have been) treated in a similar
manner. As for profiles developed on granitic rocks, mass balance relations com
bined with the available kinetic data permit the prediction of evolutionary composi
tional trends developed on basalts through rhyolites. The concepts and calculations
are similar to those presented for crystalline rocks, hence are not repeated here.
They are discussed in some detail in the literature (Nesbitt and Young, 1984; 1989).
Predicted and observed trends in A-CNK-FM diagrams
As demonstrated in the previous section, compositional trends can be predicted

with reasonable accuracy by assuming that Al is conserved within secondary prod
ucts of weathering profiles. Weathering of all feldspars therefore yields resultant
vectors emanating from the parent rock composition and directed away from the
CNK apex (Figure 6.5a, vector K5, emanating from average granite). Plagioclase
and K-feldspar weathering vectors (congruent dissolution) have the same direction
in Figure 6.5, and when the two minerals are weathered simultaneously, the vectors
are additive in the compositional space of Figure 6.5. The direction and magnitude
of the vector representing formation of secondary minerals after each feldspar is
different depending upon the type and amount of secondary product formed. The
treatment of FeOy + MgO is complicated by the variable oxidation state of Fe;
consequently, FeO and MgO may behave differently, depending on the oxidation
potential in the weathering environment. If Fe(II) is oxidized to form insoluble
Fe(III) oxides and hydroxides at the site of weathering, little Fe will be removed
from the profile whereas Mg may be removed. If the oxidation potential is suffi
ciently low, Fe may remain in the ferrous state and behave much more like Mg (that
is, may be subject to leaching). The vector representing Fe and Mg removal from
samples is always directed away from the FM apex. Fe and/or Mg removal from
average granite, for example, is represented by vector V6 (Figure 6.5). The magni
tude (length) cannot be predicted without oxidation potentials and resultant vectors
cannot be accurately calculated without kinetic information; weathering trends for
these diagrams cannot be predicted with confidence.
Despite the inability to predict trends in A-CNK-FM space, there are system
atic aspects to weathering trends of active profiles (Figure 6.5b). All indicate that
proportionally more Ca, Na and K are removed from the profiles than Fe + Mg,
thus all trends, whether granites, granodiorites or basalts, are sub-parallel to the
A1 2 0 3
Fig. 6.5. Illustrates the vectors representing weathering feldspars and mafic minerals (or components)
from some average igneous rocks. Vector K5 illustrates the result of weathering feldspar from av. granite
(Nockolds, 1954) and V6 illustrates the effects of removal of Fe and Mg from av. granite. The squares
of b represent bulk compositions for samples from the recent Mazaruni weathering profile (Lovering,
1959); the dots are data from the recent Toorongo Granodiorite profile (Markovics, 1977); the stars are
data from the recent Morvern Basalt profile (Bain et al., 1980). Abbreviations are defined in captions
to Figure 6.1.
left-hand (CNK) boundary. These provide a template against which to interpret the
trends of paleosols.
Interpretation of a Paleozoic paleosol
Wahlstrom (1948) reports the study of the Boulder Creek paleosol. It is devel
oped on the Boulder Creek Granodiorite and is unconformably overlain by the
Fountain Sandstone. The weathering profile was developed during the Carbonifer
ous Period, approximately 300 million years ago. It signifies that the granodiorite
was exposed, and while exposed, weathering proceeded to 25 m depth during the
Carboniferous. The unconformity and overlying non-marine sandstone indicate that
the profile was affected by tectonism prior to burial and that burial occurred in a
continental setting. It is likely, therefore, that some of the uppermost portion of the
profile has been eroded during tectonism and/or during the period when the profile
was inundated by fresh waters. Additional details are found in Wahlstrom (1948)
and Nesbitt and Young (1989).
Although there are many aspects of this paleosol to address, only constituents
of the feldspar minerals, Fe and Mg are discussed here. The bulk compositions
of samples from the paleosol are plotted on Figure 6.6 as crosses. The unweath-
142 H.WNesbitt
U.U ^Η· "u. ui. i.u

CaO + N a 2 0 + K 2 0 FeO T +MgO
Fig. 6.6. Predicted initial bulk compositional weathering trends and observed bulk compositional trends
for the 300 million year old Boulder Creek Granodiorite profile. The thick arrow emanating from the
cross below the feldspar join is the predicted weathering trend. The crosses represent the analyses of
samples from the paleosol and the circles (on thick vector) represent bulk compositions of samples
from the recent profile developed on the granodiorite. The dots represent bulk compositions of sample
from the Tborongo Granodiorite profile; the squares, samples from the recent Mazaruni profile and
the triangles data from the recent Ricany Granite profile (Minarik et. al., 1983). Figure modified after
Nesbitt and Young (1989).
ered granodiorite composition plots just below the feldspar join, as is true for the
Tborongo Granodiorite (Figure 6.6, dots) and the average upper continental crust
(UCC) composition (Figure 6.3a, 5). It differs little from these compositions and
direct comparison alone should provide a reasonable estimate of the compositional
trend followed by the granodiorite as it is progressively weathered. The thick vector
emanating from the fresh granodiorite (Figure 6.6) is the predicted trend [equation
(8)]. The accuracy of the prediction is confirmed by additional data from Wahlstrom
(1948) in that two samples from the recent weathering profile developed on the
granodiorite plot on the predicted trend (Figure 6.6a, circles). All paleosol samples,
however, plot on the K20-rich side of the predicted trend. An hypothetical vector
(Figure 6.6a) may be constructed where only Ca and Na are leached from the profile
(all K 2 0 originally present in the fresh rock is conserved in secondary phases). Even
for this extreme circumstance, only one of the samples from the paleosol plot on
or close the the hypothetical vector. All remaining samples contain excessive K 2 0.
The proportions of A1203 : (CaO + N 2 0) : K 2 0 in the paleosol samples cannot be
explained simply by leaching of the parent granodiorite; either large amounts of
A1203 were leached from the profile (while K 2 0 remained conservative) or K 2 0
has been added to the profile (K-metasomatism). Most likely potassium metasoma-
tism of the profile has occurred during and/or after burial diagenesis (Nesbitt and
Young, 1989).
The nature of the metasomatic reactions are known. Most recent mature weather
ing profiles contain substantial amounts of kaolin-group minerals, allophane, imog-
olite or other Al-silicates (Grant, 1963; Farmer, 1982) as alteration products of
rapidly weathered plagioclase. Plagioclase is the most abundant mineral of fresh
Boulder Creek Granodiorite and like recent profiles, the Boulder Creek paleosol
certainly contained large amounts of secondary Al-silicates. These Al-silicates now
represent only a very minor portion of the Boulder Creek paleosol. Illite, by con
trast, is exceedingly abundant, and much more abundant than would be expected for
a granodiorite profile (too much K2O), and much more abundant than observed in
recent profiles. The secondary products of plagioclase, such as the kandites, proba
bly have been converted to illite by reaction with groundwaters, the reaction having
taken place during or after burial (Nesbitt and Young, 1989; Rainbird et al., in
press):
Al2Si205(OH)4 + K+ = KAl3Si3O10(OH)2 + H+
halloysite, imogolite illite
Illite may have formed as a separate phase, may be interstratified with other sec
ondary phyllosilicates, or it may be reconstituted vermiculite; regardless, recognition
of the diagenetic reactions, based on bulk compositional data, permits evaluation of
the original mineralogy and of the diagenetic reactions which have affected the pa
leosol. Since the textures, mineralogy and bulk compositional trends of the original
profile are similar to recent profiles there is good evidence that the weathering pro
cesses which produced the Boulder Creek paleosol were similar to those operative
now, at least with regards to the feldspar components. Moreover, there is no in
dication that Pennsylvanian atmosphere or soil water composition was significantly
different from those of today. The anomalous mineralogie and bulk compositional
properties of the paleosol likely result from diagenesis; they do not reflect unique or
substantially different soil forming processes in the Pennsylvanian, some 300 million
years ago.
The data for the Boulder Creek paleosol are also plotted on Figure 6.6b, as are
trends for the Tborongo Granodiorite and the Mazaruni and Ricany Granites. The
trend for the paleosol is not systematic and is difficult to interpret by any known
weathering processes. Most likely there has been metasomatism of the profile with
introduction of Fe and/or Mg, in variable amounts, to different parts of the profile.
The addition probably occurred during diagenesis and perhaps as K-metasomatism
proceeded. The reactions may include incorporation of Mg and Fe into illite of the
profile, and possibly reconstitution of partially weathered biotites (vermiculite and/
or hydrobiotite).
144 H.WNesbitt
Interpretation of three Precambrian (Huronian) paleosols
Introduction
Weathering profiles developed from Precambrian rocks, particularly from Ar

chean and Early Precambrian rocks have received attention (Holland, 1984; Gay
and Grandstafï, 1980) because they may provide insights into the early evolution
of the atmosphere. One of the most interesting of these aspects is the abundance
of oxygen. Profiles developed on Archean rocks and preserved at the unconformity
between the Archean and Huronian sedimentary rocks of northern Ontario, Canada
have been investigated by Gay and Grandstafï (1980) and Rainbird et al. (in press).
Diagenesis of the Ville Marie and Pronto Mine Granite paleosols are examined
here.
Ville Mane paleosol
A well developed unconformity separates the Huronian Lorrain Formation from

the underlying Archean basement 3 km northwest of Ville Marie, Quebec, Canada.
The sedimentary rocks immediately overlying the granite contain fragments of the
partially decomposed granite, as well as material derived from local granitic sources.
The Ville Marie paleosol is developed on Late Archean (Algoman) Ville Marie
Granite (Rainbird et al., in press). It contains sub-equal amounts of quartz, pla-
gioclase and perthitic microcline. Rainbird et al. (in press) demonstrate that the
textural features of the altered granite are similar to textures observed in the re
cent profile developed on the Tborongo Granodiorite (Markovics, 1977; Nesbitt
et al., 1980). Compositions of samples from the profile are plotted on Figure 6.7
(squares, arrow through data indicates up-profile sequence), along with the data for
the Tborongo Profile (Figure 6.7, triangles). The predicted weathering trend for the
Ville Marie profile is sub-parallel to the left-hand boundary of the A-CN-K dia
gram (Figure 6.7, thick vector). The bulk compositional data for the Ville Marie pa
leosol plot at K2O values greater than predicted. Rainbird et al., (in press) observe
abundant illite in thin section and a dearth of kandites (kaolinite group minerals),
which are common in recent profiles, including the Toorongo profile (Nesbitt et al,
1980). Rainbird et al. (in press) argue that kandites were altered to illite during
diagenesis, a result of K-metasomatism during burial (Nesbitt and Young, 1989). K-
metasomatism was promoted by introduction of K(aq) to the profile by continental
groundwaters circulating through permeable horizons, including the unconformity
at the base of the Huronian sedimentary sequence.
Calcite is observed in thin section near the base of the profile and the bulk
compositional data display an anomalous increase in CaO precisely where calcite
is observed. The anomaly is absent from the recent profile and probably does not
result from normal weathering processes (Nesbitt and Young, 1989; Rainbird et
C a 0 + N a 2 0 + K20 FeO T +MgO
Fig. 6.7. Illustrates the compositional trends of the Ville Marie Granite paleosol (filled squares; Rainbird
et al., in press). The thick arrow is the predicted weathering trend and the arrows branching from the
predicted trend illustrate the effect of adding K 2 0 to the weathering profile. Diagram modified after
Rainbird et al. (in press). Triangles represent samples from from the recent Toorongo profile.
al., in press). The calcite and the CaO anomaly are interpreted by Rainbird et
al. (in press) to be of early diagenetic origin, the carbonate having formed at the
groundwater table during the period when the profile was forming (Nesbitt and
Young, 1989).
The data for the Ville Marie paleosol and the Tborongo profile are plotted on A-
CNK-FM diagram (Figure 6.7b). The trend shows no unusual features. Specifically,
Fe-Mg metasomatism of the profile is not apparent in this compositional space. Fe
of the Ville Marie profile apparently has behaved much like Fe in the Tborongo
profile, where it is rapidly oxidized and is immobilized (Nesbitt, 1979; Rainbird
et al., in press). Apparently, the oxidation potential within the Ville Marie soils
was sufficiently high to convert Fe(II) of the primary mafic minerals to Fe(III)
phases. The conversion was sufficiently rapid that little Fe was lost to solution
before oxidation. The oxygen fugacity equivalent to the required oxidation potential
is between 10"55 and 10"70, certainly very low values.
The compositional trends of the Ville Marie paleosol are strikingly similar to
the trends of the Boulder Creek Granodiorite paleosol (compare Figures 6.6a and
6.7a). Both paleosols contain little kaolin-group minerals and abundant illite; K-
metasomatism of each weathering profile has resulted in the conversion of kandites
to illite, thus the intermediate portions of both profiles are anomalously enriched
in K 2 0. Carbonate minerals are found in restricted zones of both profiles suggest
ing that they formed from groundwaters (capillary zone) during the formation of
146 H.WNesbitt
the profile. The striking similarities suggest that there was little difference in the
chemical conditions giving rise to weathering of granitic rocks or to the conditions
prevalent during early and burial diagenesis of the profiles. The comparison, in fact,
indicates that the chemical and environmental conditions were essentially the same,
despite the very long period separating the formation of the two profiles (approx.
2 x 109 years).
Pronto Mine Granite paleosol

The paleosol developed on the granite northwest of the Pronto Mine was sam
pled to a depth of 9 m. The bulk compositional data for the profile (Gay and
Grandstaff, 1980, table II) have been plotted on the feldspar and mafic diagram
(Figure 6.8a) as large dots (line segments connecting samples indicate up-profile
sequence). The two deepest samples (9 and 7 m depth) plot close to the feldspar
join, indicating that they are little weathered. The predicted weathering trend for
the granite is sub-parallel to the left-hand boundary of the diagram, as indicated by
the thick arrow (vector). The trend of the deepest four samples (four samples plot
ting closest to the feldspar join, Figure 6.8a) do not follow the predicted trend, but
define a trend of progressive K 2 0 enrichment. If the trend were a result of weath
ering, Na and Al would have to be removed from the freshest samples in a molar
ratio of 4:1 and a weight ratio (Na 2 0 : A1203) of 7:3. Obviously, large amounts of
Fig. 6.8. Illustrates compositional trends of the paleosol developed on an Archean granite near the
Pronto Mine (large dots; Gay and Grandstaff, 1980), and compositional trends in some recent profiles
(triangles = Toorongo Granodiorite profile; circles = the Ricany Granite profile, Bohemian Massif;
stars = Mazaruni Granite profile, New Guinea). The thick arrow is the predicted weathering trend for
the Archean granite.
Al would have to be removed to explain this trend. Although Al may be leached

under extreme weathering conditions such as in acidic environments, development
of such a trend in this mature profile is unusual because such mobility has not been
observed in recent, mature weathering profiles. Furthermore, the above ratio as
sumes that only plagioclase is affected by leaching. If the solutions were so acidic
that there was substantial leaching of Al from the profile, the aggressive solutions
certainly would have altered appreciable amounts of K-feldspar as well, thus deplet
ing the profile in K2O. In fact K2O remains very high in all samples (Figure 6.8a).
As well, the samples near the right-hand boundary plot between muscovite (sericite)
and K-feldspar. Since the profile contains little or no biotite (Gay and Grandstaff,
1980), but contains abundant K-feldspar, feldspar must have been generated in the
profile to account for the increased proportion of K 2 0 in the middle portion of
the profile. The reaction is most likely diagenetic where K has been introduced
to the profile by groundwaters during or after burial. As well, K-metasomatism
by continental groundwaters is common (Nesbitt and Young, 1989), hence the di
agenetic products and the overlying Matinenda Sandstone may be of continental
origin (Nesbitt and Young, 1989). The data (Figure 6.8a) display compositional
trends similar to those of the Pennsylvanian Boulder Creek paleosol and the Ville
Marie paleosol. Apparently the three paleosols have been affected by the same
processes.
The data for the Pronto profile are also plotted on Figure 6.8b (large dots).
Tïends for the Tborongo Granodiorite (triangles), the Mazaruni Granite (stars) and
Ricany Granite profiles (circles) are included for comparison. The Pronto trend is
much different from the recent trends and is most simply explained by introduc
tion of the components FeOr + MgO to the profile. The alternative explanation
requires AI2O3 and (CaO + Na2Û + K 2 0) to be leached from the profile in the ap
proximate ratio 3:2. Solutions responsible for removal of these elements, whether
weathering or diagenetic in origin, would be unusual and have not yet been ob
served either in the diagenetic environment (sedimentary basins) or associated with
mature weathering profiles. Most likely, the trends in Figure 6.8a and b are the re
sult of metasomatism during diagenesis, with K and Fe + Mg added to the profiles.
If correct, the ferric to ferrous iron ratio of the Pronto profile may reflect burial
diagenetic rather than the original chemical weathering conditions.
Although the trends observed in Figure 6.8 are different from recent trends, they
are similar to the trends of the 300 million year old Boulder Creek paleosol (Figure
6.6). Both profiles probably have been affected by K, Mg and Fe metasomatiza-
tion during diagenesis. Their similarity of trends and their distinct differences from
recent profiles developed on granites strongly implies that a common process or
processes affected both the Pennsylvanian Boulder Creek and the Pronto profile.
Furthermore, although the paleosols display different trends from recent profiles, it
has already been concluded that the Boulder Creek weathering profile (Figure 6.6)
developed in a manner similar to recent profiles (and subsequently was altered by
148 H.W.Nesbitt
diagenesis); extension of the arguments suggests that the Pronto Mine weathering
profile also developed in the same way as the Pennsylvanian and recent profiles.
Denison Mine basaltic paleosol
A final weathering profile developed on volcanic material is presented. Although

the mineralogy of secondary phases may be bewildering in their range of assem
blages and compositions, the bulk compositional trends observed in recent profiles
display simple, readily interpreted compositional trends (Nesbitt and Young, 1989).
The compositional trends within three recent profiles developed on basalts from
A1 2 0 3
1.0 AKa.,Gi.
0.0^ Bi. 1.0 0.0 Bi. l.o

CaO*+Na 2 0-fK 2 0 Fe0T+Mg0 Ca0*+Na 2 0+K 2 0 FeOT+MgO
Fig. 6.9. Recent weathering trends of basalts from different climatic zone. The Casino profile (Craig
and Loughnan, 1964) has developed in a sub-tropical climate, the Baynton profile (Wilson, 1978) in a
semi-arid, warm temperate climate and the Morvern profile (Bain et al., 1980) in a very humid, cool
temperate climate.
A1203
u.u ^μ· "u· I-"· i.u

CaO+Na20-fK20 FeO T +MgO
Fig. 6.10. Compositional trends within the Denison Mine basaltic paleosol (Gay and Grandstaff, 1980).
The trends are dramatically different from trends of recent profiles (Figure 6.9) but they are similar to
the trends of the Hekpoort Basalt profile (Nesbitt and Young, 1989). Note that "freshest" basalt plots
at approximately 73% AI2O3 whereas unweathered basalts plot near 40% AI2O3. The figure is modified
after Nesbitt and Young, 1989.
different climatic regimes are illustrated in Figure 6.9a and b. The Casino profile
is developed in a subtropical climate, the Baynton profile in a semiarid, warm tem
perate climate and the Morvern profile in a very humid, cool temperate climate.
The three profiles are very similar. The Denison Basalt paleosol is plotted on Figure
6.10 and the trends have no relationship to the recent trends. The "parent" basalt
plots near 70-75% A1203 whereas continental (and mid-ocean ridge) basalts plot
below the feldspar join of Figure 6.10. Obviously, the "parent" is highly altered,
probably as a result of combined weathering and diagenesis. The data of Figure
6.10a do not define a weathering trend sub-parallel to the Al-CNK boundary; all
but three samples plot near the illite field. The Denison samples plot between chlo
rite and illite compositions in Figure 6.10b and again display a trend dissimilar
to the recent trend. The Dension trends are remarkably similar to the Hekpoort
trends (Nesbitt and Young, 1989; Button, 1973) and the interpretation which were
offered for the Hekpoort trends (Nesbitt and Young, 1989) also explain the Denison
trends. After formation of the weathering profile, which includes weathering of the
"parent" material at 30 m depth, the entire profile has been subjected to extensive
K-metasomatism. Mineralogical changes accompanying K-metasomatism are con
version of kandites to illite (sericite). There is evidence for Fe-Mg-metasomatism.
As shown on Figure 6.10b, three samples plot close to the FM apex. If recent trends
can be used as a template (Figure 6.9), Fe and/or Mg has been added to the three
samples, shifting their bulk composition towards the FM apex. Fe and Mg most
150 H.W.Nesbitt
likely have been added to the other samples as well, but the effects are not imme
diately apparent on Figure 6.10b. The trends for the Denison Mine basaltic profile
and the Hekpoort trends yield convincing evidence that these paleosols suffered
extensive metasomatism and metamorphism. The Denison profile is without a rea
sonable "parent" from which compositional changes resulting from weathering can
be deduced. Furthermore, neither the Hekpoort nor the Denison profile has been
sampled in the detail required to define the important compositional trends accom
panying the initial stages of weathering. Additional, much more detailed studies of
these profiles are required before the genesis of these trends can be confidently
explained.
Certainly, there is good reason to question deductions about the nature of an
cient atmospheres which are based on ancient profiles that have undergone ex
tensive metasomatism during diagenesis and metamorphism. This is especially true
where the effects of diagenesis and metamorphism have not been given thorough
consideration or where the effects are still poorly understood. Greater insights into
these processes, however, eventually may result in the extraction of valuable in
formation about the compositions of ancient atmospheres from these and other
profiles. There remains the immediate problem of understanding better, diagenesis
and metasomatism of weathering profiles.
Summary
Feldspar minerals constitute 70-80% of the labile minerals of the upper crust.
Knowledge of the weathering of these minerals therefore permits prediction of
gross composition trends in weathering profiles developed from feldspathic rocks.
Leach rates for Ca and Na from plagioclase and K from K-feldspars are known and
can be used to predict the relative rates at which feldspars are weathered in granitic
rocks. The leach rates of Ca and Na from plagioclase, and K from K-feldspars,
of most granodiorites are similar so that mass balance considerations alone can
be used to predict the bulk compositional of granodiorites as they are weathered.
Kinetic considerations as well as mass balances are required to accurately predict
the weathering trends of granites. Although predicted and recent weathering trends
are similar, trends observed in paleosols commonly are different from either pre
dicted or observed trends. The differences most likely result from metasomatism of
the profiles during and after burial. Diagenetic reactions, particularly the effects of
metasomatism, recognized and carefully considered when deducing the conditions
under which the original profile formed.
Comparison of the Carboniferous Boulder Creek Granodiorite paleosol and the
Precambrian Ville Marie paleosol suggests strongly that the conditions under which
the weathering profiles formed were similar. Furthermore the diagenetic conditions
affecting each profile were similar. K-metasomatism, resulting in the conversion of
Al-silicates to illite has occurred, has produced anomalously high K2O values in the
intermediate portions of each profile. Carbonate minerals are found in each pale-
osol. The restricted zone over which they are found and textural relations suggest
that the carbonate has formed early in each profile and are interpreted to form in
the capillary fringe of a water table which existed during formation of the profile.
The carbonates are interpreted to be early diagenetic and markers of ancient water
tables. The similarity between the Boulder Creek and Ville Marie paleosols strongly
implies that the compositional and other environmental conditions were almost
identical during formation of both paleosols, including the conditions prevalent dur
ing weathering and diagenesis. The results of the comparison provide no evidence
that there has been substantial change to the weathering or diagenetic environments
over the period of 2 x 109 years.
There is persuasive evidence that Fe and Mg has been introduced to some of
the paleosols studied here. The kinetic data generally are not available to predict
weathering trends in A-CKN-FM compositional space, but trends of mature re
cent profiles indicate that systematic compositional trends are developed during
weathering. Extensive metasomatism and metamorphism have masked the original
weathering trends of some profiles. Extreme care must be taken where deductions
are made based on Fe and Mg trends. They may be affected substantially by dia
genesis and their abundance and distribution in profiles may reflect diagenetic and
metamorphic conditions rather than conditions which prevailed during weathering.
Additional studies into diagenesis, however, will result in a better understanding of
its affects, and possibly provide the insights required to extract information about
ancient atmospheres (and other environmental aspects) from these paleosols.
References
Bain, D.C., Ritchie, RE, Clark, D.R. and Duthie, D.M., 1980. Geochemistry and mineralogy of weath
ered basalt from Morvern, Scotland. Mineral. Mag., 43: 865-872.
Busenberg, E. and Clemency, C.V., 1976. The dissolution kinetics of feldspars at 25°C and 1 atm CO2
partial pressure. Geochim. Cosmochim. Acta, 40: 41-49.
Button, A., 1979. Early Proterozoic weathering profile on the 2200 m.y. old Hekpoort Basalt, Pretoria
Group, South Africa. Geol Res. Unit, Info. Cira, 133, Johannesburg, S.A., 19 pp.
Clarke, EW., 1924. The data of Geochemistry. U.S. Geol. Surv., Prof. Pap., 770, 841 pp.
Craig, D.C. and Loughnan, EC, 1964. Chemical and mineralogical transformations accompanying the
weathering of basic volcanic rocks from New South Wales. Aust. J. Soil Res., 2: 218-234.
Eggleton, R.A., 1987. Non crystalline Fe-Si-Al-oxyhydroxides. Clays Clay Miner., 35: 29-37.
Farmer, V.C., 1982. Significance of the presence of allophane and imogolite in podzol Bs horizons for
podzolization mechanisms: a review. Soil Sci. Plant Nutr., 28: 571-578.
Feakes, C.R. and Retallack, G J., 1988. Recognition and chemical characterization of fossil soils devel
oped on alluvium: a late Ordovician example. In: J. Reinhardt and WR. Siglio (Editors), Paleosols
and Weathering Through Time: Principles and Applications. Geol. Soc. Am., Spec. Pap., 216: 35-48.
Gay, A.L. and G rand staff, D.E., 1980. Chemistry and mineralogy of Precambrian paleosols at Elliot
lake, Ontario, Canada. Precamb. Res., 12: 349-373.
152 KW. Nesbitt
Grant, W.H., 1963. Weathering of Stone Mountain Granite, In: E. Ingersol (Editor), Clays and Clay
Minerals. Pergamon, Oxford, pp. 65-73.
Holland, H.D., 1984. The Chemical Evolution of the Atmosphere and Oceans. Princeton Univ. Press,
Princeton, N.J., 582 pp.
Karathanasis, A.D., 1988. Compositional and solubility relationships between aluminum-hydroxy inter-
layered soil-smectites and vermiculites. Soil Sci. Soc. Am. J., 52: 1500-1508.
Lovering, T.S., 1959. Significance of accumulator plants in rock weathering. Geol. Soc. Am. Bull., 70:
781-800.
Markovics, G., 1977. Chemistry of weathering of the Toorongo Granodiorite, Mt. Baw Baw, Vic. Hon
ours Thesis, Latrobe Univ., Melbourne, Vic, 36 pp. (unpublished).
Minarik, L., Absolon, K., Zdislava, K. and Klecka. M., 1983. Chemical changes of granite during weath
ering, In: S.S. Augustithis (Editor), Leaching and Diffusion in Rocks and their Weathering Products.
Theophrastus, Athens, pp. 293-306.
Nesbitt, H.W, Markovics, G. and Price, R.C., 1980. Chemical processes affecting alkalis and alkaline
earths during continental weathering. Geochim. Cosmochim. Acta, 44: 1659-1666.
Nesbitt, H.W. and Muir, I., 1988. SIMS depth profiles of weathered plagioclase and processes affecting
dissolved Al and Si in some acidic soil solutions. Nature, 334: 336-338.
Nesbitt, H.W. and Young, G.M., 1984. Prediction of some weathering trends of plutonic and volcanic
rocks based on thermodynamic and kinetic considerations. Geochim. Cosmochim. Acta, 48: 1523-
1534.
Nesbitt, H.W. and Young, G.M., 1989. Formation and diagenesis of weathering profiles. J. Geol., 97:
129-147.
Nockolds, S.R., 1954. Average chemical compositions of some igneous rocks. Geol. Soc. Am. Bull., 65:
1007-1032.
Rainbird, R.H., Nesbitt, H.W. and Donaldson, J.A., in press. Formation and diagenesis of a sub-
Huronian saprolith: comparison with a modern weathering profile. J. Geol.
Shaw, D.M., Reilly, G.A., Muysson, J.R., Pattenden, G.E. and Campbell, F.E., 1967. An estimate of the
chemical composition of the Canadian Precambrian Shield. Can. J. Earth Sci., 4: 829-853.
Taylor, R.M., 1988. Proposed mechanism for the formation of soluble Si-Al and Fe(III)-Al hydroxy
complexes in soils. Geoderma, 42: 65-77.
Taylor, S.R. and McLennan, S.M., 1981. The composition and evolution of the Continental crust: rare
earth element evidence from sedimentary rocks. Philos. Trans. R. Soc, London, Ser. A, 301: 381-
399.
Wahlstrom, E.E., 1948. Pre-Fountain and recent weathering on Flagstaff Mountain near Boulder, Col
orado. Geol. Soc. Am. Bull., 59: 1173-1190.
Weaver, CE., 1967. Potassium, illite, and the oceans. Geochim. Cosmochim. Acta, 31: 281-296.
Wedepohl, K.H., 1969. Handbook of Geochemistry, Springer, New York, N.Y., 1: 247-248.
Wilson, R.E., 1978. Mineralogy, petrology and geochemistry of basalt weathering. Unpublished B.Sc
Thesis, Dep. Geol., Latrobe Univ., Melbourne, Vic
155
Chapter 7
Soils of Alpine mountains

J.R LEGROS
Introduction
The mountains of Alpine fold belts have high relief and sharp topography. Con
sequently, altitude-related climate zones are well developed and are reflected in a
sequence of vegetation zones. As Dokuchayev (1900) first pointed out, this zonation
resembles the global zoning from the equator to the poles. Thus variations in Alpine
belts may serve as a model for examining the general, global effect, of climatic
change on the weathering of rocks.
Defining the limits in the study of soils of Alpine mountains is not easy. One may
choose a geological frame of reference and limit the study to ranges of Alpine or
Himalayan-type orogenesis. Or one may use climatic and phytosociologic definitions
and consider only regions of high altitude, characterized by coniferous forests and
Alpine meadows. These limits sometimes seem too restrictive, and since it is often
instructive to compare different situations, reference will be made here to geologic
structures and zones of lower altitudes that lie outside such frames of reference.
In the first part of this chapter, I will examine the general features of altitudinal
zonality, though without ignoring departures from the general case. In the second
part, I will present mechanisms of pedogenesis, which, without being specific to
mountains, take on a particular importance in regions of higher altitude. In the last
part, I will describe the soils of various geological and geomorphological systems. I
will show that, despite the heterogeneity of climate and vegetation, the parent mate
rial defines natural regions that are clearly identifiable in the field, so that it forms a
good organizational basis for the study of the genesis and spatial organization of the
Earth's soil mantle.
156 J.R Legros
Altitude belts according to climate, vegetation and soils
General outline
For a given latitude and longitude, the mean temperature of the air decreases as
the altitude increases, by about 0.5 to 0.6°C per 100 metre (Table 7.1).
There are variations depending on a number of factors. The temperature gradi
ent is slightly higher in equatorial regions (0.7°C per 100 m in Ethiopia), it is more
pronounced in summer than in winter, it is not rigorously constant from one altitude
zone to another for a given mountain (Escourrou, 1980), and it may be reversed
during the day to give a temperature inversion.
Paralleling the decrease in temperature, rainfall generally increases with altitude.
This is related to the fact that mountains constitute an obstacle obliging air masses to
flow upward, with consequent cooling and condensing of their moisture (Tkble 7.2).
This effect is far from consistent. In high mountains, condensation occurs well
before air masses attain the summit, leading to a maximum rainfall at some lower
point. Summits are particularly arid in the tropics. The top of Mount Cameroon
is much drier that its foothills, for example. By contrast, in temperate zones the
increase in precipitation with altitude is more regular and may persist to the summit.
These variations in temperature and rainfall are accompanied by other equally
important phenomena, such as variations in relative humidity, freezing days and
mean snow fall.
All this explains why the various plant species separate into groups having the
same affinities at various altitudes where their optimum climatic conditions are
available. This permits the distinction of altitudinal belts (zones) of vegetation (Tk
ble 7.3). In the French and Swiss Alps, these belts are arranged as follows (Richard
and Pautou, 1982; Ozenda, 1985).
Hilly Belt, characterized by oak-hornbeam mixed forest (Quercus robur, Carpinus
betulus).
Mountain Belt where, as the altitude rises, Fagus sylvatica forest gives way to
silver fir (Abies alba), maple (Acer campestris) and Norway spruce (Picea abies). In
general, this area corresponds to the fog belt, that is, having a high humidity but
little frost.
Sub-Alpine Belt characterized by conifers adapted to higher altitudes, Norway
spruce, larch (Larix decidua) and arola (Pinus cembro). At the tree line, beyond
which trees can no longer live (annual isotherm of 3 to 4°C), the sub-Alpine belt
grades into a characteristic heath-like landscape of dwarf shrubs of rhododendron
(Rhododendron ferrugineum) or juniper (Juniperus nana).
Alpine Belt primarily of Alpine meadow, virtually a continuous carpet of grasses,
with a very short growing season and abundant snow.
Nival Belt, discontinuous plant cover, principally of clumps of dicotyledons, cryp
togams and, higher up, thallophytes.
Soih ofAlpine mountains 157
TABLE 7.1
Soil thermal regime in the mountains of western Canada

Evaluation MASTa Days > 5°C
460 10.5 240
610 9.9 239
760 8.9 227
910 7.6 212
1060 6.8 201
1220 5.9 186
1370 4.8 166
1520 3.7 143
1670 3.1 99
1820 2.6 73
a
Mean Annual Soil Temperature.
b
Soil thermal period (days) greater than 5°C.
Source: Void and Marsh (1981).
TABLE 7.2
Windward and leeward rainfall according to altitude

Altitude Windward Leew
(m) (mm) (mm}
0 720
100 860
200 1000
300 1150
400 1300 1270
500 1450 1360
600 1580 1450
700 1700 1540
800 1800 1630
900 1890 1700
1000 1980 1770
1100 2060 1830
1200 2130 1890
Value of 100 m gradient 117 77
Source: Escourrou (1980).
This vertical zonality, distinguished by different species, is found in most of the

world's mountains. However, these belts depend less on altitude, than on tempera
ture. The upper and lower limits for a given belt drop as the latitude rises toward
the colder regions of the planet. A difference of one degree latitude will cause the
altitude of a belt to drop by around 90 to 110 meters (Ozenda, 1985). Thus, in the
north of Scandinavia, the Alpine belt descends to the ocean shore. By contrast, hilly
belt vegetation is found at 2000 m in North Africa.
158 IR Legros
TABLE 7.3
Vegetation belts of the western Alps
Both soil and vegetation are functions of climate, so that soil, vegetation and
climatic zones are generally congruent (Souchier, 1984). However, since soil de
pends also on vegetation and vegetation on soil, there is much feedback and the
relationship is not a simple one.
The variation of soil type with altitude is shown in Table 7.4. Here, acidic parent
materials have been chosen, since soils on carbonate parent rocks do not show much
variation with altitude (Georgiev, 1985). There is a general similarity in the nature
of the zones worldwide, at least above the hilly belt. Thus poorly differentiated
soils (Leptosols, Rankers) characterise the highest altitudes, with soils of podzolic
tendency below them. Only in the hilly belt and below do differences related to lati
tude show up: Ferralsols occur in the tropics, Chromic Cambisols in Mediterranean
regions and Cambisols in temperate zones.
Soils ofAlpine mountains 159
TABLE 7.4
Simplified relationships between soils and vegetation belts in various mountains of the northern hemi
sphere
Belts French Alps (Mont-Blanc) Carpathian Mountains The Himalayas (Nepal)

granitic parent materials on silty-clay mica on phyllites, schists and
ceous deposits gneiss
(from Legros and (from Souchier, 1979) (from Bottner, 1984;
Cabidoche, 1977) Righi and Lorphelin, 1986)
Sub-nival and 2600m 4300 m
Alpine belts Sparse Alpine meadow Alpine meadow Alpine meadow
Dystric Leptosol, Dystric Cambisol, sol Umbrie Leptosols,
Ranker alpin (on rocks) brun alpin (on screes) Ranker alpin, Cryumbrept
Transition 2300 m 2300m 3800 m
between Dwarf shrubs (Vaccinium Man-made meadow Dwarf shrubs
Alpine and uliginosum)
sub-Alpine belts Different podzolic soils Podzolic soils Thin humo-ferric
rich in humus Podzol
Upper zone of 2200 m 2200 m 3500 m
sub-Alpine belt Shrubs Shrubs (Pinus mughus Shrubs (Rhododendron
(rhododendrons) and Rhododendron antropogon, Rhododendron
kotschyï) campanulatum)
Thin humo-ferric Podzol Podzol Humo-ferric Podzol
Intermediate 1900 m 1850 m 2900 m
zone of Larches with Transitional zone Tsuga dumosa
sub-Alpine belt rhododendron undergrowth
Humo-ferric Podzol Podzol or
Cambric Podzol
Lower zone of 1700 m 1700 m
sub-Alpine belt Spruces with bilberry Spruces
Cambic Podzol, sol ocre Cambic Podzol, sol Cambic Podzol, sol
podzolique, ocre podzolique, ocre podzolique,
Umbrie Dystrochrept Umbrie Dystrochrept Dystrochrept
Upper zone of 1300 m 2800 m
mountain belt Spruces with grassy Spruces Abies spectabilis
undergrowth
Dystric Cambisol, Cambic Podzol, sol Umbrie Cambisol
sol brun ocreux ocre podzolique,
Umbrie Dystrochrept
Intermediate 1400 m
zone of Beech and silver fur Rhododendron
mountain belt mixed forest arboreum
Eutric Cambisol Umbrie Cambisol
Lower zone of 700 m
mountain belt Beech forest Lyonia ovalifolia
Eutric Cambisol Dystric Cambisol
Note: Soil terminology according to FAO, French and/or American classifications.
160 J.R Legros
Altitude (meters)
5 000 -+
Alpine meadows in dry environment
4 500 - |
4 000
3 500 H
3 000
2 500 H
2 000 H
1500 H
1000
500 H
Organic matter content (metric ton/ha)

Fig. 7.1. Relationship between altitude and organic matter content in soil (from Dobremez, 1979)
The organic content of soil is clearly related to altitude (Figure 7.1). At high
altitude, content is low due to biological productivity. A maximum concentration is
reached at intermediate altitudes since biological productivity is relatively high and
microbial breakdown relatively low (Teoharov and Ninov, 1980). At low altitude,
heightened microbial activity again produces a low organic content.
Varìations in the general pattern
The general pattern of zonation may be disrupted by a number of factors. Some

are directly related to climate (Samie et al, 1981; Choisnel, 1986; Douguedroit,
Soib ofAlpine mountains 161
1986), for example: (a) the "foehn effect" (rain shadow), which distinguishes the
very wet windward from the very dry leeward, especially in the Himalayas; (b) the
exposure, which implies major differences in terms of the amount of solar energy
received (Durand and Legros, 1981; Legros et al., 1986) and causes considerable
variation in the atmospheric water demand, resulting in significant variation in the
climatic conditions available to the vegetation (Fleury, 1986) and to the soil (Legros,
1975; Archer and Cutler, 1983; Dambrine, 1985); and (c) the snow cover, which has
a protective effect. In the southern Rocky Mountains, the soils of windy sites which
have their layer of insulation removed too soon and are chilled in respect to contact
with the atmosphere, have a mean annual "handicap" of 2.9°C (Burns, 1985).
There are other factors which disrupt the general arrangement of altitudinal
zones which are not directly related to the climate, such as (a) topography, which is
responsible for local concentration of water retention and may lead to development
of bogs (Legros and Cabidoche, 1977), and (b) the pétrographie nature of materials,
their texture and structure (scree accumulation, glacial till).
Pedologica! processes in mountainous terrain
All soil processes are active on mountains including the accumulation of organic
matter, the production of complexing acids, the weathering of minerals, and the
liberation of iron and aluminum and often their accumulation in spodic horizons.
However, I will only deal with the processes which are most typical of mountain
sites.
Rockfragmentationand soil functioning
High latitude regions are characterized by low temperature and permafrost. By

contrast, there is no permafrost in mountainous areas, and the soils are subjected
to considerable fluctuation in temperature between the day and night (Birse, 1980).
Up to 165 freeze/thaw cycles per year were counted in the French Alps at 2500 m.
This causes fragmentation of the bare rocks and the formation of screes. Thus, in
the French Alps, most of the glacial valleys are draped by a heavy layer of debris
(Figure 7.2). Moraines result when debris of this type has been transported by a
glacier.
Much work has been done on the processes of rock fragmentation, and formation
of the finer components (Legros and Pedro, 1983; Lautridou, 1984). Here, the effect
of freezing on the nature and functioning of soils is analyzed.
First, it should be noticed that most soils are not to be found on the solid rock
mass itself, but on broken down and transported products (scree accumulation,
glacial till). Fragmented substratum permits water percolation and has greater min
eral surface, hence promotes faster soil formation (Cabidoche, 1979).
162 J.R Legros
Crumbling
rock faces
u - shaped glacial valley
Fig. 7.2. Profile of valleys with scree and morainic mantles.
The importance of particle size should be emphasized. Soils in cold areas (moun
tains, northern regions) are statistically much richer in silt than soils in warm regions
(Africa). This has been verified using the soil data banks of Canada and France (Du-
manski and Legros, pers. commun.). Freezing breaks down the coarser elements to
silt size without changing the mineralogy of particles. Clays form far more readily in
hot than in cold areas since chemical reactions are accelerated by temperature in
creases (Legros, 1982). Even when volcanoes supply a large quantity offineparticles
(ash), the soils at higher altitudes (2350 m) remain richer in silt and poorer in clay,
than the soils at lower altitudes (1000 m; Chartres and Pain, 1984).
Lastly, lenticular micro-structures attributed to freezing and thawing may be
observed in cold soils (Mellor, 1986).
Solubility, particle size and mineralogy of soils
In the mountains, phenomena of solubility are especially important for several

reasons. First, the climate is often wet, the potential évapotranspiration (PET) low,
and much water percolates through the profile. Next, the carbon dioxide in the
atmosphere as well as that produced by biological activity in the soil, is readily
dissolved at low temperatures, and rain or snow melt waters are corrosive (Clement
and Vaudour, 1967).
Solubility phenomena are very important on carbonate rocks leading, for ex
ample, to solution rills typical of karsts ("Lapiaz"). T\vo main types of corrosion
surface can be distinguished. At high altitude, exposed karst surfaces show angular,
knife-like, vertical irregularities (Figure 7.3a; Quinif, 1976). Such rough morphology
demonstrates the combined action of dissolution by water and fragmentation by
freezing. By contrast, smoother karst surfaces with rounded contours are probably
»■ -■■:^$ϊ&^ψί***
■ / ? / ■ > ■ # .
'Ί j/ii
'&'ψ J;
wfJtÂ
fi
Fig. 7.3. Solution factors, a. Rough-surfaced karst ("Lapiaz") Flaine, France, b. Smooth-surfaced karst
Flaine, France.
formed within the soil mantle and uncovered by soil erosion (Figure 7.3b). Frost
shortening accounts for little in this case, and the percolation of water is hampered.
Solution alone is the effective process here.
164 J.R Legros
Solubility phenomena have a number of consequences for soils.

(a) First, surficial waters rapidly charge themselves with ions on contact with
rocks. Solutions collected from C horizons of soils on limestone already contain half
of the dissolved calcium they will have when measured at their point of emergence
after percolating through bedrock (Pochon, 1978). In the limestone parts of the
Alps, the fractured coarser elements of the upper part of the soil mantle are worn
smooth and rounded by dissolution. Deeper blocks, however, remain angular below
one meter. This confirms observations made by hydrologists specializing in karst
(Lepiller, 1979,1982; Lepiller and Tklour, 1980; Mondain, 1983).
(b) Using a computer simulation it is possible to show that, in a collection of
monomineralic particles of mixed sizes, the dissolution is expressed in terms of the
relative concentration of the coarsest fraction. This latter presents the smallest ratio
of surface to mass and thus dissolves more slowly than finer particles. This is why
in moist mountain soils over limestone, the grains of calcium carbonate disappear
from thefinesoil fraction (<2000 μτη) while the coarser limestone particles remain.
In dry environments, the fine fraction often retains a high proportion of carbonate
(Dupuis, 1969).
(c) Lastly, as a general rule, the more soluble a mineral is, the more it concen
trates in the coarser fractions (Legros, 1982; Chamayou and Legros, 1989). In moist
environments, soils over gypsiferous rock often contain gypsum only as a few large
isolated blocks. By contrast, quartz, which is much less soluble, resists dissolution
down to the level of loam and clay in nearly all soils.
Wind contrìbution
Not all soils form entirely from in situ parent material. Many receive allochtho-
nous material during development. Wind born substances, for instance, can signifi
cantly contribute to soil profiles.
(a) On a global scale, the most frequent wind-borne substance is volcanic ash. Tec
hnically active mountain areas are often zones of intense volcanic activity. The "ring
offire"rimming the Pacific Ocean is a good example of this. Sometimes ash layers
are sufficiently thick within the soil profile to be morphologically recognizable. In
other cases, only by very close mineralogie analysis can exotic materials (especially
volcanic glass) be detected. This type of allochthonous contribution has been studied
in the Canadian Rocky Mountains and the Cascade Range of Canada and the United
States (Singer and Ugolini, 1974; Pawluk and Brewer, 1975; Smith et al., 1981). The
chemical composition of volcanic ash varies, but is often rich in iron (up to 8%) and
aluminum (around 20%). Since Podzols are observed on soils highly contaminated
with ash (van Ryswyk and Okazaki, 1979), the question arises whether their forma
tion may be facilitated by ash (Smith et al., 1983). On the other hand, Bs horizons
can form on rocks containing very little iron, and Souchier (1971) has shown that
excess iron can hamper podzolization. Thus the role of ash merits further study.
(b) The second type of air-borne sediment is dust, swept up from desert regions
and carried at high altitudes by strong currents, and often trapped among high
mountains. The French and Swiss Alps for instance are regularly supplied with dust
brought from the Sahara. In general, dust particles measure less than 60 μπι, with
a mean value of around 4 μτα (Bucher et al., 1983). The composition of these
particles varies considerably, depending on the nature of the geological formations
from which they came (Paquet et al., 1984). They may be quartz (10-50 %), calcite
(10-50%), feldspars (10-30%), clays of all kinds (20-50%) and even traces of
gypsum. Illite predominates and attapulgite is sometimes represented in the clay
minerals.
The problem is to establish whether this eolian dust is capable of altering sig
nificantly the physical and chemical properties of soils. In terms of chemistry, such
modifications appear possible, even in the short term. According to Litaor (1987),
"the steady supply of calcite through wind-blown material raises the pH of the sur
face horizons close to neutrality and apparently increases the ability of the soils
to resist acidification". In terms of physical properties, the original particle size
of the soil may be changed, if the allochthonous particles mix with autochthonous
ones, but for this phenomenon to be significant, the in situ production of new fine
material must be low. This may be the case on massive, resistant limestone. The
deposition of eolian dust may also cause an additional horizon to be added at the
top of the profile. While this case is not necessarily the most common, it is obviously
the easiest to detect (Jalalian and Southard, 1986).
(c) The third type of eolian material, and possibly the most important with regard
to the Alps, was added during cold, dry periods of the Pleistocene (Dryas I, II, III).
Wide areas were denuded of vegetation and became susceptible to wind erosion.
Much topsoil material was transferred and sedimented into areas still vegetated.
This may explain, in part, the abundance of similar loams throughout the Alps, over
all kinds of substrates (Spaltenstein, 1984). This is also the case in the Cordilleras of
North America.
Principal geo-morpho-pedological systems
Here, I review the major environments as defined by a given rock, by a char

acteristic morphology, and resulting soils. Tb the extent possible, I will distinguish
between cases in wet mountain areas, and those in dry.
Calcshale soils
As their name suggests, calcshales contain both phyllosilicates (mica, chlorite)
and calcite. They also contain appreciable quantities of quartz, and often some
pyrite (FeS2). Calcite dissolves, while the sulfides are oxidized to sulfates, accel-
166 J.R Legros
erating acidification of the environment and elimination of the calcium carbon

ate (van Oort and Robert, 1988). The rock loses its cohesiveness and liberates its
residue of dioctahedral phyllosilicates and quartz. Transformation continues within
this residue. The phyllosilicates break down and go from silt to clay size particles
(Legros et al., 1980). Thus the soils become increasingly rich in fine particles as the
surface is approached.
Weathering of calcshales generates deep soil profiles (>2 or 3 m) even on some
steep slopes where there is little erosion and practically no gullies. TWo explanations
are possible: either the water runs off the surface causing no damage to the soils
because of the existence of a continuous cover of vegetation, or the water infil
trates readily. This has led to a closer examination of the porosity and hydrological
properties of these soils (van Oort and Dorioz, 1983; van Oort, 1984). There is
considerable initial porosity at the base of the profile, due to the empty spaces left
by the dissolution of carbonates, while at the B horizons the porosity is reduced to
1.7 cm3 cm - 3 by compression. The hydraulic conductivity remains above 0.8 cm per
hour in two out of three samples. All in all, the calcshale soils are not easily eroded.
On the other hand, high porosity causes the soils to become considerably enriched
in water at snow melt, causing landslides where the Atterberg liquidity limit has
been attained.
Calcshales behave differently depending on whether the climate is dry or wet.
In dry climates, for example in the southern Alps, weathering is limited. The
carbonates are dissolved out of the fine fraction (<2 mm) in the A and sometimes
the B horizons, although these generally contain a high proportion of more or less
calcareous pebbles and boulders which have resisted disaggregation and dissolution.
This is the domain of Eutric soils. At the bases of slopes, there is often a somewhat
thicker and more acid layer of weathered residue (Duchaufour and Gilot, 1966).
This organization of the soils as a function of position on the slope is reflected in
the vegetation (Labroue and Tbsca, 1977; Gensac, 1977).
In wet climates, in the northern Alps for example, water is available in great quan
tities, but its action varies considerably, depending on slope (Dorioz et al., 1983).
When the dip of the rock layers is parallel to the slope, water cannot infiltrate deeply,
thus it causes a great deal of weathering at the surface. Carbonates are dissolved
down to a depth of several meters. The soil mantle is acid, yellowish, is constituted
essentially of fine materials and is very sensitive to solifluction, generating humps
and hollows topography. These humps and hollows are very large and when the
weathered materials move, they represent, each time, landslides of hundreds, and
sometimes thousands, of tons. Circulation of water over this terrain is irregular. Hy-
dromorphic mottling may be seen in the very acid soils of the hollows: the water
flowing through them has been moving in contact with the rock along the slope and
is more or less saturated with calcium. On the humps, on the other hand, subtractive
phenomena predominate and many of the cations are leached out. The soils become
acid and, if rhododendrons take hold, true Podzols may be generated locally.
When the dip of the rock layer is opposite to that of the slope, the water in
filtrates more readily and to a greater depth. Solifluction is limited, all the more
so since the spurs of rock present at the surface constitutes natural barriers to the
movement of weathering residues.
The situation just described with regard to calcshales, is very common in the
mountains, and it applies to all rocks containing some calcium carbonate and much
weathering resistant residues of phyllite and quartz. Falling in this category are the
flyschs, the "schistes lustrés" of the southern Alps, and the calcareous micaschists, in
other words, rocks that are typical of the metamorphosed clastic wedge assemblages
of active continental margins.
Soils over limestone

Limestones containing less than 5% non-carbonate residue are very resistant to
weathering if the climate is not too wet. In France they form the plateau, cliffs
and saw-tooth ridges characteristic of the northern and southern sub-Alpine ranges
which include two principal geological levels: the Tithonian, which is the upper limit
of Jurassic sedimentation in the Alps, and the Urgonian, limestone typical of the
Lower Cretaceous (upper Barremian or lower Aptian) and parent rock of nearly all
the soils of the Vercors region of France. In North America, the Rocky Mountains
are another example of limestone terraine.
Since the residues remaining after the carbonate has been dissolved out, are
not abundant, much of the soil covers formed are thin. TWo main cases have been
observed (Duchaufour and Bartoli, 1966): (a) accumulation of a horizon very rich
in humus lying directly over a limestone slab, and (b) mixture of highly organic
material with limestone pebbles (as in the case of screes).
Pedological evolution is very different in the two cases. When the soil develops
over a slab, acidification is possible. This occurs all the more clearly when the layer
rich in humus exceeds 35 cm in thickness (Michalet and Bruckert, 1986), and acid-
loving vegetation, such as the rhododendron (Rhododendron ferrugineum L.), moves
in (Gilot, 1967). This humus, saturated with calcium below and acidified on the sur
face, is called "Tàngel". This is the "roh-humoser Humus-karbonatgesteinsboden"
of Swiss authors, and "Tkngel-rendzina" of the Federal Republic of Germany (Juch-
ler, 1988).
In screes, on the other hand, acidification is practically impossible due to the
lateral circulation of calcium-charged water and the presence of reserves of calcium
in the coarse material (Gaiffe and Schmitt, 1980), while the accumulation of organic
matter is facilitated (Blondel et al., 1986). It should also be noted that the surface
area in contact with the calcium is much greater on a fragmented substrate than on
smooth slabs (Pochon, 1978; Cabidoche, 1979).
In some cases, the soil over limestone is deep and without too many stones,
posing the question of where it derives from. TWo hypotheses are possible: accu-
168 J.R Legros
mulation of the residue after the carbonate has been dissolved out, or eolian dust.
Based on a few simplifications, it can be calculated that in 10,000 years, the
weathering of a limestone containing 5% silicate residue, evolving under an annual
rainfall of 1 or 2 m, would yield a soil around 5 to 10 cm thick (Legros et al, 1987).
The value of 10,000 years is used as a reference since it dates the retreat of the
last Pleistocene ice sheet in the Alps. On the other hand, based on extrapolation
from the current situation, the amount of Saharan dust deposited on Europe in
10,000 years is estimated at 70 cm (Coude-Gaussen and Rognon, 1983). For their
part, Amundson et al. (1989) estimated the amount of material deposited on the
Spring Mountains of Nevada over the same period of time, to be 6 cm. These
figures are not precise, but serve to show that the possible role of eolian dusts
may not be rejected out of hand, especially when the parent limestone is poor in
residue.
The hypothesis of allochthony with regard to the fine fraction, is supported by
two types of arguments (Pochon, 1978): (a) presence of wind-worn grains of quartz
in the fine material, and (b) presence of resistant minerals within the residue, but in
too low a proportion in the natural soil, as if they had been diluted by an external
source. On the basis of the quantities of chlorite present, Pochon estimates that
eolian dust makes up 3/4 of the soil mass in the Swiss Alps. By contrast, other
authors emphasize the strict correspondence between the clays in the soils and the
clays in the residues after the carbonate has been artificially dissolved out of samples
of the underlying rocks (Delmas, 1981). It is probably impossible to pick a definitive
answer. Using clay mineralogy to determine whether or not eolian additions have
been important can be difficult. Both eolian dusts and the autochthonous clays of
limestones are very variable. Chlorite, vermiculite, illite and kaolinite are all found
and each may originate externally or internally, by air-borne addition, by inheritance
or by weathering and soil-formation in situ.
The presence of relatively high concentrations of Ca2+ in the weathering en
vironment generally stabilises 2:1 sheet silicates rather than 1:1 sheet silicates.
Consequently the cation exchange capacity of derived soils tends to be high. A sec
ond consequence of high Ca2+ is that humus accumulates, since Ca2+ blocks its
mineralization (Figure 7.4; Bruckert et al., 1986).
When the clay fraction is relatively abundant, a highly differentiated microstruc
ture appears, reflecting the constitution of a complex organization between the clay,
organic matter and calcium. This structure is apparent at the base of the humus in
direct contact with a limestone slab (Party, 1982; Gaiffe, 1987).
When the soil mantle is thick (>70 cm), whether it results from residues after the
carbonate has been dissolved out, or from deposits of eolian dust, the evolution of
the soil may be considerable as shown by the particle size differentiation between A
and B horizons. Since this is a very slow process, it may be observed only in old soils
in stable-flat environments. Such a situation is visible in the sub-Alpine chains of the
French and Swiss Jura, but not in the high Alps.
Soih ofAlpine mountains 169
40 H
·· · φ <
co 30-
ö)
• ·
·· · · ·
• · · · .
20 H
10H
;**.
1 1 1 1 1 1 r
0 .. - 10 20 30 40 50 60 70
Exchangeable Ca + + of the CEC
(me/100 g)
Fig. 7.4. Relationships between exchangeable Ca 2+ and organic matter content.
The situation with regard to soils over limestone formations, may be summed up
as follows.
(a) In wet climates, carbonate is removed from the fine fraction (<2 mm) totally
and systematically, although the soils remain saturated in calcium ions (Ca 2+ ) due
to the presence of coarse forms of calcite in the environment (abundant pebbles,
or rock not far below the surface). Carbonate remaining in the fine fraction is rare,
though it occurs in very active scree slopes. Cases of strong acidification are limited
to non-stony soils, often quite thick, on a level surface.
(b) In dry climates, as in the southern Alps, the removal of carbonate is incom
plete, and some CaC0 3 remains in the fine fraction. These are mainly the rendzic
Leptosols (Rendzines), or Haplic Calcisols (sols bruns calcaires).
In both cases, accumulation of the fine material in hollows and its loss from
nearby higher areas, is the principal phenomenon; this is the karst soil chain as
described by various authors (Duchaufour and Bartoli, 1966; Bottner, 1971; Pochon,
1978; Gaifïe, 1987).
Soils over granular crystalline rocks

The first stage in the weathering of granites, gneiss and micaschists is disintegra
tion, which leads to crumbling and separation of the mineral grains. This relatively
170 J.R Legros
coarsely textured residue is preserved mineralogically unchanged in mountains, in

some cases by virtue of cold conditions, in others by virtue of dry conditions, both of
which retard.
Many mountains are to be found in natural podzolisation zones. A major prob
lem here is to understand the cycle of aluminum. Tb study this, dialysis sacks of
"test minerals" (vermiculite in particular) were implanted in soil profiles (Robert
et al., 1987). The vermiculite initially saturated with barium, can either lose its
aluminum by solubilization, or pick up aluminum ions circulating in the soil solu
tion to form a hydroxy-interlayered vermiculite. The former case indicates an acid
soil, well-drained and readily forming organo-metallic complexes; in other words,
strongly podzolising (the eluviated horizon of Podzols). The latter case, where alu
minum is trapped, corresponds to a less acid, less well-drained environment. This
may be found in the B horizons of podzols (Robert et al., 1980), especially if they
are hydromorphic (Righi and Wilbert, 1984). Intergrade phases are also abundant in
Cambisols. Assays of this kind, with test minerals, were used to determine that pod
zolisation in the Alps, at the sub-Alpine level, takes place mainly in winter, that is,
under intense cold and wet conditions (Dambrine, 1985). In summer, the aluminum
is not removed from the profile.
Another way to study podzolisation is to analyze the waters draining through the
soils, or the water from streams draining the watershed. Such analyses are difficult
because the ion concentrations are low. Unexpected results can be obtained, such as
the following.
(a) There is no more export of ions under strongly than under weakly podzolised
soils. Perhaps this is because the former, previously impoverished, no longer have
the same capacity to supply ions to the drainage waters (Dambrine, 1985).
(b) The same ionic composition is not necessarily found in the waters emerging
from the watershed, as is found in the waters draining through the soil. The under
lying parent material made of granitic sand retains the dissolved products extracted
from the superficial layers of the soil (Dupraz et al., 1986).
Lastly, podzolisation may obviously be studied by examining the evolution of the
constituent minerals of the rocks and soils.
(a) Quartz is essentially preserved intact. This is because the destruction of sil
icate minerals saturates the drainage waters with silica, thus, protecting the less
weatherable forms of Si0 2 .
(b) Feldspars may be dissolved more or less quickly, depending on their com
position; calcic plagioclases are less stable than sodic, which are less stable than
potassic.
(c) Micas and chlorites are transformed slowly, as for example, according to the
progression shown in Figure 7.5.
It is important to note that at each stage a significant portion of minerals break
down. However, the yield of secondary phases may not be large, with a given amount
of mica producing a small quantity of smectite for example. Partly, this is a function
Chlorites ^ _
. * * *·*· ^ . \/«rrv,;^.,i;t Q ^^Smectite formed
Interstratified minerals —■► Vermicuhte — ► b y weathering
I I
Micas *
Solubilization Solubilization Solubilization Solubilization
Fig. 7.5. Clay transformations in soils over crystalline rocks.
of pH, with low pH favouring a greater destruction of mica and loss of the products
in solution, than higher pH values.
It is in soils over crystalline rocks that, in mountains, the best correspondence
between the soil-climate-vegetation may be found (Figure 7.6). In the sub-Alpine
Glacier des Pèlerins

"The Pilgrim's
glacier"
Moraines and bare rocks
Teleferic station of
Plan de l'Aiguille
(2317 m)
Alpine meadow
#:::**
2200 m Ranker alpin
Dy strie leptosol
Predominance of
rhododendrons
1900 m
Larches with rhododendron
undergrowth Sol intergrade
Various podzolic
1700 m soils
Podzol nain humo-ferrugineux
Thin humo-ferric podzol
Spruces with 1 8/7

grasses or bilberry
es
undergrowth
c
V////A R
Podzol humo-ferrugineux
Humo-ferric podzol
The town of
Chamonix Sol ocre podzolique
(1030 m) Cambic podzol
l?ó>°.ó;| Granitic moraine
[ j j j t| Granitic rock
Sol crypto-podzolique brun Ekl*3 Gneiss
Dystric cambisol
Fig. 7.6. Altitudinal distribution of soils and vegetation in the northern French Alps, near Chamonix
— Mont Blanc (from Legros and Cabidoche, 1977).
172 J.R Legros
belt, the conifer forests and ericaceous shrubs are characterized by highly pod-
zolised soils. A few examples are given in Table 7.3, but others could be cited, from
the Himalayas (Sehgal et al, 1985) and Canada (van Ryswyk and Okazaki, 1979).
It is interesting to note that the true, or Humo-ferric Podzols only appear un
der environmental conditions leading to maximum acidification (Legros, 1975). Tb
achieve this, weathering conditions must include acid rock (granitoids, leptynites,
quartzo-feldspathic gneisses), cold, wet climate (especially sub-Alpine) and acidify
ing vegetation (such as ericaceous shrubs, or conifers with ericaceous undergrowth).
If any one of these conditions is lacking, in particular if the parent rock is not very
acid (Souchier, 1971), the soils will fall into one of the less evolved categories of the
Cambic Podzols or Dystric Cambisol type: the "alpiner braun Podsols" described by
Swiss authors, the "ocre-podzoliques" of French soil scientists or the "Umbrie Dys-
trochrepts" of American workers. Practically no Podzols are found over micaschists
in the Alps (TVosset, 1980).
This generalisation concerning the relationship between Podzols and parent
rocks is not absolute. Even calcite-bearing materials (for example, a sandstone with
a calcareous cement) may achieve podzolisation given sufficient weathering time to
remove the calcite, and allow the pH to drop to acid values.
Mountain Podzols show certain peculiar features relative to those of the sandy
plains of Europe or to tropical Podzols. Their horizons are very humiferous, well
defined and thin. These are the dwarf Podzols, also called "alpiner eisenhumus
Podsol" by Swiss and German authors, and "Typic Cryorthod" by authors in the
United States.
Commonly, soils intermediate between the Podzols and Umbrie Leptosols (Rank
ers) occur at the border between the sub-Alpine and Alpine belts. These are humif
erous soils, not very thick, having either a more or less distinct bleached E horizon
or a ferruginous B horizon, at the base of the soil, in contact with the rock. Some
times true Humo-ferric Podzols may be found beneath Alpine prairie, but this is
essentially due tofluctuationsin the altitude of the tree line, relative to climatic
variations and extension or regression of the pasture land (Müller, 1983).
In general, there are two types of soil over crystalline rock in the Alpine belt:
Umbrie Leptosols (Rankers) and Cambisols (sols bruns) (Bartoli and Burtin, 1979).
The former often correspond to the wetter environments and lie over slabs of hard
rock, while the latter are more related to scree slopes, drier environments, and rocks
providing weathered material which is fairly rich in clay such as micaschists.
In each belt, the topography introduces microclimatic effects, such as variations
in the temperature of the soil, solar irradiation and humidity. These result in mod
ifications in the type of vegetal cover which in turn leads to variable degrees of
podzolisation from one point to another (Blaser, 1988).
Soils over volcanic rocks

Altitudinal zoning is particularly interesting to study in the case of volcanic parent
materials since volcanoes have the double advantage of representing homogeneous
pétrographie systems (in terms of an eruptive episode) and of having regularly rising
steep slopes with significant climatic variation from foot to summit.
Numerous local sequences have been described throughout the world but often
in a fragmentary manner, because the range of altitudes observed in each case is
generally small. Synthetizing available data, it is possible to construct the following
idealized sequence.
(a) First, at the summit, Andosols form where podzolic characters are distinct.
These are the "Typic Haplohumods" (USA) or the "sols ando-podzoliques" (F).
Lulli and Bidini (1980) encountered such soils at the summits of Italian volcanoes,
at an altitude of around 1300 m with 1330 mm of rain per year and a mean annual
temperature of 7.8°C. In other, wetter regions, the true Humo-ferric Podzols may
be formed. This is the case of the island of Réunion, where the temperature on its
volcanic summits is relatively high (13°C), but still with a high rainfall (anywhere
from 1.7 to 3 m; Zebrowski, 1975).
(b) At an intermediate level well differentiated Andosols (Hydric Dystradepts)
may be found which indicate very wet climates: either around 1000 mm of precipita
tion per year if the mean annual temperature is below 8 or 9°C (as in the mountains
of Europe), or at least 1500 mm of precipitation per year if the mean annual tem
perature is in the neighbourhood of 13 or 14°C (as is the case in New Zealand, as
reported by Parfìtt et al., 1983), and up to 2000 or 2500 mm of precipitation if the
temperature exceeds 20°C (as in New Guinea: Chartres and Pain, 1984).
(c) Lower down on the slopes, the andic properties are less developed. Andic
Cambisols (sols bruns andiques, Andic Haplumbrept or Andic Eutrochrept) are
generally found. The transformation of true Andosols into less andic soils is not
only related to the climate but also to soil moisture regime. Any cultivation, which
increases évapotranspiration, may provoke such transformations (Moinereau, 1977).
(d) Lastly, at the foot of volcanoes, in a drier environment, it is possible to
find Chromic Cambisol and Chromic Luvisols (sols fersiallitiques/Ustochrepts) or
Vertisols (Chromoxererts; Caldas et al., 1981).
There is a decrease in the amount of organic matter and an increase in the
clay content from the top to the bottom of this sequence. Undoubtedly, the most
interesting phenomenon is the gradual shifting of the aluminum to silicon (Al/Si)
ratio in clay fraction reflecting the leaching of Si in soil solution. Thus, the value of
this ratio is two at higher altitudes where the soil is rich in allophane, it drops to one
when halloysite invades the profile, and falls below one when smectites appear, at
the base of the slope (Parfitt et al., 1983).
174 IR Legros
Fig. 7.7. Soil over marly limestone in wet environment (from Legros, 1986).
Otherpedolithic systems seen in mountains
Soils over marly limestone

Marly limestones represent an intermediate type between the pure limestones
and the calcshales. This is true with regard to the characteristic mixed landscape
(between rolling and steep) as well as for the nature of the associated soils. The
thickness of the mantle of weathered material, and its degree of acidification are av
erage statistically. The calcium saturation status of the exchange capacity, depends
on all factors already mentioned: depth of the parent rock, stability of the envi
ronment, annual rainfall and proportion of coarse elements containing carbonates,
but the most important factor is the topography. All of the projecting parts of the
terrain become acid: crags, ridges, the high edges of solifluction deposits, while on
the other hand, all the low points contain soils with clays saturated with calcium.
This has been demonstrated in the French Alps of the Chablais (Figure 7.7; Legros
et al., 1987).
This proves that surface or sub-surface runoff does exist in these terrains and
plays a fundamental role in soil evolution. On the other hand, acidification is
thwarted over marly limestone in mountainous areas when there is a more or less
hard limestone cliff providing an abundance of pebbles to the forming soil profile.
In this case, the soil remains saturated with calcium.
Soils over marls and heavy clays of glacial origin

The high clay content of certain soft materials limits the penetration of water and
hampers the solution of carbonate. In addition, this low permeability increases the
lateral surface runoff resulting in the regular renewal of the upper part of the soil
profile. Thus in the wetter parts of the Alps, soils over calcareous clay do not lose
their carbonates, in contrast to soils over other types of rocks containing CaC03.
On the high plateau of Peru (altitude 4500 m), carbonates are dissolved only in the
upper 70 or 80 cm of the profile (Wilcox et al., 1988).
Clayey soils which have had their carbonates dissolved out, in Alpine areas often
show a well-developed prismatic structure in the B horizons. This characteristic is
insufficient to consider them as leached soils, that is, those in which the clay has
been translocated to form a Bt horizon (Alfisols). Most probably, these are pelosols,
that is, those in which most of their properties are related to the original clay
concentration in the parent material.
If the climate is wet, or in humid hollows, a more or less conspicuous hydromor-
phy is observed.
SoiL· over gypsum

When the calcium sulfate contains substantial impurities, these accumulate on
the surface of the rock and develop a soil profile. Gypsum, being a highly soluble
salt, does not remain in the soil unless there is very little rainfall. Thus, even in the
drier southern Alps, the soils over gypsum are calcareous but not gypsiferous; they
contain calcium carbonate but not calcium sulfate. In the wetter northern Alps, the
soils over gypsum are neither calcareous nor gypsiferous, but acid. Gypsum on the
surface in any humid environment can only resist dissolution if it is in large masses
(as in blocks or boulders).
Soils over the greenstones of the Alps (serpentinitesy prasinites)

These rocks are abundant in the "inner Alps", especially in the Italian zone (Pare
National du Grand Paradis, Massif du Viso, Val D'Aoste). They are particularly rich
in iron and in magnesium. The average composition in percentage of the serpen-
tinites is as follows (Verger, 1987).
Si0 2 : 40.49 Ti0 2 : 0.65
MgO: 37.36 MnO: 0.22
Fe 2 0 3 : 10.85 Na 2 0 3 : 0.15
A120: 1.77 K 2 0: 0.10
CaO: 0.74
Weathering liberates the components of their characteristic primary minerals of
chlorite, talc, amphiboles and pyroxenes. In the "inner Alps", the environment is
relatively dry and the mineralogical transformation remains limited and very small
quantities of clay are formed.
This limited mineralogical evolution goes hand in hand with a limited evolution
in the soil morphology (Verger, 1987). This is the domain of the Leptosols and
Cambisols: "Serpentinic Cambisols" (FAO), "saure serpentin-Braunerde" (Swiss) or
176 J.R Legros
"Frigid Dystric Eutrochrept" (USA). The pH never falls below 5 or 5.5 and the rate
of saturation of the absorbent complex is very close to 100%, at the pH of the soil.
At a pH of 7, it is much lower and more variable. Magnesium makes up over 60% of
the exchange sites of the CEC.
Contrary to what one might expect intuitively, the soils over serpentines are not
very clayey (usually less than 20%) and are often richer in sands than in loams
(Verger, 1989). The soils over serpentines, in the Alps, are young. The maximum
degree of evolution observed is represented by just the beginnings of podzolisation
with the E horizon weakly developed. This can be seen under the conditions most
favourable to acidification in sub-Alpine belt with conifers with ericaceous under
growth. Some authors think that podzolisation may be favoured on these rocks by
eolian dust, or by the production of plant opal (Sticher et al, 1986; Juchler, 1988).
Along with the serpentines, in the same regions, the prasinites may occur. These
have a composition similar to that of gabbros, with nearly as much calcium as
magnesium, and almost 50% of Si0 2 . These rocks supply a large amount of silicate-
containing residue capable of providing the constituents of fairly strongly podzolised
soils.
Conclusion
This review of soils in the Alpine environment demonstrates the complexity of

their spatial organization. Altitude delimits the vegetation and soil belts. Topogra
phy also plays an important role. In stable zones, soils may become deep and slowly
but strongly transformed; in unstable zones conditions are not favourable to devel
opment of evolved soils. The topography also determines the dynamics of water and
that of calcium ions, which disappear from high points in the landscape but remain
in the hollows.
The parent rock plays an essential role, less by its initial composition than by
the characteristics of the residue it is capable of leaving through weathering and
dissolution. Thus in the northern Alps, a pure limestone can only support humic
soils, while a limestone containing a small amount of siliceous residue, can develop
Podzols on stable levelled topography.
References
Amundson, R.G., Chadwick, O.A., Sowers, J.M. and Doner, H.E., 1989. Soil evolution along an altitu-
dinal transect in the Eastern Mojave Desert of Nevada, USA. Geoderma, 43: 349-371.
Archer, A.C., Cutler, E.J.B., 1983. Pedogenesis and vegetation trends in the alpine and upper subalpine
zones of the northeast Ben Ohau Range. New Zealand. N. Z. J. Sci., 26: 127-150.
Bartoli, F. and Burtin, G., 1979. Étude de quatre séquences sol-végétation à l'étage alpin. Doc. Cartogr.,
Grenoble, XXI: 79-93.
Birse, E.L., 1980. Suggested amendments to the world soil classification to accommodate Scottish
mountain and eolian soil. J. Soil Sci., 31: 117-124.
Blaser, P. 1980. Der Boden als Standortsfaktor bei Aufforschungen in der sub-alpinen Stufe (Stillberg,
Davos — CH). Mitt. Schweiz. Anst. Forstl. Versuchswes., 33: 529-611 (English summary and figure).
Blondel, J.L·, Gaiffe, M., Contini, D. and Bruckert, S., 1986. Relations entre le caractère humifère des
soils des plateaux jurassiens et le régime hydrique induit par la perméabilité des bancs calcaires.
Pédologie, XXXVI: 155-177.
Bottner, P. 1971. Le pédogenèse sur roches-mères calcaires dans une séquence bioclimatique méditerra-
néo-alpine du Sud de la France. Thèse d'Etat, USTL, 271 pp.
Bottner, P. 1984. Reconnaissance des soils dans les séquences bioclimatiques altitudinales du Nepal
Central. Doc. Cartogr. Ecol., XXVII: 85-99.
Bruckert S., Gaiffe, M., Duquet, B., Tavany, Y. and Tavant, H., 1986. Rôle du flux de calcium sur la
stabilisation de la matière organique des sols. Ann. Sci. Univ. Franche-Comté, Besançon, 4ème Sér.,
6: 25-29.
Bûcher, A., Dubief, J. and Lucas, C. 1983. Retombées estivales de poussières sahariennes sur l'Europe.
Rev. Géol. Dyn. Géogr. Phys., 24: 153-165.
Burns, S.F., 1985. Classification and development of alpine soils, Southern Rocky Mountains, USA.
Proceedings International Meeting and Classification and Management of Soils in Mountain regions
(1984), Bulg. Soil Sci. Soc, N. Poushkarov Institute, pp. 72-85.
Cabidoche, Y.M., 1979. Contribution à l'étude des sols de haute montagne. Analyse des relations sol-
milieu dans les Pyrénées Occidentales cristallines et calcaires. Thèse 3ème Cycle, 2 volumes, 157
and 61 pp.
Caldas, E.F., Quantin P. and Tejedor-Salguero, M.L, 1981. Séquences climatiques de sols dérivés de
roches volcaniques aux Iles Canaries. Geoderma, 26: 47-62.
Chamayou, H. and Legros, J.P., 1989. Les bases physiques, chimiques et minéralogiques de la Sci. Sol.
PUF, Paris, 593 pp.
Chartres, C.J. and Pain, CF., 1984. A climosequence of soils on late quaternary volcanic ash highland,
Papua New Guinea. Geoderma, 32: 131-155.
Choisnel, E., 1986. Aspects topoclimatiques: une méthodologie d'étude en région de moyenne mon
tagne. Agrométéorologie des régions de moyenne montagne. INRA, 39: 177-195.
Clement, P. and Vaudour, J., 1967. Observation on the pH of melting snow in the Southern French
Alps. In: H.E. Wright, Jr. and W.H. Osburn (Editors), Artie and Alpine Environments. Indiana
Univ. Press, Bloomington, Ind., pp. 205-231.
Coude-Gaussen, G. and Rognon, P., 1983. Les poussières sahariennes. Rev. Rech., 147: 1051-1061.
D'Alessio, D. and Previtali, F., 1988. I podzoli della Valle Camonica (Alpi meridionali bresciane).
Natura Bresciana, Ann. Mus. Civ. Sci. Nat. Brescia, 24: 47-73.
Dambrine, E., 1985. Contribution à l'étude de la répartition et du fonctionnement des sols de haute
montagne. Massif des Aiguilles rouges et du Mont Blanc. Thèse Université de Paris VII, 265 pp.
de Coninck, F. and Debrebant, P., 1986. Diagenese des argiles dans le domaine sub-alpin: rôles respectifs
de la lithologie, de l'enfouissement et de la surcharge tectonique. Rev. Géol. Dyn. Géogr. Phys., 26:
321-330.
Delmas, M., 1981. Contribution à la pédogenèse sur calcaire urgonien à l'étage subalpin dans les
préalpes française du Nord. DEA, INAPG, Paris, 66 pp.
Dobremez, J.F, 1979. Note sur la teneur en matière organique des sols himalayens le long d'une
séquence topographique et climatique de grande amplitude. Sciences Naturelles. Ann. Univ. Savoie,
4: 49-53.
Dokuchayev, B., 1900. Zones verticales des sols. Zones agricoles. Sols du Caucase. Edition du Ministère
des Finances, Exposition Universelle de 1900 à Paris, Section Russe, 56 pp.
Dorioz, J.M., van Oort, F. and Dambrine, E., 1983. Principaux facteurs de répartition des sols sur cale-
178 J.R Legros
shales dans les alpages du Beaufortin. 108ème Congrès National des Sociétés Savantes, Grenoble.
Sciences, I: 115-126.
Dougedroit, A., 1986. Les topoclimats thermiques en moyenne montagne. Agrométéorologie des régions
de moyenne montagne. INRA, 29:197-213.
Duchaufour, Ph. and Bart oli, C, 1966. Note sur l'évolution des sols calcimorphes de l'étage montagnard
humide. Sci. Sol, pp. 29-40.
Duchaufour, Ph. and Gilot, J.C., 1966. Étude d'une chaîne de sols de l'étage alpin (Col du Galibier) et
ses relations avec la végétation. Oecol. Plant. Gauthier-Villars, 1: 253-274.
Dupraz, C , Lelong, F. and Bonneau, M., 1986. Effet des boisements résineux purs sur l'évolution de
la fertilité du sol: premiers résultats obtenus sur bassins versants expérimentaux du Mont Lozère
(France). Ann. Sci. For., 43:147-164.
Dupuis, M., 1969. Dosage des carbonates dans les fractions granulométriques de quelques sols calcaires
et dolomitiques. Ann. Agron., 20: 61-88.
Durand, R. and Legros, J.P., 1981. Cartographie automatique de l'énergie solaire en fonction du relief.
Ann. Agron., 1: 31-39.
Escourrou, G., 1980. Climat et environnement. Les facteurs locaux du climat. Masson, Paris-New York,
N.Y., 182 pp.
Fleury, Ph., 1986. Les relations entre la phénologie des espèces des prairies montagnardes (Beaufortin,
Savoie), la topographie et les variations climatiques interannuelles. DEA, Université de Grenoble,
79 pp.
Gaiffe, M. and Schmitt, A., 1980. Sols et végétation à l'étage montagnard dans les forêts du Juras
Central (Haute Vallée du Doubs, E), Sci. Sol, 4: 265-296.
Gaiffe, M., 1987. Processus pédogénétique dans le karst jurassien. Analyse de la complexation organo-
minérale en ambiance calcique. Thèse d'Etat, Université de Franche-Comté, 160 pp.
Gensac, P., 1977. Sols et groupements végétaux de la zone des schistes lustrés dans l'étage alpin. Extraits
des travaux scientifiques du Parc National de la Vanoise, VII: 29-40.
Georgiev, A., 1985. Vertical-zonal manifestations of the mountains-forest rendzinas. Proceeding of
international meeting on classification and management of soils in mountain regions, Bulg. Soil Sci.
Soc, N. Poushkarov Institute, pp. 120-121.
Gilot, J.C., 1967. Note écologique sur divers groupements à Rhododendron ferrugineum se développant
sur substrat calcaire. Oecol. Plant. Gauthier-Villars, 2: 139-162.
Jalalian, A. and Southard, A.R., 1986. Genesis and classification of some paleborolls and cryoboralfs in
Northern Utah. Soil Sci. Soc. Am. J., 50: 668-672.
Juchler, S.J., 1988. Die Böden auf Serpentinit in der subalpinen Stufe bei Davos. Thesis ETH No. 8716,
Zürich University, 174 pp.
Labroue, L. and Tosca, C, 1977. La pédogenèse à l'étage alpin. Relation avec l'évolution des versants.
Bull. Ecol., 8(3): 299-316.
Lautridou, J.P., 1984. Le cycle périglaciaire pleistocène en Europe du Nord-Ouest et plus particulière
ment en Normandie. Thèse de Doct. d'Etat, Université de Caen, 1100 pp.
Legros, J.P., 1986. Cartographie pédologique dans les Alpes humides. Exemple du Chablais. Colloque
Agro-météorologie de Moyenne Montagne, Colloq. INRA, 39: 119-127.
Legros, J.P., 1982. Lévolution granulométrique au cours de la pédogenèse. Approche par simulation
sur ordinateur. Application aux sols acides sur matériaux cristallins en zone tempérée. Thèse d'Etat,
Université de Montpellier, 436 pp.
Legros, J.P., 1975. Occurence des podzols dans l'Est du Massif Central français. Sci. Sol, 1: 37-49.
Legros, J.P., Antonioletti, R. and Genre-Grand pierre, G., 1986. Evaluation de l'énergie solaire en zone
de montagne. Colloque Agro-meteorologi e de Moyenne Montagne, Colloq. INRA, 39: 87-93.
Legros, J.P. and Cabidoche, Y.M., 1977. Les types de sols et leur répartition dans les Alpes et les
Pyrénées cristallines. Doc. Cartogr. Ecol., Grenoble, XIX: 1-19.
Soils of Alpine mountains 179
Legros, J.P., Party, J.P. and Dorioz, J.M., 1987. Répartition des milieux calcaires, calciques et acidifiés
en haute montagne calcaire humide. Doc. Cartogr. Ecol., Grenoble, XXX: 137-157.
Legros, J.P. and Pedro, G., 1983. Importance relative des processus de dissolution et de fragmentation
(chimique) lors de l'altération et de la pédogenèse dans les zones tempérées froides. Approche par
simulation sur ordinateur. Coll. Int. CNRS, Petrologie des Altérations, 1: 85-95.
Legros, J.P., Robert, M. and van Oort, E, 1980. Caractères principaux de la pédogenèse sur calcschistes
en montagne. Sci. Sol, 4: 297-312.
Lepiller, M., 1982. La capacité de stockage dans la zone non saturée des aquifères karstiques. Estimation
et conséquences pour la localisation de la dissolution dans les systèmes. Ann. Sci. Univ. Besançon,
Géologie, Mémoire, 1: 147-157.
Lepiller, M., 1979. Le rôle de l'activité biologique dans l'acquisition de la charge dissoute des eaux
d'acquifères carbonates: étude sur 12 systèmes karstiques des bauges et des bornes. Colloq. Int.
CNRS, 303: 403-411.
Lepiller, M. and Talour, B., 1980. Différenciation thermique et chimique des eaux dans un système
aquifère très karstifié à alimentation essentiellement météorique (système de la Diau, Alpes françai
ses du Nord). 105ème Congrès National des Sociétés Savante, Caen Sci. Fase, II: 95-109.
Litaor, M.I., 1987. The influence of eolian dust on the genesis of alpine soils in the Front Range,
Colorado. Soil Sci. Soc. Am. J., 51: 142-147.
Lulli, L. and Bidini, D., 1980. A climosequence of soils from tuffs on slopes of on extinct volcano in
southern Italy. Geoderma, 24: 129-142.
Mellor, A., 1986. A micromorphological examination of two alpine soil chronosequences, Southern
Norway. Geoderma, 39: 41-57.
Michalet, R. and Bruckert, S., 1986. La podzolisation sur calcaire du subalpin du Jura. Sci. Sol, 24:
363-375.
Moinereau, J., 1977. Altération des roches, formation et évolution des sols sur basaltes, sous climat
tempéré humide (Velay, Vivarais, Coiron). Thèse d'Etat, USTL, 137 pp. (with annexes).
Mondain, P.H., 1983. Le système karstique de Morette (Massif des Bornes, Haute-Savoie): précision
sur ses limites sud-Ouest et caractères géochimiques des eaux à l'exutoire. 108ème Congrès Nat. des
Sociétés Savantes, Grenoble. Sciences, 1: 311-326.
Müller, M., 1983. Bodenbildung auf Silikatunterlage in der alpinen Stufe des Oberengadins. Diss. ETH
7352, Juris Druck, Zürich, 80 pp.
Ozenda, P., 1985. La végétation de la chaîne alpine dans l'espace montagnard européen. Masson, Paris,
330 pp.
Paquet, H., Coude-Gaussen, G. and Rognon, P., 1984. Étude minéralogique de poussières sahariennes
le long d'un itinéraire entre 19 et 35 degrés de latitude Nord. Rev. Géol. Dyn. Géogr. Phys., 25:
257-265.
Parfitt, R.L., Rüssel, M. and Orbell, G.E., 1983. Weathering sequences of soils from volcanic ash
involving allophane and halloysite, New Zealand. Geoderma, 29: 41-57.
Party, J.P., 1982. Inter-relations sol-faune en milieux à forte contraintes. DEA ENSA. M, USTL, Mont
pellier, 76 pp.
Pawluk, S. and Brewer, R., 1975. Micromorphological, mineralogical and chemical characteristics of
some alpine soils and their genetic implications. Can. J. Soil Sci., 55: 415-437.
Pochon, M., 1978. Origine et évolution des sols du Haut Jura Suisse. Phénomènes d'altération des
roches calcaires sous climat tempéré humide. Thèse Université de Neuchâtel, 190 pp.
Quinif, Y., 1976. Contribution à l'étude morphologique des karst algériens de type haut-alpin. Rev.
Géogr. Phys. Géol. Dyn., XVIII: 5-18.
Richard, L. and Pautou, G., 1982. Alpes du Nord et Jura Méridional. Notice détaillée des feuilles
Annecy et Grenoble. Edit. CNRS, F, 316 pp.
180 J.R Legros
Righi, D. and Lorphelin, L., 1986. Weathering of silt and clay in soils of a toposequence in the Hi
malayas, Nepal. Geoderma, 39: 141-155.
Righi, D. and Wilbert, J., 1984. Les sols sableux podzolisés des landes de Gascogne (F). Répartition et
caractères principaux. Sci. Sol, 4: 253-264.
Robert, M., Cabidoche, Y.M., and Berner, J., 1980. Pédogenèse et minéralogie des sols de haute
montagne cristalline. Etages alpin et subalpin — Alpes et Pyrénées. Sci. Sol, 4: 313-334.
zolisation. In: D. Righi and A. Chauvel (Editors), Podzols et Podzolisation. AFES-INRA, pp. 207-
223.
Samie , C , Robert, M., Legros, J.P. and Cabidoche, Y.M., 1981. Le milieu physique en montagne. In:
l'INRA et la Montagne, Tome 1. INRA, Paris, pp. 15-28.
Sehgal, J.L., Sys, C , Stoops, G. and Tavernier, R., 1985. Morphology, genesis and classification of two
dominant soils of the warm temperate and humic region of the Central Himalaya. J. India Soc. Soil
Sci., 33: 846-857.
Singer, M. and Ugolini, E C , 1974. Genetic history of two well-drained subalpine soils formed on
complex parent materials. Can. J. Soil Sci., 54: 475-489.
Smith, C.A.S., Coen, G.M. and Pluth, D.J., 1981. Podzolic soils with luvisolic-like morphologies in the
upper subalpine subzone of the Canadian Rockies. 1 — Stratigraphy and Mineralogy. Can. J. Soil
Sci., 61: 325-335.
Smith, C.A.S., Spiers, G.A., Coen, G.M. and Pluth D.J., 1983. On the origin of Fe in some podzolic
soils formed on calcareous parent materials in the Canadian Rocky Mountains. Can. J. Soil Sci., 63:
691-696.
Souchier, B., 1971. Evolution des sols sur roches cristallines à l'étage montagnard. Thèse d'Etat,
Université de Nancy, 130 pp.
Souchier, B., 1979. Un exemple de séquence altitudinale dans les Carpathes roumaines. Comparaison
avec les Alpes Occidentales. Ann. Univ. Savoie., Sci. Nat., 4: 7-11.
Souchier, B., 1984. La pédologie dans ses rapports avec l'écologie. Evolution des concepts et applica
tions. Sci. Sol, 2: 149-165.
Spaltenstein, H., 1984. Pédogenèse sur calcaire calcaire dur dans les Hautes Alpes calcaires de Suisse.
Thèse 540, Doctorat des Sciences Naturelles, EPFL, Lausanne, 140 pp.
Sticher, H., Gasser, U. and Juchler, S., 1986. Die Böden auf Serpentinit bei Davos. Entstehung, Ver
breitung, Eigenschaften. Veröif. Geobot. Inst. Eidg. Tech. Hochsch., Stift. Rubel, Zürich, 87: 275-
290.
Teoharov, M. and Ninov, N., 1980. Some regularities in the distribution of soil cover in Ograzhden
mountain. Soil Sci. Agrochem., XV, 6 pp.
Trosset, L., 1980. Différenciation pédogenétique et organisation des sols sur micaschistes dans les Alpes
Nord Occidentales. Sciences du Sol, 4: 337-351.
van Oort, F, 1984. Géométrie de l'espace poral, comportement hydrique et pédogenèse. Thèse de
Docteur Ingénieur. INAPG, Paris, 281 pp.
van Oort, F. and Dorioz, J.M., 1983. Signification des propriétés hydriques des sols sur calcschistes en
alpage. 108ème Congrès National des Sociétés Savantes, Fase. 1(1): 101-114.
van Oort, F. and Robert, M., 1988. Rôle de l'oxydation du soufre dans les premiers stades de l'altération
superficielle des calcschistes (Beaufortain, Alpes du Nord). Conséquences générales sur l'évolution
pédologique. C. R. Acad. Sci., Paris, Sér. II, 307: 819-825.
van Ryswyk, A.L. and Okazaki, R., 1979. Genesis and classification of modal subalpine and alpine soil
pedons of South-Central British Columbia. Canada. Arct. Alp. Res., II: 53-67
Verger, J.R, 1987. Végétation et pédogenèse sur roches vertes et gneiss acide dans une séquence
altitudinale montagnarde alpine en Val d'Aoste (Italie). Thèse d'Etat, Université de Grenoble, 2
volumes, 216 and 100 pp.
Verger, J.P., 1989. Evolution granulométrique des solums et action sur la végétation. Exemple en Val
d'Aoste sur roches ultrabasiques, basiques et acides. Sci. Sol, 27:157-170.
Void, T. and Marsh, R.D., 1981. Determination of soil climate classes and soil families for a level 3
survey in the Cordillera. Can. J. Soil Sci., 61: 203-210.
Wilcox, B.P., Allen, B.L. and Bryant, EC, 1988. Description and classification of soils of the high-
elevation grasslands of Central Peru. Geoderma, 42: 79-94.
Wilke, B.M., Mishra, V.K. and Rehfuess, K.E., 1984. Clay mineralogy of a soil sequence in slope deposits
derived from hauptdolomit (Dolomite) in the Bavarian Alpes. Geoderma, 32: 103-116.
Zebrowski, C, 1975. Étude d'une climato-séquence dans l'Ile de la Réunion. ORSTOM, Sér. Pédol.,
XIII: 255-278.
183
Chapter 8
Soils of cold climate regions

LB. CAMPBELL and G.G.C. CLARIDGE
Introduction
The world's coldest climates are found in the north and south polar lands and
within these regions temperature has the greatest influence on weathering and soil
development. The Arctic and Antarctic regions are frigid because they receive little
solar radiation owing to their high latitude. In the atmosphere, low humidity and
precipitation are a consequence of the cold while significant amounts of fluid water
are present in the soils only for short periods during warm months. Thus, water
is largely unavailable for weathering and translocation of the weathering products.
With increasing latitude, chemical weathering which is the dominant weathering
process in tropical and temperate regions, declines in importance, while physical
processes become much more important in soil formation. The soils of cold climate
regions differ therefore in many ways from those of temperate regions of the earth.
Definition of cold climate lands
We consider cold climate high latitude regions as those areas in which permafrost
occurs. Permafrost describes soil or rock material which remains below 0°C over
at least two consecutive winters and the intervening summer (Brown and Kupsch,
1974). If there is little or no moisture between component particles, the soil is said
to be dry-frozen, and if water is present, the soil is ice-cemented. Because most
areas in which permafrost occurs have a positive moisture balance, the permafrost is
usually ice-cemented and forms an extremely tough deposit.
The position of the upper surface of the permanently frozen material is con
trolled by the depth of penetration or loss of supergene energy and occurs at a
depth where the maximum annual temperature is 0°C. The zone above this level
thaws annually and is called the active layer. It is usually a zone of intense physi
cal disturbance due to freeze and thaw processes. In the Arctic region where there
184 LB. Campbell and G.G.C. Clarìdge
may be greater extremes between summer and winter climates, the thickness of the
active layer may exceed 1 metre but in the Antarctic it is generally much less.
Characteristic processes of cold climate lands
Influence of ice
Glaciers have had a profound influence on cold climate lands and soils through
their widespread periodic removal of weathering mantles and extensive deposition
of new materials. In parts of the Arctic, massive ice invasions and retreats from
various centres have been widespread and rapid. Weathered mantles are commonly
stripped while extensive till sheets, ranging from lithified to unconsolidated, fine to
coarse textured and young to very old, have been deposited along with widespread
resorted alluvial or aerosolic deposits laid down during stages of rapid glacial re
treat. Many surfaces in the Arctic were covered by ice during the last glacial episode
and most exposed surfaces are less than 10,000 years old. Some areas in inland
Canada and Siberia were not covered by the last continental ice sheets, but were
probably glaciated during earlier glacial episodes. Some surfaces on Baffin Island
Fig. 8.1. A soil with a sand wedge, Yukon Valley, northern Canada.
Soils of cold climate regions 185
and probably elsewhere were weathered during interglacial periods and have been
exposed as nunataks during the last glaciation also (Bockheim, 1979).
The Antarctic ice sheets are much more stable because of the greater cold and,
as they are commonly frozen to their beds, glacial erosion is minimised. Because
the amount of water available for sediment sorting is negligible, tills and exposed
bedrock form the major parent materials.
The most important processes occurring in cold climate land soils are the annual
freeze/thaw cycles, the effects of which are manifest as ice wedges and patterned
ground (Figure 8.1). Sand is incorporated into the soil and regolith by these pro
cesses. The widespread occurrence of fossil sand wedges in northern hemisphere
cold climate lands is probably indicative of former arid conditions.
Weathering regimes
The soils of the cold climate regions can be distinguished from those of other
parts of the world by the greatly restricted weathering of mineral materials and
consequent slow release of nutrients. With increasing latitude and progressively
cooler climate, weathering diminishes. The extreme is reached in some of the most
arid Antarctic regions where weathering is confined largely to the slow release of
iron oxides while atmospheric inputs of nutrients, such as nitrates or sulphates,
remain in the soil as distinctive salt horizons through lack of leaching.
Soil biological regimes
In cold climate lands, biological activity is at a very low level and has minimal im
pact on soil weathering processes. At lower latitudes, the more favourable moisture
and temperature regimes allow the existence of Subarctic forest, but with increas
ingly severe conditions, these give way at first to scrub and grass tundra, moss or
lichen tundra, andfinallyinto cold stony deserts with very sparse lichen. As the tem
perature decreases, there is less humification of organic matter owing to diminished
levels of microbial activity, consequently, where conditions are favourable for plant
growth, peat may accumulate.
Zonation of cold climate lands
Definition of zones
As latitude increases, there are marked changes in the appearance of the cold
climate lands, most notably in vegetation patterns which are governed by climate,
but also in soil and landscape features (Figure 8.2). Proximity to oceans, altitude,
precipitation pattern and glacial history all affect soil-plant relationships so that the
SUBARCTIC TUNDRA fc SUBPOLAR | POLAR COLD

£ DESERT ■£ DESERT DESERT
o o
Coniferous forest
p
shrubs * M scattered
grasses sedges mosses plants microorganisms
1PokM T<
Wet sites Gley podzols Peat
Dry sites Podzols Arctic brown soils Polar Desert soils Cold Desert Soils
Discontinuous Continuous carbonate accumulations salinity
Permafrost
Fig. 8.2. The relation between climate, vegetation and soils in cold climate regions.
changing pattern is not strictly latitudinal. Nevertheless, cold climate regions can be
divided into a number of distinct zones, which have been described in various ways.
The forest (taiga) and tundra zones are clearly recognised, but it is only relatively
recently that subdivisions of the regions beyond the tundra have been recognised.
In Canada, the terms "Low Arctic" and "High Arctic" have been used. In the
USSR the presence of deserts in northern regions was recognised relatively early
(Gerasimov, 1956). In this chapter, we follow the zonation of Tedrow (1968), who
used the subdivision into Subarctic, T\indra, Subpolar Desert, Polar Desert and Cold
Desert.
Subarctic zone
The Subarctic zone is the zone of discontinuous forest, underlain by permafrost,

south of the tundra. It occurs in the northern fringe of Scandinavia and European
Russia, extends well to the south in Siberia, most of Alaska and northern Canada.
Its northern boundary is the tree line and its southern boundary is the occurrence
of forests with a closed canopy. It is essentially a transition zone between forest and
non-forested land.
Because of the continental nature of much of the Subarctic zone, it has long, very
cold winters and short, cool summers. The coldest winter temperatures recorded in
the northern hemisphere are from the Subarctic zone of central Siberia. Summer
temperatures are high enough, however, for trees to grow. Precipitation is relatively
low, between 300 and 400 mm per year and is generally evenly distributed through
out the year, about half falling as snow, which is not available for soil processes until
the spring thaw.
Most of the Subarctic is forested, predominantly with conifers. In Eurasia the
dominant trees are Scots pine, spruces and larch. The forest vegetation is better
developed on well-drained areas free of permafrost. In the poorly drained sites,
where permafrost is present, the forest becomes open, and an understory of willows,
birch and sedges develops. Wet sites are boggy, with a thick ground cover of various
species of mosses, low shrubs, and a scattered and stunted forest cover.
TUndra zone
The Tlindra zone is found around the margin of the Arctic ocean north of the
tree line, with the exception of some coastal areas. Its northern limit is generally
considered to be the 4.4°C July isotherm (Tfedrow and Brown, 1962), but the bound
ary is very approximate and relates more to the number of days during which the
temperature is above 5°C in any year. The Tundra zone lies between 120-200 degree
days and 15 degree days. The tundra zone has cold winters and cool summers, with
relatively low precipitation. The period between thaw and freeze-up is generally
about 90 days. During this time temperatures are almost continually above freezing
and the ground may thaw to 50 cm or more.
There is a gradation in vegetation from north to south in the Tundra zone.
The transition from forest (taiga) to tundra is relatively sharp, with only scattered
stunted trees found in sheltered sites beyond the tree line. Beyond this limit scrubby
plants and reeds occur. With increasing latitude the plant cover diminishes in height
and mosses and lichens become more important. On dry sites dwarf heath and
willows, reeds, sedges, herbs and mosses occur. On wet sites, tussocky clumps of
sedge, low shrubs and mosses, especially sphagnum, are found. The plant cover is
continuous where the site is undisturbed.
Subpolar Desert zone
The Subpolar Desert is regarded as a transitional zone between Tlindra and

Polar Desert (Tedrow, 1977). It has a cold and dry climate, with precipitation gen
erally less than 400 mm, and temperatures ranging from 8° in summer to -30° in
winter. The increasing cold and aridity is accompanied by increased intensity of
physical weathering, with wind erosion and frost shattering assuming greater signifi
cance.
Because of the severity of the climate, the vegetation consists mainly of scattered
mats of Dryas, Salix and Saxifragia, with occasional sedge-moss marshes. Vegetation
cover is generally discontinuous in the drier sites but a continuous cover can be
found on poorly drained sites. The variety of plants and their relation to specific
sites has been described by Edlund (1984).
Polar Desert zone

The Polar Desert comprises the dry barren landscapes of the Arctic region which
have only sparse ground cover. The Polar Desert zone comprises the northern
188 LB. Campbell and G.G.C. Claridge
islands of Svalbard, the islands of Severnaya Zemlya, to the north of Taimyr Penin
sular in the Siberian Arctic, the Canadian Arctic north of about 75°N and the
northern coastal fringe of Greenland. The actual boundary of the zone is much
more complex, and contained within it are inclusions of the Subpolar Desert and
Tlindra zones, occupying sites where the environment is less severe. Since the gov
erning factor of the zone is temperature, or the solar radiation input which can be
used for plant growth, the microclimate of any particular site is important. Thus,
even in Inglefield Land, north Greenland (Tfcdrow, 1970), although Polar Deserts
predominate, T\indra soils and Arctic brown soils can be found where moisture can
accumulate such as in valleyfloorsand gullies with a southerly aspect.
The essential features of the climate of the Polar Desert zone are prevailing cold,
with only a very short period in which temperatures are above freezing in summer,
and aridity. Precipitation, which mainly falls as snow, is low (generally <150 mm)
because of the influence of the almost perennially frozen Arctic Ocean. Bovis and
Barry (1974) found that the Polar Desert zone could be clearly distinguished from
the transitional zones and the Tundra zone by using various statistical methods of
calculating the water budget, taking into account net precipitation, evaporation and
radiation. In the Polar Desert zone the net water budget was zero or less, while sites
in the Subpolar Desert zone had water budgets less than 50 mm.
The Polar Desert zone is almost completely devoid of vegetation. Bare ground is
extensive with only isolated and scattered plants such as papaver or saxifragia. On
some slopes where there is a favourable supply of moisture a heath meadow may
develop, while in swampy valley floors Carex may form a continuous cover. Plant
growth is limited by water availability, the short growing season and by salinity,
a consequence of the negative water budget of most soils. Despite their appar
ent sterility, soils of the Polar Desert contain a considerable microbiological flora,
although these have a relatively minor influence on soil properties (Cameron, 1972).
Cold Desert zone
The Cold Desert zone is restricted to Antarctica, and possibly parts of northern
Greenland. In the Cold Desert, mean annual air temperatures are everywhere below
freezing, while summer maxima are close to 0°C in coastal regions and around
-30°C in inland areas (Weyant, 1966). Surface ground temperatures during summer
however, are frequently above 0°C, sometimes for several days at a time, and in
favourable situations may reach 20°C, but wide daily fluctuations are common.
Below about 40 cm, the soil temperature is seldom above 0°C. Moisture supply is
from snowfall or from drifting snow, and varies between 5 and 60 g cm"2 per year.
In inland regions, the soil is probably never wet but in coastal regions the upper few
centimetres of the soil may be moistened several times a year by snow melt. Some
moisture may be contributed by thawing, or sublimation of underground ice or
permafrost. Because of the extreme cold and aridity, the amount of water available
for weathering processes and plant growth is severely restricted. In the Cold Desert
proper, plant life is confined to the occurrence of very small patches of lichens or
mosses in favourable areas.
Soils of the Subarctic zone
The soils of the Subarctic zone that are underlain by permafrost, tend to have
impeded drainage and grade from the Podzols of well-drained sites to gley Podzols
and peats on wet sites. Profiles are characterised by a thick layer of Utter, plant roots
and fungal mycelia underlain by an O horizon of black amorphous organic matter
with some mineral matter and a mottled grey to yellow-brown silt or sand horizon
near the permafrost layer (Figure 8.3). In some areas, especially on the tops of
mounds where drainage is better, the mineral layer may have dark brown or dark red
colours, and be somewhat more friable. The red colours are due to goethite (Pawluk
and Brewer, 1975), often associated with expanding clay minerals. As drainage
becomes impeded the organic layer, consisting of fibrous and spongy remnants
of Sphagnum, with occasional fragments of trees and heath shrubs, thickens and
becomes more acid. In general, the thicker the active layer above the permafrost the
denser and taller the forest.
Msr- · ^ "*:*V»; % Λ Α p « A S
:
I^WÉÌI
r Sì":f' *
■ : ■ " * *
. ' V « ' \,'V

*
i%^Sik "
4
% *>:%Ρ^ , ^.
*%
SfcÇ* :
\, ,t
% ;- ■ »
^. ™ « β ^ ^ % ί | ^ ~'
^%T.f*4 V >U
*Α**Β . *
Fig. 8.3. A shallow soil of the subarctic, overlying permafrost, Yukon Valley, northern Canada.
190 LB. Campbell and G.G.C. Clonage
Soils underlain by permafrost, even the forested soils of the Subarctic zone are
strongly cryoturbated. Thus much of the Subarctic zone, in common with the zones
to the north, is covered by patterned ground. On lowland sites under forest the
most common feature is hummocky terrain (Zoltai and Pettapiece, 1973; Zoltai and
Tàrnocai, 1974). Hummocks are mainly circular, 40-50 cm in height and between 1
and 3 m in diameter. Soil horizons are discontinuous and fragmented, with remnants
of material from other horizons incorporated. Less than 10% of the trees grow on
the hummocks, most growing on the sides and in the troughs. Mosses form the
ground cover in the troughs while lichens dominate the tops of hummocks.
In the soils of the Subarctic zone there is evidence for three major processes: ac
cumulation of organic matter, faunal activity and physical processes due to freezing
and thawing. Chemical and mineralogical data show a slight to moderate degree of
weathering, less in the north than in the south (Pawluk and Brewer, 1975; Wilde and
Krause, 1960). This indicates that pedogenic processes active in the development of
podzolic soils are also present in these soils underlain by permafrost.
The main features of soil formation on silicate rocks in the Subarctic permafrost
are a low rate of weathering of the mineral material and slow decomposition of the
organic matter. Both of these are controlled by low soil temperatures. As a result
the soils have an acid reaction, are leached and have low base saturations. Mineral
transformations are limited and are controlled by mobile and aggressive organic
compounds on the mineral part of the solum. Where leaching is not inhibited by
the presence of ice at the base of the profile, and soil temperature rises, mineral
breakdown and translocation of the products takes place, forming podzols. Where
drainage is impeded or slowed, gleying becomes the dominant process.
Soils of the Tundra zone
Wet and dry tundra

Because most soils of the Tbndra zone are formed in heavy-textured alluvium in
low-lying situations where the permafrost is not far below the surface, these soils
are poorly drained and extremely wet when thawed. These are the true Tbndra soils.
However, where the soils are better drained such as on ridge crests, gravelly dunes
or on the edge of escarpments the soils develop in a different manner.
Soils of the wet tundra

The soils of the wet tundra are saturated with water and rich in organic mat
ter. Horizons are discontinuous and interrupted as a result of cryoturbation. An
idealised Tundra soil has a surface horizon of up to 15 cm of dark brown to black
organic matter consisting of partially decomposed plant material overlying 10-30
cm of light olive brown to grey, very wet silt loam, overlying very dark grey, per
manently frozen silt loam. Organic layers are interspersed throughout the profile
and even within the permanently frozen material (Tedrow et al., 1958; Everett and
Brown, 1982). The rate of decomposition of the organic matter in wet T\indra soils
is extremely slow. Douglas and Tbdrow (1959) found that inorganic matter incorpo
rated in such soils was up to 8000-9000 years old.
Pedological processes in wet Tundra soils are dominated by gleying and by the
organic cycle. Mineral weathering is very slow, because of the prevailing cold con
ditions and restricted movement of water. Clay minerals present reflect the compo
sition of the parent material, generally fine textured alluvium and are largely 2:1
layer silicates, although kaolinite has been reported. Where volcanic ash layers are
encountered, as in the soils of Alaska or of the Aleutian Islands, allophane may be
present. Authigenic smectites are found in some Tbndra soils.
Over much of the wet tundra landscape the genesis of the soil is related to the
cycle of development of the landscape itself. The dominant process is the formation,
migration and disappearance of thaw lakes (Everett and Parkinson, 1977). Lakes
form by the disruption of soil cover, generally in association with an ice lens, fol
lowed by the melting of sub-surface ice and the formation of a melt pool. Sediments
accumulate in the pool, while peat-forming plants grow around the edges. The lakes
eventually drain as other interconnected lakes form, leaving a very wet site where
highly organic soils and peats form over the lake sediments.
Eventually the pattern of ice wedges becomes re-established in the drained
basins, and a polygonal microrelief forms. As the wedges grow, contrasts develop
between troughs and centres with the formation of either raised-centre or raised
edge polygons. Raised areas become better drained, organic matter decomposes
more readily and mineral weathering with translocation of iron or carbonates may
occur. The soils resemble those of the dry tundra, and the plant cover becomes
sparse, mineral soil may be subject to wind erosion and subsurface ice becomes
exposed, with the initiation of a further cycle.
Soils of the dry tundra

Although wet Tbndra soils dominate in terms of area, their properties are largely
a consequence of gleying caused by impeded drainage. Well-drained soils, the dry
Tbndra soils, show best the relationship between weathering processes and environ
ment, and the relation between soil formation in cold regions and in other parts of
the world.
Well-drained soils are found on ridge tops, terrace edges, gravelly outwash, dunes
and old beach ridges. On these sites, where water can drain away during the thaw
season, the soil is dry in summer and dry-frozen in winter. The vegetation at these
dry sites is somewhat more xerophytic than that of the wet tundra, and consists of
mosses, lichens, heaths, herbs and in some places dwarf willows, while the sedges
and reeds which dominant the wet sites are much less important.
The dominant soil of the dry tundra is the Arctic brown soil (Tedrow and Hill,
1955), although Podzols, Rendzinas and other soils are found. Arctic brown soils
are considered to be related genetically to the Podzols and differ in their properties
because genetic processes weaken with decreasing temperature and precipitation.
Arctic brown soils are characterised by dark brown to yellow-brown (10YR 3/2-
10YR 3/4) colours, formed by humus staining of soil particles and mineral grains,
and a fine crumb to single grain structure. They are usually strongly acid and have
a high concentration of organic matter in the surface horizon, gradually decreasing
with depth. There are no obvious indications of horizons of accumulation, although
a dark stain is sometimes present at the surface of the ice cement, indicating some
translocation takes place. In the T\indra zone of Siberia, soils of the dry sites are
described either as Al-Fe humus soils or Podburs: the former shows more pod-
zolic features than the latter, but being essentially similar to Arctic brown soils
(Tkrgulyan, 1971).
Mineral weathering is minimal in dry Tbndra soils. Most soils of the dry Tundra
zone of Alaska, for example, contain some kaolinite but this is derived from the
parent material, sandstones of Mesozoic to Tbrtiary age. Other minerals present
include mica and slightly weathered illite showing some interstratification. Ttaces of
smectites are found in those soils which show podzolic features. However, because
of the relative youth of most of these soils, there has been little time for significant
mineral transformation.
Soils of the Subpolar Desert
Because of the transitional nature of the zone, soils resemble those of the zones
to the north and to the south. Soils resembling those of the wet T\indra are found
on poorly drained sites, which occur in narrow strips close to water courses. On well
drained sites soils resembling Arctic brown soils, but somewhat less developed, are
found. Colours are not so intense and, organic matter contents not so high as in
soils of the dry Tbndra, principally because of limited plant growth. Frost action and
cryoturbation are more common however. The soils are gleyed at depth, especially
when fine-textured. There may be some translocation of iron or mineral matter
implying that podzolisation may still affect soil development to a limited extent. The
soils are usually alkaline and salts effloresce at the surface in vegetation-free areas
such as salt boils. Some of these soils are somewhat older than those of the Tbndra
zone as parts of the Canadian Arctic Archipelago have escaped glaciation during the
last glacial episode. Both on Banks Island (Tbdrow and Douglas, 1964) and around
Cumberland Peninsula, Baffin Island (Bockheim, 1979), soils are found that may be
more than 100,000 years old.
Soils of the Polar Desert zone
Polar Desert soils are well-drained and horizons are poorly developed except for
the presence of desert pavements. A typical profile shows a virtual absence of an
A horizon, and the predominance of a brownish-yellow B horizon below the desert

pavement. Soils are commonly neutral to alkaline, contain less than 1% organic
matter and are saline. They are usually coarse-textured, consisting mainly of sand
and gravel and are structureless. Soil texture is, however, controlled largely by the
lithology of the parent material, and to a lesser extent by weathering processes.
Where plant cover is present, soils sometimes contain sufficient organic matter and
fine roots to allow a weak fine granular structure to develop.
On Ellef Ringnes Island, for example, soils were relativelyfine-textured,because
of the nature of the parent materials, sandstones or shales (Foscolos and Kodama,
1981). Consequently, the soils contain very large amounts of clay, up to 75% in
one instance and high exchange capacities. Soil pH ranged from 3 to 7. The soils
contained only traces of soluble salts. Smectite, mica and kaolinite are the principal
clay minerals, although quartz, feldspar, pyrite and jarosite also occur. The layer sil
icates were derived virtually unchanged from the parent rocks but the high acidities
indicate that weathering of pyrite, present in the parent material, has taken place
forming jarosite or natrojarosite, and producing high acidities, a point of contrast
with most soils of this zone.
Elsewhere in the Polar Desert zone of northern Canada kaolinite is found
(McDougall and Harriss, 1969; Bockheim, 1980). In contrast to the acidic, heavy-
textured soils of Ellef Ringnes Island, the soils of Melville Island or of Cornwallis
Island are mainly derived from coarse-textured glacial sediment and are themselves
coarse in texture. They are alkaline to neutral in reaction, except for the surface
horizon which, because of the influence of the small amount of raw organic mat
ter, is acid. The dominant minerals in the clay-size fractions are mica, vermiculite,
chlorite and kaolinite.
Bockheim (1979) found that, while the proportion of mica increased with depth
in some of the soils of Baffin Island, the proportion of kaolinite decreased. He
considered that this indicated that kaolinite was being formed at the expense of
mica or feldspar, while mica was also being transformed to vermiculite by hydration
and loss of potassium. Alternatively, kaolinite could have been forming directly
from solutions or gels containing aluminum and silicon during thaw periods, in
acid soils in which calcium and magnesium ion concentrations were low. Since
kaolinite concentrations were greater in the upper part of the profile, especially in
older soils, formation in situ was indicated. However, McDougall and Harriss (1969)
considered that the high concentration of kaolinite in marine sediments of the
Arctic Ocean, was a result of material transported by wind from the older kaolinite-
bearing sediments on land, and therefore an aeolian origin cannot be ruled out for
kaolinite in the soils of Baffin Island.
Soils of the Cold Desert zone
Distribution of Cold Desert soils
Cold Desert soils have recently been described in some detail by Campbell and
Claridge (1987). They occur on the scattered ice-free areas of the Antarctic conti
nent, largely on deposits of till or bedrock (Figure 8.4). Weathering and soil forma
tion range from almost nothing on young surfaces (<50,000 years) to well weath
ered (by Antarctic standards) on the oldest surfaces (Miocene or older; Campbell
and Claridge, 1975). On well-weathered surfaces, typical Cold Desert soils have a
surface stone pavement, indistinct horizon development but some staining which is
at a maximum just below the surface and which diminishes downwards. They are
also coarse in texture, commonly bouldery or pebbly sands weakly consolidated by
finer materials. They have a slightly acid to alkaline reaction and high salinity, with
salts that are principally chlorides, nitrates and sulphates of sodium, potassium and
magnesium. Organic matter contents are extremely low. While many soils contain
detectable populations of micro-organisms, many sites are devoid of plant or ani
mal life. The most distinctive feature of the soils, however, is a frozen ground or
permafrost horizon, usually within 1 m of the soil surface. The main soil features of
Cold Desert zone soils are discussed below.
Fig. 8.4. Wright Valley, Antarctica: a typical landscape of the Cold Desert.
Sous of cold climate regions 195
Soil horizons
Several somewhat indistinct horizons may occur within Cold Desert soils, includ
ing a pale coloured surface horizon, an oxidised horizon and a salt horizon. The
surface horizon commonly merges with the stone pavement and generally comprises
a layer of loose coarse-textured material derived largely through fragmentation of
surface clasts. In older soils this horizon may be up to 10 cm thick and is comprised
of strongly stained rock fragments. The fine material in the surface horizon is usu
ally pale-coloured and often forms a matrix which gives the whole horizon a weak
cohesion. Where the proportion of fine material is high, weak vesicular structure
may develop and a thin surface crust may also be present. These features are best
developed in regions where the soil surface is occasionally moistened by summer
snow melt. Oxidised horizons vary considerably in respect of colour and the inten
sity and depth of colouration. The colour is a result of slow oxidation of iron-bearing
minerals and is usually a thin coating of iron oxides on mineral grains (Claridge and
Campbell, 1986). The colour of the soil varies approximately with the amount of
iron extracted by citrate/dithionite (Table 8.1). Soil colours also related to the types
of rocks forming the parent material and also the age of the soil. When unweath-
ered, the soil is usually lithochromic, typically grey to light olive brown (5Y 5/1-2.5Y
5/4) in quartzo-feldspathic parent materials, or dark greyish brown (10YR 4/2) in
iron-rich (diabasic) parent materials. Weathered soils from quartzo-feldspathic par
ent materials may be yellowish brown (10YR 5/6), and those from iron-rich parent
materials yellowish red to dark red (5YR 4/8-2.5YR 3/6). The colour is most in
tense between 10-30 cm depth and diminishes downwards, but may persist in some
soils to >1 m. In moist hollows, where salts accumulate, horizons are dominantly
olive coloured (2.5Y 5/4-4/4).
Salt accumulations are a common feature of Cold Desert soils and are of several
types. Surface efflorescences are usually thin powdery accumulations but sometimes
may be thick encrustations, found mainly in moist sites where there is pronounced
water loss through evaporation. Calcite crusts and veneers are more widespread and
often form prominent coatings on the undersides of surface stones. Salt horizons
within profiles may be either diffuse or discrete. Diffuse salt horizons have small
concentrations of salts scattered throughout the profile with greatest concentrations
in the upper 30 cm, while discrete salt horizons are layers of salts only, up to 15
cm thick and about 20 cm below the surface (Figure 8.5). The composition and
abundance of the salts in the soils vary in relation to both source and local climate.
For example, soils of coastal regions have a high content of the chlorides of sodium,
magnesium and calcium reflecting a dominantly marine aerosolic origin. In colder
and more arid inland regions, chloride decreases in proportion, whereas nitrate
and sulphate, derived from upper atmospheric circulation become dominant. The
amount of salt present, in a well drained soil, is related to soil age.
In most soils the 0°C isotherm only penetrates between 20 and 50 cm, and thus
196 IB. Campbell and GGC Clonage
Fig. 8.5. A typical soil of the Cold Desert, in Wright Valley, Antarctica, showing a desert pavement,
coarse texture, and a salt horizon.
this is the true depth of the permafrost. The soil below this depth is dry frozen
for a further 50-150 cm, and the soils are underlain by ice-cement, which forms an
effective base to the soil profile. The depth to ice-cement is related to the availability
of moisture. In most cases, weathering is confined to the material above the ice-
cement. Where weathered material is found within the ice-cement it indicates a
change in conditions at some time in the past.
Soil particle size
The Cold Desert soils are invariably characterised by their coarse particle size.
Soils formed from till frequently contain about 60-90% of material >2 mm, much
of this being boulder-sized. Soils weathered from bedrock may have fewer very
coarse particles but still have a relatively small proportion of material <2 mm in
size. The surface horizon generally has the highest amounts of material >2 mm. In
older soils, particle size is distinctly finer than that found in young, unweathered
soils (Table 8.1). Coarse-grained rocks within profiles are sometimes disaggregated
through salt weathering processes (Wellman and Wilson, 1965), leaving ghosts, the
crumbled outlines of former clasts.
Many old weathered soils show a small but distinct increase in proportion of
both silt and clay fractions, the latter sometimes reaching more than 5%. Sand, pro-
TABLE 8.1
Selected chemical properties of four soils of the Cold Desert, showing changes in CBD extractable iron
content, conductivity and salt chemistry with degree of weathering, age and distance from the sea.
Depth < 2 mm <0.002 mm CBD PH Cond. TSS a Ions in extract (meq per 100 g)
(cm) (%) (%) Fe203b (mmho (%)
(%) cm-1)
Na Mg Cl so4 N03
Soil 1. Moderately weathered, coastal, 0.5 m.y

0- 5 51 2.1 0.2 8.9 5 2 174 9 15 10 1
5-6 64 1.6 0.2 7.7 140 49 1087 47 1530 57 4
6 - 20 64 1.8 0.3 7.9 44 15 278 40 249 33 14
40- 60 65 1.7 0.3 7.8 15 5 63 13 64 6 5
100-120 65 1.7 0.2 8.7 3 1 11 3 12 5 1
150-160 58 1.2 0.2 8.0 1 1 9 4 8 3 1
Soil 2. Moderately weathered, coastal, 2.1 m.y.
0- 2 84 2.9 0.2 9.1 3 <1 9 8 8 <1
2- 4 72 3.8 0.2 9.1 64 22 728 20 345 186 21
5-15 72 5.2 0.4 8.1 60 21 387 69 289 69 35
30-40 45 3.5 0.3 7.8 20 7 98 17 97 12 4
40-80 68 2.1 0.5 8.8 11 4 53 14 53 10 1
Soil 3. Strongly weathered, inland, >3.5 m.y
0- 3 71 1.4 0.5 6.3 0.5 0.2 1 1 <1 2 <1
3- 6 92 2.2 0.6 6.8 42 15 150 74 <1 140 90
6-12 86 2.0 0.8 6.3 29 10 150 138 <1 82 170
12-15 68 2.9 0.6 6.1 5 2 10 15 <1 15 17
25-45 60 2.9 0.6 6.8 3 1 10 14 <1 10 11
Soil 4. Strongly weathered, inland, >>3.5 m.y.
0-5 60 1.2 0.8 4.8 1.4 0.5 7 3 <1 6 <1
5-8 90 1.2 0.9 5.1 15 5 35 31 <1 38 40
8.18 90 1.0 1.6 5.4 54 20 200 123 <1 57 300
18.36 66 0.8 1.1 7.1 5 2 25 10 <1 11 20
The apparent age of the surface on which the soils has formed is indicated.
a
Total soluble salts.
b
Citrate-bicarbonate-dithionite soluble Fe2C>3.
duced by wind erosion of surface rocks, may be assimilated into the soils when it is
trapped in patterned ground depressions and worked through the soil by cryoturbic
movements.
Chemical weathering and soil mineralogy

Chemical weathering processes are operating in soils of the Cold Desert, as seen
by the red and brown staining on rocks, the oxidised horizons present in soils and
the chemistry of the soluble salts, some constituents of which are derived from rock
weathering. Chemical weathering processes also lead to the presence of authigenic
clay minerals in soils, although the processes are much slower than comparable
processes found in the Polar Desert or TUndra zones, their effects can be seen
because of the much greater age of the soils, many of which have been exposed to
weathering throughout the Quaternary.
In many soils the dominant clay size minerals are mica and illite, derived largely
from fine-grained mica from sandstone sediments or from schists. During weath
ering these minerals increase in hydration, and transform through illite to hydrous
micas and vermiculite. However, in the presence of abundant magnesium ions in
alkaline environments, such as those in undrained hollows and depressions where
moisture and salts are abundant, smectites may form.
Although kaolinite minerals have not been found in soils of the cold desert,
except where sediments containing it form part of the parent material, aluminium
hydroxides may precipitate in the interlayer space of vermiculite forming hydroxy
interlayered vermiculites or aluminium chlorites.
During the weathering of diabase, in situations where sufficient moisture is avail
able, ferromagnesian minerals such as pigeonite can decompose, forming iron-rich
nontronite and leaving a residue of iron oxides, while in the coldest and most arid
situation, where moisture is absent, this transformation does not take place and the
weathering products remain amorphous (Claridge and Campbell, 1986).
Pedogenic gradients in cold climate regions
In cold climate regions there is a general weakening of pedogenic processes

with increasing latitude, and the soils change in character, as the balance between
the various processes alter. Tbdrow (1968) described these gradients, while Ugolini
(1986) related them principally to the nature of the proton donors in soil solution.
The soils of the Subarctic zone, immediately south of the Tundra, are dominantly
Podzols. In these soils the most significant proton donors are organic acids, which
both attack soil minerals and chelate iron and aluminium. These elements can then
be translocated downwards and released as the organic acids break down. They
then recrystallise and precipitate. This podzolising process also operates but with
diminishing intensity well within the Tundra zone.
With decreasing temperature, the ability of the organic matter to decompose
to appropriate organic acids diminishes, and the principal proton donor in soil
solution becomes carbonic acid, produced by respiration of roots, and rendered
more important by the increased solubility of C0 2 at low temperatures. Carbonic
acid cannot form complexes with humus, iron or clay, so these accumulate in the
surface horizons, staining it brown. The process is termed bonification. It operates
to some extent in all soils, but its effects are usually marked by more vigorous
processes such as podzolisation. It supersedes podzolisation in colder soils and is
the principal soil forming process operating in Arctic brown soils. It persists into the
Polar Desert zone.
With increasing latitude the amount of organic matter available decreases and
brunification becomes less significant. Oxidation of ferrous iron to the ferric state,
which takes place in all soils, then becomes the major pedological process operat
ing in Polar Desert and Cold Desert soils to produce soil horizons. In cold soils
oxidation is extremely slow, because of the absence of proton donors such as or
ganic acids or carbonic acid. Dissociation of water may provide sufficient protons
for the reactions to take place. Although the reactions may be slow, sufficient time
is available for strong colours to develop especially in the very old soils of the Cold
Desert.
Other processes, unrelated to the abundance or nature of the proton donors,
also operate. One of these is particle mobility, or pervection and another is the
accumulation of salts. In soils of temperate regions and in podzols, clay illuviation
is an important process, but in Arctic brown soils, or in soils of the Polar Desert,
its place is taken by the movement of silt-sized particles, which form caps on the
upper sides of coarse fragments in the lower part of the profile (Bockheim, 1979).
The formation of silt caps is due either to upward movement of coarse particles
throughfine-texturedmaterial by the growth of ice lenses, or downward flushing
of silt (probably loessial) by thaw water. In Cold Deserts, where neither of these
processes operate, there is little evidence of silt movement.
The accumulation of salts, or salinization, is a process related to aridity. It occurs
to a limited extent in Arctic brown soils, where carbonates sometimes accumulate.
GLEY
TUNDRA
| _ PODZOL
Gleying Gleying
Podzolisation Cryoturbation
ARCTIC BROWN POLAR DESERT COLD DESERT!

PODZOL
SOIL SOIL SOIL
Podzolisation Brunification Oxidation Oxidation

Brunification Oxidation Calcification Salinisation
Salinisation
4.4°C
tree line July isotherm
Increasing Latitude (= decreasing temperature, increasing aridity)
Fig. 8.6. The relationship of soils of the polar regions to latitude and climate, and the processes operating
in them.
With increasing aridity and decreasing temperature the more soluble salts accu
mulate as well, thus Polar Desert soils have characteristic efflorescences of halite or
thenardite on the surface. The salts do not however accumulate as discrete horizons.
In the most arid soils of the Cold Desert, thick horizons of extremely soluble salts
such as sodium nitrate may form within the profile.
The relationship of the soils of the polar regions to latitude and drainage, and the
concurrent pedological processes is shown in Figure 8.6.
Summary and conclusions
In this chapter we have brought out the relationships of the soils of cold climate
regions to those of more temperate regions. Cold climate regions are dominated
by the prevailing cold temperatures and the presence of permafrost. Energy input
from solar radiation is low, so that plant growth is restricted, and decay of organic
matter is very slow. The presence of subsurface ice inhibits soil drainage so that
many soils are water saturated. However, with decreasing temperatures, there is a
decrease in precipitation and also in effectiveness of water-related processes in soils,
and consequently soils of cold climate regions are also arid.
There is a continuity of soil processes from temperate regions through to the
most extreme cold deserts of Antarctica which can be related to decreasing in
tensity of pedogenic processes with decreasing temperature. As different processes
decrease in intensity at different rates the balance between the various pedological
processes alters. TWo main groups of processes are involved, those in which acids,
or proton donors of various kinds, bring about changes in soil constituents and
downward translocation of soil constituents and those related to aridity, which cause
upward mobility of more or less soluble constituents. The limited number of soil
processes available and their very slow rates at low temperatures, result in extreme
fragility of the soils of cold climate regions. They have very limited capacity to heal
after disturbance, and in order to be able to use them without causing irreparable
harm, extreme care must be taken.
References
Bockheim, J.G., 1979. Properties and relative age of soils of southwestern Cumberland Peninsula, Baffin
Island, N.W.T., Canada. Arct. Alp. Res., 11: 289-306.
Bockheim, J.G., 1980. Properties and classification of some desert soils in coarse-textured glacial drift
in the Arctic and Antarctic. Geoderma, 24: 45-69.
Bovis, M.J. and Barry, R.G., 1974. A climatological analyses of north polar desert areas. In: T.L. Smiley
and J.H. Zumberge (Editors), Polar Deserts and Modern Man. University of Arizona Press, Tucson,
Ariz., pp. 23-31.
Brown, RJ.E. and Kupsch, W.O., 1974. Permafrost terminology. Nati. Res. Counc. Can., Tech. Mem.
111.
Cameron, R.E., 1972. Application of low latitude microbial ecology to high latitude deserts. In: T.L.
Smiley and J.H. Zumberge (Editors), Polar Deserts and Modern Man. University of Arizona Press,
Tucson, Ariz., pp. 71-90.
Campbell, LB. and Claridge, G.G.C., 1975. Morphology and age relationships of Antarctic soils. In: R.P.
Suggate and M.M. Cresswell (Editors), Quaternary Studies. R. Soc. N. Z. Bull., 13, pp. 83-88.
Campbell, LB. and Claridge, G.G.C., 1987. Antarctica: Soils, Weathering Processes and Environment.
Elsevier, Amsterdam, 368 pp.
Claridge, G.G.C, and Campbell, LB., 1986. Mineralogical changes in Cryic environments. Trans. XIII
Congr. Int. Soc. Soil Sci., 6: 1043-1052.
Douglas, L.A. and Tedrow, J.C.E, 1959. Organic matter decomposition rates in Arctic soils. Soil Sci.,
88: 305-312.
Edlund, S.A., 1984. High Arctic plants: more limits emerge. Geos, 13:10-13.
Everett, K.R. and Brown, J., 1982. Some recent trends in the physical and chemical characterisation and
mapping of Tundra soils, Arctic slope of Alaska. Soil Sci., 133: 264-280.
Everett, K.R. and Parkinson, R.J., 1977. Soil and landform association, Prudhoe Bay area Alaska. Arct.
Alp. Res., 9: 1-19.
Foscolos, A.E. and Kodama, H., 1981. Mineralogy and chemistry of Arctic desert soils on Ellef Ringnes
Island, Arctic Canada. Soil Sci. Soc. Am. J., 45: 987-993.
Gerasimov, I.P., 1956. Soil map of USSR. Acad. Sci. USSR, Moscow.
McDougall, J.D. and Harriss, R.C., 1969. The geochemistry of an Arctic watershed. Can. J. Earth Sci.
6: 305-315.
Pawluk, S. and Brewer, R., 1975. Investigation of some soils developed on hummocks of the Canadian
sub-arctic and southern Arctic regions. 2. Analytical characteristics, genesis and classification. Can.
J. Earth Sci., 55: 321-330.
Targulyan, V.O., 1971. Soil Formation and Weathering in Cold Humid Regions. Nauka, Moscow, 264
pp.
Tedrow, J.C.F., 1968. Pedogenic gradients of the polar regions. J. Soil Sci., 19: 197-204.
Tedrow, J.C.F., 1970. Soil investigations in Inglefield Land, Greenland. Medd. Groenl., 188, 96 pp.
Tedrow, J.C.F., 1977. Soils of the Polar Landscapes. Rutgers University Press, New Brunswick, N.J., 638
pp.
Tedrow, J.C.E and Brown, J., 1962. Soils of the northern Brooks Range, Alaska: Weakening of the
soil-forming potential at high Arctic altitudes. Soil Sci., 93: 254-261.
Tedrow, J.C.F. and Douglas, L.A., 1964. Soil Investigations on Banks Island. Soil Sci., 98: 53-65.
Tedrow, J.C.F., Drew, J.V. and Hill, D.E., 1958. Major genetic soils of the Arctic slope of Alaska. J. Soil
Sci., 9: 35-45.
Tedrow, J.C.F. and Hill, D.E., 1955. Arctic brown soil. Soil Sci., 80: 265-275.
Ugolini, F.C, 1986. Pedogenic zonation in the well-drained soils of the Arctic regions. Quat. Res., 26:
100-200.
Wellman, H.W. and Wilson, A.T., 1965. Salt weathering, a neglected geological erosive agent in coastal
and arid environments. Nature, 205: 1097-1098.
Weyant, W.S., 1966. The Antarctic climate. In: J.CF. Tedrow (Editor), Antarctic Soils and Soil Forming
Processes, Am. Geophys. Union, Antarct. Res. Ser., 8: 47-59.
Wilde, S.A. and Krause, H.H., 1960. Soil-forest types of the Yukon and Tanana Valleys in Subarctic
Alaska. J. Soil Sci., 11: 266-279.
Zoltai, S.C. and Pettapiece, W.W., 1973. Terrain, vegetation and permafrost relationships in the northern
part of the Mackenzie Valley and northern Yukon. Environmental-social Program, Task Force on
Northern Oil Development, Govern. Can. Rep. 73-4,105 pp.
Zoltai, S.C. and Tarnocai, C , 1974. Soils and vegetation of hummocky terrain. Environmental-social
Program, Task Force on Northern Oil Development, Govern. Can. Rep. 74-5, 86 pp.
Chapter 9
Organic soils
W. SHOTYK
Introduction
Organic soils form in oxygen-deficient, water-saturated environments such as

wetlands and peatlands (mires) from the accumulation of carbonaceous limnic sedi
ments and peat. By definition, organic soil materials contain at least 25% by weight
organic matter (Figure 9.1).
Wetlands and peatlands are the sites of organic soil formation and together
they account for almost 5% of the Earth's total land area (Figure 9.2). They ex
hibit great diversity depending on climate, topography, hydrology and geology. As
a consequence, organic soil materials vary widely in physical, chemical and biologi
cal composition (Dawson, 1956; Farnham and Finney, 1965; Everett, 1983; Hobbs,
1986).
Here a brief introduction is given to organic soils (wetlands and peatlands) and
organic soil materials (carbonaceous limnic sediments and peats), and an overview
low ash |
medium ash
PEAT
high ash
low ash
CARBONACEOUS
SEDIMENT
high ash
<
CD
cr 75
o
MINERAL SEDIMENT
0 ■ '100
Fig. 9.1. Classification of peat and organic sediments (from Andrejko et al., 1983).
204 W Shotyk
Fig. 9.2. Wetlands map of the world (from Matthews and Fung, 1987).
of organic soil formation is provided. This is followed by a summary of the geochem-

ical processes operating in organic soil environments, with a focus toward the view
that these soils are dynamic and function as a globally important, biogeochemical
link between the lithosphère, hydrosphere and atmosphere.
Classification of organic soil materials
Distinctions have been made between the various kinds of peats and organic
sediments in many of the books on peat published since the 17th century, but
these classification schemes treated peat from the point of view of an important
industrial material which could be used either as a fuel (Patin, 1663; Eiselen, 1795;
Johnson, 1866) or an agricultural soil or soil amendment (Anderson 1794; Walker,
1803; Aiton, 1805; Rennie, 1807). Systematic classification systems based upon the
genesis of organic sediments and sedimentary environments appear to have begun
with the studies of the Swedish geologist Hampus von Post (1822-1911). His system
of classifying freshwater organic sediments and organic horizons in mineral soils
(von Post, 1862; Ramann, 1888) is the basis for most modern classification schemes,
including those by Dachnowski (1919), von Post (1924), and von Bülow (1929).
The most popular classification scheme in use by geologists today is that devel
oped by Lennart von Post (1884-1951). The basis for his system (von Post, 1924)
lies clearly with that proposed by Hampus von Post published much earlier (von
Post, 1862). Among the advances made by Lennart von Post were his criteria for
Organic soils 205
classifying organic soil materials in the field (von Post, 1924). He introduced a 10
point scale for measuring by hand the degree of humification (symbol H) of peat
samples which is now commonly known as the von Post scale. In addition, he intro
duced a comparable 5 point scale for describing peat water content (B), a 4 point
scale for the relative proportion of sedge fibres (F), a 4 point scale for Erìophorum
(cottongrass) root threads (R) and a 4 point scale for wood remains (V). Lennart
von Post's (1924) system of organic sediment classification is fully illustrated in the
paper co-authored with Erik Granlund (von Post and Granlund, 1925) now available
in English (Landva et al., 1986). The Swedish system was modified slightly by von
Bülow (1929) who extended its applicability to central European mires.
Regarding the types of peat and organic sedimentary materials which constitute
organic soils, Dachnowski (1919) noted that the outstanding physical relationship is
the water content of the initial area and its effect upon the character of the series
of vegetation stages forming the sediments and upon the manner of accumulation
of plant remains. From the point of view of the conditions of accumulation, he
distinguished between transported organic materials and those formed in situ. The
transported materials are those which were formed in open water and below the
water surface; this includes his aquatic group of materials (gyttja, dy, dopplerite)
and marsh group (reed-grass peat, sedge peat, Brown Moss peat). The materials
formed in place are those which formed at and above the water level at the site
actually occupied by the peat-forming vegetation units; this includes his bog group
(Sphagnum, Erìophorum, Scheuchzeria, and dwarf shrub peats) and swamp group
(alder-willow type, birch type, deciduous forest type and coniferous forest type).
In the von Post (1924) system, the terms "sedimentary" and "sedentary" are
used for transported and residual materials, respectively, and von Bülow (1929)
used the terms "allochthonous" and "autochthonous", respectively. Other common
terms include "limnic peat" which is used to describe organic sediments formed
in open water, "terrestric peat" for sediments formed above the water table and
"telmatic peat" for peats developed in environments of intermediate wetness (Auer,
1927).
Sedimentary organic materìals

A representative classification scheme for sedimentary organic materials is given
in Table 9.1. "Dy" and "Gyttja" are Swedish terms which have no simple English
equivalents. The widespread use of these terms has given rise to considerable con
fusion in the literature (Hansen, 1959). Critical, comprehensive reviews have been
compiled by Wasmund (1930) and Grosse-Brauckmann (1961).
Dy is a gelatinous, acidic sediment formed in brown waters such as humic lakes
from the flocculation and precipitation of dissolved humic materials. The humic
materials are primarily of terrestrial origin, formed in organic-rich soils from the
decomposition of plant and animal remains and carried into lakes by percolating
206 W. Shotyk
TABLE 9.1
Sedimentary organic materials
1. Dy
2. Gyttja
2.1. Organic gyttja
2.1.1. Fine detritus gyttja (profundal gyttja)
2.1.2. Coarse detritus gyttja (littoral gyttja)
2.1.3. Algal gyttja
2.2. Calcareous gyttja
2.2.1. Lacustrine chalk
2.2.2. Calcareous
2.2.3. Detrital calcareous
2.2.4. Silty calcareous
2.3. Siliceous gyttja
2.3.1. Silty gyttja
2.3.2. Sandy-silty gyttja
2.3.3. Diatomaceous gyttja
3. Alluvial peat
Sources: Dachnowski (1919); von Post (1924); von Bülow (1929); Markowski (1976).
solutions. The dissolved humic materials are sparingly soluble, and precipitate read
ily in the presence of dissolved Ca, Fe and Al. Since waters rich in dissolved humic
materials ("brown waters") are especially characteristic of peaty terrains, intensive
formations of dy occur primarily in lakes that are embedded in bogs ("bog lakes")
or are fed by them (Ruttner, 1963).
In contrast to dy which is a chemical precipitate, gyttjas are biologically produced
sedimentary oozes. The light coloured (almost white to greenish gray) gyttjas gener
ally form in the epilimnion while the dark coloured gyttjas (chocolate brown) form
in the hypolimnion (Kindle, 1927). The four most important components of gyttjas
are (Dean, 1981):
(1) Detrital clastic material derived from those materials that are available in the
drainage basin. These are deposited mainly in the littoral zone, the region between
the shoreline and the outer limit of rooted plants.
(2) Biogenic silica which occurs mainly in the form of diatom frustules. These
frustules are the skeletal remains of diatoms, unicellular aquatic plants related to
algae.
(3) Carbonate minerals, mainly low Mg calcite, but some high Mg calcite, dolo
mite and/or aragonite, depending on the Mg : Ca ratio of the water. Calcareous
muds in the littoral zone contain a very high proportion of biogenic CaCC>3, mainly
debris from mollusks, ostracods, and charaphytes, the calcareous green algae of the
genera Chara and Nitella.
(4) Organic matter, a mixture of herbaceous allochthonous materials from the
drainage basin (pollen grains, needles, seeds, woody materials and other organic
Organic soils 207
detritus) and lipid-rich allochthonous material (largely debris from planktonic algae
and aquatic macrophytes).
Sedentary organic materials
An ecological classification scheme for sedentary organic materials (peats) is

shown in Table 9.2. In addition to the peats shown above, the Geological Survey of
Finland recognizes the following types: Trichophorum, Molinia and Menyanthes.
The botanical composition of the sediments provides a record of organic soil
development and structure and identifies the series of vegetation stages each of
which began, continued, or was terminated and replaced in response to changes in
environment such as climate or hydrology (Dachnowski, 1919; Barber, 1981; God
win, 1981). The classification scheme distinguishes peats formed under eutrophic
and mesotrophic conditions (plant nutrients abundant and moderately abundant,
respectively), and those formed under oligotrophic conditions (plant nutrient defi
cient).
The materials indicated in Tables 9.1 and 9.2 are described in detail by von
Post (1862), Ramann (1888), Dachnowski (1919,1924), von Post (1924), von Bülow
(1929), Auer (1930), Wasmund (1930), Rigg and Richardson (1938), Rigg (1940),
Hansen (1959), Grosse-Brauckmann (1961), Markowski (1976), Dean (1981) and
TABLE 9.2
Sedentary organic materials
A. Mesotrophic and eutrophic types B. Oligotrophic types

1. Li m nie peats 1. Limnic-telmatic peats
1.1. Reed peat 1.1. Sphagnum cuspidatum peat
2. Limnic-telmatic peats 1.2. Scheuchzeria peat
2.1. Cladium peat 1.3. Sedge peat
2.2. Equisetum peat 1.4. Sphagnum peat
2.3. Gfyceria peat 1.5. Cottongrass (Eriophorum) peat
2.4. Iris peat 2. Terrestric and ombrogenic peats
2.5. Thetypteris peat 2.1. Sphagnum fu scum peat
3. Telmatic peats 2.2. Sphagnum magellanicum peat
3.1. Tall sedge peats 2.3. Ericaceous (dwarf shrub) peat
3.2. Brown moss peats 2.4. Swamp peat
3.3. Sphagnum peats 2.4.1. Pine swamp peat
4. Terrestric peats 2.4.2. Spruce swamp peat
4.1. Short sedge peats 2.4.3. Birch swamp peat
4.2. Swamp forest peats
4.2.1. Alder swamp peat
4.2.2. Birch swamp peat
4.2.3. Coniferous swamp peat
4.2.4. Deciduous swamp peat
Sources: Dachnowski (1919); von Post (1924); von Bülow (1929).
208 W Shotyk
Levesque et al. (1987). Photomicrographs of peat thin sections for a wide range of
peats and organic sediments have been published by von Post and Granlund (1925),
Kivinen (1948), Landva and Pheeney (1980) and Levesque and Dinel (1982).
Classification of organic soils
Over the years many wetland and peatland classification schemes have evolved,
each suited to a particular range of physiographic and climatic regimes. Some of the
criteria which have been used to classify organic soils are shown in Table 9.3.
Much of the German and northern European literature on the subject has been
cited by Overbeck (1975) and more recently by Succow (1988), and the book by
Moore (1984) provides a summary of eastern and central European studies. Peat
land classification is discussed in many of the publications of the International Peat
Society. Some of the world's major wetland and peatland systems are described in
the book edited by Gore (1983).
American workers also have made important contributions to the subject, be
ginning with the pioneering efforts of Shaler (1888), Davis (1907), Harper (1910),
Dachnowski (1912), Huels (1915), Soper (1919) and Osborn (1921). Modern Amer
ican work is summarized in the contemporary work of Crum (1988). In Canada,
desirable features from both American and European systems have been combined
to develop a national wetlands classification scheme. Canadian wetlands and peat-
lands are illustrated and described in the book written by the National Wetlands
Working Group (1988).
TABLE 9.3
Criteria used to classify organic soils
Criterion Examples
Landform topography Kettle hole bog, floodplain swamp
Method of development Terrestrialization, paludification
Dominant surface vegetation Cedar swamp, Sphagnum bog
Trophic status Oligotrophic, mesotrophic, eutrophic
Acidity Strongly acidic, neutral, alkaline
Vegetation structure Trees, shrubs, sedges, mosses, lichens
Soil morphology Fibrisol, mesisol, humisol
Hydrology Rheophilous, soligenous, topogenous
Aerial pattern Marbloid, reticuloid, terrazoid
Peatland morphology Raised bog, sloping fen
Geotechnical properties Shear strength, Atterberg limits, shrinkage
Ecology Bog, fen, marsh, swamp
Sources: Davis (1907); Kotilainen (1927); Auer (1930); Farnham and Finney (1965); Bellamy and Bel
lamy (1966); Radforth and Bellamy (1973); Moore and Bellamy (1974); Landva et al. (1983); Hobbs
(1986); National Wetlands Working Group (1988).
Organic soils 209
The National Wetlands Working Group (NWWG) ecological classification sys

tem contains three hierarchical levels: class, form, type. There are five wetland
classes: open water, marshes, swamps, fens and bogs. Seventy wetland forms are
differentiated on the basis of surface morphology, surface pattern, water type and
morphology of the underlying mineral soil. Wetland types are classified according to
vegetation physiognomy.
The NWWG (1988) scheme distinguishes between mineral wetlands with peat
accumulations less than 40 cm, and organic wetlands (peatlands) in which at least
40 cm of peat has accumulated. Whereas fens and bogs are peatlands, swamps and
marshes may be either mineral wetlands or peatlands, depending on the thickness
of the peat accumulation. Marshes, swamps and fens are minerotrophic wetlands,
that is, they receive their mineral nutrition from flooding surface waters, by perco
lating groundwaters, or both. In contrast, the surface layers in bogs are beyond the
influence of groundwaters and surface waters, and the plants growing on the surface
receive their mineral nutrition solely from atmospheric deposition (Dau, 1823). This
condition which is characteristic of bogs is called ombrotrophic.
Marshes
Marshes are wetlands and peatlands most often associated with bodies of open
water. They are subjected to periodic or regular innundation with standing or slowly
moving water. Flood waters may bring with them an abundant supply of suspended
sediments. As a consequence, the substratum which forms in marshes may contain
abundant detrital mineral material. The detrital sediments provide abundant plant
nutrients which gives rise to a luxuriant growth of emergent sedges, grasses, rushes
and reeds, with submergent andfloatingaquatic plants in the zones of open water.
Swamps
Swamps are treed wetlands and peatlands. Like marshes, swamps are subjected
to periodic or regular flooding. Because most of the water in swamps has perco
lated through the surrounding mineral soils, the waters may be rich in dissolved
solids supplied by mineral dissolution. The vegetation is characterized by a cover of
deciduous or coniferous trees, shrubs, herbs and mosses.
Fens
Fens are meadow-like peatlands dominated by sedges, grasses, reeds, brown

mosses, and shrubs. They too are fed by solutions which have percolated through
mineral soils.
210 W. Shotyk
Bogs
The ombrotrophic condition of bogs clearly distinguishes them from marshes,
swamps and fens. They are dominated by Sphagnum mosses which thrive in acidic,
nutrient-deficient environments, but ericaceous shrubs, sedges and herbs also are
common. British bogs have been described by Godwin (1981), bogs of northeastern
North America by Johnson (1985) and German bogs by Succow and Jeschke (1986).
Formation of organic soils
Organic soils form wherever the rate of production of organic materials exceeds
their rate of decomposition. The single most important criterion for the formation
of organic soils is an abundance of water. When water covers a soil or sediment
surface and displaces air from the pores, the inward diffusion of oxygen from the
atmosphere is inhibited. In water-saturated soils and sediments containing decom
posable organic matter, oxygen is consumed by microorganisms faster than it can be
supplied by diffusion from air, leading to an anaerobic condition; it is this anaerobic
condition that severely retards the decay of plant matter.
Three environments which satisfy this moisture requirement are bodies of shal
low open water such as lakes and ponds, poorly drained mineral soils with high
water tables, and imperfectly drained soils in areas of high precipitation, low evap
oration and cool temperatures (Auer, 1927). Consequently, the two principal evolu
tionary paths leading directly to the formation of organic soils are terrestrialization,
the filling of shallow lake basins by limnic sediments and aquatic vegetation, and
paludification, the waterlogging of terrestrial soils by absorbent organic materials
(Frenzel, 1983).
Terrestrialization
Terrestrialization refers to the successive appearance of distinct plant associations
which generally follow this sequence: algae and other aquatic microorganisms, pond
weeds, pond lilies, bullrushes, amphibious sedges, herbs and shrubs, Sphagnum
mosses, coniferous trees and, finally, deciduous trees (Davis, 1913). This is illus
trated in Figure 9.3 and is described in detail in the monographs by Davis (1907)
and Dachnowski (1912), and more recently by Tkllis (1973). In Canada the process
was studied by Auer (1930) who referred to this as "progressive" mire development
because the environment evolves from a wetter condition to a drier one.
Paludification
This term was introduced by Lennart von Post to denote "a ground becoming a
bog ... owing to (an excess of) moisture, but without being inundated" (von Post,
Organic soils 211
Fig. 9.3. Schematic view of terrestrialization, the progressive filling of a shallow lake basin by limnic,
telmatic and eventually terrestric vegetation. The plant associations which progressively displace one
another belong to the following major vegetation groups: open water succession, marginal succession,
shore succession, bog succession (comprising a bog meadow, shrub bog and bog forest) and mesophytic
forest succession. (From Dachnowski, 1912.)
1937, p. 214). Auer (1930) called this "regressive" mire development because the
environment evolves from a drier condition to a wetter one. Both he (Auer, 1927)
and von Post (1937) recognized two kinds of paludification: (a) the formation of
mires on poorly drained soils with high water tables, and (b) their formation on
imperfectly drained soils and bedrock in cool regions with an excess of precipita
tion over evaporation. These paludification types have been termed "soligenous"
and "ombrogenous", respectively (von Post, 1937). Ombrogenous mires throughout
their existence are controlled entirely by the precipitation falling directly upon the
mire surface whereas soligenous mires are initiated by soil water but may become
214
H-*
CARIBOU "A"
ESCUMINAC
foËïï=/!^M Sphagnum peat, raw
Sphagnum peat, decayed
JJ Carex peat, raw —- Point where peat bog

was drilled
Fig. 9.4. Peat stratigraphy of two ombrotrophic Sphagnum bogs. a. Caribou bog, Nova Scotia, formed from the progressive accumulation of lacustrine
sediments and peat in a shallow, open water environment, b. Escuminac bog, New Brunswick, formed from the regressive accumulation of peat in
W. Shotyk
response to high precipitation, low evaporation, and cool temperatures. (From Auer, 1930.) 33
4
Organic soils 213
ombrotrophic. In Canada the majority of bogs in oceanic climatic zones are om-
brogenous (Auer, 1927). In contrast, paludification in continental climatic regions is
almost entirely soligenous (Auer, 1927; Rigg, 1940; Heinselman, 1963,1970, 1975).
The stratigraphie record as it is revealed by peat cores may be used to distinguish
which evolutionary path any given wetland may have taken (Figure 9.4).
Summary of the geochemistry of peatlands
The chemical properties ofpeats
Physical properties
For a given type of peat, bulk density increases and pore volume decreases
with increasing humification (decomposition). The calorific value (energy content)
also increases with increasing humification, while moisture content decreases (Keys,
1983).
Ultimate chemical composition

Peats vary widely in organic chemical composition, in part reflecting the diversity
of plant materials which give rise to the peat. For a given kind of peat, the con
tent of carbon (50-60%), hydrogen (5-6%), nitrogen (0.5-2%), and sulphur (0.1-
0.5%) tends generally to increase with increasing degree of decomposition, while
the oxygen content (30-40%) decreases (Naucke, 1980).
Physico-chemical properties
All peats have relatively high cation exchange capacities, ranging from approxi
mately 80 to 160 meq per 100 g (Puustjärvi and Robertson, 1975). The percent-base
saturation of peats reflects the trophic status of the peatland: peats in eutrophic
peatlands have a large percent-base saturation whereas the peats in oligotrophic
peatlands have a small percent-base saturation (relatively hydrogen saturated). The
measured pH of peat/water slurries increases with increasing percent-base satura
tion of the peat (Holmen, 1964).
The chemical properties ofpeatlands
Abundance and distribution of mineral matter

The greatest concentrations of mineral matter usually are found in the shallow
peat layers at the edge of a peatland (the "lagg zone", sometimes also called the
"rand" or the "marginal fen"), and in the basal layers, adjacent to the underlying
mineral sediments (Holmen, 1964), for several reasons. First, since these peats are
so close to the boundary with the mineral soil, considerable quantities of mineral
matter are washed or otherwise mixed into the peat. Second, many of the plants in
214 W. Shotyk
this zone are rooted in the mineral soil and thereby accumulate significant amounts
of bioessential elements.
The surface peat layer may be relatively enriched in mineral matter because of
anthropogenic inputs of pedogenic dust (Chapman, 1964; Gorham and Tilton, 1978)
and uptake of elements by growing plants.
Abundance and distribution of inorganic constituents

Aluminum and Si are the building-block elements of aluminosilicate minerals
which in many cases make up the majority of the mineral matter in peat. Thus, the
concentrations of Al and Si generally correspond with the abundance and distribu
tion of ash. Relatively insoluble trace metals (Ti, Ni, V, Cr) also correlate strongly
with the amount of mineral matter present in the peat (Shotyk, 1987).
Potassium is highly enriched in surface peat layers because of plant uptake; below
the surface both K and Na are rapidly leached from the peat (Damman, 1978).
Calcium and Mg may increase in concentration with depth in calcareous terrains
because of the supply of these metals by groundwaters. In acidic, anaerobic peats Fe
and Mn tend to become depleted whereas in alkaline, anaerobic peats they tend to
accumulate. Copper and U are easily enriched in all groundwater-fed peats ranging
in pH from acidic to alkaline. Zinc and Pb may also become enriched in alkaline,
anaerobic peats, but in acidic peats their enrichments are restricted to surface layers
and have been caused by air pollution (Shotyk, 1988; Shotyk et al., in press).
The chemical properties ofpeatland waters

Ttoo chemical characteristics which most easily distinguish peatland waters from
typical freshwaters are the abundance of dissolved organic matter and the anaerobic
condition (Shotyk, 1989a). In respect to bog surface waters, two additional charac
teristic properties are the low ionic strength and the naturally low pH (Gorham et
al., 1985; Shotyk, 1988,1989a).
Chemical composition
Minerotrophic peatland waters (in marshes, swamps, and fens) range in compo
sition from eutrophic (nutrient-rich) waters in calcareous sedimentary terrains to
oligotrophic (nutrient-poor) waters in siliceous igneous terrains. In other words, the
chemical composition of minerotrophic peatland waters reflects the composition of
local rocks and soils.
In contrast, bog waters fed solely by atmospheric deposition resemble rainwater
in respect to the concentrations of major element cations Na, K, Mg, Ca (Ramann,
1895; Gorham et al., 1985). Bog waters are often reported to be enriched in Fe,
Mn and Al, relative to the composition of typical freshwaters, apparently because
of in situ dissolution of atmospherically-derived minerals in the acidic, organic-rich,
anaerobic milieu (Shotyk, 1989a).
Organic soils 215
pH, conductivity and alkalinity

The pH of peatland waters ranges from approximately 4 to 8. Lower pH values
have been reported for oxidizing, sulphidic swamp peats, and higher values have
been found in saline fens, but these are exceptional cases. The most commonly
reported pH values are pH 4 to 4.5 for bog waters, and 6 to 6.5 for fens and swamps
(Gorham et al., 1987).
Redox chemistry
Surface waters that are visiblyflowingand pools of open water that are exposed
to the air and stirred by wind are saturated with dissolved O2, but stagnant surface
pools are oxygen-deficient. Below the water table, however, the pore waters are
anoxic (Shotyk, 1988, 1989a). Thus, two redox zones may be recognized: a shallow,
oxygenated surface layer whose thickness varies with the height of the water table,
and an anoxic zone below the water table. The strongly "reducing" condition of
the anoxic zone is confirmed by the presence of characteristic gaseous, aqueous
and solid phases: gases such as CH4, CO, H2 and H2S; decreased concentrations
of dissolved sulphate and elevated concentrations of aqueous sulphide, Fe and Mn;
the occurrence of authigenic carbonate, sulphide and phosphate minerals (Shotyk,
1988,1989a, b).
Environmental geochemistry of peatlands
Effects on the lithosphère

Dissolution ofprìmary minerals
Since peatland waters generally are anaerobic, always contain abundant dissolved
organic acids and often are acidic, there are at least three mechanisms by which
mineral weathering may be promoted: (a) reductive dissolution, in response to the
anaerobic condition, (b) ligand-promoted dissolution, in response to the abundant,
complex-forming organic acids, and (c) proton-promoted dissolution, in response to
the acidic condition of bog waters.
Experimental studies have shown that the reductive dissolution of mafic minerals
in anoxic solutions releases Fe up to ten times faster than its rate of release rates in
oxygenated solutions (Siever and Woodford, 1979). The reductive dissolution of Fe-
bearing minerals in anaerobic peatlands, the movement of these solutions by mass
flow and the subsequent precipitation of Fe in oxygenated drainage waters in the
form of lacustrine Fe ores, has been thoroughly documented and described in detail
(Liebig, 1843; Daubrée, 1845; Bischoff, 1854; Hunt, 1863; Johnson, 1866; Le Conte,
1878; Irving, 1892). The precipitation of dissolved Fe emanating from anaerobic
sediments is mediated by iron-oxidizing bacteria (Ehrenberg, 1837; Ehrlich, 1981).
The black, gelatinous muds which result are now known as "bog iron ore" and "bog
manganese ore".
216 W. Shotyk
There are two ways in which the dissolved organic acids in peatland waters
may promote mineral dissolution. First, they can form complexes with aqueous
metal species such as Al thereby increasing metal solubility (as defined by the total
dissolved metal concentration). This may help to explain why bog waters may have
high concentrations of dissolved Al, relative to typical freshwaters. Second, the
organic ligands can become adsorbed and form complexes with (for example) Al on
the surface of the mineral. This latter process, known as "surface complexation",
polarizes the critical metal-oxygen bonds, thus facilitating the detachment of Al
from the surface of the mineral and increasing the rate of mineral dissolution
(Stumm et al., 1987).
There are also many examples of proton-promoted dissolution, in response to the
low pH of bog waters. The dissolution of carbonate minerals by acidic bog waters
in limestone terrains, for example, is often rather obvious (Brown, 1869; Goodchild,
1890). The dissolution of siliceous rocks by acidic bog waters is sometimes just as
striking, and also is well established (Browne, 1858; Stremme, 1908; Weiss, 1910;
Endell, 1911; Humphreys and Julien, 1911; Niklas, 1912; Grüner, 1922). Mass bal
ance calculations by Russian workers have shown that the measured ash contents
of peats in bogs is only about 10% of the concentration anticipated from inputs of
eolian dust (Lukashev et al., 1974). Thus, approximately 90% of the mineral matter
deposited on bog surfaces appears to be dissolving and is subsequently removed by
the drainage waters.
Precipitation of secondary minerals

While many primary minerals are thermodynamically unstable and decompose in
peatland waters, a variety of secondary minerals may form (Table 9.4).
The formation of secondary mineral phases may result in considerable metal ac
cumulations. For example, peats from southern Norway contain as much as 40% by
weight Fe, because of the formation of siderite (Selmer-Olsen and Lie, 1983), and
peats from northern Canada contain up to 22% Zn, because of the formation of
smithsonite (Lee et al, 1984). Native Cu has been encountered in many cuprifer
ous peatlands (Henwood, 1857; Lovering, 1927; Forrester, 1942; Lett and Fletcher,
1980). A swamp in eastern Canada is estimated to contain as much as 300 tonnes
of Cu (Boyle, 1977) and a Russian swamp more than 6 million tonnes of vivianite
(Bgatovetal., 1986).
Organic enrichment of trace metals

The most frequently reported enrichments are those of Cu and U. In extreme
cases peats have been found to contain on a dry weight basis as much as 3% U
(Armands, 1967) and 10% Cu (Fraser, 1961). These very unusual accumulations
are generally attributed to organically-mediated processes such as ion exchange
and surface complexation by the peat, but inorganic solubility controls are also
involved.
Organic soih 217
TABLE 9.4
Authigenic minerals in peatlands

1. Iron minerals
1.1 Native iron
1.2 Siderite, FeC03
1.3 Vivianite, Fe 3 (P0 4 ) 2 8H 2 0
1.4 Iron sulphides (pyrite, FeS2, marcasite, FeS2)
2. Copper minerals
2.1 Native copper, Cu
2.2 Covellite, CuS
2.3 Chalcopyrite, CuFeS2
3. Zinc minerals
3.1 Wurztite, ZnS
3.2 Smithsonite, ZnC03
3.3 Zincic calcite, (Zn,Ca)C03
4. Other sulphur-bearing minerals
4.1 Native (elemental) sulphur
Various sources.
Effects on the hydrosphere

Dissolved organic matter
The most visible effect of peatlands on the hydrosphere is the continual export
of dissolved organic matter. The concentrations of dissolved organic carbon (DOC)
vary seasonally, ranging from 20 to 400 mg Γ 1 , averaging about 50 mg l"1 (Thur-
man, 1985). In contrast, typical river waters contain less than 10 mg Γ 1 DOC, and
groundwaters less than 1 mg Γ 1 . In boreal catchments with extensive peatland cover
the long term average of exports of dissolved organic matter (DOM) may exceed
201 km2 per year (Sallantaus, 1986).
pH and inorganic constituents

Pioneering studies of the swamps of Africa showed that all traces of sulphate
are removed when rivers pass through significant areas of swampy terrain (Beam,
1906). When surface waters pass through the papyrus (Cyperus papyrus) swamps
of Uganda, for example, the concentrations of dissolved organic matter, CO2, Fe
and NH4+ increase, while the pH, dissolved 0 2 concentration and redox potential
decrease (Carter, 1955). Similar changes have been reported for the Sudd Swamp
which extends for 500 km along the River Nile (Tailing, 1957; Bishai, 1962).
Mass balance calculations have since shown that swamps may act as a sink for
dissolved sulphur (LaZerte and Dillon, 1985). In regions with extensive wetland cov
erage and high levels of atmospheric sulphate deposition (acid rain), the generation
of alkalinity in swamps by sulphate reduction may be a factor mitigating watershed
acidification. In contrast to the removal of sulphur, swamps may act as a source of
218 W. Shotyk
dissolved iron (LaZerte and Dillon, 1985). This must be related to the dissolution
of Fe-bearing minerals described earlier, and may explain why rivers draining exten
sive peatland areas are rich in dissolved and paniculate Fe (McCrea and Merriman,
1981; McCrea et al., 1984).
Effects on the atmosphere
Peatlands as a source of methane

The occurrence of "carburetted hydrogen" in peatlands was reported by many
early writers (Aiton, 1805; Rennie, 1807). Since that time many quantitative studies
have confirmed the predominance of CH4 in "marsh gas" (Websky, 1864; Swain,
1986). Currently there is widespread interest in CH4 emissions from organic soils,
for two reasons. First, CH4 is one of the "greenhouse" gases and is increasing in at
mospheric concentration at a rate of about 1% per year (Steele et al., 1987; Ehhalt,
1988). As a radiatively active trace gas, CH4 has a significant potential to contribute
to global climate change within the next century (Ramanathan et al, 1985; Ra-
manathan 1988). Second, organic soils are generally considered to represent the
largest single source of CH4 to the atmosphere (Cicerone and Oremland, 1988).
About 60% of the total CH4 emitted from wetlands comes from temperate wetlands
concentrated between 50° and 70°N latitude.
The magnitude of CH4 emissions from organic soils depends on the relative
importance of methanogenesis in anaerobic sediments and methane oxidation in
the aerobic surface layers. Methane is produced by obligate anaerobic bacteria in
the anoxic zone (below the water table) and consumed by microaerophilic methane
oxidizing bacteria in the oxic zone (above the water table). The net flux of CH4
from organic soils to the atmosphere, therefore, represents the difference between
methane production in the anaerobic zone and methane oxidation at the surface
(Shotyk, 1989b). The measured fluxes vary seasonally, depending on the relative
thicknesses of the aerobic and anaerobic zones.
Peatlands as a sink for carbon dioxide

While peatlands act as a net source of CH4, they serve as a net sink of CO2, also a
radiatively active "greenhouse" gas. The amount of carbon stored in all peatlands is
estimated to be 3 x 10111 (Sjörs, 1980) which makes it a relatively small component
of the global carbon cycle. However, enormous peatlands have existed since the
Carboniferous (approximately 300 million years ago), giving rise to coal formations
throughout the world which today store approximately 3 x IO15 t of carbon; this
represents a significant sink for the global cycling of carbon (Bowen, 1979). Thus,
the removal and fixation of atmospheric CO2 by peatland plants and its burial and
preservation as peat and eventually coal appears to be one of the globally significant
mechanisms which buffers the concentration of CO2 in the atmosphere.
Organic soils 219
Summary
Organic soils are dynamic sedimentary environments which account for about 5%
of the Earth's total land area. They are a vital component of the biosphere and rep
resent an important biogeochemical link between the lithosphère, hydrosphere and
atmosphere. In general they are rich in slowly decomposing organic matter, mainly
anaerobic and often acidic. These geochemical properties give rise to complex inter
actions among the inorganic and organic solid, liquid and gaseous phases present.
These interactions have many ramifications for mineral sediments underlying them,
and the natural waters and gases which are cycled through them.
The chemical properties of organic soils are profoundly affected by biological
and anthropogenic processes in the surface layers, and the natural geochemical pro
cesses of primary mineral dissolution and the precipitation of authigenic, secondary
phases. The rates and mechanisms of the inorganic reactions in organic-rich, anaer
obic environments are not well known, thus we do not completely understand how
the chemical composition of the soils changes over time. More complete chemical
and mineralogical data is needed for complete peat profiles representing all the
evolutionary stages of peatland development, not just for Sphagnum bogs. Such
studies could be complemented by experimental studies of mineral dissolution and
precipitation reactions in organic-rich, anaerobic systems.
The influence of these soils on the geochemistry of natural waters is well estab
lished, but in most cases the details and possible significance are still unclear. The
redox chemistry of these environments is not well understood, and further studies
are needed to describe the redox properties quantitatively and to see how they vary
both spatially and temporally within the peatland ecosystem.
The emissions of CH4 to the atmosphere and its possible importance is beginning
to receive some attention, but emissions of other trace gases, including CO, H2,
N2O (nitrous oxide), H2S, CS2 (carbon disulphide), COS (carbonyl sulphide) and
(CH3)2S (dimethyl sulphide) may also be significant and should not be overlooked.
In respect to their function of CO2 removal and fixation through geological history,
the significance of peatland destruction by drainage and development, and the im
plications which this may have for global climate change, should also be carefully
evaluated.
References
Aiton, W, 1805. A Treatise on the Origin, Qualities and Cultivation of Moss-Earth. Niven, Napier, and
Khull, Trongate, Glasgow, 178 pp.
Anderson, J., 1794. A Practical Treatise on Peat Moss. Edinburgh, 233 pp.
Andrejko, M J., Fiene, F. and Cohen, A.D., 1983. Comparison of ashing techniques for determination
of the inorganic content of peats. In: P.M. Jarrett (Editor), Testing of Peats and Organic Soils. STP
820, Am. Soc. Test. Mat., Philadelphia, Pa., pp. 5-20.
220 W. Shotyk
Armands, G., 1967. Geochemical prospecting of a uraniferous bog deposit at Masugnsbyn, northern
Sweden. In: A. Kvalheim (Editor), Geochemical Prospecting in Fennoscandia. Interscience, New
York, N.Y., pp. 127-154.
Auer, V, 1927. Stratigraphical and morphological investigations of peat bogs of southeastern Canada.
Comm. Inst. Quae. Forestal. Finlandiae Ed. 12, 32 pp.
Auer, V, 1930. Peat bogs in southeastern Canada. Geol. Surv. Can., Mem. 162, 32 pp.
Barber, K.E., 1981. Peat Stratigraphy and Climatic Change. A.A. Balkema, Rotterdam, 219 pp.
Beam, W, 1906. Chemical composition of the Nile waters. Wellcome Res. Lab. (Khartoum), Second
Report, 1906, pp. 206-214.
Bellamy, D. and Bellamy, R., 1966. An ecological approach to the classification of the lowland mires of
Ireland. Proc. R. Irish Acad., Sect. B, 65: 237-251.
Bgatov, B.I., Matukhina, V.G., Grigoreva, T.R., Altukhov, V.M. and Rusanov, D.K., 1986. Peat vivianites
and vivianite peats of the Suzun district, Novosibirsk region (USSR). Geol. Geofiz., pp. 51-56.
Bischof, G., 1854. Elements of Chemical and Physical Geology (B.H. Paul and J. Drummond, transla
tors). The Cavendish Society, London.
Bishai, H.M., 1962. The water characteristics of the Nile in the Sudan with a note on the effect of
Eichhornia crassipes on the hydrobiology of the Nile. Hydrobiologia, 19: 357-382.
Boyle, R.W., 1977. Cupriferous bogs in the Sackville area, New Brunswick, Canada. J. Geochem. Explor.,
8: 495-527.
Bowen, H.J.M., 1979. Environmental Chemistry of the Elements. Academic Press, London, 333 pp.
Brakke, D.F, Henriksen, A. and Norton, S.A., 1987. The relative importance of acidity sources for
humic lakes in Norway. Nature, 329: 432-434.
Brown, E., 1869. The waste of rock masses by solution. Geol. Mag., 6: 379.380.
Browne, D.J., 1858. The Field Book of Manures; or, The American Muck Book. A.O. Moore, New
York, N.Y., 422 pp.
Carter, G.S., 1955. The Papyrus Swamps of Uganda. W. Heffer, Cambridge, 26 pp.
Chapman, S.B., 1964. The ecology of Coom Rigg Moss, Northumberland. II. The chemistry of peat
profiles and the development of the bog system. J. Ecol., 52: 315-321.
Cicerone, R.J. and Oremland, R.S., 1988. Biogeochemical aspects of atmospheric methane. Global
Biogeochem. Cycles, 2: 299-327.
Crum, H., 1988. A Focus on Peatlands and Peat Mosses. University of Michigan Press, Ann Arbor,
Mich., 306 pp.
Dachnowski, A., 1912. Peat deposits of Ohio. Their origin, formation, and uses. Geol. Surv. Ohio, Bull.
16, 424 pp.
Dachnowski, A., 1919. Quality and value of important types of peat material. A classification of peat
based upon its botanical composition and physical and chemical characteristics. U.S. Dep. Agric,
Bull. 802, 40 pp.
Dachnowski, A., 1924. The stratigraphie study of peat deposits. Soil Sci., 17: 107-133.
Damman, A.W.H., 1978. Distribution and movement of elements in ombrotrophic peat bogs. Oikos, 30:
480-495.
Dau, J.H.C., 1823. Neues Handbuch über den Torf. J.C. Hinrichsche, Leipzig, 240 pp.
Daubrée, M. A., 1845. Observations sur le minéral de fer qui se forme journellement dans les marais et
dans les lacs. C. R. Acad. Sci., 20: 1775-1780.
Davis, C.A., 1907. Peat, Essays on its Origin, Uses and Distribution in Michigan. In: Report of the State
Board of Geological Survey of Michigan for the Year 1906, pp. 95-395.
Davis, CA., 1913. Origin and formation of peat. In: D. White and R. Thiessen (Editors), The Origin of
Coal. U.S. Bur. Min. Bull., 38: 165-186.
Dawson, J.E., 1956. Organic soils. Adv. Agron., 8: 377-401.
Organic soils 221
Dean. W.E., 1981. Carbonate minerals and organic matter in sediments of modern north temperate
hard-water lakes. Soc. Econ. Paleontol. Mineral. Spec. Pubi., 31: 213-231.
Ehhalt, D.H., 1988. How has the atmospheric concentration of CH4 changed? In: ES. Rowland and
I.S.A. Isaksen (Editors), The Changing Atmosphere. Dahlem Konferenzen, 1988. Wiley, New York,
N.Y., pp. 25-32.
Ehrlich, H.L. 1981. Geomicrobiology. Marcel Dekker, New York, N.Y., 393 pp.
Ehrenberg, CG., 1837. Remarks on the real occurrence of fossil infusoria, and their extensive diffusion.
Taylor's Sci. Mem., 1: 400-407.
Eiselen, J.C., 1795. Handbuch oder ausführliche theoretisch-praktische Anleitung zur nähren Kenntniss
des Torfwesens. W. Viewig, Berlin, 444 pp.
Endell, K., 1911. Über die chemische und mineralogische Veränderung basischer Eruptivgesteine bei
der Zersetzung unter Mooren. Neues Jahrb. Mineral. Geol. Paläontol., 31: 1-54.
Everett, K.R., 1983. Histosols. In: L.P. Wilding, N.E. Smeck, and G.F. Hall (Editors), Pedogenesis and
Soil Taxonomy. II. The Soil Orders. Elsevier, Amsterdam, pp. 1-53.
Farnham, R.S. and Finney, H.R., 1965. Classification and properties of organic soils. Adv. Agron., 17:
115-162.
Forrester, J.D., 1942. A native copper deposit near Jefferson City, Montana. Econ. Geol., 37: 126-135.
Fraser, D.C., 1961. A syngenetic copper deposit of recent age. Econ. Geol., 56: 951-962.
Frenzel, B., 1983. Mires — repositories of climatic information or self-perpetuating ecosystems? In:
A.J.P. Gore (Editor). Mires: Swamp, Bog, Fen and Moor. General Studies. Ecosystems of the World,
Vol. 4A. Elsevier, Amsterdam, pp. 35-65.
Godwin, H., 1981. The Archives of the Peat Bog. Cambridge University Press, Cambridge, 229 pp.
Goodchild, J.G., 1890. Notes on some observed rates of weathering of limestones. Geol. Mag., 7: 463-
466.
Gore, A.J.P. (Editor), 1983. Mires: Swamp, Bog, Fen, and Moor. Ecosystems of the World, Vol. 4A,B.
Elsevier, Amsterdam, 919 pp.
Gorham, E. and Tilton, D.L., 1978. The mineral content of Sphagnum fuscwn as affected by human
settlement. Can. J. Bot., 56: 2755-2759.
Gorham, E., Eisenreich, S.J., Ford, J. and Santelmann, M.V., 1985. The chemistry of bog waters. In: W
Stumm (Editor), Chemical Processes in Lakes. Wiley, New York, N.Y., pp. 339-363.
Gorham, E., Janssens, J.A., Wheeler, G.A. and Glaser, P.H., 1987. The natural and anthropogenic
acidification of peatlands. In: T.C. Hutchinson and K.M. Meema (Editors), Effects of Atmospheric
Pollutants on Forests, Wetlands and Agricultural Ecosystems. Springer, Berlin, pp. 493-512.
Grosse-Brauckmann, G., 1961. Zur Terminologie organogener Sedimente. Geol. Jahrb., 79: 117-144.
Grüner, J.W, 1922. The origin of sedimentary iron formations: the Biwabik Formation of the Mesabi
Range. Econ. Geol., 17: 407-460.
Hansen, K., 1959. The terms Gyttja and Dy. Hydrobiologia, 13: 309-315.
Harper, R.M., 1910. A preliminary report on the Florida peat deposits. Florida State Geol. Surv., Third
Annu. Rep., 1909-1910, pp. 197-375.
Heinselman, M.L., 1963. Forest sites, bog processes, and peatland types in the glacial Lake Agassiz
Region, Minnesota. Ecol. Monogr., 33: 327-374.
Heinselman, M.L., 1970. Landscape evolution, peatland types, and the environment in the Lake Agassiz
Peatlands Natural Area, Minnesota. Ecol. Monogr., 40: 235-261.
Heinselman, M.L., 1975. Boreal peatlands in relation to environment. In: A.D. Hasler (Editor), Cou
pling of Land and Water Systems. Ecol. Stud. Vol. 10. Springer, New York, N.Y., pp. 93-103.
Henriksen, A., Brakke, D.F and Norton, S.A., 1988. Total organic carbon concentrations in acidic lakes
in southern Norway. Env. Sci. Technol. 22: 1103-1105.
Henwood, J.H., 1857. Notice of the copper turf of Merioneth. Edinburgh New Philos. J., 5: 61-63.
222 W. Shotyk
Hobbs, N.B., 1986. Mire morphology and the properties and behaviour of some British and foreign
peats. Q. J. Eng. Geol., London, 19: 7-80.
Holmen, H., 1964. Forest ecological studies on drained peat land in the Province of Uppland, Sweden.
Parts I—III. Stud. For. Suec, 16, 236 pp.
Huels, F.W., 1915. The peat resources of Wisconsin. Wis. Geol. Nat. Hist. Surv., Bull. 45, 274 pp.
Humphreys, E.W. and Julien, A.A., 1911. Local decomposition of rock by the corrosive action of pre-
glacial peat-bogs. J. Geol., 19: 47-56.
Hunt, T.S., 1863. Hydrous peroxyd of iron, or limonite. In: Geology of Canada. Geol. Surv. Canada,
Report of Progress from its Commencement to 1863. Dawson Brothers, Montreal, Que., pp. 510-
513.
Irving, A., 1892. Organic matter as a geological agent. Geol. Assoc. Proa, 12: 227-238.
Johnson, C.W., 1985. Bogs of the Northeast. University Press of New England, Hanover, 269 pp.
Johnson, S.W., 1866. Peat and its Uses, as Fertilizer and Fuel. Orange Judd, New York, N.Y., 168 pp.
Keys, D., 1983. Effect of peat type and decomposition on the calorific value of some New Brunswick
peats. In: P.M. Jarrett (Editor), Testing of Peats and Organic Soils. STP 820, Am. Soc. Test. Mat.,
Philadelphia, Pa., pp. 111-121.
Kindle, E.M., 1927. The role of thermal stratification in lake sedimentation. Trans. R. Soc. Can., IV,
Ser. Ill, 21: 1-35.
Kivinen, E., 1948. Suotiede. Werner Soderstrom Osakeyhtio, Porvoo and Helsinki, 219 pp.
Kotilainen, M.J., 1927. Untersuchungen über die Beziehungen zwischen der Pflanzendecke der Moore
und der Beschaffenheit, besonders der Reaktion des Torfbodens. Suomen Suoviljelysyhidstys, Ti-
eteellisia Julkaisuja, 7, 219 pp.
Landva, A.O. and Pheeney, P.E., 1980. Peat fabric and structure. Can. Geotech. J., 17: 416-435.
Landva, A.O., Korpijaakko, E.O. and Pheeney, P.E., 1983. Geotechnical classification of peats and
organic soils. In: P.M. Jarrett (Editor), Testing of Peats and Organic Soils. STP 820, Am. Soc. Test.
Mat., Philadelphia, Pa., pp. 37-51.
Landva, A.O., Korpijaakko, E.O. and Pheeney, P.E., 1986. Notes on the original von Post peat and peat-
land classification system. In: Proceedings, Advances in Peatlands Engineering. National Research
Council Canada, Ottawa, August 25, 26,1986, pp. 17-29.
LaZerte, B.D. and Dillon, P.J., 1985. Chemical transformations. V. Chemical transformations of stream
and soil water flowing through Sphagnum-conifer swamps of central Ontario. Muskoka '85 Confer
ence, poster presentation.
Le Conte, J., 1878. Elements of Geology. D. Appleton, New York, N.Y., 588 pp.
Lee, J., Jonasson, I.R., and Goodfellow, W.D., 1984. Metal accumulation by bryophytes in some zinc-rich
blanket bogs, Selwyn Mountains, Yukon Territory. Can. J. Bot., 62: 722-728.
Lett, R.E.W. and Fletcher, W.K., 1980. Syngenetic sulphide minerals in a copper-rich bog. Miner.
Deposita, 15: 61-67.
Levesque, M.P. and Dinel, H., 1982. Some morphological and chemical aspects of peats applied to the
characterization of Histosols. Soil Sci., 133: 324-336.
Levesque, M., Dinel, H., Lord, T. and Lortie, G., 1987. The characterization of organic soils developed
on peat and Iimnic materials in British Columbia. Tech. Bull. 1987-5E, Research Branch, Agriculture
Canada, Ottawa, Ont., 40 pp.
Liebig, J., 1843. Chemistry and its applications to Agriculture and Physiology. Taylor and Walton,
London, 400 pp.
Lovering, T.S., 1927. Organic precipitation of metallic copper. U.S. Geol. Surv. Bull., 795: 45-52.
Lukashev, K.I., Kovalev, VA. and Zhukhovitskaya, A.L., 1974. Peat bogs as a medium for the secondary
weathering of aluminum silicates. Dokl. Akad. Nauk Beloruss. SSR, 18: 357-360.
Markowski, S., 1976. Gyttja deposits at Pomerania region. In: Proc. 5th Int. Peat Congr., Poznan, Poland,
Sept. 21-25, 1976, Vol. 2, pp. 129-141.
Organic soils 223
Matthews, E., and Fung, I., 1987. Methane emission from natural wetlands: global distribution, area,
and environmental characteristics of sources. Global Biogeochem. Cycles, 1: 61-86.
McCrea, R.C. and Merriman, J.C., 1981. Water Quality in the Moose River — A Pilot Study, 1977-
1978. Report Series No. 70. Inland Waters Directorate, Ontario Region, Water Quality Branch,
Environment Canada, Burlington, Ont., 9 pp.
McCrea, R.C, Kwiatkowski, R.E., Campbell, D.E., McCarthy, P.P. and Norris, T.A., 1984. An investi
gation of contaminants and benthic communities in the major rivers of the Hudson Bay Lowland,
Ontario. Technical Bulletin No. 131. Inland Waters Directorate, Ontario Region, Water Quality
Branch, Environment Canada, Burlington, Ont., 47 pp.
Moore, P.D. (Editor), 1984. European Mires. Academic Press, London, 367 pp.
Moore, P.D. and Bellamy, D.J., 1974. Peatlands. Elek Science, London, 221 pp.
National Wetlands Working Group, 1988. Wetlands of Canada. Ecological Land Classification Series No.
24. Sustainable Development Branch, Environment Canada, Ottawa, and Polyscience Publications,
Montreal, 452 pp.
Naucke, W, 1980. Chemie von Moor und Torf. In: K. Gottlich (Editor), Moor und Torfkunde. E.
Schweizerbart, Stuttgart, pp. 173-195.
Niklas, H., 1912. Untersuchungen über den Einfluss von Humusstoffen auf der Verwitterung der Si
likate. Int. Mitt. Bodenk., 2: 214-244.
Osborn, C.C., 1921. Classification and formation of peat and related deposits. J. Am. Peat Soc, 14:
37-44.
Overbeck, E, 1975. Botanisch-Geologische Moorkunde. Karl Wachholtz, Neumunster, 719 pp.
Patin, C , 1663. Traité des Tourbes Combustibles. Bray et P. Variquet, Paris, 122 pp.
Puustjävi, V. and Robertson, R.A., 1975. Physical and chemical properties. In: D.W. Robinson and
J.G.D. Lamb (Editors), Peat in Horticulture. Academic Press, New York, N.Y., pp. 23-38.
Radforth, N.W and Bellamy, D.J., 1973. A pattern of muskeg - a key to continental water. Can. J. Earth
Sci., 10: 1420-1430.
Ramanathan, V, 1988. The radiative and climatic consequences of the changing atmospheric compo
sition of trace gases. In: F.S. Rowland and I.S.A. Isaksen (Editors), The Changing Atmosphere.
Dahlem Konferenzen, 1988. Wiley, New York, N.Y., pp. 159-186.
Ramanathan, V, Cicerone, F.J., Singh, H.B. and Kiehl, J.T., 1985. Trace gas trends and their potential
role in climate change. J. Geophys. Res., 90: 5547-5566.
Ramann, E., 1888. Die von Post'schen Arbeiten über Schlamm, Moor, Torf und Humus. Landwirtsch.
Jahrb., 17: 504-420.
Ramann, E., 1895. Organogene Ablagerungen der Jetztzeit. Neues Jahrb. Mineral. Geol. Paläontol., 10:
119-166.
Rennie, R., 1807. Essays on the Natural History and Origin of Peat Moss. Archibald Constable, Edin
burgh, 234 pp.
Rigg, G.B., 1940. The development of Sphagnum bogs in North America. Bot. Rev., 6: 666-693.
Rigg, G.B. and Richardson, C.T., 1938. Profiles of some Sphagnum bogs of the Pacific coast of North
America. Ecol., 19: 408-434.
Ruttner, F, 1963. Fundamentals of Limnology (D.G. Frey and F.E.J. Fry, translators). University of
Toronto Press, Toronto, Ont., 3rd ed., 295 pp.
Sallantaus, T, 1986. Terrestrial sources of aquatic organic matter. Luonnon TUtkija, 90: 191-197.
Selmer-Olsen, A.R. and Lie, O., 1983. Jemkarbonatutfelling (sideritt) i myr. Saertrykk av Jord og Myr,
4: 131-140.
Shaler, N.S., 1888. General account of the freshwater morasses of the United States, with a description
of the Dismal Swamp district of Virginia and North Carolina. U.S. Geol. Surv., 10th Annu. Rep.,
Part I, pp. 257-339.
224 W. Shotyk
Shotyk, W, 1987. The inorganic geochemistry of peats and the physical chemistry of waters from some
Sphagnum bogs. Ph.D. Thesis, University of Western Ontario, London, Ont., 748 pp.
Shotyk, W, 1988. Review of the inorganic geochemistry of peats and peatland waters. Earth-Sci. Rev.,
25: 95-176.
Shotyk, W, 1989a. The chemistry of peatland waters. Water Qual. Bull., 14: 47-58, 103.
Shotyk, W, 1989b. An overview of the geochemistry of methane dynamics in mires. Int. Peat J. 3: 25-44.
Shotyk, W, Nesbitt, H.W. and Fyfe, WS., in press. The behaviour of major and trace elements in
complete vertical peat profiles from three Sphagnum bogs. Int. J. Coal Geol.
Siever, R. and Woodford, N., 1979. Dissolution kinetics and the weathering of mafic minerals. Geochim.
Sjörs, H., 1980. Peat on Earth: multiple use or conservation? Ambio 9: 303-308.
Soper, E.K., 1919. The peat deposits of Minnesota. Univ. Minnesota, Minn. Geol. Surv., Bull. 16, 261
pp. plus maps.
Steele, L.P., Fraser, P.J., Rasmussen, R.A., Khalil, M.AK., Conway, T.J., Crawford, A.J., Gammon, R.H.,
Masarie, K.A and Thoning, K.W., 1987. The global distribution of methane in the Troposphere. J.
Atmos. Chem., 5: 125-171.
Stremme, H., 1908. Über Kaolinbildung. Z. Prakt. Geol., 16: 122-128.
Stumm, W, Wehrli, B. and Wieland, E., 1987. Surface complexation and its impact on geochemical
kinetics. Croat. Chim. Acta 60: 429-456.
Succow, M. 1988. Landschaftsökologische Moorkunde. Borntraeger, Berlin, 340 pp.
Succow, M. and Jeschke, L., 1986. Moore in der Landschaft. Harri Deutsch, Thun, 268 pp.
Swain, F.M., 1986. Composition of marsh gases in the central and eastern United States. Appi. Geo-
chem., 1: 301-305.
Tailing, J.F., 1957. The longitudinal succession of water characteristics in the White Nile. Hydrobiologia,
11: 73-89.
Tallis, J.H., 1973. The terrestrialization of lake basins in North Cheshire, with special reference to the
development of a "Schwingmoor" structure. J. Ecol., 61: 537-567.
Thurman, E.M., 1985. Organic Geochemistry of Natural Waters. Martinus Nijhof/W. Junk, Dordrecht,
497 pp.
von Biilow, K., 1929. Handbuch der Moorkunde. I. Allgemeine Moorgeologie. Einführung in das
Gesamtgebiet der Moorkunde. Borntraeger, Berlin, 308 pp.
von Post, H., 1862. Studier ofver nutidens koprogena jordbildningar, gyttja, dy, torf och mylla. K. Sven.
Vetenskapsakad. Handl., 4: 3-59.
von Post, L., 1924. Das genetische System der organogenen Bildung Schwedens. In: Quatrième Com
mission, Commission pour la Nomenclature et la Classifications des Sols. Comité Internationale de
Pedologie. IV Commission No. 22, Rome, 1924, pp. 287-304.
von Post, L., 1937. The geographical survey of Irish bogs. Irish Nat. J., 6: 210-227.
von Post, L. and Granlund, E., 1925. Sodra Sveriges torvtillgangar. Sver. Geol. Unders., Ser. C, No. 335.
Arsb., 19(2), 127 pp.
Walker, J., 1803. An essay on peat, containing an account of its origin, of its chymical principles, and
general properties. Prize Essays and Transactions of the Highland Society of Scotland, 2: 1-137.
Wasmund, E., 1930. Bitumen, Sapropel und Gyttja. Geol. Foren. Förhandl., 52: 315-350.
Websky, J., 1864. Beitrage zur Erkenntniss der Zusammensetzung und Bildung des Torfes. Z. Prakt.
Chem., 92: 65-96.
Weiss, F., 1910. Vorkommen und Entstehung der Kaolinerden des ostthuringischen Buntsandstein
beckens. Z. Prakt. Geol., 18: 353-367.
225
Chapter 10
Desert soils
A. WATSON
Introduction
Some 30 to 35% of the Earth's landsurface comprises areas with arid or semiarid
climates. Though the arid zone has usually been considered of marginal value for
agriculture, the growing need to expand food production in many parts of the
world is prompting ever increasing efforts to develop these regions. Particularly in
Africa, where nearly 60% of the landsurface is arid, and parts of Asia (33%), the
future exploitation of desert soils will be essential if development is to be sustained.
Desert landscapes commonly comprise regions of exposed bedrock (hamada), salt
encrusted basins and coastal flats, active eolian dunes, and occasionally areas where
intense erosion has created badlands. The materials mantling such landsurfaces are
not generally considered soils (Grossman, 1983), yet they make up as much as 40
to 60% of some of the Earth's deserts (Thomas, 1989a). For the purposes of this
review, however, many of these "non-soils" will be included in the Entisol order
of the USDA Classification (Soil Survey Staff, 1975) — the materials are mainly
Lithosols, Regosols and Fluvisols of the FAO Classification (Fitzpatrick, 1980; Lof,
1987). These two classification systems will be used extensively in this chapter. This
is in keeping with recent trends in soil classification, though it should not be assumed
that these systems are necessarily superior to others such as the Soviet (Lobova,
1967, 1968), Australian (Moore et al, 1983) and French (Fitzpatrick, 1980). Since
most of these taxonomic systems have been devised to meet the needs of regional
soil surveys, each has its merits and shortcomings.
Desert soils are broadly classified as surficial materials of arid regions exhibiting
pedogenic horizons — particularly eluvial or illuvial horizons. There are, however,
significant differences between the various definitions of "aridic" moisture regimes
and of the degree and type of pedogenesis which qualifies a material as a desert soil
rather than a non-soil or a soil of a more humid environment. For the purposes of
the USDA Classification, an aridic moisture regime is one under which soil water
is held at a tension of greater than 15 bar (15 x IO5 N m" 2 ) for most of the year
226 A. Watson
and there is no period of three months or more when water is continuously available
to plants (Soil Survey Staff, 1975). Cold temperatures do not necessarily preclude
aridity, so some soils of the Polar regions can be considered desert soils (Bockheim,
1980). However, pergelic soils are usually excluded from the main orders of desert
soils. Boyadgiev (1982) defined an aridic moisture regime as one under which the
ratio of mean annual precipitation to mean annual évapotranspiration is less than
0.40. He noted, however, that this can vary greatly for the same location depending
on whether the index is calculated using mean annual or mean monthly data. In
general, aridic moisture regimes have a growing season of from zero to 103 days
compared with 67 to 250 days for Mediterranean (xeric) regimes (Boyadgiev, 1982).
The characteristic features of pedogenesis in deserts result largely from aridity,
though other influences peculiar to these environments are sometimes significant in
soil formation. Yet the main classification systems incorporate very few diagnostic
criteria for identification of desert soils. For example, in the USDA Classification,
Aridisols are defined in part by exclusion from other orders; they cannot have oxic
or spodic horizons (distinguishing them from Oxisols and Spodosols, respectively).
In effect, while the main orders of desert soil are usually associated with aridic
moisture regimes, soils of several other orders are not necessarily excluded from
desert environments.
Poorly developed desert soils generally exhibit only an ochric (pale) epipedon; the
presence of additional pedogenic horizons is the main criterion for classifying them
as more highly developed soils (Aridisols, Solonchaks, Xerosols and Yermosols).
Typically, well developed desert soils have an ochric epipedon and one or more hori
zons which are either (a) natric or argillic, (b) salic, or (c) calcic, petrocalcic, gypsic,
petrogypsic or cambic, or they have a duripan. Few of these features are diagnostic
of desert soils alone. Natric, argillic and cambic horizons are not climatically specific.
Siliceous duripans appear to form in a variety of environments ranging from humid
tropical to hyperarid (Summerfield, 1983a, b; Watson, 1989) and the formation of
calcic and petrocalcic horizons has been reported from Greenland (Swett, 1974) to
Guatemala (Blount and Moore, 1969). Nevertheless, pedogenic calcretes are most
commonly found in semiarid and arid settings. The geographical distribution of soils
with gypsic, petrogypsic and salic horizons is generally even more environmentally
specific. Because of the high solubility of the minerals, such soil horizons occur only
in arid and hyperarid regions.
Usually, desert soils contain only small amounts of organic material owing to
the sparsity of vegetation cover. However, soils with mollic epipedons (MoUisols of
the USDA Classification) do occur in areas with aridic moisture regimes. Indeed,
they form intergrades with Aridisols (Gilè, 1977). Other surficial features such as
gravelly pavements and vesicular crusts which are common desert phenomena are
not useful criteria for soil classification (Nettleton and Peterson, 1983). Though they
can be significant geomorphic features, valuable in the differentiation of landscape
units, they occur on a variety of terrain and rarely have any universal association
Desert soils 227
with a specific soil group. It has been suggested that certain clay minerals may be
diagnostic of arid zone pedogenesis (Boyadgiev, 1982) but these can be inherited
from parent materials and occur as authigenic minerals only under very specific
edaphic conditions.
The main classification systems currently in use employ chiefly morphological and
mineralogical criteria to differentiate and subdivide the higher taxa of desert soils.
Though the names of soil units sometimes have strong genetic connotations, these
are not intended to be explicit. For example, Luvic Xerosols and Luvic Yermosols of
the FAO Classification have clayey B horizons but these are not necessarily illuvial
as the unit names imply. Indeed, for the purposes of the USDA classification, the
Soil Survey Staff (1975) deliberately abandoned the use of terms such as "sierozem"
and "solonchak" in an attempt to produce a nomenclature free of genetic impli
cations. However, such an approach to soil taxonomy and classification does not
provide a very useful foundation upon which to build a model of desert soil forma
tion which is widely applicable. The soils of desert regions are often perceived as
immature but morphologically stable because of the small amount of leaching and
low level of biological activity. Parent materials may be only minimally altered. Yet
many desert landscapes are extremely ancient, so pedogenesis may be very advanced
though its character can be markedly different from that in more humid regions. In
many arid environments, the presence of water-soluble salts has a profound effect
on geomorphic processes and pedogenesis. The influence of salt weathering on the
granulometrie properties of surface materials; the effects of salt accumulation on
soil mineralogy; and the role of soluble salts in diagenetic processes, all contribute
to the unique nature of desert pedogenesis.
Global climatic change, particularly during the Quaternary Period, has brought
about major fluctuations in the extent of the Earth's deserts. Vast areas of Africa,
Asia and Australia experienced intense aridification during the late Pleistocene.
Hence, though desert soils have been accorded comparatively little attention in
soil science literature owing to the limited agricultural potential of the areas they
occupy, the influence arid zone pedogenesis has had on the soils of semiarid regions
may have been greatly underestimated.
Classification of desert soils
Background
The USDA and FAO Soil Classification systems subdivide desert soil types us
ing similar but not identical criteria. In the USDA system, three main soil orders
commonly occur in deserts: Entisols, Aridisols and Vertisols. If semiarid regions are
included, Alfisols and Mollisols also become significant (Dregne, 1976). In the FAO
system, most desert soils fall within seven primary categories — Lithosols, Regosols,
228 A. Watson
TABLE 10.1
USDA, FAO, Soviet and Australian soil classification systems for the highest taxa of arid-zone soils
USDA CLASSIFICATION
Order Suborder Great group Symbol1
Aridisols Argids (Aa) Durargids Aa1
Nadurargids Aa2 SOVIET CLASSIFICATION
Natrargids Aa3
Paleargids Aa4 Type Subtype Genera
Haplargids Aa5 Grey-brown des Typical Ordinary
Orthids (Ao) Salorthids A0I ert soils Solonetz
Paleorthids Ao2 On compact rock
Durorthids Ao3 On loess
Gypsiorthids Ao4 Gypsic
Calciorthids Ao5 Low-carbonate Ordinary
Camborthids A06 Solonetz
Vertisols Torrerts V1 On compact rock
Mollisols Various M Gypsic
Alfisols Various Al Arid On compact rock
On loose material
Entisols Torripsam- Primitive desert Typical
ments E1 soils Residual humus
Torrifluvents E2 Saline
Torriorthents E3 Formerly irrigated
Sandy desert Typical
FAO CLASSIFICATION soils Weakly developed
Class Soil unit Correlation2 Crusted
Solonchaks Orthic Ao1 Palaeosols
Mollic M Takyrs Typical Ordinary
Takyric Various Sandy
Gleyic Various Formerly irrigated
Xerosols Haplic Ao3/Ao6 "Khaks" Leached
Calcic Ao5 (Ao2) Solonetz
Gypsic
Luvic
Ao4
Aa/Aa1/Aa5
Palaeosols _
Yermosols Haplic A03/A06 AUSTRALIAN CLASSIFICATION
Calcic Ao5 (Ao2) Great group Correlation2
Gypsic Ao4 Solonchaks Ao1/Aa3/Aa5
Luvic Aa/Aa1/Aa5 Grey-brown and red calcareous
Takyric A06 (Aa3) soils Ao5 (Aa5/Ao2)
Vertisols Chromic V1 Red and brown hardpan soils Aa1/Ao3
Solonetz Orthic Aa2/Aa3/AI Grey, brown and red clays V1
Mollic M
Gleyic Aa2/Aa3/AI Solonetz Aa3/AI
Fluvisols Various E2 Solonized brown soils A05/AI (Ao2)
Regosols Various E1/E3 Calcareous red earths Aa5 (Ao2)
Lithosols - E3/Non-soil Red earths AI (Torrox/Aa5)
Alluvial soils E2
Note ■: used here for correlation with other class- Lithosols Aa5/Ao6
ification systems. Calcareous sands E1
Note 2: generally, the closest equivalent - less likely Siliceous sands E1/E3
counterparts are in parentheses. Earthy sands E1 (Aa5)
Sources: Lobova (1967, 1968); Soil Survey Staff (1975); Fitzpatnck (1980); Moore et al. (1983); Lof
(1987); Malan'in (1987); and others.
Fluvisols, Solonchaks, Xerosols, Yermosols and Vertisols. The correlations between

these orders of the two systems, as well as between the lower taxonomic categories,
suborders and great groups of the USDA system, and soil units of the FAO system,
are shown in Tkble 10.1. It should be stressed, however, that since the criteria used
to classify soils differ, and the number of subdivisions within taxa varies, it is often
impossible to ascribe specific counterparts.
Moreover, because the USDA Classification has evolved to meet the needs of
North American soil surveys, some soils in other parts of the globe fit badly in the
Desert soils 229
overall scheme. For example, the highly weathered desert soils of Australia are 'Very
poorly catered for in the US Soil Taxonomy" (Moore et al., 1983, p. 263). This is
also true of many soils of hyperarid regions.
In contrast, the FAO Classification has been developed specifically to provide a
system suitable throughout the world. As yet, there is little information on the char
acteristics of different soils at all but the highest taxonomic levels. Hence, the FAO
system is of little practical value in detailed soil surveys. In comparison, the USDA
system has been developed in an attempt to rationalize the huge volume of empirical
data available at the lowest taxonomic levels, the soil families and series. Neverthe
less, the classification of a specific pedon using the USDA system often requires
data on the soil moisture regime, temperature and, perhaps, micromorphology. In
most deserts, this information is not readily available so very comprehensive soil
surveys employing this system are not feasible.
The USDA Soil Classification system

Three of the great groups of the Entisol order occur in warm deserts: Tor-
riorthents, Torrifluvents and Tbrripsamments. Respectively, these soils are devel
oped on recent erosion surfaces, fluvial sediments and eolian sands. By definition
they have an ochric epipedon, but no other pedogenic horizonation is present other
than perhaps exhumed, relict soil crusts in the case of some Torriorthents.
Aridisols have one or more pedogenic horizons in addition to the ochric epipe
don, this distinguishes them from Entisols. Aridisols make up about 36% of the
soils of the Earth's desert whereas Entisols account for about 40% (Dregne, 1976;
Table 10.2). Aridisols are divided into two suborders: Argids, which have an illuvial
clay horizon, and Orthids, which comprise all other Aridisols. Argids are subdivided
into five great groups and Orthids into six (Table 10.1). The definitions of these
great groups and the distinctions between them have been detailed elsewhere (Soil
Survey Staff, 1975).
Vertisols of the USDA Classification are differentiated from Aridisols (particu
larly Argids) by having more than 30% clay in a subhorizon within 0.50 m of the
surface, or by having cracks more than 10 mm wide or extending below 0.50 m, or
by having gilgai. Vertisols, mainly of the Tbrrert suborder, make up about 4.0% of
the Earth's desert soils lying outside the Polar regions (Dregne, 1976). Alfisols and
Mollisols make up about 7.0 and 12% of desert soils (Dregne, 1976; Table 10.2) but
are largely confined to semiarid regions (Figure 10.1).
The FAO Soil Classification system

Lithosols, Regosols and some Fluvisols of the FAO Classification equate most
closely with non-soils and Entisols of the USDA Classification (Lof, 1987). Most
well developed desert soils which fall within the Aridisol order equate with Solon-
chaks, Xerosols and Yermosols.
230 A. Watson
TABLE 10.2
Areal extent of different soil orders (USDA) in the Earth's arid and semiarid regions excluding the
Polar regions
Area in 000's km 2 and % of total land area (in parentheses)

ORDER AFRICA ASIA AUST. N. AMER. S. AMER. SPAIN
Aridisols 4,890 5,920 2,765 1,950 790 185
(16.4) (13.6) (36.3) (8.1) (4.5) (36.6)
Entisols 10,320 4,855 2,285 350 1,175 69
(34.6) (111) (30.0) (1.4) (6.7) (13.7)
Vertisols 245
(0.8)
780
(1.8)
760
(10.0)
100
(0.4)
- ■
Mollisols 120
(0.4)
2,850
(6.5)
- 1,790
(7.4)
490
(2.8)
-
Alfisols 2,085
(7.0)
- 440
(5.8)
165
(0.7)
380
(2.2)
-
Total area 17,660 14,405 6,250 4,355 2,835 254
(59.2) (33.0) (82.1) (18.0) (16.2) (50.3)
TOTAL % OF ALL % LAND

AREA ARID SURF.
SOILS
Aridisols 16,570 35.9 11.3
Entisols 19,149 41.5 13.1
Vertisols 1,885 4.1 1.3
Mollisols 5,475 11.9 3.7
Alfisols 3,070 6.6 2.1
The total area of some of the soil orders includes additional regions in parts of the semiarid zone of
Europe. Note that there are significant differences, especially in the extent of Aridisols and Entisols,
between the data presented here and that shown in Figure 10.1. (Source: Dregne, 1976.)
Fig. 10.1. Extent of Aridisols and distribution of soil orders (USDA) within regions receiving less than
250 mm of mean annual rainfall (after Wilding et al., 1983). Note that the areal extent of the soil orders
shown here reflects some marked divergences from the data presented in Table 10.2.
Desert soils 231
Fig. 10.2. Takyric soils developed in a clayey basin in southern Mauritania.
Solonchaks are salic soils which have no diagnostic pedogenic horizons other than
an epipedon and a cambic, calcic or gypsic horizon. They are subdivided into three
units: Orthic, Mollic and Takyric Solonchaks which have characteristic epipedons
(being ochric, mollic and clay-pan types, respectively). Xerosols and Yermosols dif
fer from each other only in the degree of development of the ochric epipedon; in
Xerosols it is weak, in Yermosols it is very weak (Fitzpatrick, 1980). Xerosols are
subdivided into four units: Haplic, Calcic, Gypsic and Luvic. While Xerosols are
found mainly in semiarid regions, Yermosols are true arid-zone soils (the name
being derived from yermo, the Spanish for "desert"). Five units are distinguished.
The first four are morphologically similar to the subdivisions of the Xerosols. The
fifth unit — Tàkyric Yermosols — includes most non-saline soils of desert clay pans
(takyrs or playas; Figure 10.2).
Vertisols of the FAO Classification are defined in essentially the same way as
those of the USDA Classification. However, their subdivision differs; in the FAO
Classification this is on the basis of chroma. Most arid-zone Vertisols fall within the
Chromic Vertisol unit, many equating with Tbrrerts of the USDA Classification. In
addition to these soil units, some Solonetz soils of the FAO Classification qualify as
desert soils — particularly those of the Gleyic Solonetz unit.
Desert soil classification in the USSR and Australia
The classification system employed in the USSR identifies four types of desert
soil with subtypes and genera distinguished on the basis of parent material, salinity,
carbonate content and residual features (Lobova, 1967, 1968; Table 10.1). Lobova
(1967, p. 277) stated that "the marked effect of rocks in the desert is so little
232 A. Watson
superceded by biological processes that the variations of soils due to differences

in the bedrock must be assigned a higher taxonomic level." Furthermore, Malanin
(1987, p. 5) observed that "the bulk composition (of sandy desert soils) indicates
the absence of distinct differentiation of chemical elements as a result of the weak
expression of soil-forming processes." Hence, many of the genera of the Soviet
Classification equate with Entisols of the USDA Classification and Regosols of the
FAO Classification.
A considerable amount of detailed soils mapping has been accomplished in Aus
tralia where the arid zone comprises about 80% of the continent. Correlations
between the great groups of the Australian Classification (Stace et al, 1968; North-
cote, 1979) and other systems cannot be precise (Table 10.1). Indeed, two great
groups — calcareous red earths and red earths — are characterized as soils domi
nated by sesquioxides. Such highly weathered soils are peculiar to many Australian
and some African deserts. They cannot be readily accommodated in the USDA
Classification other than as Haplargids or Camborthids, great groups which are
catch-alls for non-specific Aridisols.
Soil mapping in deserts
The current status of soils mapping in the Earth's arid zones is very variable
from region to region and country to country. Often, soil surveys of arid lands
are of little more than esoteric interest and generally not very detailed. However,
with increasing population pressure, there is a growing demand for exploitation of
marginal land particularly in the arid zones of Africa and Asia. Moreover, there
is also a renewed awareness of the potentially disastrous effects of desertification
in areas bordering the Earth's arid regions. The susceptibility of desert soils to
erosion by wind and water is often severe. Furthermore, if irrigation schemes are not
carefully monitored, soil salinization is a major threat. In effect, an understanding of
the physical and chemical properties of desert soils and of the nature of pedogenesis
in arid regions is a prerequisite for prudent environmental management.
Desert soils have been mapped at the global scale but only at the highest tax
onomic levels and at very small scales, so the resolution is poor. A comparison of
two maps which employ the USDA Soil Classification system reveals many marked
discrepancies, notably in the differentiation of Entisols and Aridisols (Dregne, 1976;
Wilding et al., 1983, p. 389; compare Figure 10.1 and Table 10.2). For most of Africa
and Asia these soils maps are little more than rough estimates of the distribution
of soils which are most likely to occur under the prevailing climatic and geomor-
phic conditions. Particularly in Soviet Central Asia, such simplistic and disparate
interpretations are unwarranted. Notwithstanding the acknowledged difficulties in
correlating the soil types of different classification systems, Central Asia has been
studied in sufficient detail to merit more accurate mapping of the soils using the
USDA Classification (Bespalov, 1964; Lobova, 1967; Evstifeev, 1980). In Africa
Desert soils 233
north of the Equator, there have been comparativley few detailed soil surveys.
However, with burgeoning international assistance for development of the region,
particularly the Sahel, our knowledge of the soils in even the most remote areas is
rapidly increasing (Vogg, 1981; Krumkachev, 1984).
The soils of the arid zones of Australia and North America have been mapped
in considerably more detail than those of Africa and much of the Middle East. Yet
even in these areas information on soil chemistry and mineralogy is often scant.
Hence, there remain many unanswered questions concerning the nature of pedo-
genesis in deserts; this lacuna in our understanding has major implications for the
use of desert soils as palaeoenvironmental indicators. There are major differences
of opinion concerning the origins of clay horizons in some soils (Argids) and the
sources of water-soluble salts which characterize many arid-zone soils. The follow
ing discussion will examine some of these diverse viewpoints and will provide a
geomorphological perspective on a few of the perplexities of desert soil science.
Parent materials
Desert landscapes are widely perceived as being made up predominantly of mo

bile dunes. In reality, most of the Earth's warm arid zones comprise extensive grav
elly or sandy plains and mountainous terrain. Indeed, in the south-western United
States, sand seas make up less than 1.0% of the arid zone (Table 10.3). Though eo-
lian processes account for much of the erosion and sedimentation in deserts, fluvial
processes and colluviation can have a marked influence on the landscape (Harvey,
1989; Oberlander, 1989; Reid and Frostick, 1989). This is because eolian sediments
are often made up of reworked alluvium and colluvium. Hence, fluvial processes
and mass-wasting are critical phenomena in the evolution of the terrain. Similarly,
TABLE 10.3
Geomorphological components of arid-zone landscapes expressed as percentages of the total area of

the various arid regions
Unit SW USA Sahara Libya Arabia Australia

Mountains 38.1 43 39 47 16
Low angle bedrock
surfaces 0.7 10 6 1 14
Alluvial fans 31.4 1 1 4 Ί
Γ
River plains 1.2 1 3 1
Dry watercourses 3.6 1 1 1
Badlands 2.6 2 8 1
Play as 1.1 1 1 1 1
Sand seas 0.6 28 22 26 38
Desert flats and
playa margins 20.5 10 18 16 18
Recent volcanics 0.2 3 1 2
Source: Thomas (1989a).

234 A. Watson
while hydrological basins such as sabkhas and desert pans account for only a small
proportion of most deserts, their importance as sources of deflated salt and dust is
profound.
Weathering processes
The first step in the formation of most sediments and soils is the disintegration
of bedrock by weathering. Owing to the paucity of surface water and soil moisture
in most desert settings, mechanical weathering generally prevails over chemical pro
cesses such as dissolution, oxidation and reduction, and hydration and hydrolysis.
However, it would be incorrect to assume that all chemical weathering processes are
insignificant — biochemical weathering, for example, can have a strong influence
on the chemistry of desert soil (Whitford and Freckman, 1988). Moreover, while
it is commonly assumed that mechanical processes of rock disintegration such as
pressure release, wetting and drying, insolation and salt weathering produce coarse
debris and little silt and clay, this is another misconception. In many arid environ
ments, silt and clay-sized materials are formed during weathering; indeed, it can
be an important component in atmospheric fallout and large volumes of this eolian
dust may form extensive deposits of loess (Pye, 1987; Thomas, 1989b).
The relative importance of the various mechanical weathering processes occur
ring in arid environments usually depends on highly localized conditions such as
moisture and salt availability, and lithology. Fracturing by pressure release resulting
in rock exfoliation is commonly associated with plutonic rocks but any rock which
has been deeply buried can experience this form of weathering as lithostatic pres
sure is released during unloading. Periodic wetting by rainfall, dew or fog, or even
atmospheric humidity, and subsequent desiccation can have a marked influence on
the disintegration of rocks and minerals in deserts. Though the effects are most
pronounced on claystones and shales, even crystalline rocks can deteriorate in time.
Owing to the increased coefficient of expansion when rock contains moisture, rel
atively small fluctuations in temperature can produce much greater strain when a
rock is wet rather than dry (Goudie, 1989). Indeed, it has been argued that cyclic
heating and cooling alone has little effect on the integrity of most rocks (Schattner,
1961; Roth, 1965). At present, there is little experimental evidence to support the
contention that rocks and boulders may split in response to the large temperature
fluctuations that are common in deserts. However, laboratory simulations of such
insolation processes suggest that they may cause microfracturing of mineral grains
(Goudie, 1989). This could be an important precursor for the inception of more
pronounced disintegration involving other mechanisms.
Salt weathering
Salt weathering occurs in a wide variety of environments ranging from the cold
and dry climate of Antarctica to moist conditions in temperate zones but it is in
Desert soils 235
warm desert settings that its intensity and pervasiveness are most evident (Cooke,
1981; Goudie and Watson, 1984; Sperling and Cooke, 1985). Three main mecha
nisms have been identified: hydration, crystal growth and thermal expansion. Several
common salts occur in different stages of hydration under natural conditions. For ex
ample, the transition from hydrous sodium sulphate (thenardite — Na2SO4.10H2O)
to the anhydrous phase (mirabilite — Na2SC>4) occurs readily above about 32°C.
The pressure generated by volumetric changes during hydration can fracture rocks
within which the salts have precipitated. The conversion from anhydrite (CaS0 4 )
to gypsum (CaS0 4 .2H20) can generate some of the highest pressures (Winkler and
Singer, 1972) but the increase in volume results from the introduction of water of
crystallization. In a closed system, where water is present prior to hydration, swelling
does not necessarily occur. Under natural conditions, volumetric expansion during
the hydration of anhydrite may often be related to the presence of swelling clays
(Henke, 1976) or sulphide oxidation (Dougherty and Barsotti, 1972). The latter
process can produce an eightfold increase in volume.
It is likely that crystal growth during the hydration or precipitation of salts is
the major mechanism involved in salt weathering. In the case of calcium sulphate
minerals, for example, the conversion of the hemihydrate (2CaS0 4 .H20) to gyp
sum involves a theoretical decrease in volume of about 7.0%, but acicular crystal
growth results in a net increase (Budnikofï, 1928). Salt crystallization may result
from evaporation of solvent, falling temperature, or changing chemistry of the so
lution. When it occurs in the pores and fissures of a rock, fracturing or granular
disintegration can be extremely rapid (Goudie and Watson, 1984; Goudie, 1989).
It has also been reported that calcite crystallization can accomplish quartz-grain
breakage in pedogenic calcretes (Braithwaite, 1989). This has important implica
tions for our understanding of the mechanisms involved in the illuvial accretion of
pedogenic mineral horizons because it indicates that crystallization can displace host
materials.
Probably of lesser significance than salt hydration and crystal growth is the role of
thermal expansion of invasive salts on the disintegration of rock (Mortensen, 1933;
Cooke and Smalley, 1968). Certainly, many common salts have higher coefficients
of thermal expansion than granite, for example, but the volumetric changes are not
great (Goudie, 1989).
The influence of salt weathering on desert landscapes and surficial materials is
still widely underestimated despite the increasing amount of evidence indicating
that it is extremely pervasive. Not only is salt weathering implicated in the forma
tion of tors and pedestal rocks, etchplains and desert pans, but it is also of major
importance in the comminution of particulate rock debris. It has been shown exper
imentally that silt-sized quartz particles can be formed by salt weathering of sand
grains (Goudie et al, 1979; Pye and Sperling, 1983). Accumulations of such ma
terial in saline basins may provide important sources of eolian dust when they are
eroded. Much of the atmospheric dust deflated from the Sahara (Prospero et al.,
236 A. Watson
1970; Morales, 1979, pp. 8-11; Prospero, 1981) and Central Asia (Blank et al, 1985)
is probably formed by salt weathering and eolian abrasion (Pye, 1987).
Extrinsic materials
Silt and clay
While it would be incorrect to assume that authigenic clay minerals are ab
sent from desert soils, silt and clay-sized materials are often derived from external
sources rather than being autochthonous (Yaalon and Ganor, 1973). Atmospheric
dust fallout is one of the main sources of fine-grained clastic material in desert
soils. The mobilization of fine particulate material is achieved by highly turbulent
airflow over bare, often disturbed, landsurfaces (Middleton, 1989). The material
is transported in suspension within the airstream. Pye (1987) has pointed out that
active accumulation of eolian dust occurs mainly where rainfall is sufficient to sup
port dust-trapping vegetation. Hence, while many desert loess deposits are found
peripheral to the Earth's arid regions (Thomas, 1989b), they are uncommon within
hyperarid areas. This notwithstanding, eolian dust influences the character of soils
and morphology of the landscape in many deserts. Indeed, some have contended
that widespread phenomena such as desert varnish and desert pavements may be
attributed to dust deposition. It has even been suggested that mineral enrichment in
some bauxites may be ascribed to eolian inputs (Brimhall et al., 1988).
Salt
In most instances, water-soluble salts occurring within the soil zone are also
allochthonous. Particularly in Australia, there has been considerable debate on the
provenance of the salts which create saline soils. It appears unlikely that they are
derived from solutes in rainfall because while sodium chloride is usually the most
common salt in the soils, chloride levels in rainwater decrease rapidly away from
the coast. However, the deposition of saline aerosols probably results in a slow
net accumulation of salt in the solum if rainfall is insufficient to flush out the
saline minerals periodically (Yaalon, 1971). In addition, the weathering of bedrock
and deflation of saline sediments from evaporative basins can contribute to soil
salinization (Isbell et al, 1983). The phenomenon is most pronounced in regions
which have experienced prolonged aridity. In Australia, for example, actual chloride
concentrations in soils about 400 km inland can be accounted for if 2.0 kg of salt per
ha per year were deposited over about 725,000 years (assuming that no depletion
has occurred and that the rate of deposition has remained constant) (Isbell et al.,
1983). Elsewhere, some extremely high deposition rates have been reported: 10 to
20 kg of salt per ha per year 50 km inland from the coast of Israel (Yaalon, 1964b),
Desert soils 237
100 kg of salt per ha per year in the Namib Desert (Boss, 1941), and 475 kg of salt
per ha per year in Kalmyckaja, west of the Caspian Sea (Shumakov and Mikhovich,
1960). Clearly, such massive inputs of water-soluble material have a major influence
on soil and ground-water chemistry when the salts are mobilized by leaching.
Even in hyperarid regions, rainfall can be sufficient periodically to flush some
of the water-soluble salts through the soil zone and into ground water. Once the
solutes reach the phreatic zone, subsurface flow carries them to topographic basins.
If the water-table in these basins lies close to the landsurface, evaporation will
precipitate the dissolved minerals or salts which speciate from the solutes. In turn,
the saline evaporites deflated from these basins may be reintroduced into soils on
the surrounding landscape.
Clastic and clayey soils
For the most part, desert soils are developed on coarse clastic materials such as
mechanically weathered rock, poorly-sorted fluvial sediments and dune sand. So,
since cambic soil horizons, which are one of the diagnostic features of incipient
pedogenesis, can develop only in materials finer than very fine sand (Grossman,
1983), most desert soils must be classified as Entisols (Lithosols, Regosols or Flu-
visols of the FAO Classification). In humid settings, such soils would be expected
to evolve into Inceptisols as chemical weathering, especially vermiculitization and
kaolinitization, and pedogenesis progress (Allen and Fanning, 1983). But in the arid
zone, where clay authigenesis is severely restricted, the development of first a cam
bic horizon and subsequently an argillic, salic, calcic or gypsic horizon or a duripan
would generally necessitate the introduction of allochthonous material. This is usu
ally accomplished by deposition of aerosols and airborne dust at the surface or by
hydromorphic processes if there is a shallow water-table.
The origin of desert soils exhibiting clayey pedogenic horizons has been the
subject of considerable debate, particularly in the North American arid zone where
Argids (Haplic or Luvic Xerosols and Yermosols of the FAO Classification) are a
common suborder. Here, most soil scientists have contended that the clay horizons
are illuvial but relict, having formed in the late Pleistocene when the climate of the
southwest United States was generally wetter than today (Gilè, 1966). The absence
of clay skins, which are diagnostic of active clay illuviation (Gilè and Grossman,
1968; Nettleton et al., 1969) and the fact that the maximum clay content is often
below the depth of 100 year wetting (Nettleton et al., 1975), support the view
that these are fossil horizons, preserved because of current aridity. However, some
weakly developed soils occurring in Holocene sediments exhibit incipient illuvial-
clay horizonation (Gilè, 1975). It is also possible that clay horizons lying below
the depth of 100 year wetting are accumulating under the current climatic regime
as a result of clay translocation to still greater depths at times of very infrequent
238 A. Watson
wetting (Nettleton and Peterson, 1983). Periodicity in rainfall trends over intervals
of tens or hundreds of years may have a strong influence on pedogenesis; this may
not be evident if only short-term meteorological data are available. Furthermore,
the presence of sodium salts may enhance clay eluviation (Peterson, 1980). So the
occurrence of an illuvial clay horizon is not necessarily indicative of an old (late
Pleistocene) soil or landsurface.
An important type of clayey desert soil which is especially widespread in Central
Asia is associated with clay pans (takyrs) which are subject to seasonal inundation
(Lobova, 1981; Figure 10.2) or prolonged flooding in the case of khaks (Lobova,
1967). The soils comprise the Tkkyric Solonchaks and Tàkyric Yermosols of the FAO
Classification. Their classification in the FAO and Soviet systems is based more upon
geomorphological rather than specifically pedological criteria. By definition the soils
occur only in depressions filled withfine-grainedsediments but their morphological
features are very varied. They commonly have salic or gypsic horizons but some
exhibit only a cambic horizon. Hence, the soils do not equate with any specific great
group of the Aridisols nor are they exclusive to either of the suborders.
While clay authigenesis is weak in desert soils (Buoi, 1965; Lobova, 1981) and ex
trinsic clay sources are usually dominant, arid-zone pedogenesis does exhibit certain
characteristic traits of clay mineralogy (Folkoff and Meentemeyer, 1987). Because
of the nature of weathering processes, most clay formation in deserts occurs in
the solum rather than in saprolite (Nettleton and Peterson, 1983). About 75% of
the Argid soil series and 65% of the Orthids in the United States are dominantly
montmorillonitic, the others are mostly of mixed-clay mineralogy. Though mont-
morillonite dominates, it is readily converted to the mixed-layer clays (Allen and
Fanning, 1983) and to clay mica as a result of plant activity (Nettleton et al., 1973).
The products of more advanced feldspar weathering, such as gibbsite and kaolinite,
are rare except where ancient weathering profiles are preserved. Illite is generally
stable under arid conditions and may even occur as an authigenic mineral formed
from smectite by the addition of potassium derived from atmospheric inputs (Singer,
1989).
Another group of clayey or silty soils which has been widely reported despite
being rare in many arid regions has been loosely termed red desert soil. Though
probably equating with Haplargids or Camborthids, these soils are difficult to clas
sify because they also resemble Oxisols (Torrox?) of the USDA system and Ferral-
sols of the FAO Classification. In Central Australia, most red earths are associated
with ancient lateritic weathering profiles, so their characteristic physical properties
may well be inherited from the saprolitic parent materials (Northcote and Wright,
1983). Nonetheless, in some arid regions, reddening of cambic soil horizons (Gilè,
1966; Lobova, 1981) and surficial materials such as dune sand can be pronounced.
In the case of desert dune sand, the phenomenon results from the dehydration of
yellowish hydrous ferric oxide (derived from the weathering of ferromagnesian sili
cates) to red hematite. This stains clays adhering to the surface of sand grains (Folk,
Desert soils 239
1976). The clay cutans themselves are probably of illuvial origin, being derived from
clay eluviated from atmospheric dust deposits rather than authigenic clays. Hence,
though chemical weathering accounts for the formation of the red pigment, its
fixation depends upon an external source of clay (Walker, 1979). This would char
acterize the surficial materials as cumulative soils. While the intensity of the red
coloration has been used as an indicator of age (Besler, 1976), many other factors
can influence the hue and chroma. These include the quantity and mineralogy of
the iron oxides (Pye, 1983) and clays, the size, shape and provenance of the sand
grains, and moisture conditions which affect the rate of weathering as well as the
stability of the sand surface (Walker, 1979). Indeed, rubefied sands may lose their
redness as they age because further weathering can result in green-darkening (Folk,
1976). Some red beds found in the rock record have been interpreted as former
desert soils which underwent pedogenic or early diagenetic reddening. However,
Pye (1983) pointed out that such soil profiles are rarely more than 1.5 m thick, so
most thick red-bed sequences probably originate as soils or weathering profiles in
wetter environments.
Saline and sodic soils
Soils exhibiting accumulations of water-soluble minerals are typical of most arid

environments; they are also found in semiarid settings. Salorthids have been de
scribed from Spain (Gumuzzio et al, 1982). Though Typic Salorthids and Solon-
chaks are usually restricted to hydrological basins or littoral settings where evap-
orite minerals are abundant, saline or sodic soils are much more widespread. In
Australia, for example, about 5.0% of the landsurface (about 400,000 km2) has
saline soils (>0.10% NaCl in sands and loams, >0.20% NaCl in clays, or >0.30%
NaCl in the B horizon) and 25 to 30% of the soils are sodic having an exchangeable
sodium percentage (ESP) greater than 6.0 (Isbell et al., 1983). The salinity of desert
soils determines not only the type of vegetation they are able to support but also
the character of pedogenesis. Salinity affects the rate of weathering and clay authi-
genesis, and sodicity strongly influences the mobility of clays and the stability of the
solum (Yousaf et al., 1987; Goldberg et al., 1988).
Surficial and near-surface accumulations of water-soluble minerals in the regolith
have been ascribed three main origins: most are either subaqueous, phreatic or va
dose evaporites (Watson, 1983,1989). The first type is found in littoral, lagoonal and
lacustrine settings in the arid zone where the evaporation of bodies of saline water
commonly precipitates minerals such as gypsum and halite which accumulate as sed
imentary strata. Particularly in volcanic regions, however, certain highly soluble min
erals (nitratine (NaN0 3 ), natron (Na2CO3.10H2O) and trona (NaH(C0 3 ).2H 2 0),
for example) can form thick accumulations in evaporative basins (Stoertz and Erick-
sen, 1974; Eugster, 1986). Because many of these salt deposits are toxic to all higher
plant life, they cannot be considered true soils.
240 A. Watson
Salt crusts of phreatic origin are common in desert basins and ephemeral drain
age networks where the ground water contains dissolved salts. Saline soils develop
when the water-table approaches close enough to the landsurface for the ground
water to be subject to evaporation. This usually occurs within about 2.0 m of the
surface but the depth varies according to air temperature and the textural properties
of the surficial materials. The precipitated minerals may be derived from bedrock or
may be leached from surface deposits by rainwater percolating to the water-table.
Vadose evaporites are water-soluble salts introduced into the solum and regolith
in meteoric water or in moisture moving laterally as throughflow within the soil
zone (Conacher, 1975; Risacher, 1978). Subsequent evaporation precipitates the
dissolved minerals; net accretion occurs only if periodic wetting is insufficient to
flush the accumulated salt to the water-table. While diffuse salt accumulation is
common in many desert soils, in all but hyperarid environments occasional thorough
wetting of the soil zone results influshingand prevents the build-up of discrete salic
horizons unless there is a high rate of salt supply. In general, clayey soils are more
likely to be persistently saline because they are less readily leached. The main salt
sources are atmospheric fallout (eolian sand and dust, and aerosols) and cumulative
deposits of colluvial or alluvial sediments at the surface. Locally, bedrock or ancient
lacustrine evaporites may exert a strong influence on soil salinity or gypsum content.
In many arid regions, the use of saline irrigation water leads to soil salinization when
management techniques are poor. While high salinity is itself toxic to most plants,
the evaporative accumulation of certain elements such as selenium and arsenic may
poison the soil still further.
Sodic soils (those having a high ESP) are not necessarily saline but it is likely
that many examples of such soils were formerly saline. Certainly, factors such as
the soil parent materials, as well as clay content and mineralogy, can have a major
impact on sodicity. However, it has been suggested that the sodicity of many soils
in Africa (Watson et al., 1984) and Australia (Isbell et al., 1983), for example, may
have originated under former, more arid climates. It has been proposed that some
saline, argillic soils may have evolved into Natrargids and Solonetz soils in response
to the amelioration in climate since the beginning of the Holocene because leaching
of soluble salts, deflocculation of colloids, and downward translocation of clay oc
curred (Isbell et al., 1983). High sodicity is most common in relatively impermeable
materials but it is uncertain whether this is because such materials are most suscep
tible to initial salinization or because low permeability is itself a function of high
sodicity. In some saline soils, a sodic horizon occurs immediately above the salic
horizon (Eisenberg et al., 1982). Presumably, leaching of water-soluble salts during
wetter than average years creates this residual sodicity. It should be pointed out,
however, that this apparent relationship between sodicity and prior salinity is far
from universal. In North America, for example, some Natrargids are attributed to
recent salinization of older Argids in response to Holocene aridification (Nettleton
and Peterson, 1983).
Desert soils 241
Soil crusts
Desert soils exhibiting indurated mineral horizons composed predominantly of

silica (termed silcretes or duripans), calcium carbonate (calcretes), gypsum (gyp-
cretes) or halite (salcretes) are widespread in arid regions. Indeed, some desert
soil classifications are for the most part based upon the type and intensity of such
mineral cementation (Boyadgiev, 1982). These soils generally comprise the Orthid
suborder of the USDA Classification, soils which are probably most characteristic
of active pedogenesis in deserts. The soils often form in response to prolonged
accumulation of a specific mineral in the illuvial B horizon of the solum or at a
shallow water-table. In this respect they differ from the uncemented clay crusts of
most takyric soils and vesicular crusts found on many loessial soils in deserts. They
are also distinct from subaqueous evaporite crusts because these generally occur as
surface features unless buried by subsequent sedimentation (Warren and Kendall,
1985). Certain types of these mineral crusts, particularly silcretes and calcretes, are
not formed exclusively in arid environments and those that are may have a variety
of origins. Hence, assumptions about the nature of pedogenesis or palaeoenviron-
ments based upon the occurrence of specific types of duricrust must be tempered
with caution.
The distribution, nature and origins of calcretes and silcretes are dealt with
elsewhere in this book and by Watson (1989). Here, the characteristics of gypsum,
halite and some other evaporite crusts will be discussed. These soils usually fall
within the Gypsiorthid and Salorthid great groups of the USDA Classification, and
Gypsic Xerosol and Yermosol and various Solonchak units of the FAO system.
Soils exhibiting petrogypsic horizons occur in parts of most of the Earth's deserts
but are most extensive in Central Asia, North Africa and the Middle East, Australia,
and the Namib Desert. Generally, they are found in areas with less than about 250
mm of annual rainfall (Watson, 1985). Often their distribution is intimately linked
with the occurrence of evaporative basins. Soils with pedogenic salic horizons which
are well cemented are far less extensive than pedogenic gypsum crusts. They have
been reported mainly from hyperarid regions such as the Namib Desert (Watson,
1983) and the Atacama Desert (Ericksen, 1981) where mean annual rainfall is less
than about 25 mm. Again, their development is commonly associated with high
rates of salt deposition in atmospheric fallout (Orlova, 1980; Ericksen, 1981) or fog
moisture (Watson, 1983). Examples of halite crusts developed at the water-table are
also less common than phreatic gypsum crusts but they have been reported from
some hypersaline sabkhas (Shearman, 1963; Arakel, 1980).
Petrogypsic soil horizons may be of phreatic or vadose origin. Though exhuma
tion of these subsurface accumulations results in micromorphological and chemical
modification, the two main genetic forms are readily distinguished. Ground-water
crusts which can reach thicknesses approaching 5.0 m are typically composed of
large, lenticular gypsum crystals up to about 0.50 m in diameter. The development
242 A. Watson
of this crystal form and habit is probably related to the presence of other ions —
particularly sodium — in the precipitating solutions (Edinger, 1973), though com
plex organic compounds may also have a strong influence (Cody, 1979; Cody and
Cody, 1988). Typically, these crusts have higher sodium concentrations than the va
dose forms (Watson, 1985). The large crystals often incorporate fragments of host
material; such poikilitic inclusions are very rare in the crusts of vadose origin and
probably attest to uninterrupted crystal growth under stable hydrochemical condi
tions (Watson, 1988). The pedogenic gypsum horizons which form in the vadose
zone are usually of illuvial origin; they rarely exceed a thickness of 2.0 m unless they
are polygenetic. The gypsum crystals within these horizons range in size from lentic
ular or euhedral forms about 1.0 mm long, to crystallites less than about 50 μηι in
diameter which form cryptocrystalline masses of alabastrine gypsum. The develop
ment of alabastrine textures is often associated with exhumation of illuvial horizons.
Periodic wetting and drying results in the dissolution of larger crystals and reprecip-
itation of alabastrine material from evaporating soil moisture. In the vadose zone,
the migration of ions to crystal-growth sites is severely restricted owing to the lim
ited availability of moisture, so only minute crystallites develop. This phenomenon,
akin to sparmicritization in calcretes (Kahle, 1977), leads to a reduction in the con
centration of ions such as sodium and strontium in the exposed crusts. This is proba
bly because these ions are coprecipitated within the gypsum crystal structure during
the growth of larger crystals (Kushnir, 1980) but since the alabastrine crystallites
experience little growth following nucleation, further coprecipitation is inhibited
(Watson, 1985).
There is comparatively little information on the chemistry and micromorphology
of pedogenic halite crusts. Some data is available from salcretes which occur in
association with nitrate beds in the Atacama Desert (Ericksen, 1981). The mineral
ogy of the salcretes appears to be dependent on the chemistry of surficial deposits
of soluble minerals in fog water and volcanic dust, and to a lesser extent on the
chemistry of the host material. As with gypsum crusts, the crystalline habit of halite
in salcretes varies according to the depositional environment. In phreatic settings,
granular textures predominate (Arakel, 1980), while rapid, sporadic precipitation in
vadose settings creates saccharoidal textures made up of small crystals. Halite pre
cipitated from evaporating surface water usually exhibits the cubic habit and hopper
crystals with negative-crystal inclusions of mother brine are common (Lowenstein
and Hardie, 1985). Halite efflorescences and some beach salcretes (Pye, 1980) de
velop when saline soil moisture or shallow ground water is drawn to the landsurface
during intense evaporation. The powdery deposits usually exhibit fibrous or cubic
crystal habits (Eswaran et al., 1980).
A common feature of illuvial petrogypsic and halite horizons is the development
of columnar structures (Figure 10.3). The vertical columns are up to about 0.75 m
in diameter and extend through the whole horizon. They have been attributed to
volumetric fluctuations resulting from chemical dehydration (Hunt et al., 1966) or
Desert soils
Fig. 10.3. Columnar gypsum crust exposed beneath a thin cover of loess, southern Tlinisia.
Fig. 10.4. Desert pavement of basalt cobbles and pebbles overlying a well indurated gypsum crust,
central Namib Desert.
thermal contraction (Kocurek and Hunter, 1986) but a more likely cause is periodic
swelling and shrinkage in response to wetting and subsequent desiccation (Watson,
1985).
Though the origin of desert soil crusts has long been debated, a consensus has
evolved that most are either hydromorphic (phreatic) or illuvial (vadose). Phreatic
accretion has been attributed to chemical interactions (Pouget, 1968), but the evap
oration of ground water is the most likely mechanism. Similarly, gypsum and halite
horizons in the vadose zone have been ascribed a variety of origins. The belief
244 A. Watson
that they are buried sedimentary strata deposited in lake basins (Stainier, 1912),
as alluvium (Butzer and Hansen, 1968) or eolian dust (Coque, 1955, 1962; Jessup,
1960; Tucker, 1978) are common misconceptions. The widespread occurrence of
these soils on undulating landscapes, and beneath gravelly surface pavements (Fig
ure 10.4), strongly suggests that the gypsic and salic horizons are illuvial. Illuvial
accretion occurs when soil moisture containing water-soluble salts leached from sur
face evaporates. Provided there is net salt accumulation even taking into account
infrequent wetting which exceeds the moisture storage capacity of the soil, well-
cemented subsurface crusts containing up to about 90% gypsum or halite can form.
Clearly, passive void-filling crystallization alone cannot account for such high con
centrations of introduced material. The mode of accretion of such horizons has been
the subject of considerable debate; in the case of some silcretes and calcretes, chem
ical replacement of host material may be involved. However, there is little evidence
for this in gypcretes (Watson, 1985). Upward heaving of the overburden owing to
subsurface displacive crystallization in the illuvial horizon is probably the main ac-
cretionary mechanism. However, pseudoanticlines — domes resembling structural
folds in bedrock — which commonly occur in pedogenic calcretes (Watts, 1977) and
gypcretes, attest to horizontal compression as well.
Under certain conditions, two-tiered illuvial horizons may develop in some soils.
Examples of gypsum crusts beneath calcretes (Horta, 1980; Figure 10.5), halite
crusts beneath gypsum crusts (Watson, 1983, 1985) and crusts of still more soluble
salts beneath halite crusts (Ericksen, 1981) can be attributed to the different vertical
mobilities of the leached salts (Yaalon, 1964a). These duplex soil crusts form only
where both minerals are present in surface inputs, and where the moisture regime
Sand with calerete nodules
Calerete
/ Gypsum — ' ..^"

^ Base
Fig. 10.5. TWo-tiered calcrete-gypsum crust, southern Tunisia.

Desert soils 245
is wet enough to mobilize sufficient quantities of the less soluble mineral but not so
moist that the more soluble salt is leached through the solum (Drever and Smith,
1978). Since any single wetting of the soil will mobilize minerals in quantities pro
portional to their different solubilities, the mineralogy of the salt horizons might be
expected to reflect this balance. However, because the depth of individual wetting
incidents is variable, the more soluble salts are periodically flushed from the upper
soil and reprecipitate at depth (Nettleton and Peterson, 1983; Watson, 1983). Addi
tional factors such as the mineralogy of the surface deposits (and host material) may
promote chemical interactions which augment the differentiation of the salt hori
zons. For example, owing to the common ion effect (Wigley, 1973) gypsum solubility
decreases in the presence of calcium carbonate; on the other hand, the presence of
sodium chloride enhances gypsum solubility.
Diagenetic features of desert soils
Many desert soils exhibit mineralogical and morphological features which are
perhaps more correctly characterized as the products of diagenesis rather than pe-
dogenesis. The two phenomena are not mutually exclusive, so certain processes
such as clay authigenesis may be a feature of both the later stages of pedogenesis
and early diagenesis. While many clay minerals in desert soils are either inherited
from parent materials or introduced at the surface (Buoi, 1965; Allen and Fanning,
1983), some authigenic clays may be diagnostic of arid-zone pedogenesis (Boy-
adgiev, 1982). The origin and significance of palygorskite (attapulgite) and sepiolite
in desert soils has received considerable attention in recent years. In some soils,
these clays may be inherited (McLean et al, 1972; Shadfan and Dixon, 1984) or
of hydromorphic origin (Hodge et al., 1984), but there is compelling evidence that
under alkaline conditions they may be authigenic (Yaalon and Wieder, 1976; Singer,
1984). Watts (1980) attributed palygorskite formation in calcretes to interactions
between montmorillonite and magnesium liberated from high-magnesian calcite as
well as authigenesis from solutions rich in magnesium and silica. It is important to
note, however, that the range of environmental conditions under which palygorskite
and sepiolite form is very narrow (Brinkman, 1982), and palgorskite may readily
weather to montmorillonite (Bigham et al., 1980). Under similar alkaline conditions
but when sodium rather than magnesium activity is high, the sodium zeolite anal-
cime may form (Brinkman, 1982). Analcime may persist as a metastable mineral
even if pH and sodium activity fall into the range of montmorillonite stability. These
same processes probably occur in gypsic soils (Yaalon and Wieder, 1976; Reheis,
1987) but comparatively little work has been undertaken on the clay mineralogy of
such soils.
Uranium mineralization is an unusual but economically significant diagenetic
feature of some calcretes and gypsum crusts (Carlisle et al., 1978; Arakel and
246 A. Watson
McConchie, 1982; Briot, 1983). Most commonly it is associated with supergene

enrichment in the phreatic zone. Indeed, some major supergene ore bodies may
be primarily pedogenic features (Brimhall et al., 1985; Brimhall and Dietrich,
1987).
In general, the nature of diagenesis in most desert soils other than Calcior-
thids, Durorthids and related Paleorthids is poorly known. In part this is because
soils containing horizons of minerals more soluble than calcite and silica (mainly
Gypsiorthids and Salorthids) are very susceptible to degradation when the climate
becomes wetter. While calcretes and silcretes can persist under semiarid, even in
subhumid conditions, well developed petrogypsic and halite crusts are confined to
regions which have experienced uninterrupted aridity throughout the Quaternary.
Indeed, some common diagenetic features of gypsum crusts, such as alabastrine
overgrowths, develop only during the exhumation of illuvial horizons and this is of
ten precursive to their degradation as a result of enhanced leaching (Watson, 1988).
Despite this vulnerability, some of these crusts create geomorphological features
which can persist long after the mineral horizons have dissipated (Figure 10.6).
Notwithstanding the impediments to their preservation, pedogenic gypsum crusts
have been reported from the rock record (Wright et al., 1988). It is likely that many
other examples of lithified desert soils have simply been misinterpreted owing to
the current lack of awareness of their prevalence and morphological characteristics.
The discovery of ancient silcretes and calcretes in the rock record has been made
possible only by the dissemination of geochemical and morphological information
Fig. 10.6. Vegetation polygons formed by grass growth (Stipagrostis gonatostachys) in sand-filled fissures
in surficial sands cemented by calcium carbonate and gypsum, central Namib Desert. It is likely that the
assuring is a result of volumetric changes which occurred during the desiccation of former gypsum-rich
soils.
Desert soils 247
obtained from modern analogues. Certainly, many more types of desert palaeosols
await identification, though the potential for preservation is undoubtedly greatest
for siliceous and calcareous soils.
Desert soils and geomorphology
Though locally it may be of major importance, the role of surface runoff in shap
ing desert landscapes is often far less influential than mass-wasting or eolian pro
cesses. This can result in the prolonged preservation of landforms which would nor
mally degrade in more humid environments. For example, the persistance of ancient
alluvial fans (Harvey, 1989) or relict pediments is a common feature of many arid
and semiarid regions. Such landforms may owe their preservation to encrustation
by palaeosols which can create armoured landscapes (Goudie, 1984; Milnes, 1986),
but, on the other hand, the stability of such surfaces contributes to the development
of very mature soils. However, neither soils nor landscape can be regarded as static,
independent phenomena: the two evolve in tandem (Tïicart, 1978). Topography and
geomorphology play major roles in the evolution of desert soil associations, not
least in their influence on the parent materials produced by weathering or by ero
sion and sedimentation. Topographic basins often form hydrological sinks wherein
the evaporation of saline water precipitates salts which contribute to soil salinization
on a regional scale when they are deflated and subsequently redeposited in eolian
dust. On a smaller scale, the movement of saline soil moisture as throughflow down
desert slopes can produce distinctive catenas with basal Salorthids (Conacher, 1975)
or sodic soils (Nyamapfene, 1986). It has been reported that marked differences in
soil development may occur even in response to slope aspect because of variations
in infiltration rates and evaporation (Wieder and Yaalon, 1985). Similar influences
have been implicated in the formation of some gypcretes (Bureau and Roederer,
1961) and differential rates of salt weathering (Goudie, 1986).
The intimate association between desert soils and landscape has attracted partic
ular attention in Australia and the North American arid zone. Markedly different
soils occur on landscape units of different age (Gilè, 1966, 1975; Wopfner and
TWidale, 1967). In part this is a response to variations in the maturity of the soils
but the different palaeoenvironmental conditions that the various soils have expe
rienced during the course of their evolution also has an effect. While Argids in
North America may be relicts from the wetter late Pleistocene, Gypsiorthids of
the Australian arid zone may have formed under more arid conditions when saline
sediments where deflated from pans (Jessup, 1960).
Certain desert landforms such as flatirons and mesa-and-butte formations are
often associated with landscapes encrusted with silcrete or calerete which are un
dergoing dissection. But these topographic features can form wherever there is an
erosion-resistant caprock or substrate. Other common landforms such as desert
248 A. Watson
pavements, desert gilgai and pseudoanticlines, and perhaps desert varnishes are
more intimately linked with pedological processes.
The formation of gravelly desert pavements (Figure 10.4) has been attributed to
several processes: deflation of fine particles leaving a lag of coarse material on the
surface, upward movement of gravel-sized fragments through the solum as a result
of volumetric changes in the host material caused by wetting and drying (Jessup,
1960; Cooke, 1970), or eluviation of fine-grained atmospheric deposits and their
illuvial accretion beneath the gravelly surface (McFadden et al., 1987). No single
process can account for all desert pavements and some are undoubtedly formed by
a combination of these mechanisms.
Desert gilgai are soil hummocks resembling those commonly developed in Verti-
sols but which may occur in Aridisols — though the presence of gilgai is one
criterion for differentiating the two orders (Soil Survey Staff, 1975). The gilgai re
ported from Australia by Oilier (1966) are associated with soils rich in swelling clays
(montmorillonite-group minerals) but illuvial gypsic horizons are also present. So,
the hummocky surfaces could be pseudoanticlines produced by displacive gypsum
accretion rather than true vertisolic gilgai.
Desert varnish is a thin coating of iron and manganese-rich clay found on rock
surfaces and pebbles in some arid regions (Dorn and Oberlander, 1982). Made up
of microlaminae from 0.25 to 20 μιη thick (Perry and Adams, 1978; Dorn, 1984),
the coatings only rarely exceed a total thickness of 0.50 mm. Though desert varnish
usually forms above the soil surface and, therefore, is not a pedogenic phenomenon,
its origin often seems to be intimately linked with soil biota (Scheffer et al., 1963).
Numerous origins have been proposed. These range from chemical or organic rock
weathering phenomena (Engel and Sharp, 1958) tofixationof atmospheric deposits
on the rock surfaces (Potter and Rossman, 1977; Allen, 1978). The latter model
has recently gained considerable support from studies of microorganic fixation of
manganese by algae and fungi (Krumbein and Jens, 1981) or bacteria (Dorn and
Oberlander, 1981, 1982). If the model is valid, mineralogical variations between
microlaminae may reflect fluctuations in the chemistry of atmospheric fallout re
sulting from significant environmental changes (Dorn, 1984, 1986). Hence, though
desert varnish plays an almost insignificant role in the evolution of desert land
scapes, it may have a great potential in the field of palaeoenvironmental research.
Some varnishes appear to be of great age, so they may be able to provide palaeoen
vironmental data for the past several hundred thousand years. More investigations
of varnishes from beyond the southwest United States, as well as studies of the re
gional significance of environmentalfluctuationsdeduced from individual varnishes
are essential (Watson, 1989).
The most significant geomorphic impact that desert pedogenesis has is probably
through its influence on the erodibility of the landsurface. While soil crusts can
preserve ancient landscapes by retarding erosion, other properties of certain desert
soils can exacerbate erosion. Eluviation of clay from surface horizons can lead to
Desert soils 249
Fig. 10.7. Badlands developed in colluvial materials in Swaziland, southern Africa. The colluvium typ
ically has a very high exchangeable sodium content which renders it extremely susceptible to erosion
once the protective soil cover is disrupted. It is likely that the clay mineralogy (particularly a high smec
tite content) and the soil chemistry of the material are inherited from palaeosols which formed as the
colluvium accumulated under arid conditions during the late Pleistocene (between about 30,000 and
10,000 years before present).
rapid deflation of sandy topsoil if the protective vegetation cover is disturbed. In

other cases, saline and sodic soils may be susceptible to accelerated erosion by
runoff owing to enhanced dispersion of clays in the presence of sodium. High ex
changeable sodium levels can have a marked effect on the dispersiveness of weakly
cohesive soils and other surficial materials which are aggregated by clays (Stocking,
1979; Watson et al., 1984,1987). The presence of smectites appears to have a strong
influence on sodicity and clay dispersion because sodium ions bonded weakly in the
clay structure are readily released upon wetting (Goldberg et al., 1988). Severe soil
erosion and gullying may occur in areas where sodic desert palaeosols come under
the influence of wetter conditions because of climatic change (Figure 10.7).
Desert palaeosols
Ancient and relict soils and soil crusts are often used as evidence of climatic
change in the arid zone and adjacent semiarid regions (Figure 10.8). However, two
fundamental difficulties have confounded many of these palaeoenvironmental in
vestigations. First, the precise environmental bounds within which different types of
pedogenesis occur are often poorly understood. Second, the potential for accurate
dating of many palaeosols is very limited. It has been widely assumed that pedogen
esis in deserts occurs mainly during slightly moister interludes, while greater aridity
generally promotes eolian erosion (Jessup and Norris, 1971 — in South Australia;
250 A. Watson
Fig. 10.8. Palaeosols exposed by gully erosion in Swaziland, southern Africa. The lighter material form
ing the gully wall is grey-coloured, sodic colluvium which contains calcareous palaeosols dating from
the late Pleistocene. The darker material on the gully floor is an older colluvial deposit which has been
reddened and depleted in clay as a result of pedogenesis, probably under warm, humid conditions.
This occurred prior to the onset of the arid conditions which saw the deposition of the overlying grey
colluvium.
Wieneke and Rust, 1973, 1975 — in the central Namib Desert). However, if the
mode of formation of the soils is unknown, such interpretations are at best over-
simplistic aad at worst erroneous. In the Namib Desert, for example, Gypsiorthids
of both vadose and phreatic origin are widespread (Watson, 1985, 1988); while the
formation of illuvial crusts may be retarded under hyperarid conditions, the devel
opment of hydromorphic crusts could be enhanced. Hence, the different forms of
Gypsiorthids can have very diverse palaeoenvironmental implications.
Considerable advances have been made in the dating of calcareous palaeosols
using 14C (Williams and Polach, 1971; Magaritz et al., 1981) and ^ T h - ^ U (Ku et
al., 1979; Schlesinger, 1985) techniques. However, other types of desert soil are less
amenable to radiometrie dating. One major potential pitfall exists because many
of the diagnostic soil horizons are of illuvial origin. So, the dating of material in
corporated within them will misrepresent the true age of the soil if the inclusion
is considerably older. The same applies to human artefacts found within illuvial
soil horizons. The age of some soils can be estimated from the time required to
accumulate the quantity of introduced material currently found in the illuvial hori
zon based on present rates of deposition at the surface and the leaching potential
(Goudie, 1973). However, the accuracy of such calculations is dependant on the pre
cise measurement of actual net accretion rates (depositional inputs minus periodic
losses owing to leaching beyond the solum). Moreover, since for the purposes of
the calculations it is generally assumed that accretion rates have been constant, the
Desert soils 251
significance of any palaeoenvironmental variations is lost. This particular problem

has confounded attempts to estimate the age of pedogenic argillic horizons in the
southwest United States using this approach because in most cases the horizons do
not seem to be forming at present (Gilè and Grossman, 1968). Nevertheless, using
such procedures it has been estimated that the gypsum soils of the Namib Desert
are at least 100,000 years old (Watson, 1988) and similar soils in Sinai (Dan et al.,
1982) and Wyoming (Reheis, 1987) may be several hundred thousand years old. Be
cause such soils are highly susceptible to dissolution if the climate becomes wetter
(Yaalon and Ward, 1982), their great antiquity has profound palaeoenvironmental
implications.
Summary
The characteristic features of most desert soils can be attributed primarily to

aridity. The accumulation of water-soluble salts such as calcite, gypsum, and halite
within the solum is a result of ineffective leaching because of low rainfall and some
times high evaporation rates. The lack of moisture also severely constrains chemical
weathering, so the clay fraction of many desert soils is dominated by minerals
commonly associated with juvenile weathering profiles in more humid settings. In
addition to these direct influences of aridity on pedogenesis, there are numerous
secondary factors. The soluble salts and clay-sized particles which accumulate as il-
luvial horizons within soil profiles are often derived from distinctly desertic sources.
Atmospheric fallout of dust deflated from evaporative basins often distributes sol
uble minerals over the surrounding landscapes. Similarly, much fine-grained clastic
material, desert loess, may originate in saline depressions as the product of mechan
ical salt weathering. In more restricted settings, atmospheric salts may be deposited
on the landsurface in saline fog moisture.
Within the Earth's arid zones, well developed soils (those which conform to the
Aridisol order of the USDA Classification or are Solonchaks, Xerosols or Yer-
mosols of the FAO Classification) are not always widespread. Entisols, which usually
correspond to Lithosols, Regosols and Fluvisols of the FAO Classification, often
predominate. Indeed, materials such as active dune sands and indurated salt crusts,
which can be extensive features of the landscape, should be classified as non-soils
because under natural conditions, they are unable to support plant growth. Many of
the true soils are cumulative: their development is, or has been, dominated by the
introduction of water-soluble salts or fine-grained clastic material through leaching
of surficial deposits and subsequent illuviation in the solum. In some areas, desert
pavements may be formed by this same process. The formation of desert varnish
may also be closely linked with atmospheric dust deposition.
Desert soils, particularly siliceous duripans and petrocalcic horizons, can be per
sistant features in arid and semiarid landscapes. They have also been widely re-
252 A. Watson
ported from the rock record. However, most other desert soils are very vulnerable
to degradation through the dissolution of minerals or translocation of clays when
the climate ameliorates. Some common sodic soils such as the Natrargids of the
USDA Classification and Solonetz soils of the FAO Classification, may be former
saline soils which have been modified by leaching under moister climatic conditions.
Indeed, desert palaeosols are probably widespread in many semiarid regions espe
cially in much of Africa and Australia where during the late Pleistocene the arid
zones appear to have been far more extensive than today.
Desert soils and their associated geomorphological phenomena can provide valu
able palaeoenvironmental information in regions where other sources of data such
as archaeological and palynological evidence are scarce. However, some characteris
tic desert soils such as pedogenic silcretes, calcretes and gypsum crusts, for example,
have a variety of forms which develop under markedly different moisture conditions
ranging from vadose to phreatic and lacustrine. The various forms of each soil type
may be associated with very different physical environments. Hence, they must be
used with circumspection in palaeoenvironmental reconstructions. The use of iso
topie data to help determine the origin of certain desert soils (Sofer, 1978; Vieille-
fon, 1980) offers great potential for future studies. Similarly, once the regional
significance of desert varnish has been ascertained, radiometrie dating (Knauss and
Ku, 1980; Dorn et al, 1986) and isotopie studies (Dorn and DeNiro, 1984) may
prove an extremely powerful tool for palaeoenvironmental research.
This brief examination of desert soils has hardly touched upon their agricultural
potential. While saline soils are toxic to most plants and sodic soils generally have
poor moisture regimes and low chemical fertility (Northcote, 1983), desert soils
are being increasingly exploited for agriculture. Juvenile soils, especially Tbrriflu-
vents (Fluvisols of the FAO Classification), which have not experienced pronounced
salinization, are widely cultivated wherever the availability of irrigation water and
markets provide an economic incentive (for example Bashour et al., 1983). Though
these soils may be deficient in certain trace elements, the major hindrances to their
cultivation are salinization and toxification, and depletion of fossil water resources.
The growing need to expand and intensify agricultural production, particularly in
Asia and Africa, will necessitate still further exploitation of soils in the arid zone. Tb
date, the implementation of many agricultural aid projects in the Third World has
had little positive impact on local economies. The mismanagement of soils leading
to irreversible salinization or catastrophic erosion has been a recurrent problem.
In the future, the success of such development schemes will depend upon their
ability to meet long-term economic goals without detriment to the environment.
Tb achieve this, a thorough understanding of the inherent instabilty as well as the
potential productivity of arid-zone soils and palaeosols is crucial.
Desert soils 253
References
Allen, B.L· and Fanning, D.S., 1983. Composition and soil genesis. In: L.P. Wilding, N.E. Smeck and
G.F. Hall (Editors), Pedogenesis and Soil Taxonomy, I: Concepts and Interactions. Developments in
Soil Science, 11A Elsevier, Amsterdam, pp. 141-192.
Allen, C.C., 1978. Desert varnish of the Sonoran Desert — optical and electron probe microanalysis. J.
Geol., 86: 743-752.
Arakel, AV., 1980. Genesis and diagenesis of Holocene evaporitic sediments in Hutt and Leeman
Lagoons, Western Australia. J. Sediment. Petrol., 50: 1305-1326.
Arakel, A.V. and McConchie, D., 1982. Classification and genesis of calerete and gypsite lithofacies
in paleodrainage systems of inland Australia and their relationship to carnotite mineralization. J.
Sediment. Petrol., 52: 1149-1170.
Bashour, 1.1., Al-Mashhady, AS., Devi Prasad, J., Miller, T. and Mazroa, M., 1983. Morphology and
composition of some soils under cultivation in Saudi Arabia. Geoderma, 29: 327-340.
Besler, H., 1976. Wasserüberform te Dünen als Glied in der Landschaftsgenese der Namib. Mitt. Basi.
Afr. Bibliogr., 15: 83-106.
Bespalov, N.D., 1964. Soils of Outer Mongolia. Israel Program for Scientific Translations, Jerusalem,
320 pp.
Bigham, J.M., Jaynes, WF. and Allen, B.L., 1980. Pedogenic degradation of sepiolite and palygorskite
on the Texas High Plains. J. Soil Sei. Soc. Am., 44: 159-167.
Blank, M., Leinen, M. and Prospero, J.M., 1985. Major Asian eolian inputs indicated by the mineralogy
of aerosols and sediments in the western North Pacific. Nature, 314: 84-86.
Blount, D.N. and Moore, C.H., 1969. Depositional and non-depositional carbonate breccias, Chiantla
Quadrangle, Guatemala. Bull. Geol. Soc. Am., 80: 429-441.
Bockheim, J.G., 1980. Properties and classification of some desert soils in coarse-textured glacial drift
in the Arctic and Antarctic. Geoderma, 24: 45-69.
Boss, G., 1941. Niederschlagsmenge und Salzgehalt des Nebelwassers an der Küste Deutsch Südwest
afrikas. Bioklim. Beibl., 8: 1-15.
Boyadgiev, T., 1982. Classification of arid soils. Soviet Soil Sci., 3: 60-67.
Braithwaite, C.J.R., 1989. Displacive calcite and grain breakage in sandstones. J. Sediment. Petrol., 59:
258-266.
Brimhall, G.H., Alpers, C.N. and Cunningham, A.B., 1985. Analysis of supergene ore-forming processes
and ground-water solute transport using mass balance principles. Econ. Geol., 80: 1227-1256.
Brimhall, G.H. and Dietrich, WE., 1987. Constitutive mass balance relations between chemical com
position, volume, density, porosity, and strain in metasomatic hydrochemical systems: results on
weathering and pedogenesis. Geochim. Cosmochim. Acta, 51: 567-587.
Brimhall, G.H., Lewis, C.J., Ague, J.J., Dietrich, W.E., Hampel, J., Teague, T. and Rix, P., 1988. Metal
enrichment in bauxites by deposition of chemically mature eolian dust. Nature, 333: 819-824.
Brinkman, R., 1982. Clay transformation: aspects of equilibrium and kinetics. In: G.H. Bolt (Editor),
Soil Chemistry, B. Physico-chemical Models. Elsevier, Amsterdam, pp. 433-458.
Briot, P., 1983. Lenvironnement hydrogéochimique du calerete uranifière de Yeelirrie (Australie Occi
dentale). Miner. Deposita, 18: 191-206.
Budnikoff, P.P., 1928. Die Beschleunigen und Verzögerer der Abbindegeschwindigkeit des Stuckgipses.
Kolloidzeitschrift, 46: 242-249.
Buoi, S.W., 1965. Present soil-forming factors and processes in arid and semi-arid regions. Soil Sci., 99:
45^9.
Bureau, P. and Roederer, P., 1961. Contribution à l'étude des sols gypseux du Sud-tunisien: croûtes et
encroûtements gypseux de la partie Sud du Golfe de Gabès. Bull. Assoc. Fr. Etude Sol, 150-176.
254 A. Watson
Butzer, K.W. and Hansen, C.L., 1968. Desert and River in Nubia: Geomorphology and Prehistoric
Environments at the Aswan Reservoir. University of Wisconsin Press, Madison, Wise, 562 pp.
Carlisle, D., Merifield, P.M., Orme, A.R. and Kolker, O., 1978. The Distribution of Calcretes and
Gypcretes in Southwestern United States and their Uranium Favorability. Based on a Study of
Deposits in Western Australia and South West Africa (Namibia). University of California, Los
Angeles, Calif., Open File Report 76-022-E, 274 pp.
Cody, R.D., 1979. Lenticular gypsum: occurrences in nature, and experimental determinations of effects
of soluble green plant material on its formation. J. Sediment. Petrol., 49:1015-1028.
Cody, R.D. and Cody, A.M., 1988. Gypsum nucleation and crystal morphology in analog saline terres
trial environments. J. Sediment. Petrol., 58: 247-255.
Conacher, A.J., 1975. Throughflow as a mechanism responsible for excessive soil salinisation in non-
irrigated, previously arable lands in the Western Australian Wheatbelt: a field study. Catena, 2: 3 1 -
68.
Cooke, R.U., 1970. Stone pavements in deserts. Ann. Assoc. Am. Geogr., 60: 560-577.
Cooke, R.U., 1981. Salt weathering in deserts. Proc. Geol. Assoc, 92: 1-16.
Cooke, R.U. and Smalley, I.J., 1968. Salt weathering in deserts. Nature (London), 220: 1226-1227.
Coque, R., 1955. Les croûtes gypseuses du Sud-tunisien. Bull. Soc. Sci. Nat. Tlinis., 8: 217-236.
Coque, R., 1962. La Tunisie Présaharienne: Étude Géomorphologique. Armand Colin, Paris, 476 pp.
Dan, J., Yaalon, D.H., Moshe, R. and Nissim, S., 1982. Evolution of reg soils in southern Israel and
Sinai. Geoderma, 28: 173-202.
Dorn, R.I., 1984. Cause and implications of rock varnish microchemical laminations. Nature, 310: 767-
770.
Dorn, R.I., 1986. Rock varnish as an indicator of eolian environmental change. In: W.G. Nickling
(Editor), Eolian Geomorphology. Allen and Unwin, Winchester, Mass., pp. 291-307.
Dorn, R.I., Bamforth, D.B., Canili, T.A., Dohrenwend, J.C., Tlirrin, B.D., Donahue, DJ., Juli, A.J.T.,
Long, A., Macko, M.E., Weil, E.B., Whitley, D.H. and Zabel, T.H., 1986. Cation-ratio and accelerator
radiocarbon dating of rock varnish on Mojave artifacts and landforms. Science, N.Y., 231: 830-833.
Dorn, R.I. and DeNiro, M.J., 1984. Stable carbon isotope ratios of rock varnish organic matter: a new
paleoenvironmental indicator. Science, N.Y., 227: 1472-1474.
Dorn, R.I. and Oberlander, T.M., 1981. Microbial origin of desert varnish. Science, N.Y., 213: 1245-
1247.
Dorn, R.I. and Oberlander, T.M., 1982. Rock varnish. Prog. Phys. Geogr., 6: 317-367.
Dougherty, M.T. and Barsotti, N.J., 1972. Structural damage and potentially expansive sulphide miner
als. Bull. Assoc. Eng. Geol., 9: 831-837.
Dregne, H.E., 1976. Soils of Arid Regions. Developments in Soil Science, 6. Elsevier, Amsterdam, 788
pp.
Drever, J.F. and Smith, C.L., 1978. Cyclic wetting and drying of the soil zone as an influence on the
chemistry of groundwater in arid terrains. Am. J. Sci., 278: 1448-1454.
Edinger, S.E., 1973. The growth of gypsum. An investigation of the factors which affect the size and
growth rates of the habit faces of gypsum. J. Cryst. Growth, 18: 217-224.
Eisenberg, J., Dan, J. and Koyumdjisky, H., 1982. Relationships between moisture penetration and
salinity in soils of the northern Negev (Israel). Geoderma, 28: 313-344.
Engel, CG. and Sharp, R.P., 1958. Chemical data on desert varnish. Bull. Geol. Soc. Am., 69: 487-518.
Ericksen, G.E., 1981. Geology and origin of Chilean nitrate deposits. U.S. Geol. Surv., Prof. Pap., 1188:
1-37.
Eswaran, H., Stoops, G. and Abtahi, A., 1980. SEM morphologies of halite. J. Microsc, 120: 343-352.
Eugster, H.P., 1986. Lake Magadi, Kenya: a model for rift valley hydrochemistry and sedimentation? In:
L.E. Frostick, R.W Renaut, I. Reid and J.J. Tiercelin (Editors), Sedimentation in the African Rifts.
Blackwell Scientific, Oxford, pp. 177-189.
Desert soils 255
Evstifeev, Y.G., 1980. Extra-arid Gobi soils. Probi. Desert Dev., 1980 (2): 17-26.
Fitzpatrick, E.A., 1980. Soils: Their Formation, Classification and Distribution. Longman, London, 353
pp.
Folk, R.L., 1976. Reddening of desert sands: Simpson Desert, N.T., Australia. J. Sediment. Petrol., 46:
604-615.
Folkoff, M.E. and Meentemeyer, V, 1987. Climatic control of the geography of clay mineral genesis.
Ann. Assoc. Am. Geogr., 77: 635-650.
Gilè, L.H., 1966. Cambic and certain non-cambic horizons in desert soils of southern New Mexico. Proc.
Soil Sci. Soc. Am., 30: 773-781.
Gilè, L.H., 1975. Holocene soils and soil-geomorphic relations in an arid region of southern New
Mexico. Quat. Res., 5: 321-360.
Gilè, L.H., 1977. Holocene soils and soil-geomorphic relations in a semiarid region of southern New
Mexico. Quat. Res., 7: 112-132.
Gilè, L.H. and Grossman, R.B., 1968. Morphology of the argillic horizon in desert soils of southern
New Mexico. Soil Sci., 106: 6-15.
Goldberg, S., Suarez, D.L. and Glaubig, R.A., 1988. Factors affecting clay dispersion and aggregate
stability of arid-zone soils. Soil Sci., 146: 317-325.
Goudie, A.S., 1973. Duricrusts in Tropical and Subtropical Landscapes. Clarendon, Oxford, 174 pp.
Goudie, A.S., 1984. Duricrusts and land forms. In: K.S. Richards, R.R. Arnett and S. Ellis (Editors),
Geomorphology and Soils. Allen and Unwin, London, pp. 37-57.
Goudie, A.S., 1986. Laboratory simulation of "the wick effect" in salt weathering of rock. Earth Surf.
Process., 11: 275-286.
Goudie, A.S., 1989. Weathering processes. In: D.S.G. Thomas (Editor), Arid Zone Geomorphology.
Belhaven, London, pp. 11-24.
Goudie, A.S., Cooke, R.U. and Doornkamp, J.C., 1979. The formation of silt from quartz dune sand by
salt-weathering processes in deserts. J. Arid Environ., 2: 105-112.
Goudie, A.S. and Watson, A., 1984. Rock block monitoring of rapid salt weathering in southern Tlinisia.
Earth Surf. Process., 9: 95-99.
Grossman, R.B., 1983. Entisols. In: L.P. Wilding, N.E. Smeck and G.F. Hall (Editors), Pedogenesis
and Soil Taxonomy, II: the Soil Orders. Developments in Soil Science, 11B, Elsevier, Amsterdam,
Oxford, New York, Tokyo, pp. 55-90.
Gumuzzio, J., Batile, J. and Casas, J., 1982. Mineralogical composition of salt efflorescences in a Typic
Salorthid, Spain. Geoderma, 28: 39-51.
Harvey, A.M., 1989. The occurrence and role of arid zone alluvial fans. In: D.S.G. Thomas (Editor),
Arid Zone Geomorphology. Belhaven, London, pp. 136-158.
Henke, K.F., 1976. Magnitude and rate of heave in tunnels in calcium sulphate bearing rocks. Bull. Int.
Assoc. Eng. Geol., 13: 61-64.
Hodge, T. Tbrchenek, L.W and Oades, J.M., 1984. Occurrence of palygorskite in ground-water rendzi-
nas (Petrocalcic calciaquolls ) in south-east South Australia. In: A. Singer and E. Galan (Editors),
Palygorskite-Sepiolite: Occurrences, Genesis and Uses. Developments in Sedimentology, 37. Else-
vier, Amsterdam, pp. 199-210.
Horta, J.C. de O.S., 1980. Calerete, gyperete and soil classification in Algeria. Eng. Geol., 15:15-52.
Hunt, C.B., Robinson, T.W., Bowles, WB. and Washburn, A.L., 1966. Hydrological basin, Death Valley,
California. U.S. Geol. Surv., Prof. Pap. , 494Β, 138 pp.
Isbell, R.F., Reeve, R. and Hutton, J.T, 1983. Salt and sodicity. In: Division of Soils, CSIRO, Soils: An
Australian Viewpoint. CSIRO, Melbourne/Academic Press, London, pp. 107-117.
Jessup, R.W, 1960. An introduction to the soils of the south-eastern portion of the Australian arid zone.
J. Soil Sci., 11: 92-105.
256 A. Watson
Jessup, R.W. and Norris, R.M., 1971. Cainozoic stratigraphy of the Lake Eyre basin and part of the arid
region lying to the south. J. Geol. Soc. Aust., 18: 303-331.
Kahle, CF., 1977. Origin of subaerial Holocene calcareous crust: role of algae, fungi and sparmicritisa-
tion. Sedimentology, 24: 413-435.
Kocurek, G. and Hunter, R.E., 1986. Origin of polygonal fractures in sand, uppermost Navajo and Page
Sandstones, Page, Arizona. J. Sediment. Petrol., 56: 895-904.
Knauss, K.G. and Ku, T.L., 1980. Desert varnish: potential for age dating via uranium-series isotopes. J.
Geol., 88: 95-100.
Krumbein, WE. and Jens, K., 1981. Biogenic rock varnishes of the Negev Desert (Israel) an ecological
study of iron and manganese transformation by cyanobacteria and fungi. Oecologia, 50: 25-38.
Krumkachev, L.I., 1984. Genetic properties of some soil groups of the Republic of Mali. Soviet Soil Sci.,
16: 24-33.
Ku, T.L., Bull, W.B., Freeman, S.T. and Knauss, K.G., 1979. Th 2 3 0 -U 2 3 4 dating of pedogenic carbonates
in gravelly desert soils of Vidal Valley, southeastern California. Geol. Soc. Am. Bull., 90:1063-1073.
Kushnir, J., 1980. The coprecipitation of strontium, magnesium, sodium, potassium and chloride ions
with gypsum. An experimental study. Geochim. Cosmochim. Acta, 44: 1471-1482.
Lobova, E.V., 1967. Soils of the Desert Zone of the USSR. Israel Program for Scientific Translations,
Jerusalem, 405 pp.
Lobova, E.V., 1968. Classification of desert soils in the sub-boreal belt. In: V.A. Kovda and E.V. Lobova
(Editors), Geography and Classification of Soils of Asia. Israel Program for Scientific Translations,
Jerusalem, pp. 1-33.
Lobova, E.V., 1981. Some characteristics of soil formation in arid zones of the world. Probi. Desert
Dev., 1981 (1): 19-22.
Lof, P., 1987. Soils of the World. Elsevier, Amsterdam, 1 p.
Lowenstein, T.K and Hardie, L.A., 1985. Criteria for the recognition of salt-pan evaporites. Sedimen
tology, 32: 627-644.
McFadden, L.D., Wells, S.G. and Jercinovich, M.J., 1987. Influences of eolian and pedogenic processes
on the origin and evolution of desert pavements. Geology, 15: 504-508.
McLean, S.A., Allen, B.L. and Craig, J.R., 1972. The occurrence of sepiolite and attapulgite on the
southern High Plains. Clays Clay Miner., 20: 143-149.
Magaritz, M., Kaufman, A. and Yaalon, D.H., 1981. Calcium carbonate nodules in soils: 1 8 0 / 1 6 0 and
13
C/ 12 C ratios and 14 C contents. Geoderma, 25:157-172.
Malanin, A.N., 1987. Desert sandy soils of southern Kazakhstan. Probi. Desert Dev., pp. 1-5.
Middleton, N.J., 1989. Desert dust. In: D.S.G. Thomas (Editor), Arid Zone Geomorphology. Belhaven,
London, pp. 262-283.
Milnes, A.R., 1986. Armoured landscapes. Geol. Today, 2: 73-74.
Moore, A.W, Isbell, R.F. and Northcote, K.H., 1983. Classification of Australian soils. In: CSIRO —
Division of Soils, Soils — an Australian Viewpoint. Academic Press, London, pp. 253-266.
Morales, C , 1979. Saharan Dust. Wiley, Chichester, 279 pp.
Mortensen, H., 1933. Der "Salzsprengung" und ihre Bedeutung für die regionklimatische Gliederung
der Wüsten. Petermanns Mitt., 79: 130-135.
Nettleton, W.D., Flach, K.W. and Brasher, B.R., 1969. Argillic horizons without clayskins. Proc. Soil Sci.
Soc. Am., 33: 121-125.
Nettleton, W.D., Nelson, R.E. and Flach, K.W, 1973. Formation of mica in surface horizons of dryland
soils. Proc. Soil Sci. Soc. Am., 37: 473-478.
Nettleton, WD. and Peterson, F.F., 1983. Aridisols. In: L.P. Wilding, N.E. Smeck and G.F Hall (Editors),
Pedogenesis and Soil Taxonomy, II: the Soil Orders. Developments in Soil Science, HB. Elsevier,
Amsterdam, pp. 165-215.
Desert soils 257
Nettleton, W.D., Witty, J.E., Nelson, R.E. and Hawley, J.W, 1975. Genesis of argillic horizons in soils
of desert areas of the southwestern United States. Proc. Soil Sei. Soc. Am., 39: 919-926.
Northcote, K.H., 1979. A Factual Key for the Recognition of Australian Soils. Rellim Tech. Pubi.,
Glenside, S.A., 124 pp.
Northcote, K.H., 1983. Central southern region. In: CSIRO — Division of Soils, Soils — an Australian
Viewpoint. Academic Press, London, pp. 211-217.
Northcote, K.H. and Wright, M.J., 1983. Sandy desert region. In: CSIRO — Division of Soils, Soils
— an Australian Viewpoint. Academic Press, London, pp. 173-178.
Nyamapfene, K.W., 1986. Some relationships between topography and sodic soils in Zimbabwe. Z.
Geomorph., N.F., 30: 246-252.
Oberlander, T.M., 1989. Slope and pediment systems. In: D.S.G. Thomas (Editor), Arid Zone Geomor-
phology. Belhaven, London, pp. 56-84.
Oilier, CD., 1966. Desert gilgai. Nature (London), 212: 581-583.
Orlova, M.A., 1980. Role of the eolian factor in the salt regime of desert solonchaks. Probi. Desert
Dev., pp. 73-77.
Page, W.D., 1972. The Geological Setting of the Archaeological Site at Oued el Akarit and the Paleo-
climatic Significance of Gypsum Soils, Southern Tunisia. Ph.D. Thesis, Dep. Geol. Sci., University of
Colorado, Boulder, Colo., I l l pp. (unpublished).
Perry, R.S. and Adams, J.B., 1978. Desert varnish: evidence for cyclic deposition of manganese. Nature
(London), 276: 489-491.
Peterson, F.F., 1980. Holocene desert soil formation under sodium salt influence in a playa margin
environment. Quat. Res., 13: 172-186.
Potter, R.M. and Rossman, G.R., 1977. Desert varnish: the importance of clay minerals. Science, N.Y.,
196: 1446-1448.
Pouget, M., 1968. Contribution à l'étude des croûtes et encroûtements gypseux de nappe dans le Sud-
tunisien. Cah. ORSTOM, Sér. Pédol., 6: 309-365.
Prospero, J.M., 1981. Arid regions as sources of mineral aerosols in the marine atmosphere. Geol. Soc.
Am., Spec. Pap., 186: 71-86.
Prospero, J.M., Bonatti, E., Schubert, C. and Carlson, T.B., 1970. Dust in the Caribbean atmosphere
traced to an African dust storm. Earth Planet. Sci. Lett., 9: 287-293.
Pye, K., 1980. Beach salerete and eolian sand transport: evidence from north Queensland. J. Sediment.
Petrol., 50: 257-293.
Pye, K., 1983. Red beds. In: A.S. Goudie and K. Pye (Editors), Chemical Sediments and Geomorphol-
ogy. Academic Press, London, pp. 227-263.
Pye, K., 1987. Eolian Dust and Dust Deposits. Academic Press, London, 334 pp.
Pye, K. and Sperling, C.H.B., 1983. Experimental investigation of silt formation by static breakage
processes: the effect of temperature, moisture and salt on quartz dune sand and granitic regolith.
Sedimentology, 30: 49-62.
Reheis, M.C., 1987. Gypsic soils on the Kane alluvial fans, Big Horn County, 'Wyoming. U.S. Geol. Surv.
Bull., 1590C, 39 pp.
Reid, I. and Frostick, L.E., 1989. Channel form, flows and sediments in deserts. In: D.S.G. Thomas
(Editor), Arid Zone Geomorphology. Belhaven Press, London, pp. 117-135.
Risacher, F, 1978. Genèse d'une croûte de gypse dans un bassin de l'Altipiano bolivien. Cah. ORSTOM,
Sér. Géol., 10: 91-100.
Roth, E.S., 1965. Temperature and water content as factors in desert weathering. J. Geol., 73: 454-468.
Schattner, I., 1961. Weathering phenomena in the crystalline of the Sinai, in the light of current notions.
Bull. Res. Counc. Isr., 10G: 247-265.
Scheffer, F, Meyer, B. and Kalk, E., 1963. Biologische Ursachen der Wüstenlackbildung. Z. Geomorph.,
N.F., 7: 112-119.
258 A. Watson
Schlesinger, W.H., 1985. The formation of caliche in soils of the Mojave Desert, California. Geochim.
Shadfan, H. and Dixon, J.B., 1984. Occurrence of palygorskite in the soils and rocks of the Jordan
Valley. In: A. Singer and E. Galan (Editors), Palygorskite-Sepiolite: Occurrences, Genesis and
Uses. Developments in Sedimentology, 37. Elsevier, Amsterdam, pp. 187-198.
Shearman, D.J., 1963. Recent anhydrite, gypsum, dolomite and halite from the coastal flats of the
Arabian shore of the Persian Gulf. Proc. Geol. Soc, 1607: 63-64.
Shumakov, VS. and Mikhovich, A.I., 1960. The phenomenon of impulverization in the area of Elista.
Soviet Soil Sci., 1960: 781-782.
Singer, A., 1984. Pedogenic palygorskite in the arid environment. In: A. Singer and E. Galan (Edi
tors), Palygorskite-Sepiolite: Occurrences, Genesis and Uses. Developments in Sedimentology, 37.
Elsevier, Amsterdam, pp. 169-175.
Singer, A., 1989. Illite in the hot-aridic soil environment. Soil Sci., 116: 146-155.
Sofer, Z., 1978. Isotopie composition of hydration water in gypsum. Geochim. Cosmochim. Acta, 42:
1141-1149.
Soil Survey Staff, 1975. Soil Taxonomy. USDA, Washington, D.C., 574 pp.
Sperling, C.H.B. and Cooke, R.U., 1985. Laboratory simulations of rock weathering by salt crystalliza
tion and hydration processes in hot, arid environments. Earth Surf. Process., 10: 541-555.
Stace, H.C.T., Hubble, G.D., Brewer, R., Northcote, K.H., Sleeman, J.R., Mulcahy, M.J. and Hallsworth,
E.G., 1968. A Handbook of Australian Soils. Rellim Tech. Pubi., Glenside, S.A., 435 pp.
Stainier, X., 1912. Les tufs gypseux et calcaires du Bas-Sahara. Bull. Soc. Belge Géol. Paléontol. Hydrol.,
26: 90-120.
Stocking, M.A., 1979. Catena of sodium-rich soil in Rhodesia. J. Soil Sci., 30: 139-146.
Stoertz, G.E. and Ericksen, G.E., 1974. Geology of salars in northern Chile. U.S. Geol. Surv., Prof. Pap.
, 811, 65 pp.
Summerfield, M.A., 1983a. Petrology and diagenesis of silcrete from the Kalahari Basin and Cape
coastal zone, southern Africa. J. Sediment. Petrol., 53: 895-909.
Summerfield, M.A., 1983b. Silcrete as a palaeoclimatic indicator: evidence from southern Afrcia. Palaeo-
geogr. Palaeoclimatol. Palaeoecol., 41: 65-79.
Swett, K. 1974. Calerete crusts in an Arctic permafrost environment. Am. J. Sci., 274: 1059-1063.
Thomas, D.S.G., 1989a. The nature of arid environments. In: D.S.G. Thomas (Editor), Arid Zone
Geomorphology. Belhaven Press, London, pp.1-8.
Thomas, D.S.G., 1989b. Eolian sand deposits. In: D.S.G. Thomas (Editor), Arid Zone Geomorphology.
Belhaven Press, London, pp. 233-261.
Tricart, J., 1978. Le sol dans l'environnement écologique. Rev. Géomorph. Dyn., 27: 113-128.
Tucker, M.E., 1978. Gypsum crusts (gypcrete) and patterned ground from northern Iraq. Z. Geomorph.,
N.F., 22: 89-100.
Vieillefon, J., 1980. Approche de l'Intensité de la Dynamique Actuelle des Accumulations Gypseuses
dans les Sols au Moyen Dosage de la Teneur en Tritium de l'Eau de Constitution du Gypse.
République Tunisienne, Ministère de l'Agriculture, Direction des Ressources en Eau et en Sol
— Division des Sols/ORSTOM, Mission Tunisie, E-S 181, 22 pp.
Vogg, R., 1981. Bodenressourcen arider Gebiete. Stuttg. Geogr. Stud., 97, 224 pp.
Walker, T.R., 1979. Red colour in dune sand. In: E.D. McKee (Editor), A Study of Global Sand Seas.
U.S. Geol. Surv., Prof. Pap., 1052: 61-81.
Warren, J.K. and Kendall, C.G.St.C, 1985. Comparison of sequences formed in marine sabkhas (sub-
aerial) and salina (subaqueous) settings — modern and ancient. Bull. Am. Assoc. Pet. Geol., 69:
1013-1023.
Watson, A., 1983. Evaporite sedimentation in non-marine environments. In: A.S. Goudie and K. Pye
(Editors), Chemical Sediments and Geomorphology. Academic Press, London, pp. 163-185.
Desert soils 259
Watson, A., 1985. Structure, chemistry and origins of gypsum crusts in southern Tbnisia and the central
Namib Desert. Sedimentology, 32: 855-875.
Watson, A., 1988. Desert gypsum crusts as palaeoenvironmental indicators: a micropetrographic study
of crusts from southern Tbnisia and the central Namib Desert. J. Arid Environ., 15: 19-42.
Watson, A., 1989. Desert crusts and rock varnish. In: D.S.G. Thomas (Editor), Arid Zone Geomorphol-
ogy. Belhaven Press, London, pp. 25-55.
Watson, A., Price Williams, D. and Goudie, A.S., 1984. The palaeoenvironmental interpretation of col-
luvial sediments and palaeosols of the late Pleistocene hypothermal in southern Africa. Palaeogeogr.,
Palaeoclimatol., Palaeoecol., 45: 225-249.
Watson, A., Price Williams, D. and Goudie, A.S., 1987. Reply to "Is gullying associated with highly sodic
colluvium? Further comment to the environmental interpretation of southern African dongas."
Palaeogeogr., Palaeoclimatol., Palaeoecol., 58: 123-128.
Watts, N.L., 1977. Pseudo-anticlines and other structures in some calcretes of Botswana and South
Africa. Earth Surf. Process., 2: 63-74.
Watts, N.L., 1980. Quaternary pedogenic calcretes from the Kalahari (southern Africa): mineralogy,
genesis and diagenesis. Sedimentology, 27: 661-686.
Whitford, WG. and Freckman, D.W., 1988. The role of soil biota in soil processes in the Chihuahuan
Desert. In: E.E. Whitehead, CF. Hutchinson, B.N. Timmerman and R.G. Varady (Editors), Arid
Lands Today and Tomorrow. Westview Press, Boulder, Colo., pp. 1063-1073.
Wieder, M. and Yaalon, D.H., 1985. Catenary soil differentiation on opposite-facing slopes related to
erosion-deposition and restricted leaching processes, northern Negev, Israel. J. Arid Environ., 9:
119-136.
Wieneke, E and Rust, U., 1973. Klimamorphologische Phasen in der zentralen Namib (Südwestafrika).
Mitt. Geogr. Ges. München, 58: 79-96.
Wieneke, F. and Rust, U., 1975. Zur relativen und absoluten Geochronologie und Reliefentwicklung an
der Küste der mittleren Südwestafrika. Eiszeitalter Ggw., 26: 241-250.
Wigley, T.M.L., 1973. Chemical evolution of the system gypsum-calcite-water. Can. J. Earth Sci., 10:
306-315.
Williams, G.E. and Polach, H.A., 1971. Radiocarbon dating of arid-zone calcareous palaeosols. Geol.
Soc. Am. Bull., 82: 3069-3086.
Wilding, L.P., Smeck, N.E. and Hall, G.F, 1983. Pedogenesis and Soil Taxonomy, II: the Soil Orders.
Developments in Soil Science, HB. Elsevier, Amsterdam, 410 pp.
Winkler, E.M. and Singer, P.C., 1972. Crystallization pressure of salts in stone and concrete. Bull. Geol.
Soc. Am., 83: 3509-3513.
Wopfner, H. and Twidale, C.R., 1967. Geomorphological history of the Lake Eyre Basin. In: J.N.
Jennings and J.A. Mabbutt (Editors), Landform Studies from Australia and New Guinea. Cambridge
University Press, Cambridge, pp. 119-143.
Wright, V.P., North, C.P., Hancock, PL., Curtis, M. and Robinson, D., 1988. Pedofacies variations across
an arid alluvial basin: a case-study from the Upper Triassic of SW Britain. Int. Assoc. Sedimentol.,
Eur. Meet., Leuven, pp. 227-228.
Yaalon, D.H., 1964a. Downward movement and distribution of ions in soil profiles with limited wetting.
In: E.G. Hallsworth and D.V Crawford (Editors), Experimental Pedology. Butterworth, London, pp.
159-164.
Yaalon, D.H. 1964b. Airborne salts as an active agent in pedogenic processes. Trans. 8th Int. Congr.
Soil Sci., Bucharest, 5: 997-1000.
Yaalon, D.H., 1971. Factors controlling the deposition and distribution of airborne salts over the land-
surface. Bull. Am. Metall. Soc., 52: p. 1136.
Yaalon, D.H. and Ganor, E., 1973. The influence of dust on soils during the Quaternary. Soil Sci., 116:
146-155.
260 A. Watson
Yaalon, D.H. and Ward, J.D., 1982. Observations on calerete and recent calcic horizons in relation to
land forms, central Namib Desert. Palaeoecol. Afr. Surround. Isl., 15: 183-186.
Yaalon, D.H. and Wieder, M., 1976. Pedogenic palygorskite in some arid brown (Calciorthid) soils in
Israel. Clay Miner., 11: 73-80.
Yousaf, M., Ali, O.M. and Rhoades, J.D., 1987. Clay dispersion and hydraulic conductivity of some
salt-affected arid land soils. J. Soil Sei. Soc. Am., 51: 905-907.
261
Chapter 11
Quaternary soil chronosequences in various

environments — extremely arid to humid tropical
P.W BIRKELAND
Introduction
Jenny (1941,1980) defined the factors of soil formation: climate, organisms (veg
etation), parent material, topography and time. He reasoned that if one studies
soils on landscapes or deposits of different age in a particular area and keeps the
influence of the first four factors reasonably constant, the variation of the soils is a
function of time. Common examples are the buildup of organic carbon or clay with
time. Comparison of soils within an area constitutes a chronosequence, and if there
is a mathematical relationship, by use of regression analysis, these are chronofunc-
tions. This review will include both chronosequences and chronofunctions.
Vreeken (1975) categorizes chronosequences into four different types. The most
common type is a post-incisive sequence, such as a stepped river-terrace sequence
where each soil in the sequence began to form when a particular terrace formed, and
older soils are associated with higher (older) terraces. It is assumed that younger
soils progress to the older soils in the sequence, so a plot of a soil property vs. time
tracks soil development.
The most common chronosequences studied are post-incisive ones (Jenny, 1941,
1980; Stevens and Walker, 1970; Bockeim, 1980; Birkeland, 1984). The main prob
lem with such chronosequences is proving that each older soil went through the
pedogenic pathway indicated by properties of the younger soils (Vreeken, 1975).
For this to be strictly true, the soils in the sequence have to be monogenetic that
is, soil-forming factors must have remained essentially constant through time. This
has hardly been true as most soils are old enough to have experienced variation in
the soil-forming factors, and in some places the changes, such as climatic variation
during glacial-interglacial cycles, were quite dramatic (Wright, 1983). Hence, soils
that are older than early Holocene probably are polygenetic products, that is the
soil-forming factors have varied with time. Not always are such variations in climate
(with resulting vegetation changes) easily read from the soil record, however.
262 P.W Birkeland
TABLE 11.1
Soil horizon nomenclature

A organic matter accumulation; vesicular surface horizon in deserts (Av)
Bh pedogenic accumulation of humus in the subsurface, usually with Al and/or Fe
Bw sufficient alteration to produce rubification (reddening due to pedogenic Fe, but does
not meet criteria for spodic horizon
Bs pedogenic Fe accumulation that probably meets spodic horizon criteria
Bt pedogenic clay accumulation (argillic horizon)
Bk pedogenic CaCC>3 accumulation; stages I and II of Gilè et al. (1981)
By pedogenic gypsum accumulation
Bz pedogenic accumulation of salts more soluble than gypsum, usually halite
K major accumulation of pedogenic CaCÜ3; stages III—VI of Gilè et al. (1981) and
Machette (1985)
m used to denote strong cementation by salt
Source: Birkeland (1984).
Post-incisive chronosequences will be the focus of this chapter. Climatic changes

have influenced some soil properties to varying degrees in some chronosequences,
and these will be discussed where appropriate. For each area discussed, the present
mean annual precipitation (MAP) and temperature (MAT) are given. Dating de
posits within each chronosequence is a major problem, as numerical dates are few.
Most ages are estimates based on chemical, biological, geomorphic and correlation
dating methods (Coiman et al., 1987).
Many of the examples used here are for parent materials that are coarse textured,
such as gravelly and sandyfluvialand marine deposits and gravelly sandy tills; colour
is usually 10YR hue or more yellow. Coarse-textured parent materials have the
advantage in that one can be fairly certain that many accumulations of silt and clay
in the B horizons are pedogenic and not parent material layering.
Three general soil-forming processes are important to understanding the devel
opment of chronosequences (Simonson, 1978). Common additions from the atmo
sphere are solids (silt, clay and carbonates), and ions (such as Ca2+, Na + , SO^)
dissolved in rainwater. Common transformations are the formation of clay-size par
ticles by weathering of larger grains, the alteration of clay minerals to other clay-
mineral species, the release of iron by weathering, and the weathering of apatite,
which shows up as a loss of phosphorus LPa (acid-extracted phosphorus; Walker and
Syers, 1975). Finally, transfers within the profile result in accumulation of B-horizon
silt and clay, Fe (dithionite extract: Fed; oxalate extract: Feo; see Parfitt and Childs,
1988), Al, CaC03, gypsum and halite.
The emphasis here is mainly on progressive development of soils. Johnson and
Watson-Stegner (1987) discuss the many processes that reverse the direction of soil
development, and it is important to keep these in mind in explaining trends in some
chronosequences.
Quaternary soil chronosequences in various environments 263
A powerful tool in the analysis of soil development is to numerically express

profile development on the basis of field properties. The profile development index
(PDI) of Harden (Harden, 1982; Harden and Taylor, 1983) is used because soil
properties are compared to parent material or assumed parent material properties.
Points are assigned for the differences between soil and parent material values,
summed to give profile values and normalized so that development is rated between
0 and 1, with 1 being maximum development. There commonly is a close correlation
between PDI and soil laboratory data.
The organization of this analysis is along a bioclimatic gradient from limited ef
fective soil moisture (extremely arid) to excess moisture (humid). A heavy emphasis
will be on U.S. Geological Survey reports generated from a project on chronose
quences. Soil horizon nomenclature used is given in Table 11.1.
Extremely arid to semi-arid soils with salt accumulations
In areas of limited precipitation, soluble salts accumulate in the soil, in addition

to silt and clay, and the amounts, kinds and depths of salt accumulations vary with
climate. An example of such a climatic transect is southern Israel where the kind
of salt and the accumulation depth vary with MAP (Figure 11.1; Dan and Yaalon,
1982). Above about 50 cm MAP the soils are generally free of salts. Soils forming in
areas of less than 50 cm MAP accumulate CaCC>3 at depth, and the depth to the top
of the Bk horizon is related to the amount of precipitation. Salts more soluble than
carbonate accumulate in soils with less than about 30 cm MAP; these salts usually
have a maximum concentration beneath the Bk horizon, and the zone of maximum
concentration approaches the surface in progressively drier climates. Finally, below
about 10 cm MAP, both By and Bz horizons occur at shallow depths, and CaC0 3
accumulation is minor.
Rapid soil development takes place in the extremely arid environment of south
ern Israel (Dan et al., 1982). Much of the pedogenic salt, silt and clay is considered
to be of atmospheric origin (Gerson et al, 1985; Gerson and Amit, 1987). An exam-
MEAN ANNUAL PRECIPITATION, CM

^0 10 20 30 40 50
c
DEPTH, CM
O
o
200
\
Fig. 11.1. Relation between kind of pedogenic salt, depth to the top of the horizon of salt accumulation,
and precipitation in southern Israel (from Dan and Yaalon, 1982).
264 P.W Birkeland
CLAY, %
<lu GYPSUM,%
4
ψ. RECENT
ο 50 50-
I4KA
50J
YOUNGER
;:/ 14 KA
50J 5CM
Fig. 11.2. Data for a soil chronosequence in southern Israel (from Amit and Gerson, 1986). a. Variation
in coarse and fine clay with depth, for soils of two ages. b. Variation in gypsum with depth for a young
soil and the oldest soil in the sequence.
pie of such pedogenesis is that for aflightof river terraces that range from recent to
an estimated 14 ka (thousands of years) (MAP 3 cm, MAT 2°C; Amit and Gerson,
1986; Gerson and Amit, 1987). The soils start with less than 5% of both coarse (1-2
/im) and fine (<1 μπι) clay, and in about 14 ka coarse clay content approaches 10%
and fine clay 15%;finesilt accumulation increases from about 1% in the parent ma
terial to a maximum of 25% near the surface in the same time span (Figure 11.2).
Gypsum and halite also accumulate in these soils, and gypsum reaches a concen
tration of about 8% in the oldest soils (Figure 11.2). Salt weathering has shattered
many stones in most of the salt accumulation horizons.
Pre-Holocene soils in southern Israel have still greater accumulations of silt and
gypsum (Dan et al., 1982; Gerson and Amit, 1987). For soils with estimated ages
of several hundred ka, maximum silt content is 43% near the surface and 24%
at depth in the B horizon, and the maximum concentration of gypsum is 43%
in a Bym (petrogypsic) horizon. Bt horizons commonly are non-gravelly in spite
of the presence of both overlying well-developed gravel pavement and underlying
gravelly By and Bz horizons. Quite likely the desert pavement clasts remain at the
surface as the Bt horizon thickens due to the infiltration of atmospheric silt and clay
(McFadden et al., 1987). In places, non-gravelly Bt horizons form in an estimated 50
ka (Gerson and Amit, 1987).
The rates of accumulation of silt, clay and salts can change with time for various
reasons (Gerson and Amit, 1987). Some are extrinsic and include factors in the
source area such as wind direction and degree of protective vegetation cover. Other
reasons are site specific and include the efficiency of the surface to trap dust, with
higher efficiencies associated with vegetation and/or a rough gravelly surface. In
time, as smooth desert pavement forms from an initially rough surface, the efficiency
of the surface as a dust trap lessens. Yet other site factors are the formation of a
surface crust and the plugging of near-surface voids with silt and clay; both of
these diminish the capacity of the soil to vertically transmit water and thereby
carry solids and salts to depth in soils. Also important is climatic change, as one
toward more effective soil moisture could move the more soluble salts to still greater
depths in-the profile, or out of the soil entirely. Hence, high concentrations of
pedogenic gypsum and halite close to the surface must argue for long-term extreme
aridity.
Although gypsic soils are not common in the United States of America (U.S.A),
Reheis (1987a) has described a chronosequence of gypsic soils in north-central
Wyoming (MAP 6.5 cm; MAT 7.2°C). The gypsum, silt and clay are shown to be
of atmospheric origin. The accumulation of pedogenic gypsum with time is about 9
g gypsum cm"2 for 250 cm thick soils per 100 ka. Percentages of gypsum are less
than 0.2% in 5 ka soils and exceed 50 for individual horizons, some of which are
petrogypsic, in about 410 ka. Accumulation of atmospheric gypsum in dust traps
placed in the field is sufficient to account for this rate of pedogenic gypsum. Reheis
warned, however, that not all trapped gypsum necessarily gets into the soils, as some
may be carried off by wind to other sites before a rainfall event moves it into the
soil.
The study of Reheis (1987a) brought attention to several processes in gypsic
soils that are important to predicted progressive pedological variations in chronose
quences. A common observation is that even though both CaC03 and gypsum are
available to accumulate, in extremely arid environments gypsum accumulation dom
inates. One explanation for this is that although both may be deposited at the
surface of the soil, the solubility differences are such that mainly gypsum goes into
solution and is moved into the soil; although some CaC03 also does, most of the
remaining CaCC>3 could be removed by wind erosion. Also shown to be important is
the common ion effect, with the higher Ca2+ concentrations associated with higher
gypsum contents markedly inhibiting CaCO^ solubility. Gypsum buildup also could
inhibit infiltration of clays of atmospheric origin by increasing Ca2+ concentrations
to levels where clay particles areflocculatedand their downward movement greatly
diminished. Such flocculated clays could be removed from the site by winds. Fi
nally, although present climate and climatic change are important in explaining the
position of gypsic horizon(s) in profiles, gypsum, being hygroscopic, can, if in high
enough concentrations, absorb some increases in effective soil moisture related to
a climatic change toward more moist conditions; the result could be that the By
horizon is not shifted to greater depths.
The coastal desert of Peru also is in an extremely arid climate (MAP 1 cm; MAT
18°C), but chronosequences along the coast differ from the trends in Israel. El Nino
rains are frequent along the northern coast, but not along the southern coast, and
this affects soil morphology (Noller et al., 1987). The southern coast has gypsic soil
chronosequences somewhat similar to those in southern Israel. In contrast, the part
of the northern coast impacted by El Nino rains (4-40 cm per event; usually within 6
months) is characterized by soils with deep (>1 m) translocation of atmospheric silt
266 P.W. Birkeland
and clay, and salt accumulations are much less and at greater depth than expected
from the mean annual climatic data. Thus, soils can be used to estimate both the
duration of El Nino events and the length of coastline affected.
Much research on soils with carbonate accumulation has been undertaken in the
western United States. The key work on these soils is that of Gilè et al. (1981)
in southern New Mexico, and the regional study of Machette (1985). In general,
pedogenic carbonate accumulates in soils where MAP is less than about 50 cm.
Considerations of the amount of weathering that is required to release sufficient
Ca 2+ to produce the carbonate accumulations are several orders of magnitude short
of the amounts of carbonate present in the soils, so an atmospheric origin is invoked
and substantiated by dust-trap and rainfall chemistry data. Substantial silt and clay
in the soils also are of atmospheric origin.
Detailed soil-geomorphic studies in many areas have been used to develop a
pattern of carbonate stage buildup with time (Figure 11.3; Machette, 1985). Of
the areas studied, the slowest long-term rate of change in morphological stage is
for a cool, aridic mountain valley in Colorado, and the most rapid rate is for the
warm, semiarid area of Roswell-Carlsbad, New Mexico. Long-term rates of mass
accumulation of pedogenic carbonate were calculated from soils whose ages can be
estimated, and there is a close correlation between the rate of stage buildup and the
rate of carbonate accumulation (Figure 11.3).
Machette (1985) proposed a model of carbonate accumulation that takes into
account late Quaternary climatic change, and suggests that interglacial rates could
perhaps be two times that for glacial times for some areas. The long-term rates,
however, average out these shorter variations. Two broad groups of soils are defined
based on the relative amounts of moisture and Ca 2+ influx: influx-limited soils and
moisture-limited soils (Figure 11.4). The line separating the two groups is where
there is a balance between Ca2+ supply and moisture, with all of the available Ca 2+
exceeding the capacity of the soil to accumulate carbonate, so the potential rate
of accumulation is not realized. In contrast, the influx-limited field is where there
is sufficient moisture to accumulate Ca 2+ in the soil, but some of the carbonate
can be leached from the soil during either relatively long moist intervals or spring
snowmelt events. Finally, if excess moisture is sufficiently high, the soils remain
carbonate-free, no matter what the Ca 2+ influx.
The model of Machette (1985) can be used to suggest changes in the rates of
CaC0 3 accumulation with climatic change, here denoted as wetter (glacial) and
drier (interglacial; Figure 11.4). In some soils the carbonate accumulates only in
the interglacial climates and is partly removed in the glacial climates. Other soils
stay within either the influx-limited or moisture-limited fields with climate change,
and the rate of CaC03 accumulation increases in the former and decreases in the
latter during interglacial climates. McFadden and Tinsley (1985) have made good
progress in modelling both CaCOß accumulation in semiarid soils and the influence
of climate change on the depth of carbonate accumulation; the artificial soil has
Quaternary soil chronosequences in varìous environments 267
carbonate accumulations that mimic soils in the field that have undergone climate
change.
Bt horizons seem to form rapidly in the aridic moisture regime in the U.S.A
In the well-studied Las Cruces, New Mexico area (MAP 20 cm; MAT 16°C), Bt
horizons 20 cm thick with 5YR hue and 12% clay have formed in gravelly alluvium
in the late Holocene; Bt horizons require a longer time to form in non-gravelly
materials (Gilè et al., 1981). Irrespective of parent material, the clay seems to be of
atmospheric origin. The most rapid development of a Bt horizon, however, seems
to have been attained in the Lubbock, Texas area, where a minimal Bt has formed
in 450 years (Holliday, 1988). Here rapid development is ascribed to "mud rains", a
common occurrence in the area.
0.01-
S5N&
STAGE
i._^SJAGEII_.^;
0.1-
CD
O 0.2-
STAGE m I'·'· · · . · . · · . * · ' \
X
(Π 0 5- ·:/"Λ.--^
QL
<l
1.0- · · ■ · · · ' · . · /. ·
Ci · · ; - . · *.
>- |· STAGE ET ')-■-■-/ r-, ^ i . . . .
2.0-
<
O
5 0- \]L:'j STAGE ΤΓ fi-. :.sTAGE3r -:\ :
r 0.22 - /:·. :
0.26
·' ·' ' · · .· '
0.51
· .·"
0.1
■/ \*
0.14
: · "*
RATE OF C a C 0 3 ACCUMULATION , G / C M 2 / 1 0 0 0 YR
Fig. 11.3. Estimated time required to attain various maximum stages of carbonate morphology at se
lected localities in the western United States, and estimated rate of pedogenic carbonate accumulation
(after Machette, 1985; McFadden, 1982).
Geographic area Annual climatic means Weather Climate

Precipitation (cm) Temperature (°C) ^ ' type
(1) Upper valley of 24.6 6.6 Buena Vista Cool arid

Arkansas River, Colo.
(2) Colorado Piedmont, 37.6-47.2 9.7-11.0 Denver, Boulder Temperate
Colo. semi-arid
(3) Albuquerque, 20.5 13.1 Albuquerque Warm
N. Mexico semi-arid
(4) San Acacia, 21.2 14.7 Socorro Warm
N. Mexico semi-arid
(5) Las Cruces, 20.4 15.5 N.M. State Univ. Warm
N. Mexico semi-arid
(6) Roswell-Carlsbad, 35.5-32.8 15.3-17.2 Roswell-Carlsbad Warm
N. Mexico semi-arid
(7) Vidal Junction, Calif. 11.5 22.1 Parker, Ariz. Hot arid
and Overton, Nev.
(8) Beaver, Utah 30.0 8.4 Beaver Temperate
semi-arid
268 EW Birkeland
CONTINUAL LOSS
PEDALFERS
LOSS
ce
D
K
o
O t/î
-0.5 —\
■04-
03-
Z
5
LOW MODERATE HIGH
Ca++ INFLUX
EXPLANATION
1. Buena Vista, CO 5. Las Cruces, NM

2. Boulder-Denver, CO 6. Roswell-Carlsbad, NM
3. Albuquerque, NM 7. VidalJunction, CA
4. San Acacia, NM 8. Beaver, UT
Fig. 11.4. Plot of noncalcic (pedalfer) and calcic (pedocal) soils along moisture and Ca 2+ flux gradients.
Calcic soils are further subdivided into influx-limited and moisture-limited. Horizontal and vertical lines
are rates of CaCÜ3 accumulation (g c m - 2 per 1 ka). Arrows depict the estimated change in rate of
accumulation from the relatively wet to the relatively dry climates (point of arrow) of the glacial-
interglacial cycles (from Machette, 1985). Numbered localities, given in the explanation, match those in
Figure 11.3
Soil in temperate regions (<100 cm MAP)
At rainfalls higher than that at which CaC0 3 accumulates, soil chronosequences

have a morphology dominated by the pedogenic accumulation of clay and Fe. Here
we will discuss mainly chronosequences in the mountains and basins of the western
U.S.A.; the mountains are relatively moist and eflfective soil moisture decreases with
decreasing altitude. Climate and availability of CaCC>3 in dust are important in
determining whether or not the soils are calcic in these areas.
High elevation soils on tills
Holocene chronosequences have been reported in cirques in many alpine areas.

In general for the western U.S.A., A horizons have reached maximum development
and profiles are well differentiated A/B/Cox or A/Bt/Cox profiles in the areas of
higher rainfall after 10 ka of soil formation (Mahaney, 1978; Shroba and Birkeland,
1983; Birkeland et al., 1987). In the Front Range, Colorado, latest Holocene till
have 7 to 10% clay and 0.5% Fed and, after about 10 ka of pedogenesis, these
values increase to maximum values of 14% clay and 1% Fed. If one compares
various alpine and arctic areas worldwide, Feo accumulation and P depletion are
quite rapid in the wetter alpine soils, especially New Zealand, and in places the
accumulation and depletion are more rapid than in soils adjacent to the mountains
at lower elevation (Figure 11.5; Birkeland et al., 1990; Shroba and Birkeland, 1983).
Podzolized soils with A/E/Bs/Cox profiles, some of which could be Spodosols,
form rapidly in some relatively wet alpine and sub-alpine environments under both
tundra and boreal forest vegetation (Ugolini, 1986). In southern Norway (MAP 225
cm; MAT 5°C), for example, after 230 years of soil formation, E horizons several
APPROXIMATE AGE, YEARS

10
—i 1 1 1 — 1 —i
10,000 -Pa 2
100 r Feo N
N r^ N N 2
NN 3 tf ,
X
Ld
X a N N2 K
LU N
Q
1,000
N w
[ N w
B
N WSB B
o N KWS S < - K
, N B _l
Ld
_l
w Έ S
Q_
ÜJ
ID
o w
Q KW o
\WSB,
<
KWB B 1 NWyVSBK/ WBW B^ B
<J0
Fig. 11.5. Variation in Feo accumulation index and Pa depletion index with time for several alpine areas
and one arctic area (from Birkeland et al., 1990). The areas are: N = New Zealand; K = Khumbu
Glacier area, Nepal; W = Wyoming; S = California; B = Baffin Island (arctic). Letters are underlined
and not underlined; this denotes samples from different localities within that general area. The number
indicates the number of data points that plot at that position. Accumulation index is the %-cm of Feo
greater than parent material values, and depletion index is the %-cm of Pa less than parent material
values. To make the Feo calculation, for each horizon subtract the parent material % from the present
%, multiply the latter by horizon thickness, and sum for the profile. For Pa, the horizon % is subtracted
from the parent material %, and the rest of the calculation is the same as above.
270 P.W Birkeland
centimetres thick have formed. B horizons extend to about 15 cm depth, and the
ratio of Fed in the B horizon to that in the E horizon is about 1.5, signifying
substantial translocation of Fe (Mellor, 1985). Similar data are obtained at slightly
drier and colder sites near the Arctic Circle in Norway (Ellis, 1980). Thie Spodosols
are better developed than the above soils, and are common throughout Scandinavia
and northeastern North America on till deposited in the late Pleistocene and early
Holocene.
Soils on tills at mountain fronts

Soils formed from tills on moraine crests have been studied at numerous areas
in the western U.S.A., from the Sierra Nevada on the west (MAP 40-45 cm, MAT
7-8°C) to the Rocky Mountains on the east (MAP 40-90 cm, MAT - 1 to 6°C).
Clay and Fe accumulation, as well as rubification (reddening), increase at slow
rates across the region. In general, soils on 20-ka old tills have A/Cox and A/Bw/
Cox profiles, and some have Bt horizons (Shroba and Birkeland, 1983). Colors com
monly are no redder than lOYR, but some are 7.5YR; clay contents are commonly
between 3 and 10%, and in some places less, and Fe commonly decreases with depth
and can be as high as 0.5% at the surface. Soils formed on the crests of moraines
about 140 ka old have, where best expressed, A/Bt/Cox profiles. B-horizon colour
hues can be as red as 7.5YR and some are 5YR. Clay in the Bt horizons ranges from
9 to 20%. Fed displays no trend with time in the Sierra Nevada, but in places in the
Rocky Mountains, Fed can be about twice as abundant in the older soils as in the
younger soils; the maximum amount of Fed is 0.6%.
In many places the crest soils of the older moraines are no better developed than
the crest soils of the younger moraines, leading one to suspect that the older ones
have been eroded. Tb evaluate that, catena studies have been conducted (Swanson,
1985; Berry, 1987; Birkeland and Burke, 1988). Soil development usually is better
expressed in the downslope soils in the catenas, especially in those formed on the
older moraines. For example, the catena weighted mean PDI values (method in
TABLE 11.2
Weighted mean soil profile development index values for catenas formed on moraines of two ages,
western U.S.A.
Locality Catena weighted mean PDI values Reference
20 ka moraine 140 ka moraine
Central Idaho 0.18 0.28 Berry (1987)
Western Wyoming 0.14 0.25 Swanson (1985)
Eastern California 0.18 0.37 Birkeland and
0.16 0.21, 0.14 Burke (1988)
Eastern Nevada 0.16 0.29 Birkeland (unpub. data)
Birkeland and Burke, 1988) are 1.6 to 2.0 times greater for the older moraine
soils than for the youngest ones (Table 11.2). These data suggest that when one is
estimating moraine ages by relying solely on moraine-crest soils, the result might be
an underestimation of age.
SoiL· onrìverand manne deposits

Soils in a river-terrace chronosequence in the Great Valley of California (MAP
30 cm; MAT 16°C) range in age from 0.2 ka to 3 Ma (Harden, 1987). The youngest
soil with a Bt horizon is 40 ka, the Bt is 68 cm thick, has lOYR hue, and clay and Fed
maxima of 17 and 0.5%, respectively. In contrast, after 3 Ma of soil development,
the Bt horizon is over 7.5 m thick and has lOYR hue, 63% clay and 3.4% Fed.
McFadden and colleagues (McFadden, 1982, 1988; McFadden and Hendricks,
1985; McFadden and Weldon, 1987) have studied soil chronosequences across
southern California from aridic to xeric moisture regimes. If we compare noncalcic
soils in the wetter part of the transect (MAP 40-78 cm; MAT 16°C), relative to those
in the Great Valley, mentioned above, Bt horizons form more rapidly. After about
10-15 ka, Bt horizons are about 100 cm thick and have 7.5YR hue, 10% clay and
1.4% Fed. Bt horizons in soils between 300 and 700 ka can be as much as 5 m thick
and have 2.5YR to 10R hue, 39% clay and 2.9% Fed.
Sequences of marine terraces exist along the California coast, and one chronose
quence study on an island near San Diego spans from about 2.7 ka to over 1 Ma
[MAP 17 cm; MAT 16°C (Muhs, 1982)]. The youngest deposit is an alluvial fan and
has an 18 cm thick Bt horizon with high enough exchangeable Na to be a natric
horizon; maximum hue is 10YR, clay content 36% and Fed 0.4%. Soils less than
200 ka are Alfisols or Mollisols with Bt horizons, whereas some older soils have
sufficient clay and shrink-swell related properties to be Vertisols. The oldest soil has
a 77 cm thick Bt horizon that rests on bedrock, maximum clay content of 70%, Fed
of 1% and hue of 7.5YR; this is the only soil in the chronosequence that is not
10YR. Atmospheric additions (silt and clay) from continental California seem to
have been important in producing the surface horizon of all soils (Muhs, 1983) and
probably have greatly influenced the textural properties of subsurface horizons.
Tkking the southern California sites together, the rate of soil development is
most rapid in alluvial deposits on the coast near Ventura (MAP 36 cm; MAT 15°C;
Harden et al., 1986). The time of initial Bt horizon recognition is not known, but
after 40 ka of development the Bt is at least 43 cm thick, has a hue of 7.5YR and a
maximum of 34% clay. The oldest soils in the sequence are between 80 and 105 ka
and have a Bt horizon over 2.1 m thick, 5YR hue and a maximum of 28% clay.
Soil chronosequences along Rock Creek, southern Montana, change from non
calcic soils along the mountain front (MAP 64 cm; MAT 6°C) to calcic soils down
stream (MAP 37 cm; MAT 8°C; Reheis, 1987b). At the mountain front 33 cm thick
Bt horizons form in 20 ka, they are 7.5YR hue and 11% maximum clay. Soils in the
272 P.W Birkeland
drier downstream area also have Bt horizons by 20 ka, they are 47 cm thick, have
1UYR hue and the clay maximum is 32%. In contrast, the 2 Ma soil has a thinner
Bt (34 cm), lesser maximum clay (26%) and the same hue. This example shows that
progressive development of all properties is not always obtained.
Chronosequences in humid areas (>100 cm MAP)
Few chronosequences have been studied in moist, hot tropical areas. Millions
of years of soil formation seem to be required to form some of the thick Ox-
isols present. One chronosequence study for a river-terrace sequence in Colombia
describes thick (over 3 m), red (2.5YR), clay-rich (61% maximum) B horizons
with about 5-8% Fed for the older soils of the chronosequence (Alexander and
Holowaychuk, 1983); ages, however, are not estimated. Better age control comes
from soils formed on dated carbonate marine-terrace deposits on Barbados, West
Indies (MAP 100+ m; MAT 26+°C; Muhs et al., 1986). The soils are on limestone
and are shown to be formed from a combination of dust from Africa, ash from a
nearby volcano, and weathering of non-carbonate bedrock. The 70-cm thick soil of
the 125 ka terrace has a lOYR A horizon over a 2.5YR C horizon, is high in clay
(64-94%) and has 0.3-1.8% Fed. The next older soil, on a terrace about 180 ka, has
less clay and Fed, but has 5YR hue.
Soils formed from marine and fluvial deposits in Maryland and Virginia (MAP
112 cm; MAT 15°C) display good trends with time in a chronosequence that ranges
in age from about 30 ka to 1 Ma (Markewich et al., 1987). Although soils about 30
ka lack a Bt horizon, those between 60-125 ka have a Bt horizon 120 cm thick, with
7.5YR hue, 22% clay and 1.8% Fed. Soils estimated to be about 1 Ma have a Bt
over 232 cm thick, 2.5YR hue, 48% clay and 3.5% Fed.
Soil formation in well drained soils formed from barrier island sediments in
northern South Carolina (MAP 137 cm; MAT I7°C) follows a different pathway
from those mentioned above in Maryland and Virginia (Markewich et al., 1986).
The combination of sandy parent materials with little inherited clay, excessive
drainage and low percentage of weatherable minerals (parent sands are mainly
quartz), results in the development of Spodosols with properties that show little
relationship with age, even though the deposits compared are quite old: the younger
one is about 200 ka and the older one about 2.5-3.0 Ma. Horizons of clay ac
cumulation are all less than 70 cm thick, maximum clay is 16% and the reddest
hue (5YR) is for a Bh horizon. In contrast to the above well-drained sites, soils
formed in back-barrier deposits have greater clay contents and thicker B horizons
(Bt).
Along the subtropical east coast of Australia near Brisbane (MAP 120-170 cm;
MAT 25°C), podzols reach exceptional thicknesses (Thompson, 1983; Thompson
and Bowman, 1984). Six dune systems are recognized, and the E and underlying Bh
horizons progressively increase in thickness with dune age; the oldest soil has an E
horizon about 11 to over 20 m thick. Quartz sand parent material limits the amount
of clay that can form, but the oldest soils have up to 15% clay in the B horizon.
Dating is difficult, and one estimate for the oldest soil is 125 ka or older.
These data show the importance of parent material on the kind of soil that forms
in well-drained tropical areas. In quartzitic sands, giant podzols, many of which
are too thick to be classified as Spodosols, form, whereas in finer-grained materials
with a mix of weatherable minerals, thick, red soils form, that are enriched in some
combination of iron, aluminum and clay.
Pedogenic iron buildup is rapid on the heavily vegetated, wet and cool west
coast of South Island, New Zealand. Smith and Lee (1984) studied a Holocene soil
chronosequence formed from river deposits (MAP 640 cm; MAT 11°C). The oldest
soil is an estimated 1.5 ka, and the profile is A/E (9 cm)/Bsg (42 cm)/Cg. A thin iron
pan occurs at the top of the B horizon, and the bulk of the B horizon has a 10YR
hue and a maximum of about 1.3% Feo.
A river-terrace chronosequence that extends farther back in time is also present
on the west coast of the South Island (MAP 190 cm; MAT 11°C) and displays iron
buildup followed by iron depletion (Campbell, 1975; Ross et al., 1977). After about
18 ka the soil is an A/Bw/C profile with 10YR hue and a maximum of 2% Feo in the
58-cm thick Bw horizon. In marked contrast, the 130+ ka soil has an A/E/Bhs (with
iron pan) profile, the E horizon is gray (2.5-5Y or 5GY) and can be as much as 53
cm thick, and Feo is less than 0.1% in any horizon. If these soils indeed indicate a
genetic pathway, then iron accumulation is followed by iron depletion; it is thought
that a decreasing pH with time and gleying are partly responsible for the shift from
overall iron accumulation to depletion with the formation of a thin iron pan at
depth.
Comparative data on clay and iron buildup and PDI in soils in different
climates
There are some data available with which to compare clay and iron buildup in
soils in various environments. Using bulk density data, it is possible to calculate the
mass of clay or iron per cm2 of the soil (calculate value for each sampled horizon
and sum for depth of described profile). Much of the data used here have not been
corrected for parent material clay or iron; that is, the parent material amounts have
not been subtracted from the soils so that only pedogenic amounts are reported.
This distinction is important because pedogenic amounts will plot at the origin for
0 year; in contrast, if parent material amounts are not subtracted from the data set,
the estimated parent material amounts are plotted at 0 year. Because of this it is
difficult to know the shape of many accumulation curves for the first 10-20 ka of soil
formation.
274 P. W Birkeland
YEARS
Fig. 11.6. Clay accumulation with time in various environments. Group A includes San Clemente Island,
California (Muhs, 1982), Ventura, California (Harden et al., 1986), the San Joaquin Valley (Harden,
1987), the Cowlitz River area, Washington, Dethier, 1988) and Barbados (Muhs et al., 1986). Group
B includes the Rocky Mountains (Swanson, 1984, Colman and Pierce, 1986), the Rocky Creek area,
Montana (Reheis, 1987b), the arid and semi-arid areas of McFadden's (1982) southern California
transect, and the Sierra Nevada (Berry, 1990). The intermediate amounts include the eastern U.S.A.
coastal Plain (Markewich et al., 1987) and the xeric soils (not shown) of McFadden's (1982) transect.
Only data for Ventura, Cowlitz River area, Rocky Mountain data of Colman and Pierce (1986) and the
data of McFadden (1982) have been corrected for parent material amount of clay.
The clay accumulation data seem to fall mainly into two groups (Figure 11.6).
High amounts (group A) are characteristic of such diverse areas as the coast and
San Joaquin Valley of California, Cowlitz River area, Washington, and Barbados. In
contrast, low amounts (group B) are characteristic of the Sierra Nevada, the arid
and semi-arid soils of McFadden's (1982) transect, Rock Creek, Montana and the
Rocky Mountains. The coastal plain soils of the eastern U.S.A. and the xeric soils of
McFadden's (1982) transect have intermediate amounts, but the rate of increase in
the coastal plain soils is rapid enough to intersect group A at about 1 Ma. Factors
involved in the different amounts and rates are the amount of parent material clay,
influx of atmospheric clay and weathering of various primary minerals to clay; the
relative degree of these factors vary with the area and is not always addressed by the
workers.
The Fed accumulation data also plot in two groups (Figure 11.7). Greater ac
cumulation (group A) is present at Ventura, California, Cowlitz River area and
the xeric soils of McFadden's (1988) transect. Lower accumulations (group B) are
present in the San Joaquin Valley, the semi-arid and arid soils of McFadden's (1988)
transect and the Rocky Mountains (Swanson 1984, 1985). The soils of the coastal
plain of the eastern U.S.A. have intermediate values at about 100 to 200 ka, and
trend into group A at about 1 Ma. In contrast, the New Zealand data (note that it is
for Feo, not Fed) fall in group B for the younger soils, but then the curve reverses as
Fe is lost from the profile.
Quaternary soil chronosequences in varìous environments 275
IOr
o
oc
a.
•.NZ RAIN FOREST

Ό
^ 0.1
10° \0* 10° |0C
YEARS
Fig. 11.7. Fed accumulation with time in various environments. Group A includes soils at Ventura,
California, Cowlitz River area, Washington, and xeric soils of McFadden's (1988) southern California
transect. Group B includes soils of San Joaquin Valley, the arid and semi-arid soils of McFadden's
(1988) transect, the Rocky Mountains (Swanson, 1984, 1985) and Barbados (see Figure 11.6 for other
references). New Zealand data are for the west coast soils of the South Island and are for Feo (Camp
bell, 1976); Feo probably makes up a major proportion of the pedogenic iron in these soils. Only the
data of McFadden (1988) have been corrected for parent material values.
Three comments can be made about these data. One is that, in general, the soil
chronosequences with high clay accumulation also have high Fe accumulation; this
supports the findings of McFadden (1988,), only here the data are for a wider range
of environments. Secondly, one would intuitively expect the wetter environments
included here (these of eastern U.S.A., New Zealand) to have much higher amounts
of Fe, compared to the relatively dry soils of the western U.S.A., but this does not
appear to be the case. Third, grouping the soil data into envelope curves glosses
over the individual accumulation rates reported by the various workers. Trends from
most of the areas follow the outline of the envelope curves in a general way. PDI
increases in most areas with time. One would intuitively think that plots of PDI
versus time would be different in different climatic regimes. However, in one study
that included climates ranging from arid to humid, most of the data cluster together
(Figure 11.8; Harden and Taylor, 1983). The clustering was best when only the four
best field properties in each area were used. When eight properties were used, PDFs
of the aridic soil chronosequence plotted below those of the other areas.
Soil development, steady state and thresholds
Some of the early work on soil-geomorphic studies in Quaternary stratigraphy

seemed to indicate that soils reach a steady state with time. This implies that, after a
certain period of time, soil morphology and individual properties no longer change,
so that plots of soil properties vs. time have a 0 slope (Figure 11.6; Birkeland, 1984).
276 P.W Birkeland
2,80
co
<u
x> M
2,24 M
<r M 2C
2V2M C C
öfj V
C CO M,V MM
•H
CO
<υ
•H
1,68
2P C PP
3 4-J C C
^Ι M
M PC
O <υ 1,12 M PC
τ—1
α M
00 ο 3Μ
o μ
r-\ Cu MC
0.56
»\
M
Q
PM
10" 10' 10e
Years
Fig. 11.8. Plot of profile development index (PDI) against age for four chronosequences (Figure 11.5;
Harden and Taylor, 1983). The symbols are for the location of the particular soils: M = Merced, Cali
fornia (San Joaquin Valley); V = Ventura, California; C = Las Cruces, New Mexico; P = Pennsylvania.
The number refers to the number of data points that plot at that position.
Data for many of the chronosequences reviewed here, as well as that for the PDI
(Harden and Taylor, 1983), challenges that concept, as most properties that develop
slowly in soils, such as clay, Fe and CaC0 3 still increase in amounts in soils after
several 100's of thousands of years, or even 1 or more Ma (Markewich et al, 1987).
Perhaps only the more rapidly formed soil properties, such as organic carbon or
base saturation, reach a steady state.
The geomorphic concept of thresholds has been introduced to explain the be
haviour of some soil systems (Muhs, 1984; Birkeland, 1984). Muhs (1984) defines a
pedogenic threshold as a limit of stability of soil morphology that is exceeded either
by an intrinsic (within the soil) change in morphology, chemistry or mineralogy, or
by an extrinsic change, a change in one of the external soil-forming factors, however
subtle. An example of exceeding an intrinsic soil threshold is the San Clemente
Island example, where the buildup of clay in a soil first formed marked A/Bt/C
profiles, but when the clay content became so high that the soil turbated, the soil
converted to a Vertisol and the marked horizon contrast was lost. Another is the
New Zealand example where under the initial geochemical regime, Fe accumulated.
In time, however, the pH fell to a value where Fe was depleted (Campbell, 1975;
Ross et al., 1977).
The threshold concept has considerable application in soil chronosequence stud
ies. Most soil property versus time curves are smooth and either linear, log or power
functions (Bockheim, 1980). If thresholds occur in soils, accumulation curves could
initially rise as a smooth curve, but later rapidly steepen as a threshold is exceeded.
McFadden and Weldon (1987) report on such a possibility in southern California
where a soil property such as Fed might initially slowly increase in a linear fashion
as silt and clay are added to the soil as eolian dust. In time, the accumulation of
sufficient fines so alters the internal profile environment that the accumulation of
pedogenic Fed and clay are greatly accelerated. We need to recognize such thresh
olds, and the key to this is reliable numerical ages for terraces and soils close to
one another in age. In contrast, if soils are far apart in age, regression analysis to
produce chronofunctions will produce smooth curves and obscure the evidence of
thresholds.
Applications
TWo current applications of soil chronosequence work are in the use of soils to
estimate the ages of fault movements, and in studying the age and paleoenvironment
of ancient paleosols.
Estimating ages offault movements

Many studies in the western U.S.A. have used soils to help estimate past fault
movement. McCalpin (1982) and Berry (1990) used surface soils to estimate the
ages of moraines offset by normal faults along the fronts of large normal-fault-
bounded ranges. Buried soils formed in alluvium or colluvium along normal fault
scarps can be used to estimate the duration of quiescence between faulting events
(Swan et al., 1980; Nelson and Van Arsdale, 1986; Machette, 1988). In other places,
the lack of such buried soils suggests rather continuous fault movement (Berry,
1990). Long-term rates of strike-slip fault movement along the San Andreas fault
zone have been estimated using soil development to estimate ages of offset Quater
nary deposits (Harden and Matti, 1989). Finally soil-age estimates of river terraces
have been used to date both terrace deformation and the growth of anticlines
(Rockwell et al., 1984).
Using chronosequence data in studies ofpaleosols

Buried soils (paleosols) are reported to occur in sedimentary deposits of many
ages (Wright, 1987; Reinhardt and Sigleo, 1988). They are used to recognize un
conformities, estimate the duration of time represented by an unconformity and
estimate the paleoenvironment at the time of the unconformity. The recognition of
pre-Quaternary paleosols is not easy. Biological traces in rocks indicate that a land
surface existed and, therefore, some sort of a soil must have been present. Many
field pedological features commonly recognized in contemporary soils, however,
278 P.W Birkeland
may or may not be observed in paleosols. These include an organic-matter-enriched

A horizon, structure, vertical variation in coloration and clay, clay films, and car
bonate accumulation morphologies common in aridic soils. Many features that have
been ascribed to possible pedogenesis could also have formed through burial diage-
nesis. Sorting out the effects of these two processes is difficult, should be addressed
more in publications, and should be the focus of further research.
Assuming that a feature is a soil and possesses a morphology similar to con
temporary soils, an age can be tentatively assigned. For example, an age can be
estimated by comparison with clay accumulation curves (Figure 11.6) or carbonate
accumulation stages (Figure 11.3). Extraction of specific pedogenically significant
elements, such as Fed, might not work for comparison because pedogenic Fe could
be too altered following burial, to survive and allow for an unambiguous pedological
interpretation. Additionally, total chemical analysis may not give useful chronose-
quence trends, because trends in these data seem to require highly leaching environ
ments over a long duration of time (>105 year) for pedological trends to overprint
the parent material signal. I have been impressed with the difficulty of using such
data for age estimation. Therefore, prior to undertaking this type of analysis, I
suggest one should first review in detail the total chemistry data for soil chronose-
quences in a variety of environments in U.S. Geological Survey Bulletins 1589 and
1590.
The other common use of paleosols is to infer paleoenvironment. Of the papers
reviewed here, paleoenvironment could be broadly inferred from the kind of soluble
salt in the profile as well as its depth relation, and perhaps the depth of clay accu
mulation in the soils and clay mineralogy, as long as these features survive burial
diagenesis. Classification of paleosols by Soil Taxonomy has been proposed to aid in
paleoenvironmental reconstruction (Retallack, 1983), but it appears that too many
of the key classification criteria have been lost on burial diagenesis for this approach
to be used without other supporting criteria.
References
Alexander, E.B. and Holowaychuk, N., 1983. Soils on terraces along the Cauca River, Colombia: I.
Chronosequence characteristics. Soil Sci. Soc. Am. J., 47: 715-721.
Amit, R. and Gerson, R., 1986. The evolution of Holocene reg (gravelly) soils in deserts — An example
from the Dead Sea region. Catena, 13: 59-79.
Berry, M.E., 1987. Morphological and chemical characteristics of soil catenas on Pinedale and Bull Lake
moraine slopes in the Salmon River Mountains, Idaho. Quat. Res., 28: 210-225.
Berry, M.E., 1990. Late Quaternary glaciation and faulting along the eastern escarpment of the central
Sierra Nevada, California. Ph.D. Thesis, University of Colorado, Boulder, Colo.
Birkeland, P.W., 1984. Soils and Geomorphology. Oxford University Press, New York, N.Y., 372 pp.
Birkeland, P.W, Burke, R.M. and Shroba, R.R., 1987. Holocene alpine soils in gneissic cirque deposits,
Colorado Front Range. U.S. Geol. Surv. Bull., 1590-E, 21 pp.
Birkeland, P.W. and Burke, R.M., 1988. Soil catena chronosequences on eastern Sierra Nevada moraines,
California, U.S.A. Arct. Alp. Res., 20: 473-484.
Birkeland, P.W., Burke, R.M. and Benedict, J.B., 1990. Pedogenic gradients for iron and aluminum
accumulation and phosphorus depletion in arctic and alpine soils as a function of time and climate.
Quat. Res., 32:193-204.
Bockheim, J.C., 1980. Solution and use of chronofunctions in studying soil development. Geoderma, 24:
71-S5.
Campbell, A.S., 1975. Chemical and mineralogical properties of a sequence of terrace soils near
Reefton, New Zealand. Ph.D. thesis, Lincoln College, University of Canterbury, Canterbury, N.Z.,
447 pp.
Coiman, S.M., and Pierce, K.L., 1986. Glacial sequence near McCall, Idaho: Weathering rinds, soil
development, morphology and other relative-age criteria. Quat. Res., 25: 25-42.
Colman, S.M., Pierce, K.L. and Birkeland, P.W, 1987. Suggested terminology for Quaternary dating
methods. Quat. Res., 28: 314-319.
Dan, J., and Yaalon, D.H., 1982. Automorphic saline soils in Israel. Catena, Suppl., 1: 103-115.
Ellis, S., 1980. Physical and chemical characteristics of a podzolic soil formed in Neoglacial till, Okstin-
dan, northern Norway. Arct. Alp. Res., 12: 65-72.
Gerson, R., Amit, R. and Grossman, S., 1985. Dust availabilty in desert terrains, a study in the desert of
Israel and the Sinai. Insitute of Earth Sciences, Hebrew University of Jerusalem, Jerusalem.
Gerson, R. and Amit, R., 1987. Rates and modes of dust accretion and deposition in an arid region:
The Negev, Israel. In: L. Frostick and I. Reid (Editors), Desert Sediments: Ancient and Modern.
Geol. Soc. Spec. Pubi., 35: 157-169.
Gilè, L.H., Hawley, J.W. and Grossman, R.B., 1981. Soils and geomorphology in the Basin and Range
area of southern New Mexico — Guidebook to the Desert Project. New Mexico Bur. Min. Miner.
Res., Mem. 39, 222 pp.
Harden, J.W, 1982. A quantitative index of soil development from field descriptions: Examples from a
chronosequence in central California. Geoderma, 28: 1-28.
Harden, J.W, 1987. Soils developed in granitic alluvium near Merced, California. U.S. Geol. Surv. Bull.
1590-A, 65 pp.
Harden, J.W and Matti, J.C., 1989. Holocene and late Pleistocene slip rates on the San Andreas fault in
Yucaipa, California, using displaced alluvial-fan deposits and soil chronology. Geol. Soc. Am. Bull.,
101: 1107-1117.
Harden, J.W. and Taylor, E.M., 1983. A quantitative comparison of soil development in four climatic
regimes. Quat. Res., 20: 342-359.
Harden, J.W, Sarna-Wojcicki, A M . and Dembroff, G.R., 1986. Soils developed on coastal and fluvial
terraces near Ventura, California. U.S. Geol. Surv. Bull., 1590-B, 34 pp.
Holliday, V.T., 1988. Genesis of a late-Holocene soil chronosequence at the Lubbock Lake archaeologi
cal site, Texas. Ann. Assoc. Am. Geogr., 78: 594-610.
Jenny, H., 1941. Factors of Soil Formation. McGraw-Hill, New York, N.Y., 281 pp.
Jenny, H., 1980. The Soil Resource: Origin and Behaviour. Springer, New York, N.Y., 377 pp.
Johnson, D.L. and Watson-Stegner, D., 1987. Evolution model of Pedogenesis. Soil Sci., 143: 349-366.
Machette, M.N., 1985. Calcic soils of the Southwestern United States. In: D.L. Wiede (Editor), Soils
and Quaternary Geology of the Southwestern United States. Geol. Soc. Am., Spec. Pap., 203:
1-21.
Machette, M.N., 1988. Quaternary movement along the La Jencia fault, central New Mexico. U.S. Geol.
Prof. Pap., 144U, 82 pp.
Mahaney, W.C., 1978. Late-Quaternary stratigraphy and soils in the Wind River Mountains, western
Wyoming. In: WC. Mahaney (Editor), Quaternary Soils. Geo Abstracts, Norwich, pp. 223-264.
Markewich, H.W, Pavich, M.J., Mausbach, M.J., Stuckey, B.N., Johnson, R.G. and Gonzalez, V, 1986.
280 P.W Birkeland
Soil development and its relation to the ages of morphostratigraphic units in Horry County, South
Carolina. U.S. Geol. Surv. Bull. 1589-G, 61 pp.
Markewich, H.W., Pavich, M.J., Mausbach, M.J., Hall, R.L., Johnson, R.G. and Hearn, P.P., 1987. Age
relations between soils and geology in the coastal Plain of Maryland and Virginia. U.S. Geol. Surv.
Bull. 1589-A, 34 pp.
McCalpin, J.P., 1982. Quaternary geology and neotectonics of the west flank of the northern Sangre de
Cristo Mountains, south-central Colorado. Q. Colo. School Mines, 77, 97 pp.
McFadden, L.D., 1982. The impacts of temporal and spatial climatic changes on alluvial soils genesis in
southern California. Ph.D. Dissertation, Univ. Arizona, Tucson, Ariz., 430 pp.
McFadden, L.D., 1988. Climatic influences on rates and processes of soil development in Quaternary
deposits of southern California. Geol. Soc. Am., Spec. Pap. 216, pp. 153-177.
McFadden, L.D. and Hendricks, D.M., 1985. Changes in the content and composition of pedogenic iron
oxyhydroxides in a chronosequence of soils in southern California. Quat. Res., 23: 189-204.
McFadden, L.D. and Tinsley, J.C., 1985. Rate and depth of pedogenic-carbonate accumulation in soils:
Formulation and testing of a compartment model. Geol. Soc. Am., Spec. Pap., 203: 23-41.
McFadden, L.D. and Weldon, R.J., II, 1987. Rates and processes of soil development on Quaternary
terraces in Cajon Pass, California. Geol. Soc. Am. Bull., 98: 280-293.
McFadden, L.D., Wells, S.C. and Jercinovich, M.J., 1987. Influences of eolian and pedogenic processes
on the origin and evolution of desert pavements. Geology, 15: 504-508.
Mellor, A., 1985. Soil chronosequences on neoglacial moraine ridges, Jostedalsbreen and Jotunheimen,
southern Norway: a quantitative pedological approach. In: K.S. Richards, R.R. Arnett and S. Ellis
(Editors), Geomorphology and Soils. George Allen and Unwin, London, pp. 289-308.
Muhs, D.R., 1982. A soil chronosequence on Quaternary marine terraces, San Clemente Island, Cali
fornia. Geoderma, 28: 257-283.
Muhs, D.R., 1983. Airborne dust fall on the California Channel Islands, U.S.A. J. Arid Environ., 6:
223-238.
Muhs, D.R., 1984. Intrinsic thresholds in soil systems. Phys. Geogr., 5: 99-110.
Muhs, D.R., Crittenden, R.C. and Rowland, T.R., 1986. A soil chronosequence on Uranium-series dated
coral terraces, Barbados, West Indies (abstract). Am. Quat. Assoc, Progr. Abstr., pp. 156.
Nelson, A.R. and Van Arsdale, R.B., 1986. Recurrent late Quaternary movement on the Strawberry
normal fault, basin and Range-Colorado Plateau transition zone, Utah. Neotectonics, 1: 7-37.
Noller, J.S., Wells, L.E. and Birkeland, P.W., 1987. El Nino and soil development in northern coastal
Peru (abstract). Geol. Soc. Am., Abstr. Progr., 19: 790.
Parfitt, R.L. and Childs, C.W, 1988. Estimation of forms of Fe and Al: A review and analysis of
contrasting soils by dissolution and Moessbauer methods. Aust. J. Soil Res., 26: 121-144.
Reheis, M.C., 1987a. Gypsic soils on the Kane alluvial fans, Big Horn County, Wyoming. U.S. Geol.
Surv. Bull., 1590C, 39 pp.
Reheis, M.C., 1987b. Soils in granitic alluvium in humid and semiarid climates along Rock Creek,
Carbon County, Montana. U.S. Geol. Surv. Bull., 1590D, 71 pp.
Reinhardt, J. and Sigleo, W.R. (Editors), 1988. Paleosols and weathering through geologic time: Princi
ples and applications. Geol. Soc. Am., Spec. Pap. 216,181 pp.
Retallack, G.J., 1983. Late Eocene and Oligocene paleosols from Badlands National Park, South
Dakota. Geol. Soc. Am., Spec. Pap. 193, 82 pp.
Rockwell, T.K., Keller, E.A., Clark, M.N. and Johnson, D.L., 1984. Chronology and rates of faulting of
Ventura river terraces, California. Geol. Soc. Am. Bull., 95: 1466-1474.
Ross, C.W, Mew, G. and Searle, P.L., 1977. Soil sequences on two terrace systems in the north Westland
area, New Zealand. N. Z. J. Sci., 20: 231-244.
Shroba, R.R. and Birkeland, P.W., 1983. Trends in late-Quaternary soil development in the Rocky
Mountains and Sierra Nevada of the western United States. In: S.C. Porter (Editor), The Late
Pleistocene, 1: Late-Quaternary Environments of the United States. University of Minnesota Press,

Minneapolis, Minn., pp. 145-156.
Simonson, R.W., 1978. A multiple-process model of soil genesis. In: W.C. Mahaney (Editor), Quaternary
Soils. Geo Abstracts, Norwich, pp. 1-25.
Smith, S.M. and Lee, W.G., 1984. Vegetation and soil development on a Holocene river terrace se
quence, Arawata Valley, south Westland, New Zealand. N. Z. J. Sci., 27: 187-196.
Stevens, PR. and Walker, T.W, 1970. The chronosequence concept and soil formation. Quat. Rev. Biol.,
45: 333-350.
Swan, F.H., III, Schwartz, D.P. and ClufF, L.S., 1980. Recurrence of moderate to large earthquakes
produced by surface faulting on the Wasatch fault zone, Utah. Bull. Seismol. Soc. Am., 7D: 1431-
1462.
Swanson, D.K., 1984. Soil catenas on Pinedale and Bull Lake moraines, Willow Lake, Wind River
Mountains, Wyoming. Master's Thesis, Univ. Colorado, Boulder, Colo., 147 pp.
Swanson, D.K., 1985. Soil Catenas on Pinedale and Bull Lake moraines, Willow Lake, Wind River
Mountains, Wyoming. Catena, 12: 329-342.
Thompson, C.H., 1983. Development and weathering of large parabolic dune systems along the sub
tropical coast of eastern Australia. Z. Geomorphol., Suppl.-Bd., 45: 205-225.
Thompson, C.H. and Bowman, G.M., 1984. Subaerial denudation and weathering of vegetated coastal
dunes in eastern Australia. In: B.G. Thorn (Editor), Coastal Geomorphology in Australia. Academic
Press, Sydney, N.S.W, pp. 263-290.
Ugolini, F.C, 1986. Pedogenic zonation in the well-drained soils of the Arctic regions. Quat. Res., 26:
100-120.
Vreeken, WJ. van, 1975. Principal kinds of chronosequences and their significance in soil history. J. Soil
Sci., 26: 378-394.
Walker, T.W and Syers, J.K., 1976. The fate of phosphorus during pedogenesis. Geoderma, 15: 1-19.
Wright, H.E., Jr. (Editor), 1983. Late Quaternary Environments of the United States. University of
Minnesota Press, Minneapolis, Minn., 648 pp.
Wright, VP. (Editor), 1987. Paleosols — Their Recognition and Interpretation. Princeton University
283
Chapter 12
Weathering in humid regions, with emphasis on

igneous rocks and their metamorphic equivalents
E MACIAS and W. CHESWORTH
Introduction
This chapter focuses on the weathering of igneous rocks and their metamorphic
equivalents. The factors to be considered are: nature of the parent rock, environ
mental conditions, tectonic relations and time. Of the three major weathering trends
discussed by Chesworth (this book) only the acid trend will be treated since it is the
only one for which a large body of data is available.
Igneous and metamorphic rocks are formed under conditions very different from
those that prevail at the earth's surface. One might therefore expect that they would
weather more readily than sedimentary materials, which even when they have un
dergone diagenesis have still not departed far from the pressures and temperatures
of the weathering zone. Generally however, expectations in this regard are not ful
filled. Soils form more readily on sedimentary rocks than on other types and the
reason is obviously hydrodynamic. For chemical weathering to take place to any
significant degree, water must circulate through the rock, and the open structure of
most sedimentary materials is more conducive to this than is the restricted porosity
of most igneous and metamorphic rocks. As a kind of confirmation, the sedimentary
rocks that tend to weather least readily are the massive ones such as the Carbonif
erous Limestone of England and Wales — the so-called "Mountain Limestone", a
name which itself implies a topographic feature resistant to weathering. By contrast,
the igneous rocks most susceptible are those with an open structure such as the
non-welded pyroclastics. Again, in the metamorphic regime the importance of hy
drodynamics is shown in that a vertical disposition of foliation encourages a more
facile descent of aqueous solutions and a more rapid weathering, than a horizontal
foliation.
Figure 12.1 is a general model of a weathering profile. A weathering front ad
vances downwards into the rock, preserving the overall structure and volume, to
produce a saprolite. Alteration (principally by hydrolysis) is promoted by meteoric
precipitations, with the system H 2 0 - C 0 2 determining that pH of incoming solu-
284 F. Marias and W Chesworth
UNSTABLE SECONDARY PHASES WEATHERING

PHASES PRODUCED REACTIONS
Virtually all phases Interstratified sheet silicates, Hydrolysis with influence of
unstable in organic- vermiculites, possibly organics. Mica degradation,
rich surfacehorizon, kaolinite and gibbsite. destruction of earlier formed
Biogeochemical in well drained soils. kaolinite or gibbsite. Synthesis
(pedochemical) of last two inhibited.
Alteration
μ·-·..>:.·;Λ>··.Λ."·... vi.-,·:·;»
b-r'Biotic Front rap
""^^Front^^ I I I
Amorphous aluminosilicate Hydrolysis, organic influence
1:1 sheet silicates, Iminor. Destruction of plagioclase
gibbsite.
Fig. 12.1. A generalised weathering profile.
tions, will generally be acid. Important changes are in increase in surface area,
an access of free water, and a development of porosity to provide a circulatory
system. Jackson and Sherman (1953) consider this to be geological weathering.
Subsequently or penecontemporaneously, pedological weathering begins and a soil-
forming front descends into the saprolite. Rock structure is now destroyed with
similar mineralogical and chemical changes to those that produced the saprolite,
and like the latter also brought about by hydrolysis. Where the parent rock is
coarsely crystalline, fragmentation of the saprolite produces what the French call an
"arène" (the formative process being referred to as arenisation). The fragmentation
is promoted by hydrolysis and hydration at grain boundaries. In some cases the soil
may have formed on an ancient saprolite (Fitzpatrick, 1980).
Finally, when the porosity is sufficiently advanced to allow the establishment of
organisms, their secretions and breakdown products, inspire a more rigorous weath
ering and the formation of soil types characteristic of the prevailing environmental
conditions and vegetation.
The geological factor
The properties and evolution of soils, especially in the early stages, are strongly
affected by the interrelated factors of chemistry and mineralogy, and by texture and
structure of the rocks from which they are derived. Some authors consider that
soil morphology, especially where the parent rock was metamorphic, is commonly
related to the original structure of the rock (Hart et al., 1985).
The importance of the chemical-mineralogical factor is stressed by Nesbitt and
Weathering in humid regions 285
Young (1984). They give, as the average mineralogy of the upper crust, 23.2%
quartz, 39.9% plagioclase, 12.9% orthoclase, 8.7% biotite, 5% muscovite and a bal
ance of other phases of about 3%. Since quartz is relatively resistant to weathering,
feldspars become the dominant primary mineral group of the weathering system
and exert an overall influence on the direction of change.
Since the pioneering work of Goldich (1938) it has been accepted that the nature
of the primary mineral, not only influences the nature of the secondary product,
but also the rate of weathering. In general, weathering rate follows the sequence
of minerals in Bowen's reaction series with olivine and calcic plagioclase altering
more rapidly than biotite and alkali feldspar (Goldich, 1938). Lasaga's (1984) in
vestigations of mineral-weathering kinetics provide support for this view. Jackson
(1968) devised a number of weathering stages incorporating primary and secondary
minerals, into a series which is regarded as a stability sequence of phases in the
most active part of the weathering system, the clay-fraction. In fact, it is not a sta
bility sequence (Chesworth, 1977) but a persistence sequence. As such, it may be an
indication of decreasing weathering rate.
The tectonic prehistory of a weathering system may have important large scale
effects. Orogenic massifs clearly influence climate (Tardy, 1986) as the western
Cordillera of North America shows, with abundant rainfall on the seaward side,
and relatively little in the orographie shadow. Recent orogenies are also important
in that the rapidly eroding slopes of young fold mountains ensure that weathering
systems remain relatively young and poorly developed (Leptosols), so that derived,
rapidly buried sediments (greywackes for example) tend to be compositionally im
mature. Where such sediments are deposited on land, as in the Ganges-Bramaputra
valleys, they become an important source of fertile soil because of this compositional
immaturity (Chesworth, 1982). Erhart (1967) in particular has drawn attention to
the radically different ways in which the weathering system evolves when orogenic
forces have been recently active (rhexistasis), compared with those of more stable
landscapes where the biological factor predominates (biostasis).
In areas where climate is fairly homogeneous the geological factor is readily
apparent. In Galicia (N.W Spain) for example, deep, well developed soils (humic
Cambisols with a B horizon) are found on pre-Hercynian granites. By contrast,
poorly developed, skeletal soils (Leptosols), very sensitive to erosion, have formed
on late Hercynian granites. In the same region, the effect of geological structure is
seen in the alteration of eclogites, which are weathered only along joints and faults
(Macias, 1986). Weathering along joints commonly leaves core stones as remnants,
often in the configuration of a tor.
Parent rock as a soil forming factor

Parent rock is one of the classic soil-forming factors. For the German school of
the 19th century it formed the basis of the classification of soil, since soil was consid-
286 F. Marias and W. Chesworth
ered to be no more than a pathological condition of rock. By contrast the Russian

school emphasised climate and promoted the view that rocks weathering under the
same climate and vegetation produced the same types of soil. Given enough time
soil properties became independent of parent rock. This view was helped by the
fact that the Russians studied soils on fairly uniform, recent, sedimentary materi
als (thereby minimising or holding constant, the parent rock effect), and that the
great size of Russia covers several climatic zones. For similar reasons it was easy
to transfer the Russian viewpoint to North America. The excesses of this so-called
zonal theory of soil formation meant that soils that did not fit into the system
(in many cases because of parent material differences), were labelled azonal and
intrazonal.
The parent rock effect can often be masked by the effects of climate and of
vegetation, and although the alteration products of diverse rocks may converge
when weathering in a given edafological zone, the parent material effect is seldom
completely obliterated. However it is in areas with homogeneous climates, where
sedimentary covering is sparse and lithological variability pronounced, that parent
rock may clearly play a considerable role in determining soil properties. This is
not to deny that the effects will decrease as a given weathering episode progresses.
Chesworth incorporates this into his idea that in the acid trend in weathering the
composition of the weathering system evolves towards a residua system of weather
ing (Chesworth, this book), though even with this tendency, there exist properties
(even in thefinalstages) determined by the original chemistry and mineralogy of the
rocks. Brewer (1964) drew attention to one of the most important — the maximum
amount of clay that can form in the final stages is a direct function of the original
composition (Tkble 12.1).
Tàvernier and Eswaran (1977) show similar relations based on clay producing
potential. Even in the evolutionary end stages of the residua system of weather
ing a parent rock effect may show up in that basic and ultrabasic rocks tend to
evolve towards gibbsitic assemblages, while more acid compositions tend to produce
quartzitic ones. Oxic horizons seem to develop more readily on basic and ultrabasic
materials than on others (Buoi and Eswaran, 1978). Other properties, especially
TABLE 12.1
Percentage of kaolinite potentially formed by the weathering of selected rock types
Rock type Kaolinite (%)

Amphibole-pyroxene schist 99
Amphibole granite 81
Biotite granite 33-55
Orthoclase-quartz porphyry 46
Albite-muscovite gneiss 44
Source: Brewer (1964).
those concerned with aspects of soil fertility, also show a persistent parent material
effect (van Raij and Peech, 1972).
In earlier stages of pedogenesis, Ca-rich rocks promote the formation of mollic
horizons even in regions with high rainfall; Mg-rich rocks give hypermagnesic soils
with effects on the Ca/Mg balance that affect the biomass adversely. Clay floccu-
lation, podzolisation on acid rocks, and inherent fertility, all tend to show a clear
relationship with the mineralogy and chemistry of the parent material.
The micromorphology and texture of soil are determined by grain size and the
relationship between the soil plasma of clay-sized particles, the larger skeletal grains
and the voids [the terminology is due to Brewer (1964)]. These factors are in turn
dependent on the relative proportions of resistant to alterable minerals in the parent
rock, as well as on the conditions of alteration (Figure 12.2). Again, Fe-rich rocks
tend to be more easily altered than Fe-poor ones, and therefore develop the plasma
more rapidly.
Time
For convenience, time can be considered a geological factor. At one extreme the
effects of weathering can produce rudimentary soils within times of the order of a
hundred or a few hundred years. At the other extreme, it takes millions of years
to produce a Ferralsol. In humid climatic zones, the effect of the parent material
becomes more difficult to detect as reaction time increases. On any time-scale but
especially on geological ones, no true equilibrium can exist at the surface of the
earth, because soil forming factors constantly change. Despite this, the persistence
CLAY
100 0
Basic rocks
Aλ I
Schists
Granitic rocks
| Textures of Galician
^^^ soils (Cambisols)
\ Ì Textures produced by
extreme alteration (Ferralsols)
50/ Ss.50
^ Y / ^.—-^——^
—M
·// V
—u
/ \\
' *\ /) /' \\
f / \
\ V···. ■>£...■···· /
0/ Λ / \
100 50 y .. \ 01 nn
SAND SILT
Fig. 12.2. Textures produced in soils from different parent materials.
288 E Macias and W. Chesworth
of certain weathering systems, particularly in regions undisturbed by recent tec-

tonism or glaciation, indicates that a kind of time-independent state of inertia (a
pseudo-equilibrium or a pseudo steady state perhaps) may be achieved for a time.
However, the longer the time of weathering, the more likelihood there is that the
weathering environment (particularly the climatic factor) will change, and the type
of weathering will change with it. For this reason most long term weathering systems
will be polygenetic.
Rock-water interactions in the soil zone
For the most part, weathering is essentially a form of incongruent solution, which
Pedro and Sieffermann (1979) summarise as:
primary minerals + attacking solution = secondary minerals + leachate
The secondary phases may be produced from the primary by degradation (where
the secondary phase inherits structural regimes from the primary phase, as in biotite
to vermiculite transformations), or they may result from the total destruction of the
primary mineral structure (as in the case of feldspar to allophane).
The process is strongly influenced by the nature of the primary minerals as
already indicated, and by a complex of environmental factors (climate, biospheric
influences, hydrodynamics and so on). The environmental factors can be considered
as acting through the aqueous phase, the sine qua non of weathering.
The aqueous environment

Without doubt, weathering takes place in arid regions, but it is largely physical in
nature. For a notable degree of chemical weathering to take place, water must be
present, and the amount of water, its composition and its hydrodynamics, are crucial
in determining the predominant mechanisms and the direction of weathering. In a
sense, the aqueous phase constitutes the environment of weathering (Chesworth,
this book), and its chemical properties, particularly pH and Eh (or pe), define what
changes are possible in the solid parts of the system.
Temperature is an obviously important environmental factor, since it clearly af
fects reaction rate. From polar to tropical regions reaction rate increases by 5 orders
of magnitude (Ugolini, 1986; Nahon, 1986). Furthermore, an increase in temper
ature may lead to a change from the stability field of one secondary mineral into
another. This will cause a different weathering reaction to take place (Fritz and
Tkrdy, 1973; Bourrié, 1981) than at lower temperature. High temperature (as well as
low activity of H4S1O4) favours the production of gibbsite and may help to explain
the zonality of soil mineralogy (Pedro, 1968). Of course, in any given zone, other
factors than temperature will operate and will themselves lead to rate differences
TABLE 12.2
Weathering rates under different environmental conditions
Locality Type of rock or mineral Rate Reference

(μπι per 1,000 years)
Baffin Island (Canada) Hornblende 0.1-0.5 Locke (1986)
Montana Hornblende <1 Hall and Martin (1986)
Western U.S.A. Basalts 10-20 Colman and Pierce (1981)
Southern Alps (N.Z.) Arenites 1,000 Whitehouse (1986)
Central Europe Granites 6,000-13,000 Paces (1986)
Chad Granite, Gneiss 13,500 Gac (1979)
New Caledonia Acid rocks 10,000-16,000 Trescases (1973)
New Caledonia Ultrabasic/basic rock 29,000-47,000 Trescases (1973)
Ivory Coast Granites 14,000 Boulangé (1983)
Ivory Coast Granites 5,000-50,000 Leneuf(1959)
West Africa Vol can ics 10,000-50,000 Nahon (1986)
Source: Colman and Dethier (1986).
of several orders of magnitude. Thus, although Table 12.2 can be interpreted as

showing the effects of temperature on weathering rates, it must be recognised that
lithology, structure, geomorphology and biotic influences, are all factors, and all
determined rates are site-specific.
Elsewhere (Chesworth, this book), Eh is considered with pH as one of the two
master variables of the weathering system. The highest oxidation states are con
trolled by atmospheric pC>2. The lowest states are controlled by reactions in the
biomass, especially decay, and by hydrodynamic conditions. In general, Eh varies
from —0.3 to 0.8 volts, with most values falling between 0.2 to 0.7 volts. The higher
values obtain in well drained materials, the lower values in waterlogged ones (aquic
environments, Gleysols for example). In regions of local and seasonal waterlogging,
Eh (and pH)fluctuationsoccur and acid-clays may form by the process of ferrolysis
(Brinkman, 1970).
The composition of weathering solutions is very variable, and interaction with
soil-solids, and with atmospheric and biospheric gases and solids, causes a continual
modification. Rainfall has a pH generally in the range 5.5 to 6.2, but this may be
locally affected by such factors as the presence of dust, of aerosols from the sea,
or by contamination due to human activities. The precipitation can enter the soil
directly, but usually it comes into contact with vegetation, picking up compounds
not only originating in the phytomass, but also those produced by dry deposition. All
of this may increase the pH by 1 or 2 units (Calvo, 1979). In the upper horizons of
soil, where C0 2 is usually from 5-22, and rarely as much as 400 times atmospheric
concentrations (Holland, 1978), pH may again be lowered. In addition organic
acids may be an important source of protons. Vedy and Bruckert (1982) record
acetic, formic, lactic, oxalic, malic and citric acids in concentrations between 3
and 92//gl" 1 in podzols of the Vosges mountains. In Galician soils that support
eucalyptus and pine,fivephenolic acids have been detected.
The importance of the pH of the aqueous phase arises from the fact that hy
drolysis is perhaps the most characteristic chemical reaction in the weathering zone.
Pedro (1979) has erected a classification of weathering by hydrolysis, with five divi
sions essentially as a function of pH.
(a) Acido-complexolosis. Low temperature and low biological activity, allows or
ganic matter to accumulate and concentrations of organic components (fulvic acids)
to build up in the aqueous phase to 10"2-10"5 N, generating a pH in the range
3-5 (Robert et al., 1987) However, what type of acid predominates, is an impor
tant determinant of what happens to the mineral fraction (Robert et al., 1987).
Strongly complexing acids such as oxalic and citric lead to the destruction of all
minerals except quartz. Octahedral Al is solubilised as a metallo-organic complex.
Tkrtaric, salicilic and most acids in the fulvic fraction, provoke changes that are in
termediate between complete destruction and the formation of interstratified clays.
Finally, acids such as galacturinic, lactic and acetic, with little or no tendency to
wards complexation, act in similar fashion to CO2, and lead to the formation of
hydroxyaluminum vermiculites from micas, and to the neoformation of allophanes,
imogolite, kaolinite and/or gibbsite (Macias et al., 1987).
According to Pedro (1979) acido-complexolosis leads to true podzolisation, where
Al and Fe are mobilised downwards in a profile, relative to Si. Organic acidity and
complexing ability are the important factors, with C02-acidity playing no role. All
cations other than Si, Al and Fe, tend to leave the system completely.
(b) Acidofysis. Here acidity is controlled by dissolved C0 2 , and a pH between
4.5 and 6 is obtained. Acidolysis is the common reaction in organic horizons
where organic matter decomposition and root-respiration actively raise the partial
pressure of CO2. Elements of hydrolysate, oxidate and resistate types accumulate
(Chesworth, this book), while others leave the system in solution. A representative
reaction (Stumm and Morgan, 1981) is:
cation-aluminosilicates + H 2 C0 3 + H 2 0 =
HCOJ + H4S1O4 + cations + aluminosilicates
The result is that the weathering system moves towards compositions in the system
Si02-Al203-Fe 2 03-H 2 0 and to mineralogical assemblages dominated by oxides,
hydroxides and aluminosilicates, particularly 1:1 clays.
(c) Hydrolysis (sensu stricto). In this case solutions are weakly acid to slightly al
kaline, and they tend to have an abundance of the so-called basic cations liberated
by the primary minerals and/or the exchange complex. Compared with the environ
ment of acidolysis, water commonly has a longer residence time in the weathering
system, and in dryer areas calcite may precipitate. Smectites are the common sec
ondary aluminosilicate.
(d) Salinofysis. Solutions here, are slightly alkaline, and have a high ionic strength.
This is a type of alteration common in semi-arid to arid climatic zones, where
there is a strong tendency for the aqueous phase to evaporate. The result is the
precipitation of salts of strong acids (sulphates and chlorides for example) and the
formation of various smectites such as beidellite, nontronite and stevensite (Gac
and Tkrdy, 1980).
(e) Alkalinofysis. The aqueous solution is salty and strongly alkaline, with a pH of
9 or more. As with salinolysis the evaporitic minerals form, but Fé, Al and Si may
reach relatively high concentrations in solution (Sokolova and Tàrgulyan, 1977).
This allows the formation of an albic horizon (as with podzols) and the secondary
formation of Na-smectites and alkaline zeolites.
Lower in a given profile the pH of the soil solution changes as a function of
its initial value, the type of reaction in Pedro's (1979) scheme, contact time, and
xeric type. There is a tendency for pH to move towards neutrality, with dissolved
CO2 determining hydrogen ion concentration even in podzolic regions (Ugolini and
Dahlgren, 1987; Macias et al., 1987; Robert et al., 1987). Examples of soil solution
compositions are given in Table 12.3.
The contact time of the solution is dictated by the hydrodynamic conditions. In
turn this will determine the extent to which the composition of the aqueous phase
is modified and what secondary products form. In leaching systems, drainage can be
considered as the driving force of weathering (Millot, 1970), a fact that is obvious
if we consider weathering in terms of Pedro and Sieffermann's (1979) equation.
Removing the leachate creates a mass action effect favouring the breakdown of
the primary phases. In humid zones where drainage is free, subtractive weathering
processes lead to the accumulation of Si, Al and Fe and the elimination of all
other major cations. Under dryer conditions the aqueous phase may develop high
salinity and, depending on the seasonal availability of water, may move upwards
or downwards in the profile to deposit salts at the surface or at a specific depth
determined by the water balance. The effect of residence time is clearly shown
by Wollast's (1967) experiments on the alteration of feldspar. In his experiments
the aqueous phase became saturated in Al within minutes, but saturation with
respect to Si took more than 24 hours. Thus systems with a low residence time
will favour the formation of gibbsite rather than an aluminosilicate and, although in
Jackson's (1968) scheme, gibbsite is an indicator of an advanced stage of weathering,
a low residence time for the aqueous phase may lead to its formation in incipient
weathering systems (Macias, 1981), with aluminosihcates forming in those parts of
the system (fine pores and topographic lows, for example) where residence time is
longer.
This raises the question of scale. The hydrodynamics and therefore the compo
sition of the aqueous phase in the microsystems of the weathering zone, in essence
the front line of the weathering system, will be very different from Pedro and Sief
fermann's (1979) freely flowing leachate. The latter, in the macrosystem, has water
TABLE 12.3
Examples of the composition of the weathering solution in different environments
Locality Soil type Horizon pH Electrical Ca Mg Na K Si Al

(FAO, 1985) conductivity
(μπΐθΐ l " 1 )
(mho c m - 1 )
Cold
Washington State, Podzol A 180 80 40 40 150
U.S.A.1 Bh/Bs 20 80 10 10 270
Denmark 2 Podzol A 3.6-4.4 - 11-52
Bh-Bs 4.9-7.4 - 0-63
Galicia, Cambisol A 3.6-6.6 7xl0-8-10-6 12-300 21-987 204-7739 8-569 18-295 0-2
N.W Spain 3>4 (on granite) B/C 4.3-6.6 5xl8-8-10-6 2-163 25-662 191-9130 5-141 23-150 0-259
Andosol A 4.7-7.2 4xl0-7-2xl0-7 20-550 250-325 141-643 17-182 16-546 0-35
(on gabbro) B/C 4.4-6.1 3xl0-7-9xl0-7 0-170 2-74 92-740 5-662 12-227 0-133
Temperate
Cambisol (on A 6.4-7.5 10-4-3xl0- 68-360 113-1040 278-1209 2-100 64-311 0-2
serpenti nite)
Range observed5 1250-2500 2000-4000 50-1000
Range observed 6 750-7500 200-2000 20 30-300 350-1800 0-4
Tropical dry
Chad 7 Vertisol A 7.5 1740 860 1335 685 443
E Macias and W. Chesworth

Brine 7-9.9 170-11000 140-9500 180-5 x IO5 70-6 x IO5 159-1050
Tropical wet
Queensland, Ferra lsol A 4.1-7.9 2xl0-7-10-6 10-640 50-860 700-860 100-790
Australia 8 B 4.6-6.6 3 x l 0 ~ 8 - 3 x l 0 »-7 5-100 5-100 130-400 10-180
Brasil9 Ferra lsol (avg. of - 19 12 27 10
six soils)
Sources: Ugolini and Dahigren (1987); 2 Petersen (1976); 3 Fernandez-Marcos (1985); 4 Calvo (1979); 5 Reisenaver (1966); 6 Muurman and Kontz (1972);
7
Gac and Tardy (1980); 8 Gillman and Bell (1978); 9 Weaver and Blom (1977).
Weatherìng in humid regions 293
acting clearly as a mobile component and the weathering system is thermodynami-

cally open. On the micro-scale, along grain boundaries in coherent rock for example
(Meunier and Velde, 1979), the weathering system is essentially closed in a ther-
modynamic sense, even to water. In the larger pores of the microsystem (>3 μιη
diameter), bioactivity will in general be high and contact time will be short. In
smaller pores the inverse is true (Fernandez-Marcos et al., 1979), and a relatively
high pH, with ready availability of cations, explains why smectites form at grain
boundaries in solid rock (Meunier, 1977; Meunier and Velde, 1979; Proust, 1983).
Abrasion pH (Stevens and Curran, 1948) approximates the pH of the fine pores and
it can be used as an indication of the extent of alteration in a rock (Figure 12.3;
Romero et al., 1987).
The importance offeedback

Because of the law of mass action, the behaviour of the secondary products has
a feedback effect on the progress of weathering. Whether the leachate is eliminated
pH
10
6h
Fresh Saprolite Leptosol Umbrie Cambisol Dystric Cambisol

rock (Lithic Udorthent) (Entic/Lithic (Typic Haplumbrept)
Haplumbrept)
Fig. 12.3. Abrasion pH as a measure of degree of weathering and clay production.

quickly or slowly will obviously affect the tendency of the primary phases in Pe
dro and Sieffermann's (1979) reaction, to break down. Most leachates leaving well
drained sites in humid climatic zones are in equilibrium with kaolinite (Tkrdy, 1971;
Drever, 1982). If in fact kaolinite forms as a secondary solid (monosiallitisation in
Pedro's (1968) terminology), the kaolinite effectively buffers the system with regard
to H + and species of Al and Si in solution. In excessively well drained sites, gibbsite
may be the predominant secondary solid (allitisation), whereas in conditions of poor
drainage 2:1 sheet silicates may persist (bisiallitisation). Tb these three processes,
Gavaud (1977) adds a fourth, sialitisation, or the formation of secondary amorphous
phases as in podzolisation and andosolisation. Each type of solid produced will give
different buffering effects to influence the progress of the weathering reaction.
Finally, provided the ionic strength of the aqueous solution is not high enough
to cause flocculation, the secondary clay products may disperse and be translocated
away from the reaction site. Again, the feedback effect of removing the secondary
product will be to induce more breakdown of the primary phases.
Weathering of igneous and metamorphic rocks in humid regions
The following discussion concerns the weathering of igneous and metamorphic

rocks by hydrolytic alteration. This follows the acid trend described by Chesworth
(this book). He recognises three separate tendencies:
(a) ferrallitic weathering, characteristic of stable landscapes in the humid tropics,
affecting igneous metamorphic and sedimentary materials;
(b) andosolic weathering, characteristic of Cenozoic volcanic regions, and typical
of the alteration of glassy pyroclastics; and
(c) podzolic weathering, characteristic of humid areas with coniferous or heath-
land vegetation, affecting well drained siliceous materials of igneous, metamorphic
and sedimentary provenance.
Igneous rocks, on extrusion or exposure, are fresh additions to the surface of
the earth and therefore begin entirely new weathering episodes, uncomplicated (in
principle) by earlier ones. Since igneous rocks may originate by the anatexis of sedi
mentary rocks, it could be argued that some occult sedimentary characteristics could
be inherited from an earlier weathering cycle. An abundance of normative corun
dum in granitoid rocks has been interpreted in this way (White and Chappell, 1983).
However, the high temperature of the melt-crystal processes by which igneous rocks
form means that obvious features that might have been inherited from previous
cycles of weathering, have been obliterated. Igneous rocks therefore constitute a
logical starting point in weathering studies.
Igneous rocks
Crystallization from a melt produces a thermodynamic system with a mineral as

semblage clearly out of equilibrium with the low pressure, low temperature environ
ment of the earth's surface. Weathering is an attempt to produce a new equilibrium
state. Generally, as Goldich (1938) discovered, the higher temperature minerals
weather more rapidly than those of lower temperature. Whether a new equilibrium
is reached, and how quickly, are both unpredictable, but the tendency to move to
wards the goal of equilibrium is undeniable. At the outset then, we need to define
this goal, an exercise that is best accomplished by considering weathering processes
in the humid tropics or near tropics. Here a relatively high temperature, ready avail
ability of water and long, undisturbed periods of weathering (on the stable craton
for example), result in as near an approach to equilibrium mineral assemblages as
can be expected on the terrestrial land surface.
The ferrallitic trend
The compositional spread of igneous rocks is shown in Figure 12.4. The popu
lation covered, ranges from ferromagnesian-rich, ultrabasic and basic rocks, to acid
rocks, rich in alkali feldspars. In this projection, adapted from Velde (1985), R2 =
Mg2+, Mn 2+ , Fe2+ ; R3 = Fe 3+ , Al 3+ , Ti4+ ; M = Na + , K + , 2 x Ca 2+ . The MR3 pole
represents the bulk composition of feldspars, the 2R3 pole represents minerals such
as gibbsite, kaolinite and pyrophyllite, and the 3R2 pole represents serpentine and
talc. Illite solid solutions lie roughly half way between 2R3 and MR3, and smectites
cover a zone in the bottom half of the traingle ranging in composition from beidel-
lite, on the 2R3-MR3 sideline, to saporite on the 3R2-MR3 sideline. The evolution
during ferrallitic weathering of chosen rock types is also shown (Figure 12.4), and a
clear tendency of all rocks to change towards compositions rich in one or both of the
components AI2O3 and Fe2Ü3 is obvious. With H2O and S1O2, which are implicit
in the diagram, these components constitute a four component system that can be
considered a residua system of weathering (Chesworth, this book). The weathering
trends shown in Figure 12.4, are all developed under tropical or subtropical type cli
mates where humus tends to decay rather than accumulate. In general the sequence
passes from a stage of incipient weathering, with bulk compositions not notably
different from the parent materials, through a stage where 2:1 sheet silicates are
common, to an end stage made up of the components Si0 2 , AI2O3, Fe 2 0 3 and Η2Ο,
and phases such as quartz, goethite, kaolinite or gibbsite.
Of the massive volcanic rocks the mineralogical evolution of weathering basalt is
best known. Kaolinite predominates in well drained sites with a high rainfall (Sher
man and Uehara, 1956; Singer, 1966; Swindale 1966). Where drainage is excessive,
gibbsite is prominent. Smectite is the commonest phase after kaolinite and is gen
erally saponitic or nontronitic (Weaver and Pollard, 1973). Rarely, dioctahedral and
2R3 3R2
Fig. 12.4. Weathering trends of selected igneous rocks and their metamorphic equivalents. Rocks repre
sented: nepheline syenite, anorthite, granite, granodiorite, diorite, gabbro, peridotite, dunite, rhyolite,
andésite, basalt, quartz-feldspathic gneiss, amphibolite and charnockite.
trioctahedral intergrades are found (Curtin and Smilie, 1981). Vermiculite (Smith,
1957; Briner and Jackson, 1970) and interstratified montmorillonite-vermiculite-
illite (Bain and Russell, 1980) have also been reported. Allophane is common (Sief-
fermann and Millot, 1969).
The weathering of granitoid rocks along the ferralitic trend has been extensively
studied (Dejou et al., 1977; Chesworth, 1979; Gilkes and Suddhiprakarn, 1981; Melfi
et al., 1983). Mineralogical evolution of the residuum is clearly towards kaolinite-
dominated assemblages, though in rapidly drained environments with low H4S1O4
activity, gibbsite may form (Dejou et al., 1970). Meilhac and Tardy (1970) present
elaborate schemes for the breakdown of the primary phases. They show plagioclase
producing vermiculite and smectite by four different routes, one directly and three
via sericite. Kaolinite and gibbsite are shown as being produced in three ways:
by the degradation of the secondary 2:1 phases already formed, by means of a
non-crystalline intermediary or directly. Similarly, Meilhac and T&rdy (1970) show
several routes by which biotite produces a kaolinite or gibbsite end stage. Chlorite,
vermiculite and smectite are possible intermediaries. As with basalt, the continuous
removal of components by leaching inevitably leads to end stages represented by
simple compositions in the system SiC>2-Al203-Fe2C>3-H20.

The weathering of ultrabasic rocks leads to similar end points with goethite,
kaolinite and gibbsite predominating in the most weathered materials (Pacquet et
al., 1982). Unusual features of the weathering of this type of rock is the production
of talc and, in places, Ni-rich phases, such as the smectite pimelite (Colin et al.,
1985).
An interesting aspect of this trend as suggested in the last section, is the degree
to which an equilibrium mineralogy is achieved. Assemblages predicted by the ther-
modynamic examination of equilibria in the system SiO2-Al203-Fe2O3-H2O are
commonly found. For example, a tie figure goethite-kaolinite-water, exists which
does not allow gibbsite and quartz to coexist in stable equilibrium (Chesworth,
1975). If they were to occur together, kaolinite should form until one or other is
used up. Delvigne (1965) and Clemency (1973) both give examples to support this
contention, where gibbsite is produced in quartz-absent systems, while kaolinite
forms where quartz is present. Where gibbsite and quartz do exist together, as in
the weathered Marcolès Granite, this may be explainable in terms of the gibbsite
forming in local equilibrium in a microsystem essentially separate from the quartz-
bearing parts of the macrosystem (Chesworth and Dejou, 1980).
The achievement of a near equilibrium mineralogy normally requires a long
period of change (more than 100,000 years) and many intermediate stages are
possible. A common sequence on plutonic rocks is:
Leptosol —> Umbrie Cambisol —> Distric Cambisol —
►
Ferralic Cambisol —
► Ferralsol
Commonly progress is more rapid on basic than on acid rocks. Also, basic (and
ultrabasic) rocks often produce soils with a mollic horizon. Dryer climates retard
the process, and on the more basic rocks, soils with vertic properties tend to form.
According to some authorities, clay tanslocation is also possible in ferralitic soils,
to produce Ultisols (Soil Survey Staff, 1975). Such a process is well accepted in
Alfisols (Soil Survey Staff, 1975), but authors with long experience in tropical re
gions deny that clay translocation occurs in these soils (Isbell, 1980; Buurman and
Soepraptohardjo, 1980; Herbillon, 1981).
Outside the humid tropics, a close approach to equilibrium is not expected. In
pergelic soil environments (MAST < 0°C) chemical weathering is retarded and the
major product is allophane (Campbell and Claridge, 1968; Evans and Chesworth,
1989). THoctahedral smectite and chlorite have also been reported (Sokolov and
Gradusov, 1979).
The andosolic trend

Leaving aside welded tuffs, pyroclastics (fragmental and pumiceous rocks) have
two characteristics that set them apart from most rocks considered in the last sec-
tion. First, they allow a more ready circulation of water, and second, having chilled
rapidly during explosive eruptions, they tend to contain a high proportion of glass.
On stable landscapes, weathering proceeds rapidly, a feature related to the first
named characteristic, and the typical weathering product produced, an Andosol,
is related to the second of these features. Andosolisation, like ferrallitisation, is a
hydrolytic weathering trend which drives the composition of the weathering rock
towards the residua system of weathering (Figure 12.4).
Andosols are typically [although not exclusively (Rodeja et al. (1985)] formed in
regions of recent volcanic activity. They have a low bulk density (less than 1 and as
low as 0.5 g cm" 3 ). Water and phosphate retention is commonly high and the clay
fraction is dominated by Al-rich allophanes which impart afluflytexture. Imogolite
is a characteristic, though not ubiquitous phase. Al-humus complexes dominate the
upper horizons of some andosols (Leamy, 1983).
In a humid climate, weathering of pyroclastic materials is rapid, with volcanic
glass the most vulnerable phase. Allophane forms, provided Al is not bound up in
organic complexes. Whether or not halloysite or imogolite forms at the expense of
allophane, probably depends upon the thermodynamic activity of Si in solution. A
high Si activity would favour halloysite, a lower one, imogolite.
All soils are ephemeral, but Andosols seem especially so. Older volcanics in the
Massif Central and in the Canaries do not show andosolic characteristics. On the
whole they are ferrallitic (Tejedor-Salguero et al., 1984; Chesworth et al., 1983).
Garcia-Rodeja and Macias (1984) records the evolutionary sequence:
Leptosol —» Andosol —
► Cambisol —> Ferralsol
as one favoured by a relatively dry climate. Andosols also show a gradational rela
tionship with the third type of acid soil, Podzol. Ugolini et al. (1988) show that soils
that could not be distinguished as being one or the other on chemical or morpholog
ical grounds were in fact undergoing podzolisation under Abies mariesii vegetation
and andosolisation under Miscanthus sinensis grassland. Ugolini and his coworkers
were able to make the distinction on the basis of the chemistry of the aqueous
phase.
The podzolic trend

Under cool to temperate climates, humus tends to accumulate in the A horizon
more rapidly than the microbial population can break it down. The boreal forest of
North America and Eurasia, as well as temperate zone rain forests (in the western
cordillera of the Americas) are examples where such a process is active. Podzolisa
tion invariably accompanies this phenomenon.
The salient chemical feature is the vertical separation of a Si-rich albic horizon
from an Al-Fe-organic rich spodic horizon. At the present time, a lively controversy
is proceeding regarding the origins of the separation. Does the Al and Fe separate
Weatherìng in humid regions 299
TABLE 12.4
Secondary phases in the system S1O2-AI2O3-H2O reported for three weathering trends.
Phase Ferrallitic Andosolic Podzolic

Allophane 0 c c
Imogolite - c 0
Boehmite r - -
Gibbsite c 0 r
Halloysite 0 c -
Kaolin ite c 0 0
Opal <- uniquitous in soils as phytoliths
c = common, c) = occasional, r = rare, - indicates not recorded.
from Si strictly by inorganic means, or is organic matter implicated? The historical

background to this controversy can be found in Peterson (1976) and Farmer (1982).
The characteristic parent material of the Podzol is quartz-rich, though granite
or quartzite, weathering in place does not necessarily (or often) produce a Podzol.
Only if drainage is free will podzols form.
Typically, the albic horizon of the boreal Podzol is quartz rich, with additional
alkali feldspar in younger Podzols. These represent the least weatherable of the
common minerals. Any ferromagnesian silicates tend to break down first. Given
enough time, only quartz will be left. The horizon of mineralogical neoformation,
the spodic, usually contains a smorgasbord of secondary phases mixed in with the
primary ones (Tkble 12.4). This is obviously a disequilibrium situation, far removed
from the simple weathering products of long term ferrallitisation.
The presence of imogolite is a crucial piece of evidence for the inorganic hypoth
esis of podzolisation according to Farmer and his coworkers. They believe that it
indicates that a hydroxylaluminum silicate sol transports Al and Si in a form that
precipitates imogolite or imogolite-like allophanes. In fact, the presence of a specific
mineral is no indication of the mechanism by which it got there, and the classic view
whereby Al and Fe are transported as organo-metallic complexes, still provides the
most generally accepted mechanism (Buurman and van Reuwijk, 1984; Chesworth
and Macias, 1985). In general the total possible time of soil formation is 10,000 years
or less (the time since the last ice sheet withdrew in the Boreal zone). Consequently,
the albic horizon still contains weatherable minerals, especially feldspars. Beidellitic
smectite may also be present.
Metamorphic rocks
On weathering, metamorphic rocks develop chemical trends similar to those
already described. Chemically, the weathering of quartzo-feldspathic schists and
gneisses, and of charnockites, is virtually the same as the weathering of granites,
while the weathering of greenschist facies greenstones, amphibolites and eclogites
300 F. Marias and W Chesworth
shows the same kind of trends as the weathering of basalt and gabbro. Figure
12.4 shows this in terms of the MR3-3R2-2R3 triangle of Velde (1985). Miner-
alogically of course, there are differences. Quartzo-feldspathic gneisses and amphi-
bolites may contain almandine garnet, staurolite, cordierite, Al 2 Si0 5 polymorphs
and other phases that are atypical of igneous rocks, and their breakdown prod
ucts may produce different pathways towards the usual mineralogical end point
made up of combinations of quartz, goethite, kaolinite and gibbsite. However, on
the whole, the macroscopic changes recorded are not notably different from those
seen amongst the minerals of the igneous rocks. Eggleton and Smith (1983) for
example show cordierite weathering like feldspar to halloysite. Parisot et al. (1983)
show garnet weathering to amorphous and cryptocrystalline iron hydroxides, which
finally age to goethite, a normal sequence of change for a ferromagnesian min
eral. Even serpentinised peridotites weather as expected in producing nontronitic
smectites on the way to producing a goethite-kaolinite assemblage (Ducloux et
al., 1976).
Discussion
The three typical weathering trends of humid regions, all lead to end points in the
system Si0 2 -Al203-Fe 2 03-H20. In Velde's (1988) diagram (Figure 12.4), which
can be considered as a projection from Si0 2 , these end points, the Al- and Fe-rich
end points of Ferralsols, Andosols and of podzolic B (spodic) horizons, all project
towards the 2R3 apex. Weathering trends plotted on this diagram pass through com
positions where illites, smectites ranging from beidellite to nontronite (depending
to a large degree on parent material), or mixed layer phases, may form. But at
the apex, kaolinite and goethite are the commonly dominant products, with gibb
site forming in environments where the silica activity in the aqueous phase is low.
However, the mineralogy of these advanced weathering products varies between the
three trends, and in this regard it is especially seen in phases with compositions that
fall in the system Si0 2 -Al 2 0 3 -H 2 0 (Table 12.4).
Solubility diagrams for a selection of mineral phases in this system. Manley et
al. (1987) and Macias et al. (1987) show that gibbsite, kaolinite and quartz, singly
or in the pairs gibbsite-kaolinite or kaolinite-quartz, are stable in the presence of
water under earth-surface conditions. As we have seen these are the minerals most
commonly encountered in ferralitic materials, and the diagrams support the con
tention that humid tropical weathering comes close to reaching phase equilibrium
in the weathering zone. By contrast Andosols and Podzols are dominated in the
clay fraction, by phases that are unstable in an aqueous environment, with respect
to gibbsite, kaolinite and quartz. Allophanes dominate the andosolic environment
and are common in podzolic B horizons. In both of these trends this is a sign of a
relatively youthful stage of formation and a chemical process far from equilibrium.
This lack of equilibrium is emphasized, especially in the case of Podzols, by the fact
that not just allophane, but a virtual smorgasbord of phases occurs. These may pre-
cipitate ab initio or they may result from aging of allophanic gels, in effect forming a
series of steps down towards the equilibrium mineralogy (Sposito, 1985).
The case of imogolite is an important one. The movement of imogolite-like sols
downwards in a profile has been claimed as a major mechanism for the trans-
port of Al during podzolisation (Farmer, 1982). The presence of imogolite in pod-
zolic B horizons has been used as evidence to support this claim. The first point
to make is that such evidence is inadmissible: the presence of a mineral such
as imogolite in a physico-chemical system tells us nothing about the process by
which it got there. All that is necessary is that local conditions and composition
should be conducive to its formation, and it could form directly from solution by
metastable equilibrium or by aging of a phase less stable than it, allophane for
example.
Conclusions
In the most active chemical weathering environments, that is those with a humid
climate, igneous and metamorphic rocks display one of three types of behaviour:
they may follow trends of ferrallitisation, andosolisation or podzolisation. A stable
tectonic environment in low latitudes favours ferrallitisation. Recent volcanism pro-
ducing glassy pyroclastics favours andosolisation. Freely draining, acidic materials,
beneath an accumulating humus derived from coniferous or ericaceous vegetation,
favours podzolisation. Gradations exist between these types as specific factors such
as SiC>2-content of parent materials, crystallinity of constituent phases, drainage
characteristics of the terrain, and the way that balance of production versus miner-
alisation of organic matter varies.
The parent rock effect is greatest during earlier stages of pedogenic evolution.
As weathering progresses a chemical winnowing removes all but the most refractory
components, plus water. This means that weathering in humid regions produces
deposits progressively enriched in the same elements: the résistâtes, hydrolysates
and oxidates. It also means that the terminal mineral assemblage will be one or more
of the minerals quartz, kaolinite, gibbsite and goethite, with quartz and gibbsite
showing a mutual antipathy.
This terminal assemblage, common in tropical and subtropical areas, approx-
imates the thermodynamically expected assemblage in the system S1O2-AI2O3-
Fe2C>3-H20, under earth-surface conditions. More common is the occurrence of
thermodynamically unstable combinations of phases such as 2:1 sheet silicates, hal-
loysite, imogolite and allophane, which represent ephemeral steps on the way to
assemblages of decreasing free energy.
302 F. Macias and W Chesworth
References
Bain, D.C. and Russell, J.D., 1980. Swelling minerals in a basalt and its weathered products from
Morven, Scotland. 1. Interstratified montmorillonite-vermiculite-illite. Clay Miner., 15: 445-451.
Boulangé, B., 1983. Les formations latéritiques du Côte d'Ivoire. Doctoral Thesis, Univ. of Paris. 341
pp. (unpublished).
Bourrié, G., 1981. Geochemistry of spring waters. Seasonal variation and aluminum control. Dev.
Sedimentol., 35: 459-473.
Brewer, R., 1964. Fabric and Mineral Analysis of Soils. Wiley, New York, N.Y., 470 pp.
Briner, G.R and Jackson, M.L., 1970. Mineralogical analyiss of clays in soils developed from basalt in
Australia. Isr. J. Chem., 8: 487-499.
Brinkman, R., 1970. Ferrolysis, a hydromorphic soil-forming process. Geoderma, 3: 199-206.
Buoi, S.W. and Eswaran, H., 1978. The micromorphology of Oxisols. Proc. 5th Working Meeting on Soil
Micromorphology, pp. 325-347.
Buurman, P. and van Reuwijk, L.P., 1984. Protoimogolite and the process of Podzol formation. J. Soil
Sci., 35: 447-452.
Buurman, P. and Soepraptohardjo, M., 1980. Oxisols and associated soils on ultramafic and felsic vol
canic rocks in Indonesia. Versi. Landbouwkd. Onderz., 889: 71-92.
Calvo, R.M., 1979. Efectos de la vegetación (Q. robur y p. radiata) sobre las propiedades del suelo: ciclos
del agua y elementos. Doctoral Thesis, Universidad de Santiago de Compostela.
Campbell, LB. and Claridge, G.G.C., 1968. Soils in the vicinity of Edisto Inlet, Victoria Land, Antarctica.
N. Z. J. Sci., 11: 498-528.
Chesworth, W, 1973. The residua system of chemical weathering. A model for the chemical breakdown
of silicate rocks at the surface of the Earth. J. Soil Sci., 24: 69-81.
Chesworth, W, 1973. The parent material effect and the genesis of soils. Geoderma, 10: 215-225.
Chesworth, W, 1975. Soil minerals in the SÌO2-AI2O3-H2O. Phase equilibrium model. Clays Clay
Miner., 23: 55-60.
Chesworth, W, 1977. Weathering stages of the common igneous rocks, index minerals and mineral
assemblages at the surface of the earth. J. Soil Sci. 28: 490-497.
Chesworth, W, 1979. The major element geochemistry and the mineralogical evaluation of granitic rocks
during weathering. In: L.H. Ahrens (Editor), Origin and Distribution of the Elements. Pergamon
Press, Oxford, pp. 305-313.
Chesworth, W, 1982. Late Cenozoic geology and the second oldest profession. Geosci. Can., 9: 54-61.
Chesworth, W. and Dejou, J., 1980. Are considerations of mineralogical equilibrium relevant to pedol
ogy? Evidence from a weathered granite in central France. Soil Sci., 130: 290-292.
Chesworth, W, Dejou, J., de Kimpe, C, Macias, F., Cantagrel, J.M., Larroque, P., Garcia, P., Garcia-
Rodeja, E. and Quantin, P., 1983. Etude de paléosols rouges développés sur les basaltes miocènes
du Massif Central français. Principales caractéristiques physico-chimiques de ces pédogenènes.
ORSTOM. Sér. Pédol., 20:189-208.
Chesworth, W. and Macias, F, 1986. pe, pH and podzolisation. Am. J. Sci., 258: 128-146.
Clemency, C.V., 1973. A quantitative geochemical, mineralogical and physical study of some selected
rock weathering profiles from Brazil. U.S.A.I.D. Rep. A D-A047642, 128 pp.
Colin, F, Noack, Y., Trescases, J.J. and Nahon, D., 1985. Laltération latéritique débutant des pyroxenes
de Jacuba, Niqulandia, Brazil. Clay Mineral., 20: 93-113.
Coiman, S.M. and Dethier, D.P. 1986. An overview of rates of chemical weathering. In: S.M. Colman
and D.P. Dethier (Editors), Rates of Chemical Weathering of Rocks and Minerals. Academic Press,
London, pp. 1-8.
Colman, S.M. and Pierce, K.L., 1981. Weathering rinds on basaltic and andesitic stones as a Quaternary
age indicator, western United States. U.S. Geol. Surv., Prof. Pap., 1210.
Curtin, D. and Smilie, G.W., 1981. Composition and origin of smectite in soils derived from basalt in
Northern Ireland. Clays Clay Mineral. 29: 277-284.
Dejou, J., Guyot, J., Pedro, G., Chaumont, C. and Antoine, H., 1970. Etude sur le problème de l'origine
de gibbsite dans les arènes granitiques des pays tempérés. Répartition et évolution des minéraux
secondaires dans un profil d'altération développé sur le granite de Marcolès (Cantal). Sci. Sol, 2:
15-22.
Dejou, J., Guyot, J. and Robert, M., 1977. Evolution superficielle des roches cristallines et cristallophyl-
liennes dans les regions tempérées. INRA, Paris, 464 pp.
Drever, J.I., 1982. The geochemistry of natural waters. Prentice-Hall, Englewood Cliffs, N.J.
Ducloux, J., Meunier, A. and Velde, B., 1976. Smectite, chlorite and a regular interlayered chlorite-
vermiculite in soils developed on a small serpentinite body, Massif Central, France. Clay Mineral.,
11: 121-135.
Eggleton, R.A. and Smith, K.L., 1983. Silicate alteration mechanisms. In: D. Nahon and Y. Noack
(Editors), Petrologie des Altérations et des Sols. Sci. Geol. Mém., 71: 45-54.
Erhart, H., 1967. La Genèse des sols en tant que phénomène géologique. Masson, Paris, 2ième édition.
Evans, L.J. and Chesworth, W, 1985. The weathering of basalt in an Arctic environment. Catena, Suppl.
7: 77-85.
Farmer, V.C., 1982. Significance of the presence of allophane and imogolite in Podzol Bs horizons for
podzolization mechanisms: a review. Soil Sci. Plant Nutr., 28: 571-578.
Fernandez-Marcos, M.L., 1985. Contribucion al estudio de equilibrios solidoliquido en suelos de Galicia.
Doctoral Thesis, University of Santiago, Santiago, 495 pp.
Fernandez-Marcos, M.L., Macias, F. and Guitian, F, 1979. A contribution to the study of the stability of
clay minerals from the soil solution composition at different pF values. Clay Miner., 14: 29-37.
pp.
Fritz, B. and Tardy, Y., 1973. Etude thermodynamique du sistème gibbsite, quartz, kaolinite, gaz car
bonique. Application à la Genèse des podzols et des bauxites. Sci. Geol. Bull., 26: 339-367.
Gac, J.Y. and Tardy, Y., 1980. Géochimie d'un paysage tropical. Le bassin du lac Tchad. In: Y. Tardy
(Editor), Géochimie des interactions entre les eaux, les minéraux et les roches. SARL, Elements,
Tarbes, pp. 199-239.
Garcia-Rodeja, E., 1985. Sobre la classificacion de los andosoles y su existencia en materiales no vol-
canicos de Galicia (N.E. Espana). An. Edaf. Agrobiol., 44: 1651-1661.
Garcia-Rodeja, E. and Macias, F, 1984. Caracterizacion de suelos acidos (Podsoles-Andosoles-Suelos
aluminicos) de Galicia: relacion con los procesos edafogeoguimicos. Congreso Nacional de Ciencia
del Suelo, Madrid, pp. 589-602.
Garcia-Rodeja, E., Silva, B. and Macias, F, 1987. Andosols developed from non-volcanic materials in
Galicia, NW Spain. J. Soil Sci., 38: 573-591.
Gavaud, M., 1977. Essai sur la classification génétique des sols. Cah. ORSTOM. Sér. Pédol., 15: 63-87.
Gilkes, R.J. and Suddhiprakan, A., 1980. Scanning electron microscope morphology of deeply weathered
granite. Clays Clay Miner., 28: 29-34.
Gillman, G.P. and Bell, L.C., 1978. Soil solution studies on weathered soils from tropical North Queens
land. Aust. J. Soil Res., 16: 67-77.
Goldich, S.A., 1938. A study in rock weathering. J. Geol., 46: 17-23.
Hall, R.D. and Martin, R.E., 1986. The etching of hornblende grains in the matrix of alpine tills and
periglacial deposits. In: S.M. Colman and D.P. Dethier (Editors), Rates of Chemical Weathering of
Rocks and Minerals. Academic Press, London, pp. 101-128.
Hart, D.M., Hesse, P.P. and Mitchel, P.B., 1985. The inheritance of soil fabric from joints in the parent
rock. J. Soil Sci. 36: 367-372.
Herbillon, A.J., 1981. Degree of weathering and surface properties of clays. In: D.J. Greenland (Editor),
304 F. Macias and W. Chesworth
Characterization of Soils. Clarendon, Oxford, pp. 80-96.

Holland, H.D., 1978. The Chemistry of the Atmosphere and Oceans. Wiley, New York, N. Y. 351 pp.
Isbell, R.F., 1980. Genesis and classification of low activity clay alfisols and ultisols: a selective review.
In: B.K.G. Theng (Editor), Soils of Variable Charge. N. Z. Soil Sci. Soc, Lower Hutt, pp. 397-410.
Jackson, M.L., 1968. Weathering of primary and secondary minerals in soils. 9th International Congress
of Soil Science. Adelaide, S.A., 4: 281-292.
Jackson, M.L. and Sherman, G.D., 1953. Chemical weathering of minerals in soils. Adv. Agron., 5:
219-318.
Lasaga, A.C., 1984. Chemical kinetics of water-rock interactions. J. Geophys. Res., 89: 4009-4025.
Leamy, M.L·, 1983. Proposed revision of the Andisol proposal. ICOMAND Circ. Lett., 5: 3-29.
Leneuf, N., 1959. Laltération des granites calco-alcalins et des granadiorites en Côte d'Ivoire forestière
et les sols qui en sont dérivés. Thèse Sci., Paris, 210 pp.
Locke, W.W., 1986. Rates of hornblende etching in soils on glacial deposits, Baffin Island, Canada. In:
S.M. Coiman and D.P. Dethier (Editors), Rates of Chemical Weathering of Rocks and Minerals.
Academic Press, London, pp. 129-145.
Macias, F, 1981. Formation of gibbsite in soils and saprolites at temperate humid zones. Clay Miner.
16: 43-52.
Macias, F, 1982. El material de origen su formación e influencia en las propiedades de los suelos de
Galicia. An. Edaf. Agrobiol., 41: 1747-1768.
Macias, F., 1986. Suelos de Galicia. Enc. Geografia de Galicia. Xuntanza, pp. 456-508.
Macias, F., Fernandez-Marcos, M.L. and Chesworth, W, 1987. Transformations mineralogiques dans
les podzols et les sols podzoliques de Galice (NW Spain). In: D. Righi and A. Chauvel (Editors),
Podzols and Podzolisation. AFES-INRA, pp. 163-177.
Manley, E.P., Chesworth, W, and Evans, L.J., 1987. The solution chemistry of podzolic soils from the
eastern Canadian Shield: a thermodynamic interpretation of the mineral phases controlling soluble
Al 3 + and H4SÌO4. J. Soil Sci., 38: 39-51.
Meilhac, A. and Tardy, Y., 1970. Genèse et évolution des séricites, vermiculites et montmorillonites au
cours de l'altération des plagioclases en pays tempérées. Bull. Serv. Cart. Geol. Alsace-Lorraine,
Strasbourg, 23: 145-161.
Melfi, A.J., Cerri, C.C., Kronberg, B.I., Fyfe, WS. and MacKinnon, B., 1983. Granitic weathering: a
Brazilian study. J. Soil Sci., 34: 841-851.
Meunier, A., 1977. Les mécanismes de l'altération des granites et le rôle des microsystémes. Thèse,
Université de Poitiers. Mém. Soc. Géol. Fr., 1980, No. 40.
Meunier, A. and Velde, B., 1979. Weathering mineral facies in altered granites: the importance of local
small scale equilibria. Mineral. Mag., 43: 261-268.
Millot, G., 1964. Géologie des argiles. Masson, Paris, 499 pp.
Muurman, R.P. and Kontz, F.R., 1972. Research results. U.S. Army Cold Regions Res. Eng. Lab., Spec.
Rep., 171, pp. 48-74.
Nahon, D.B., 1986. Evolution of iron crusts in tropical landscapes. In: S.M. Colman and D.P. Dethier
(Editors), Rates of Chemical Weathering of Rocks and Minerals. Academic Press, London, pp. 169-
191.
Nahon, D. and Colin, F, 1982. Chemical weathering of orthopyroxenes under lateritic conditions. Am.
J. Sci., 282: 1232-1243.
Nahon, D., Paquet, H. and Delvigne, J., 1982. Lateritic weathering of ultramafic rocks and the concen
tration of nickel in the western Ivory Coast. Econ. Geol., 77: 1159-1175.
Nesbitt, H.W. and Young, G.M., 1984. Prediction of some weathering trends of plutonic and volcanic
rocks based on thermodynamic and kinetic considerations. Geochem. Cosmochem. Acta, 48: 1523-
1534.
Paces, T, 1986. Weathering rates of gneiss and depletion of exchangeable cations in soils under envi
ronmental acidification. J. Geol. Soc, 143: 673-677.
Paquet, H., Duplay, J. and Nahon, D., 1982. Variations in the compositions of phyllosilicates in a
weathering profile on ultrabasic rocks. Proc. 7th Int. Clay Conf., pp. 595-603.
Parisot, J.C, Delvigne, J. and Groke, MC.T, 1983. Petrogarphical aspects of the supergene weatheirng
of garnet in the Serra des Carajas (Para, Brazil). In: D. Nahon and Y. Noack (Editors), Petrologie
des altérations et des sols. Sci. Geol., 73: 141-148.
Pedro, G., 1968. Distribution des principaux types d'altérations chimiques à la surface du globe.
Présentation d'une esquisse géographique. Rev. Géog. Phys. Géol. Dyn., X: 457-470.
Pedro, G., 1979. Caractérisation générale des processus de l'altération hydrolitique. Sci. Sol, pp. 93-105.
Pedro, G. and Sieffermann, G., 1979. Weathering of rocks and formation of soils. In: ER. Siegel (Editor),
Review of Research of Modern Problems in Geochemistry. UNESCO, Paris.
Pedro, G., 1985. Les grandes tendances des sols mondiaux. Cultivar, 184.
Petersen, L., 1976. Podzols and pozolisation. DSR Forlag, Copenhagen, 293 pp. Proust, D., 1983.
Mécanismes de l'altération supergène des roches basiques. Etude des arènes d'orthoamphibolite du
limousin et de glaucophanite de l'île de Groix (Morbihan). Thèse, Univ. Poitiers, 260 pp.
Reisenauer, H.M., 1966. Mineral nutrients in soil solution. In: PL. Altman and D.S. Dittmer (Editors),
Environmental Biology. Federation of American Societies for Experimental Biology, Bethesda, Md.,
pp. 507-508.
zolisation. In: D. Righi and A. Chauvel (Editors), Podzols et Podzolisation. AFES-INRA, pp. 207-
223.
Romero, R., Taboada, T., Garcia, C. and Macias, E, 1987. Utilización del pH de abrasion corno un
indice del grado de evolución de la alteración y edafogénesis en suelos graniticos de la provincia de
La Coruna. Cuad. Lab. Xeol. Laxe, 11: 171-182.
Sherman, G.D. and Uehara, G., 1956. The weathering of olivine basalt in Hawaii and its pedogenic
significance. Soil Sci. Soc. Am. Proc, 20: 337-340.
Sieffermann, G. and Millot, G., 1969. Equatorial and tropical weathering of recent basalts from Camer
oon: allophanes, halloysite, metahalloysite, kaolinite and gibbsite. Int. Clay Conf., Tokyo, pp. 417-
430.
Singer, A., 1966. The mineralogy of the clay fraction from basaltic soils from Galilee, Israel. J. Soil Sci.,
17: 136-147.
Smith, J., 1957. A mineralogical study of weathering and soil formation from olivine basalt. J. Soil Sci.,
8: 225-239
Soil Survey Staff, 1975. Soil Taxonomy. USDA, Washington, D.C., 574 pp.
Sokolov, I.A. and Gradusov, B.P., 1978. Soil formation and weathering on basic rocks in a cold, humid
climate. Soviet Soil Sci., 10: 11-22.
Sokolova, TA. and Targulyan, V.O., 1977. Dissolution and translocation of clays in a sod-podzolic soil.
In: Problems of Soil Science. USSR Acad. Sci. Nauka, pp. 479-492.
Sposito, G., 1985. Chemical models of weathering in soils. In: J.I. Driver (Editor), The Chemistry of
Weathering, Nato Advanced Study Inst., Ser. C, pp. 1-18.
Stevens, R.E. and Curran, M.K., 1948. Simple field test for distinguishing minerals by abrasion pH. Am.
Mineral., 33: 31-49.
Stumm, W and Morgan, J.J., 1970. Aquatic chemistry: an introduction emphasizing chemical equilibria
in natural waters. Wiley, New York, N.Y., 280 pp.
Swindale, L.D., 1966. A mineralogical study of soils derived from basic and ultrabasic rocks in New
Zealand. N.Z. J. Sci., 9: 484-506.
Tardy, Y., 1971. Characterization of the principal weathering types by the geochemistry of waters from
some European and African cristalline massifs. Chem. Geol. 7: 253-271.
306 F. Macias and W. Chesworth
Tardy, Y., 1986. Le cycle de l'eau. Climats, palaéoclimats et géochimie globale. Masson, Paris, 201 pp.
Tavernier, R. and Eswaran, H., 1977. Soil property changes as a function of soil genesis in the humic
tropics. Proc. Climatrops, pp. 92-100.
Tejedor-Salguero, M.L., Garcia-Lopez, L. and Fernandez-Caldas, E., 1984. Sueolos ferraliticos Canarios.
Cong. Int. Suelos Volcanicos, Ser. Inf., 13: 735-758.
Trescases, J.J., 1973. Involution géochimique supergène des roches ultrabasiques en zone tropicale.
Formation des gisements nickelifères de Nouvelle Calèdonie. Mém. ORSTOM, 78, 259 pp.
Ugolini, F.C., 1986. Processes and rates of weathering in cold and polar desert environments. In: S.M.
Ugolini, F.C. and Dahlgren, R., 1987. The mechanism of podzolisation as revealed by soil solution
studies. In: D. Righi and A. Chauvel (Editors), Podzols et Podzolisation. AFES-INRA, pp. 195-203.
Ugolini, F.C, Dahlgren, R., Shoji, S., and Ito, T., 1988. An example of andosolization and podzolization
as revealed by soil solution studies, southern Hakkoda, northeastern Japan. Soil Sci., 145: 111-125.
Van Raij, B. and Peech, M., 1972. Electrochemical properties of some Oxisols and Alfisols of the
Tropics. Soil Sci. Soc. Am. Proc, 36: 587-593.
Vedy, J.C. and Bruckert, S., 1982. Soil solution: composition and pedogenic significance. In: M. Bonneau
and B. Souchier (Editors), Constituents and Properties of Soils. Academic Press, London, pp. 184-
213.
Velde, B., 1985. Clay Minerals: A Physico-chemical Explanation of their Occurrence. Elsevier, Amster
dam. 427 pp.
Wada, K., 1985. The distinctive properties of Andosols. In: B.A. Steward (Editor), Advances in Soil
Science, Vol 2. Springer, Berlin, pp. 173-229.
Weaver, CE., and Pollard, L.D., 1973. The Chemistry of Clay Minerals. Elsevier, Amsterdam, 213 pp.
Weaver, R.M. and Bloom, PR., 1977. Solution activities of aluminum and silicon in highly weathering
soils that contain gibbsite and kaolinite. Soil Sci. Soc. Am. J., 41: 814-817.
White, A.J.R. and Chappell, B.W., 1983. Granitoid types and their distribution in the Lachalan Fold
Belt, Southeastern Australia. Geol. Soc. Am. Mem., 159: 21-34.
Whitehouse, I.E., McSaveney, M.J., Knuepfer, PL.K. and Chinn, T.J.H., 1986. Growth of weathering
rinds on Torlesse sandstone, Southern Alps, New Zealand. In: S.M. Colman and D.P. Dethier (Ed
itors), Rates of Chemical Weathering of Rocks and Minerals. Academic Press, London, pp. 419-
435.
Wollast, R., 1967. Kinetics of the alteration of K-feldspar in buffered solutions at low temperature.
309
Chapter 13
Calerete
A.R. MILNES
Introduction
In the semi-arid and arid regions of many countries there are substantial accumu-
lations of carbonate in the form of hardpans, zones of nodular or pisolitic structures,
variously hardened mottles, and even layers or horizons of unconsolidated materi-
als. In many cases the carbonate accumulations are associated with calcareous soils.
They occur in Neogene calcarenite dune complexes along the southwestern coast of
Africa (Knox, 1977), on the Channel Islands of California (Johnson, 1967), along
the western and southern coasts of Australia (Read, 1974; Fairbridge and Teichert,
1952; Crocker, 1946; Warren, 1983; Gill, 1975) and in the Mediterranean region
(Goudie, 1973; Dan, 1977). They may occur on limestones of various types, for
example Cretaceous to Early Tfertiary limestones and chalks of the Mediterranean
region (Esteban, 1976; Dan, 1977) and Pleistocene reef limestones on Barbados
in the Caribbean Sea (James, 1972). In many regions, calcretes occur over non-
calcareous sediments or bedrock of igneous, metamorphic or sedimentary origin.
On the High Plains of Texas and New Mexico, for example, well-developed calcretes
mantle non-calcareous alluvial sands of the Pliocene Ogalalla Formation (Reeves,
1970, 1976). In South Australia, calcretes overlie granites and norites (Milnes and
Hutton, 1983). All of these calcretes are considered to be the result of pedogenic
processes variously influenced by local geological, geomorphological and biological
conditions.
The term "calerete" is now generally accepted for the above mentioned forms
of carbonate accumulation. It was introduced by Lamplugh (1902); for carbonate-
cemented gravels, along with "siterete" and "ferricrete" for silica and iron oxide
cemented materials respectively. A variety of other names have been used in the
literature for terrestrial carbonate accumulations, including "caliche", "croûte cal-
caire", "kankar", "kunkar", "nari" and even "travertine", but they can have very
specific connotations (Goudie, 1973; Reeves, 1976). Goudie (1973) defined calerete
as a form of duricrust, which is produced by terrestrial processes within the zone of
310 A.R. Milnes
weathering. He considered duricrusts to be composed of minerals, such as calcite

or various phases of silica, that "have accumulated in and/or replaced a pre-existing
soil, rock, or weathered materials, to give a substance which may ultimately develop
into an indurated mass." Because of the wide range of carbonate accumulations, it
is preferable to use the term calerete in its broadest sense (Netterberg, 1978; Milnes
and Hutton, 1983). More detailed descriptors are necessary for discussing specific
facies or features.
In many cases, calcretes are comprised of various facies or horizons which have a
systematic vertical distribution in a profile. Examples are well illustrated in the pa-
pers of Read (1974), Arakel (1982), Esteban and Klappa (1983), Dan (1977), James
and Choquette (1984) and others. The term "calerete profile" is thus useful. It has
been clearly demonstrated that there is a catenary sequence of facies in calerete pro-
files (Ruellan, 1971; Yair et al, 1978; Phillips and Milnes, 1988), and this has an
important bearing on understanding facets of the development of the profile.
As in the case of silcretes (this book), information about the origin of calcretes
has come largely from detailed studies of profiles. There is an extensive literature
on the subject and a great diversity of ideas. There are also several comprehensive
reviews of the character and distribution of calcretes, one of the most useful in
terms of diagnostic features being that of Esteban and Klappa (1983). In prepar-
ing this chapter, however, the aim has been to concentrate on the development of
particular ideas and concepts dealing with the formation of calcretes in pedogenic
environments. Some concepts appear as common threads throughout the interna-
tional literature and point to a coherent and consistent appreciation of particular
processes. Other observations indicate subtle differences between calcretes in dif-
ferent regions, as well as the way in which they are perceived, and serve to delineate
characteristics on which more critical studies might be carried out. The objective in
this chapter is to draw attention to some of the highlights.
Character and distribution of calcretes
Gilè (1961) suggested terms including laminar, platy, blocky, bedded and massive
to describe structural units in calcic (ca) horizons in New Mexico, as well as names
like nodular, cylindroidal, concretionary, filamentary, veined and flaky for the many
forms of carbonate segregations. Netterberg (1967, 1969, 1980) proposed a field
classification for the various types of calerete in South Africa, and used the terms
calcified soil, powder calerete, glaebular-accretionary-nodular calerete, honeycomb
calerete, hardpan calerete and boulder calerete. He suggested that the forms from
calcified soil to hardpan calerete represented morphological stages of development,
whereas he thought that boulder calerete formed through degradation of hardpan
calerete. Esteban and Klappa (1983) proposed a similar classification of morpho-
logically distinct horizons in calerete profiles typical of the western Mediterranean
Calerete 311
regions and parts of Texas. Their forms of calerete included hardpan, platy, nodular,
chalky and transitional.
Goudie (1973) comprehensively reviewed the character and distribution of
calcretes in the context of the duricrust concept. He discussed terminology and
favoured the Netterberg classification of calerete types. He also summarised data
on chemical composition and various concepts of the origin of calcretes. It is clear
from his review that calerete is very widely distributed in the semi-arid and desert
regions of the world. Subsequently, Goudie (1983) proposed that calcretes devel
oped preferentially on low angle pediments in regions of warm climate with limited
precipitation. He discussed models of calerete formation in terms of pedogenic
and non-pedogenic environments, but did not provide information on the specific
features of calcretes that might indicate one set of conditions or the other.
There have also been other valuable studies. For example, Reeves (1976) pub
lished an extensive summary of existing knowledge of calerete (caliche) with empha
sis on North America and, in particular, the southwestern U.S.A. He discussed min
eralogy, geochemistry, morphology and genesis of calcretes, referring to calcretes
of both groundwater and pedogenic origin. He also reviewed concepts relating to
the source and movement of calcium. Esteban and Klappa (1983) described and
illustrated forms and characteristics of calcretes in considerable detail, especially in
relation to examples from the Mediterranean region and the southern U.S.A. Dan
(1977) described the widespread distribution of calcretes in Israel over a variety
of calcareous and non-calcareous rock and sediment substrates. However, he re
stricted the local term nari to exposures of thick and hard calcareous crusts on soft
calcareous rocks. He suggested that the source of the calcium carbonate for calerete
formation was partly eolian dust. Furthermore, Dan commented on the dissolution
of carbonate dissolved at the landsurface and its precipitation from infiltrating wa
ters at depth, the influence of permeability of the substrate on the character of the
carbonate accumulation zone, and the degradation of calcretes following erosion of
overlying soil horizons and the formation of karstic features.
Milnes and Hutton (1983) examined the distribution, character and stratigraphie
relationships of pedogenic calcretes in Australia. They adopted a broad geomorphic
subdivision comprising coastal and near-coastal regions where calcretes are associ
ated primarily with Neogene calcarenite dune complexes, and inland regions where
calcretes occur essentially over non-calcareous continental sediments and bedrock.
Chemical and mineralogical data provided an important basis for their interpreta
tions of the origin of calerete profiles.
Studies such as these have added important information to our general knowl
edge of calcretes. They have also provided a framework in which to assess the
general characteristics of carbonate accumulations encountered in the field, and
broad concepts that can be tested following the acquisition of site-specific data.
However, it is attention to details of the calerete profiles that provides information
critical to an understanding of their origin.
312 A.R. Milnes
Development of concepts of the origin of calere tes
Studies in North America
Interpretation of macroscopic forms and morphogenetic sequences

The results of investigations on the macroscopic features of calcretes in the
desert regions of New Mexico appear to have had the most impact on our concepts
of calerete formation (Gilè, 1961; Gilè et al., 1965, 1966). Some of the interpreta-
tions proposed had actually been expressed in the literature by other researchers,
in some cases more than a decade before. Nevertheless, from studies of profiles on
remnant landsurfaces, Gilè and his colleagues were able to demonstrate the pro-
gressive accumulation of secondary carbonate with time, leading to the formation
of calcite-dominant horizons on the oldest surfaces (Figure 13.1). They argued that
the secondary calcite originated largely from eolian dust which accumulated at the
soil surface, dissolved on wetting, infiltrated the soil profile as calcium bicarbonate
in solution via pore spaces accessible to percolating water and precipitated in these
spaces as a result of evaporation (Gilè et al., 1966; Gilè et al., 1981). The depth
Fig. 13.1. Schematic diagram shewing morphological stages in the development of a carbonate horizon
in gravelly (top) and non-gravelly (bottom) materials. Carbonate accumulations are marked in black.
1 = Thin, discontinuous pebble coatings: filaments or faint coatings. 2 = Continuous pebble coatings
and êome inter-pebble fillings: few to common nodules. 3 = Many inter-pebble fillings: many nodules
and internodular fillings. 4 = Laminar horizon overlying plugged horizon in both gravelly and non-
gravelly materials (after Gilè et al., 1966).
Calerete 313
at which the carbonate accumulated was considered to be the depth of most fre
quent wetting. In gravelly sediments of the New Mexico desert regions, as well as in
other areas from which studies have subsequently been reported (Aristarain, 1970;
Sowers, 1988), calcite precipitated on the undersides of clasts, presumably due to
the localisation of water films in these places and formed "bearded" or "cupped"
pebbles and cobbles.
Irrespective of the character of the host soil or sediment, Gilè et al. (1966) were
able to identify a sequence of events leading to calerete formation. Continuing pre
cipitation of secondary calcite eventually filled all available pore spaces and sealed
("plugged") the upper part of the carbonate accumulation horizon against further
infiltration (Figure 13.1). At this stage of development, it looked as if skeleton
grains had been forced apart and the bulk density of the material had increased
(Gilè et al., 1981). Evaporation of thin layers of water that ponded on the surface
of the plugged horizon was proposed to account for the formation of the character
istic laminar horizon. This fades is composed almost entirely of secondary calcite
and occurs at the top of the most mature calerete profiles. Gilè et al. (1966) in
terpreted the multi-layered character of the laminar horizon to indicate periodic
accretion of calcite over a long period of time. In addition, they suggested that
overlying horizons were actually displaced upwards by the thickness of the laminar
horizon.
Variations in the distribution and associations of morphological facies within
calerete profiles were considered by Gilè et al. (1966) to indicate polygenetic evolu
tion resulting from changes in climate or geomorphic environment. Soil-filled pipes
containing cylindroidal and nodular carbonate cut the hardpan horizon in some
profiles, for example. Thus, they suggested that periods of carbonate accumulation
during dry climatic regimes were interrupted by phases when carbonate dissolution
and regularflushingwith water produced karst features.
Elaboration of the processes and support for an eolian origin for the carbonate
Reeves (1970) described calcretes forming the "caprock" of the southern High
Plains of Texas and eastern New Mexico, and proposed a subdivision into young
incompetent and powdery calcretes, nodular mature calcretes, and old calcretes
containing authigenic silica. He suggested that desert loess was the predominant
source of carbonate for development of thick calerete profiles, which formed under
aggrading soil conditions through downward percolation of bicarbonate-laden soil
water. He also pointed to the fact that rapid plugging and induration of part of the
calerete profile, followed by the formation of the laminar zone, effectively preserved
underlying horizons in a chalky and unconsolidated state.
Gardner (1972) studied a thick caprock caliche on Mormon Mesa, southwestern
Nevada, and tested some of the concepts discussed by Gilè et al. (1966) particularly
the source of the calcite and its mode of emplacement in the profile. The indurated
horizon of this calerete is a dense, grey sandy limestone about 1.2 m thick. It
314 A.R. Milnes
overlies a transition zone of white powdery calerete, also about 1.2 m thick, above
friable red quartz sand. Although there is evidence in the near surface horizons of
physical breakup of the hardpan and re-cementation of the fragments by illuviated
carbonate, there are none of the features such as the laminar and nodular horizons
that are typical of the pedogenic calcretes described by Gilè et al. (1966) and Reeves
(1970).
Gardner wanted to discover whether the indurated caprock (hardpan) horizon
could have formed simply by filling of voids in the parent alluvial sand with sec
ondary calcite. On the basis of bulk density and insoluble residue data, he concluded
that the calerete contained considerably more calcite than could be accommodated
in the void space of the alluvial sand. Pétrographie observations showed detrital
quartz grains "floating" in a calcite matrix, but Gardner considered expansion of the
original detrital framework unlikely. Instead, he proposed that the calerete profile
formed primarily under slowly aggrading eolian conditions, following the suggestion
of Brown (1956). Supporting evidence included the higher content of silt and clay
in the insoluble residue of the calerete compared with the underlying alluvial sand,
and the dominance of palygorskite in the calerete compared with palygorskite plus
significant amounts of montmorillonite and kaolinite in the sand. Gardner (1972)
remarked that some of the calcite present in the profile could have been derived
from dissolved calcium and bicarbonate in rainwater, but concluded that the major
contribution was from eolian dust which also contained palygorskite. He suggested
that evaporative precipitation of calcite through repeated wetting and drying of the
soil in a desertic environment produced the hardpan calerete.
Alternative views of the origin of the chalky or powdery calerete horizon have
been proposed, principally by James (1972) in relation to calerete crusts devel
oped on Pleistocene reef limestones on Barbados. He suggested that the powdery
horizon is actually the intensely brecciated, dissolved and recrystallised substrate
limestone. Carbonate, partly derived from this in situ degradation process and partly
contributed by sea-spray, is redistributed in the profile and precipitated as micrite
in voids and fractures, around particles and clasts, and in crusts. Rabenhorst and
Wilding (1986) also concluded that the massive facies of petrocalcic horizons on
Cretaceous limestone on the Edwards Plateau, Tbxas, formed through a process of
alteration of the limestone in situ.
These differing interpretations provide a strong challenge to those of us inter
ested in the detailed processes of calerete formation. It is clear that the integration
of catenary and other geomorphic observations with data obtained from the profiles
themselves are required to focus on the problem.
The time frame

Gardner (1972) calculated from the weight of secondary calcite in an average unit
column of the Mormon Mesa calerete (averaging about 2.5 thick) and the solubility
of calcite that the profile developed over at least 400,000 years and possibly as long
Calerete 315
as 2.5 Ma. Despite his broad assumptions, the technique provided at least a gross
indication of age.
Direct dating techniques using radiocarbon or uranium series methods on the
secondary calcium carbonate in calcretes have proved to be difficult. Conventional
radiocarbon analyses cover a very limited time span of about 40,000 years BP, and
are commonly in error (Williams and Polach, 1971; Callen et al., 1983). Uranium-
series analyses, on the other hand, suffer from the problems of contamination with
detrital mineral phases and open system behaviour, which may result in the gain or
loss of parent and daughter isotopes of the decay reaction (Ku et al., 1979). Other
techniques have recently been described, for example in the highly arid Death
Valley of California. Here, carbonate accumulation horizons occur in soils formed
on alluvial fan deposits of various ages. Dorn (1988) determined the ages of the fan
surfaces using accelerator mass spectrometry radiocarbon dating of organic matter
extracted from desert varnish on the rock pavements, as well as cation-ratio dating
of the varnish. In the Hanaupah Canyon fan (Dorn, 1988), a calcic horizon in soil on
the Q3a surface, which is approximately 50,000 years old, is characterised by calcite-
cupped or bearded pebbles and cobbles. On the other hand, substantial calcretes
have formed in soils on the Q2a and Ql fan surfaces, dated at 170-180,000 years
and > 800,000 years, respectively.
In the case of the Kyle Canyon alluvial fan in southern Nevada, Sowers (1988)
and Amundson et al. (1989) pointed out that of all the properties of soil materials
on the various fan surfaces, carbonate accumulation showed the greatest change
with age. Their estimates of the age were less direct than those of Dorn (1988).
Nevertheless, their Surface 3 has a calcic soil in which authigenic calcite coats
the bottom of clasts and partly fills voids, and this was considered to be 5-15,000
years old. In some places, the carbonate horizon is plugged and overlain by a
thin, discontinuous horizon of laminar calerete. On the other hand, Surfaces 1 and
2 have complex calerete horizons which can form caprocks. Paleomagnetic data
for samples of massive calerete suggest that Surface 1 is older than 730,000 years
(Bruhnes-Matuyama magnetic reversal) whereas Surface 2 is younger than this.
As pointed out by Gilè et al. (1966), the morphology of calcretes may not specifi
cally reflect their age. Nevertheless, in alluvial fan surface sequences in geomorphic
and pedologie environments like those described by Dorn (1988) and Sowers (1988)
calcretes of complex morphology and depth up to 0.6 m can develop in less than
800,000 years and, as illustrated by the Q2a surface of the Hanaupah Fan in Death
Valley, in as little as 170-180,000 years. However, the rate and degree of forma
tion of calcretes is clearly influenced by many other factors, including the source
of the carbonate and the method by which it accumulates, the nature and stability
of the geomorphic and geological environments and, as pointed out by Amundson
et al. (1989), climatic factors and biological influences such as the structure of the
vegetation community.
316 A.R. Milnes
The Mediterranean region
Catenary concept
As part of a comprehensive study of the calcareous soils of Morocco, Ruellan
(1971) stressed the importance of studying lateral as well as vertical variations in
the characteristics of calerete horizons in a catenary sense. In terms of morpholog
ical forms, he recognised moderately differentiated and well differentiated calerete
profiles in which there were massive and nodular non-platy crusts, as well as croûte
calcaire, dalle compacte and pellicule rubanée forms of platy crusts. However, there
was frequently a consistent lateral variation from horizons with semi-indurated car
bonate mottles and nodules to non-platy calerete crusts, platy crusts and eventually
compact hardpans. These changes correlated with an increase in carbonate content,
and were interpreted to indicate significant lateral migration of carbonate in the
catena. Ruellan used this concept of lateral migration to explain the occurrence of
calcretes over non-calcareous substrates.
In Israel, Yair et al. (1978) studied the variation in thickness of calerete (nari)
developed on chalk in a catena, and concluded that it was related to the distribution
of moisture during infiltration and runoff under semi-arid conditions. In particular,
thin calerete horizons on convex slopes were considered to be due to less effective
infiltration and significant runoff and erosion, whereas thick profiles on low-angle
slopes and in basal footslopes reflected enhanced infiltration. Dan (1977) found that
calerete on hill crests and steep slopes on chalk displayed elements of the substrate
rock but lacked a laminated surface horizon. Downslope, however, the calerete
thickened and developed the profile characteristics described initially by Yaalon and
Singer (1974). In the footslope position, the calerete was covered by a soil mantle
and merged laterally with the calcic horizons of gravelly soils. A similar pattern was
noted in those localities where calerete had developed in eolianite. In this case the
calerete usually capped hilltops and steep slopes, but merged downslope with calcic
horizons in soils and paleosols.
Biological influences
As in the case of catenary studies, European researchers have been notable, al
though by no means unique, in documenting the influence of biological processes
in the formation of calcretes. Klappa (1979a) described various features including
micro-borings, calcified algal filaments and fungal hyphae, and micro-spherulitic
structures and spheres of biological origin in laminar calcretes. In conjunction with
field observations of exposed calerete surfaces in the coastal regions of the western
Mediterranean, he suggested that the laminar calcretes were formed in successive
cycles of lichen colonisation which promoted calcite precipitation and eventual ce
mentation of mineral and organic materials.
In other studies of calcretes, Klappa (1979b, 1980) identified a wealth of micro-
fossils of fungi, algae, actinomycetes and the root-hairs and roots of higher plants,
Calerete 317
many of which were encrusted with or completely replaced by micritic low-Mg

calcite. Preservation of the morphology of the micro-organisms indicates that cal
cification occurred either during the life of the organism or immediately after its
death and before there was significant decay of the organic matter. Klappa (1979b)
pointed out that the presence and abundance of the microfossils confirmed the fact
that calcretes once functioned as soils. Some of the characteristic macromorpho-
logical features of calcretes such as brecciation, platy structure, tubular or channel
structures and vertically elongate nodules, as well as part of the fossilised micro
organism assemblage, were attributed by Klappa (1980) to the presence of the roots
of higher plants which are commonly preserved as rhizoliths.
Other significant indicators of microbiological activity in calcretes include Mi-
crocodium (Esteban, 1974) which Klappa (1979b) considered to be the result of cal
cification of mycorrhizae, needle fiber calcite which Callot et al. (1985) recognised
as microcrystals that formed within fungal hyphae, and alveolar texture (Esteban,
1974) which Esteban and Klappa (1983) suggested was the product of coalesced
millimetric rhizoliths.
Australian studies
Distribution,fieldrelationships and macromorphological forms
The pedogenic origin of calcretes mantling widespread eolianite dune sequences
that mark Tertiary and Quaternary coastal marine barriers was recognised at an
early stage by Crocker (1946) and Fairbridge and Tbichert (1952) who proposed
downward leaching of calcium carbonate and subsequent precipitation in the illuvial
B horizons of soils. In the inland regions of South Australia, Crocker described
calerete profiles consisting of massive, cemented horizons normally underlain by
progressively less calcareous pisolitic, rubbly or marly material. He recognised large
lateral variations in facies and the preservation of pisolitic or nodular fabric in
many of the massive horizons. Moreover, he proposed that the calcium carbonate
originated as loess, winnowed from the dune complexes of the coastal regions and
widely deposited over continental landscapes where it was subjected to dissolution
by rainfall. The resulting solutions percolated into existing soil profiles and precipi
tated secondary calcite at depth. The concept of an eolian source for the calcareous
material was widely accepted, together with the suggestion that carbonate accumula
tion zones formed through illuvial processes in pedogenic environments (Crawford,
1965; Lawrence, 1975; Churchward, 1963; Read, 1974; Milnes and Hutton, 1983).
However, Crocker (1946) also mentioned the possible addition to soils of substantial
amounts of calcium from cyclic salts in rainwater.
Read (1974) and Arakel (1982) introduced more detail through their studies of
calcareous paleosols on Pleistocene eolian calcarenite in the Shark Bay region and
at Hutt and Leeman Lagoons, Western Australia (Figure 13.2). They each recog
nised several specific horizons, including pisolitic soil at the landsurface, together
320
Fig. 13.2. Distribution of calcretes and calcareous soils in southern Australia (after Northcote et al., 1975).
A.R. Milnes
Calerete 319
pisolitic loose soil

pisolitic soils laminar calerete
breccia calerete cm
laminar calerete -0
massive calerete
:mäßm -10
massive calerete
[-20
mottled horizon I-30
skeletal mottled calerete
iμft grainstone
(Tamala Eolianite) host sediments
Fig. 13.3. A. Sketch of idealised calerete profile (after Read, 1974). B. Sketch of typical calerete profile
at Hutt and Leeman Lagoons (after Arakel, 1982).
with laminar calerete, massive calerete and a zone of secondary carbonate mottles
successively deeper in the profile (Figure 13.3A, B). Arakel identified a further
facies, called breccia calerete, between the laminar and massive horizons. Their
work showed that there were many variations in facies depending on the degree of
weathering and the local topography and erosional history.
Read (1974) and Arakel (1982) suggested that calerete profiles were formed
essentially as described by Gilè et al. (1966). Read (1974) added the important
component of geomorphic processes, whereas Arakel (1982) discussed details of
the various stages of calerete development, focussing on the influence of topogra
phy on the circulation of infiltrating waters once the plugged horizon had formed.
They agreed that at depth in the profile, where movement of skeleton grains was
restricted, secondary calcite was precipitated from downward percolating solutions
as micritic coatings around grains to initiate the formation of ooliths. Plugging of
the horizon was achieved by continued deposition of calcite in pores. Read (1974)
suggested that coalescence of micrite coatings on skeleton grains in the upper part
of the soil profile was prevented by gradual movement of the grains downslope
from dune crests due to soil creep. Loss of surface soil on dune crests eventually
exposed indurated carbonate accumulation horizons and initiated their breakup. As
the clasts shed onto dune flanks, were moved downslope with accompanying soil
particles, they were progressively coated with secondary calcite. Coated clasts oc
casionally accumulated in hollows and swales to produce gravel or cobble deposits
which Read called lithoclast breccias. In complex calerete profiles, early-formed
calerete facies were buried by prograding eolianite, transported soil materials or
colluvial sediment, and the cycle of carbonate dissolution, water percolation and
precipitation of secondary calcite at depth was repeated.
320 A.R. Milnes
In parts of the profile, Read (1974) pointed to the preferential formation of

calcite laminae in depressions, on the floors of voids and alongfissures.Following
Gilè et al. (1966) he suggested that these laminae recorded the accumulation and
subsequent evaporation of waterfilmsenriched in calcium bicarbonate. He reported
comparatively uniform rinds of calcite laminae on clasts, noting the absence of
bearded or cupped clasts in contrast with many of the North American calcretes.
Chemical and mineralogical characteristics

Hutton and Dixon (1981) studied aspects of the chemistry and mineralogy of
calerete profiles in the inland regions of South Australia. They discovered, in many
sites, a progressive decrease in Ca/Mg ratio which related to an increase in the abun
dance of dolomite relative to calcite with depth. In some profiles, the lower horizons
of powdery calerete contained only dolomite, together with high concentrations of
the acicular layer-silicate clay minerals, palygorskite and sepiolite.
Variations in the chemistry and mineralogy of calerete profiles from both the
ancient coastal province and the inland regions were studied in more detail by
Wilson (1981), Milnes and Hutton (1983), Milnes et al. (1987) and Phillips and
Milnes (1988). Calcretes within the Neogene calcarenite dune sequences are typi
cally calcitic, as are the calcarenites themselves (Figure 13.4a). However, there are
substantial accumulations of primary dolomite in the common coastal lagoons, shal
low embayments and ephemeral lakes in this province (von der Borch et al., 1985).
Many calcretes in the inland regions, away from the ancient coasts, contain high
concentrations of dolomite (Figure 13.4b), especially in the lower parts of the pro
files. Calcite, together with the detrital minerals (primarily quartz plus layer-silicate
clays), systematically increase in abundance up-profile.
Milnes and Hutton (1983) and Wilson (1981) suggested that the unconsolidated
calcareous (calcitic as well as dolomitic) horizons were predominantly eolian sed
imentary deposits in which facies like the massive and nodular calcretes had de
veloped. An eolian origin for elements within the calerete profiles is further sub
stantiated by the pétrographie and paleontological studies of Milnes and Ludbrook
(1986) and the particle size data of Phillips and Milnes (1988). This is essentially
the model developed by Crocker (1946) in southern Australia and by Gilè et al,
(1966) Reeves (1970) and Gardner (1972) in the southwestern U.S.A. However,
Milnes and Hutton (1983) pointed out from major and trace element chemical data
(specifically Si/Ti ratios) that non-carbonate components of hardpan and nodular
calcretes in some profiles appear to have a different provenance from those in the
underlying calcareous sediment. In other profiles, Ca/Mg, Si/Ti and Ti/Zr ratios in
dicate significant differences between the contemporary soil mantle and the calerete
horizons beneath. Using these data in combination with field observations, Milnes
and Hutton (1983) Wilson (1981) and Milnes et al. (1987) argued that carbonate
pans and horizons of hard carbonate nodules and pisolites in many profiles marked
disconformities that represent periods of subaerial exposure and diagenesis with
Calerete
calcite calcite
acid-soluble CaO acid-soluble CaO
100 100
hardpan calerete,
calcareous nodules,
coated clasts, mottles
u neon soli dated

calcareous
«j\ calcretes sands & silts
o\
u neon so li dated dune
calcarenites
100 100 100 100

acid-soluble MgO acid-insolubles acid-soluble MgO aoid-inso lubles
magnesite quartz + aluminosi h cates magnesite quartz + alu mi nosiü cates
Fig. 13.4. a. Triangular composition diagram, with apices corresponding to acid-soluble CaO, acid-soluble MgO and acid-insoluble residue, showing
the chemical and inferred mineralogical composition of calcretes and associated bioclastic dune sediments from the Neogene coastal province, South
Australia, b. Triangular composition diagram showing the chemical and inferred mineralogical composition of calerete facies from the inland regions of
South Australia.
323
322 A.R. Milnes
associated erosion and transport. Their experience was consistent with that of Read
(1974) who described multiple calcretes that were formed by burial of old profiles
by prograding dunes or transported soils followed by a repetition of carbonate accu
mulation processes, or by the development of new calcretes in profiles truncated by
erosion.
Identification of biological influences

Following the European experience, Phillips et al. (1987) and Phillips and Self
(1987) identified important biological influences in the formation of South Aus
tralian calcretes. They described an extensive community of soil micro-organisms
preserved in fine detail by calcite, which exhibits a variety of crystal habits. The
excellent preservation offilamentousand other biogenic structures is surprising, as
is thefinedetail of secondary calcite crystals, given the chemical reactivity of soil en
vironments. Needle-fiber calcite was clearly identified as structures within mycelial
hyphae that are released upon lysis of the organic sheaths (Phillips and Self, 1987).
The remains of terrestrial gastropods (Bothryembrion sp.) and the pupal cases of
burrowing insects are also widely recognised as being characteristic of many calerete
profiles (Milnes and Hutton, 1983; Milnes and Ludbrook, 1986).
Assessments of thetimeframe
Williams and Polach (1971) investigated conventional radiocarbon techniques
for determining the age of Late Quaternary calcareous paleosols along the eastern
margin of Lake Tbrrens, in arid, inland South Australia. Their results indicated that
relative ages were acceptable, but absolute ages were suspect and were considered
to be between 500-7000 years too old, even if samples were rigorously selected.
Bowler and Polach (1971) found that radiocarbon dates of calcareous paleosols in
the semi-arid zone were too young. More detailed work on arid zone pedogenic
carbonate caused Callen et al. (1983) to conclude that radiocarbon dates do not
provide reliable estimates of the age of pedogenesis.
Milnes et al. (1987) pointed out that calerete profiles show consistent trends from
relatively high concentrations of organic carbon in surface soils to uniformly low
values in the underlying calcretes and sediments. This is to be expected because a
high organic carbon flux to the surface soil through biological processes outweighs
losses due to degradation processes, but only the more resistant organic matter
survives in the deeper parts of the profile. In a reconnaissance study, Milnes et al.
(1987) measured the degree of racemisation of several amino acids isolated from
whole samples collected from two calerete profiles. The degree of racemisation is
determined from the relative concentrations of the D- and L-enantiomers, and has
a direct relationship to time elapsed since the death of the organism from which
the organic matter was derived (Rutter et al., 1979). Although the data could not be
used to assign ages, the results showed a consistent increase in D/L ratios with depth
for several amino acids, consistent with increasing age. D/L ratios for alanine and
Calerete 323
aspartic acid changed most with depth. Many environmental factors are known to
influence the rates of racemisation of amino acids. However, more detailed studies
are likely to provide at least relative age data within profiles and between sites.
Characteristics of selected calerete profiles in South Australia
Calcareous soils and calcretes cover vast tracts of land across southern Australia
(Figure 13.2). In areas adjacent to the modern coasts, particularly along the Great
Australian Bight and southeastwards along the coast of South Australia, outcrops
of Eocene to Pliocene limestones provide an indication of the extent of former Ter-
tiary marine basins or embayments marginal to the cratonic areas of the continent.
Extensive calcarenites with associated calcretes constitute the coastal dune barriers
of these former bays (Milnes and Ludbrook, 1986). The widespread development
of calcareous soils and calcretes inland, well beyond the coasts, reflects a complex
reworking and redistribution of carbonate in local environments.
Studies have concentrated on detailed field observations and systematic sample
collection through profiles, followed by chemical and mineralogical analyses, micro-
morphological examinations of thin sections or polished-thin sections and scanning
electron microscope investigations.
Calcretes of the Neogene coasts
Together with thin sheets of calcareous silts and fine sands, prominent calcretes
occur towards the top of thick sections of eolian calcarenite (Figure 13.5). The
calcarenites are typically cross-bedded, variably cemented sands composed almost
exclusively of rounded fragments of marine fossils. The first appearance of calcretes
in the sequence, followed by the interval in which there are dominantly calcretes
interspersed with unconsolidated deposits of oolitic sands and silts, signifies the
decline of the marine sediment source and the increasing dominance of terrestrial
environments (Milnes and Ludbrook, 1986).
The first major calerete in the sequence is usually a complex assemblage of
horizons. A typical example at High Cliff, on the western coast of Eyre Peninsula,
is up to 5 m thick and consists of a partly indurated, reddened paleosol sequence
at the base and highly indurated, clast-rich slope deposits above. Thin layers of
lithoclasts occur in the paleosol and represent lag gravels on former landsurfaces.
The upper part of the paleosol sequence contains scattered lithoclasts and rhizoliths.
The lithoclasts are subrounded to stellate in shape due to dissolution, and have thin
coatings of brown micrite. They include dark brown to black carbonaceous calerete
and micrite-cemented bioclastic grainstone. Warren (1983) described similar clasts.
Indurated horizons above the paleosol are essentially beds of oolitic sand littered
with a variety of calerete clasts, and represent mixed slope deposits. The clasts vary
324 A.R. Milnes
Fig. 13.5. View of coastal cliff section through cross-bedded, bioclastic grainstone dunes with interbed-
ded red paleosols (indicator) and overlying sequence (beginning at arrow) of calcareous sand and silt
sheets with associated calerete crusts. Locality: Cape Spencer, southern Yorke Peninsula. Height of cliff
approximately 80 m.
considerably in size up to several centimeters in diameter and, in the uppermost

horizons, are usually rounded due to accretionary laminar deposits of calcite. Many
of these coated clasts, which Seisser (1973) termed "intraclasts", have cores of
brown to black carbonaceous calerete. Rhizoliths are common features and indicate
that the deposits survived in the pedogenic environment sufficiently long to support
plant growth. Distinctive laminar calcite deposits drape the upper surfaces of each
of the deposits. In some cases, dissolution and the formation of a karstic topography
predates the formation of the laminar calerete.
The top of the calerete profile is marked by a complex karstic surface (Figure
13.6). An initial phase of dissolution of the indurated horizon produced mounds and
hollows with isolated boulders which were subsequently coated by laminar deposits
of calcite. A later period of water percolation, perhaps post-dating formation of the
paleosol in the overlying bioclastic sand, leached the calcitic matrix of the horizon to
produce dissolution channels and voids. Coated clasts in the near surface zones were
liberated and, together with materials from the overlying paleosol, subsequently
collapsed or were washed down into channels and voids.
Higher in the sequence, and also inland of the dune barriers, calcretes mantling
Calerete 325
Fig. 13.6. Karstic upper surface of calerete complex. Dissolution hollows and channels coated by laminar
deposits of calcite are filled with coated lithoclasts liberated from the calerete matrix. Note abundant
black carbonaceous lithoclasts coated by calcite laminae in the calerete. Material above is part of a
paleosol developed in bioclastic sand. Light coloured material in places over calerete is modern wind
blown beach sand. Hammer handle 15 cm long.
oolitic sand dunes and sheets are neither as thick nor as complex. Sections have
been studied in detail in several locations.
Poochera quarry
Macromorphological description
A quarry profile on the western side of Highway 1, north of Poochera, is a typical
example (Figure 13.7a). The quarry is floored on a calerete. The overlying oolitic
sand dune is itself capped by a variably cemented zone. In places, stalactite-like
structures marking preferential zones of water percolation and carbonate precipita
tion project down into the unconsolidated sediment. The dune sediment consists of
well-sorted, sand-sized, rounded, micrite-coated ooliths (about 0.3 mm in diameter)
with cores of microfossil fragments or quartz grains. The fabric is grain supported
and highly porous. Bridges and menisci of micrite between ooliths is the cement
which is responsible for hardening the upper zones. However, the cementation is
not uniform and the horizon has a cavernous appearance due to presence of many
channels, hollows and small pipe-like structures. Laminar deposits of calcite drape
the upper surface of the horizon and line the walls of pipes and channels.
Above is a calcite-impregnated deposit of mixed origin formed of scattered litho
clasts of calerete and micrite-impregnated skeletal grainstone in a matrix in which
326 A.R. Milnes
^ ► Ä ^
",-fe.ri
ifefe
::iv*4£^
modern soil
brecciated calerete pan
oolitic fine sand with

lithoclasts
MÌI221 —
MÌI222 — laminar calerete
MÌI223 — lithoclast-rich calerete
MN224 —
MÜ225 — cemented oolitic sand
MÎI226 —
MH227 — oolitic sand
Fig. 13.7. a. View of calerete profile in Poochera quarry, western Eyre Peninsula, b. Sketch of calerete
profile showing macromorphological organisation and approximate locations of samples collected. Black
areas are carbonised plant roots.
Calerete 327
ooliths with cores of marine microfossil fragments, quartz and feldspar are ce
mented by micritic calcite. The upper part of the calerete is considerably disrupted
through dissolution and the effects of plant roots (Figure 13.7b). Carbonised roots
are quite common throughout the profile. The overlying unit is a deposit of oolitic
fine sand which has the form of a sheet thickening towards the south. It contains
scattered, hard, angular lithoclasts (up to 4 cm diameter, with thin coatings of mi-
critic calcite) which increase in abundance upwards towards a calerete pan.
The pan is discontinuous and changes laterally from plate-like or biscuit-shaped
masses to concentrations of angular clasts (Figure 13.7b). This form of the cal
erete is a consequence of degradation through dissolution by percolating waters
and physical breakup by penetration of roots from the native Eucalyptus vegeta
tion. The calerete consists of scattered, poorly-sorted and vari-shaped subrounded
to amoeboid clasts, usually of micrite-cemented bioclastic grainstone, some of which
are black and carbonaceous. They have thin coatings of brown to grey micrite and
are set in a matrix which consists of scattered ooliths and quartz grains cemented
by comparatively porous brown micrite. The irregular upper surfaces of plates and
clasts are coated by laminar deposits of calcite forming crusts up to 1 cm thick.
The laminae vary in colour from brown to dark grey and contain scattered silt-size
quartz grains typical of illuvial material. The dark grey colour is typical of laminae
containing high concentrations of fossilfilamentousmicro-organisms.
The overlying soil mantle is sandy and contains abundant plant debris, particu
larly roots and root fragments. Much of this soil material has infiltrated the calerete
horizon through cracks and root channels; it fills the irregular dissolution hollows
in the upper surfaces of the plates and biscuits, and surrounds lithoclasts that have
been broken from the pan (Figure 13.8). Considerable bioturbation and mixing of
the soil mantle is probably responsible for the scatter of calerete clasts above the
pan. Such activities are indicated by the abundant burrows and pupal cases of insects
like Leptopius. In parts of the region, these pupal cases remain as a lag gravel after
deflation of the soil.
Chemical analyses of representative bulk samples collected at 5-10 cm intervals
through the profile provide the data plotted in Figure 13.9. Although comparatively
heterogeneous, samples were sieved into <2 mm and gravel fractions for the pur
pose of reconnaissance chemical and mineralogical analysis. Samples of cemented
oolitic sand, lithoclasts of various types and calerete pans werefinelyground.
CaC03 and MgC03 concentrations (acid soluble Ca and Mg expressed as car
bonates) in the grainstone dune and its hardened crust, and the calerete caprock,
exceed 70 and 5 weight percent, respectively (Figure 13.9a). MgC03 concentra
tions in these samples are significantly enriched relative to CaC03, suggesting that
secondary dolomite may be responsible for cementation. The <2 mm fractions of
the overlying materials systematically decrease in CaC03 and MgC03 to the soil
328 A.R. Milnes
Fig. 13.8. Contact between upper calerete pan and modern soil. Note laminar deposits of calcite coating
karst surface of pan, common lithoclasts reworked into the base of the soil, the abundance of plant
roots, and insect burrows with hardened walls.
surface. Lithoclasts separated as gravel from each sample in the profile, and the
upper calerete pan, are enriched in CaC03 and depleted in MgC03 relative to
the unconsolidated sediments, indicating impregnation by secondary calcite. Fig
ure 13.9b displays these data in terms of acid-soluble CaO (representing calcite),
acid-soluble MgO (dolomite) and the acid-insoluble component of the samples (es
sentially quartz and layer-silicate clay minerals). Thus, the location of samples on
the triangular diagram provides a guide to their mineralogy.
Unconsolidated samples of the basal oolitic sand dune plot towards the CaO (cal
cite) apex of the diagram. Consolidated samples contain higher concentrations of
dolomite. The calerete crust has a different composition, with a significantly greater
concentration of acid-insoluble components, consistent with its origins as a mixed,
reworked sediment. Samples of the <2 mm fraction of the overlying oolitic fine
sand and the modern soil follow a consistent trend up-profile of decreasing carbon
ate and increasing acid-insoluble material composed of quartz plus aluminosilicate
clays. As this change represents the evolution of the profile through time, it indi
cates that the terrestrial environment has been progressively depleted in calcareous
source material.
Mineralogical data
X-ray diffraction studies of <2 mm and gravel fractions, as well as calerete crusts,
indicate that the mineral assemblage is dominated by calcite, quartz and dolomite.
Aragonite is a minor to trace component in the lower parts of the profile and
Calerete
á \
calcite
CaO
100
calerete pan . cemented oolitic sands
politic sands
lithoclast gravels
cemented oolitic sand gravel oolitic fine sand
with lithoclasts
^ nodule concentrate
Mi 1221
dolomite
MÌI222 <« cemented oolitic sand llthoclasts surface
50 ^J/ soil
MÌI223 calerete pan mantle
MÌI224 " * ^ # cemented
►""" oolitic sand
oolitic sand
-I L _L
50 -I
10 20 30 40 -L· L· 80 -I
60 70
I
90 100 CaasCaCOs 10
° 100
1 2 3 4 5 6 7 8 9 10 Mg as MgC03 MgO acid-insolubles
o/_ Ï Ë É ^ Ë É , , Ê É « magnesite quartz + clays
% acid soluble
Fig. 13.9. a. Plot of acid-soluble Ca (as CaC03) and Mg (as MgC03) concentrations with depth in the Poochera quarry calerete profile, b. Triangular
composition diagram for chemical analyses of samples collected through the profile.
323
330 A.R. Milnes
reflects the presence of the residual marine microfossil fragments. TYace amounts of
feldspar also occur.
Variations in mineralogical composition parallel the chemical data summarised
in Figure 13.9. Dolomite is conspicuous in samples of the oolitic sand at the base of
the profile, and is relatively more abundant in the variably cemented zone (Figure
13.10) immediately below the calerete crust. The crust itself also contains significant
dolomite (Figure 13.9). In those samples where the unconsolidated sediment (<2
mm) could be analysed separately from the gravel, mineralogical data indicate that
oolitic grainstone clasts contain more.
Piednippie
Sections examined at other localities have provided similar data to that obtained
at Poochera. A quarry site at Piednippie, also adjacent to Highway 1, has exposed a
somewhat thicker sequence of oolitic sands and calerete pans (Figure 13.11).
Samples collected systematically through the sequence consistently contain 70-
80% CaCC>3 (Figure 13.12a). However, MgCC>3 concentrations progressively de
crease up profile from the dune encountered at 140-150 cm. As in the Poochera
profile, calerete pans marking the tops of oolitic sand horizons at 80-90 cm and
25-30 cm are enriched in CaC03 relative to MgC03, consistent with cementation
Fig. 13.10. SEM image of cemented horizon overlying basal oolitic dune sand, Poochera quarry calerete
profile. Ooliths (left and right) cemented by micritic calcite (A) associated with abundant calcified
fossils of soil micro-organisms (arrows), and residual void and channel cutans (B) of very fine-grained,
precipitated dolomite (sample ÌÐ224). Bar scale 40 ìðé.
Calerete 331
/►ll^^^^l
Fig. 13.11. View of profile in Piednippie quarry showing sequence of oolitic sands and calerete pans.
Hammer (centre) 33 cm long.
by secondary calcite. Although the variation in chemical composition of samples

from the Piednippie profile is considerably less than at Poochera (Figure 13.12b),
the basal dune is similar in composition. Moreover, hardened zones are enriched
in Mg, suggesting cementation by secondary dolomite. The lower calerete crust at
85 cm is enriched in CaO and depleted in MgO. The oolitic sand horizon and its
calerete crust between 25-80 cm depth have a very narrow composition range and,
together with the soil mantle at the top of the profile, contain progressively more
acid-insoluble material (quartz and aluminosilicate clay), as shown infigure13.12b.
This is the same pattern as in the upper horizon of the Poochera section.
Calcretes of the inland areas

Calerete profiles have been studied in detail at several localities, for example
in the western part of the Murray Basin at Bakara Well (Wilson, 1981) and along
the western margin of Fleurieu Peninsula (Phillips and Milnes, 1988). In many
cases the profiles overlie Quaternary non-marine alluvial and lacustrine sediments,
but some occur on bedrock subcrop which includes granites. The major difference
between these substrates and those of the Neogene coasts is that the former are
non-calcareous.
Caliph
Examples of calerete profiles in this locality, in the western part of the Murray
Basin, were studied in shallow roadside quarries. They consist, typically, of thick
334
Mi 1233 - calerete pan
50
unconsolidated and
cemented oolitic dune sand
M11239 - calerete pan
lower calerete
100 r—
upper calerete
<2mm rnr |#
rocK
fraction surface soil
E
o
« Ca ® Ca mantle
• Mg Q Mg
150 l·-
o.
8
Ö ö MÌI246 - cemented oolitic
dune sand
200 h-
MÜ247 - cemented oolitic
dune sand
100 100
250 MgO acid-in solubles
magnesite quartz □ clays
MH249 - cemented oolitic

dune sand
300 r--l I I I I
io 20 30 40 so 60 70 80 90 loo % acid soluble Ca as CaC0 3
1 2 3 4 5 e 7 e 9 i o n 12 13 % acid soluble Mg as MgC0 3
A.R. Milnes
Fig. 13.12. a. Plot of acid-soluble Ca (as CaC0 3 ) and Mg (as MgC0 3 ) concentrations with depth in the Piednippie quarry calerete profile, b. Triangular
composition diagram for chemical analyses of samples collected through the profile. Material (quartz and aluminosilicate clay), as shown in Figure 13.12b.
This is the same pattern as in the upper horizons of the Poochera section.
Calerete 333
Profile CAQ1 Profile CAQ2
lithoclast-rich soil mantle

laminar calerete j+ v
hardpan calerete soil mantle
laminar calerete
brecciated
hardpan calerete
laminar-structured horizon
laminar-structured
carbonate-mottled horizon carbonate-mottled horizon
Fig. 13.13. a. View of calerete profile in roadside quarry near Caliph, western Murray Basin. Thin soil
with abundant lithoclasts overlies karst surface in partly brecciated hardpan calerete. Carbonate-mottled
zone grading down into carbonate-free clayey sand at the base. b. Sketches of CAQl and CAQ2 profiles
from which samples were collected, showing details of macromorphological features. Scale bars in 10
cm intervals.
hardpans overlying carbonate-mottled clayey sands (Figure 13.13). The hardpans

display a complex assemblage of forms. Laminar calerete on the uppermost surface
of the hardpan is draped over a karst surface in which hollows are filled by a variety
of lithoclasts in a matrix of contemporary soil material. The clasts range from large
angular fragments derived by brecciation of surface parts of the hardpan, to centi-
334 A.R. Milnes
metric rounded pisoliths. Examination of the hardpan reveals horizons consisting of

a similar assemblage of materials, now completely indurated by secondary carbonate
minerals and overlying former karst surfaces marked by laminar calcite horizons.
Beneath the hardpan is a partly consolidated, laminar-structured, carbonate-rich
horizon which is underlain by a horizon of variably cemented carbonate mottles
and residual patches of reddish-brown clayey sand. This has a gradational lower
boundary with the carbonate-free sediment substrate.
Chemical analyses of representative samples collected systematically through the
profile were used to characterise the gross chemical and mineralogical variations.
At Caliph site 1, the carbonate mineral impregnating and variably cementing the
substrate sediment is exclusively dolomite. The increasing abundance of illuviated
carbonate up-profile to the base of the hardpan complex corresponds to increasing
concentrations of dolomite (Figure 13.14a). Hardened mottles and nodules have
a similar chemical and mineralogical composition to that of unconsolidated mate
rial, consistent with dolomite being responsible for the cementation. Phillips and
Milnes (1988) reported a similar pattern of dolomite distribution in calerete profiles
along the western margin of Fleuleurieu Peninsula. The hardpan horizon, however,
contains increasing concentrations of acid-soluble Ca up-profile, indicating that its
formation has involved the precipitation of calcite from infiltrating bicarbonate-rich
waters. In contrast, the surface soil horizon is almost entirely composed of quartz
and layer-silicate clay minerals.
Mineralogical data indicate that the reddish-brown clayey sand in the carbonate-
mottled horizon of the profile consists dominantly of quartz and layer-silicate clays
(illite, kaolinite and randomly interstratified material) together with dolomite (Fig
ure 13.15a). Calcite is not detected, but minor concentrations of halite occur. As
sociated carbonate-rich zones are substantially enriched in dolomite, but no calcite
is present (Figure 13.15b). Bulk samples of the overlying hardpan are progressively
depleted in dolomite up-profile because of dilution by introduced calcite. Quartz
and layer-silicate clay minerals remain major constituents. Halite is a minor com
ponent of most samples. The laminated zones, for example draping the uppermost
surface of the hardpan, contain only calcite and minor quartz plus clay minerals.
The calerete profile at Caliph site 2 has a similar morphology to that at site 1 (Fig
ure 13.13b). However, chemical data for samples collected systematically through
the profile show that dolomite is not a major component (Figure 13.14b). The trend
up-profile is consistently one of Ca enrichment, due to calcite impregnation of both
the substrate sediment and the complex hardpan calerete capping the profile.
Klein Point
This profile is typical of the upper part of sections exposed in the modern coastal
cliffs along the eastern side of Yorke Peninsula, where calcretes and calcareous
sands overlie a thin Pleistocene clay sequence on top of Miocene marine lime
stones.
Calerete
calcite
CaO
100
Caliph calerete profile CAQ1 Caliph calerete profile CAQ2

hardpan calerete &
laminar structured zone
hardpan calerete
lower laminar
dolomite structured zone
<2mm fractions
nodule from upper

mottled zone
nodule from
mottled zone
<2mm fractions
surface
/sal
100 100 100 100

MgO acid-in solubles MgO acid-insolublee
magnesite quartz + clays magnesite quartz + days
Fig. 13.14. Triangular composition diagrams for chemical analyses of samples collected through Caliph calerete profiles, a. CAQl; b. CAQ2.
337
336 A.R. Milnes
Fig. 13.15. a. SEM image of broken surface of fragment of clayey sand zone in sample CAQ1-122,
showing smooth cutan of very fine grained layer-silicate clay and dolomite partly encapsulating larger
anhedral dolomite crystals and crystal clumps. The morphology of the larger crystals suggests physical
translocation down profile. Bar scale 100 /xm. b. SEM image of broken fragment of carbonate-rich zone
in sample CAQ1-122 showing pore spaces filled by paniculate, anhedral crystal clumps of dolomite
that appear to have been physically translocated down-profile. In places the crystal clumps are coated
by smooth deposits of very fine grained precipitated dolomite which appear to be responsible for the
cementation. Bar scale 100 ìðé.
Macromorphological description
The profile was sampled in a roadside quarry. The base of the section is marked
by a hardpan calerete capped by a laminar crust, and this is overlain by a bouldery
horizon containing large calerete lithoclasts evidently derived by physical breakup
of parts of the hardpan (Figure 13.16a). The top of this horizon is marked by a
discontinuous layer of isolated, large, subrounded boulders. Above, a calcareous
fine sand littered with calerete lithoclasts is capped by a calerete pan. At the site
sampled, the calerete consists largely of dislocated plates. Laterally, however, there
are remnants of hardpan with associated boulder-sized lithoclasts. The overlying soil
horizon contains few lithoclasts except in the zone immediately above the calerete
pan.
The variation in acid soluble Ca and Mg (expressed as weight percent CaC03
and weigth percent MgC03, respectively) through the profile is illustrated in Figure
Calerete 337
T—I
surface soil Lh&
-i0
modern soil mantle j ^ T ^ ^ <2mm fraction

50 h
calerete
calerete pan
50 j ^L^gravels >2mm
calcareous fine
sand with 100 h >,
lithoclasts f.a "<2mm fraction
Mg e
p? ' >2mm gravel

Mg a "
boulder horizon 100
Q. ß? ® rock
Mg®
Q
150 h
boulder horizon
basal calerete
rock
calcite
acid-soluble CaO 200
100
basal calerete
J I
10 20 30 40 50 60 70 80 90 100
basal hardpan calerete
% acid soluble
Ca, Mg as CaC0 3 , MgC0 3
gravels and rock samples
dolomite
50
<2mm soil fractions
surface soil mantle
ιοσ 100
acid-soluble MgO acid-insolubles
magnesite quartz + aluminosilicates
Fig. 13.16 a. Sketch of macromorphological organisation of features in calerete profile, Klein Point,
Yorke Peninsula. Depth scale in cm. b. Plot of acid-soluble Ca (as CaC03) and Mg (as MgCC>3)
concentrations with depth in the profile for < 2 mm, gravel ( > 2 mm) and rock samples, c. Triangular
composition diagram for chemical analyses of samples collected through the profile.
13.16b. These data reveal a progressive decrease in total carbonate up-profile from
values near 80% at depth to concentrations of 40% in the surface soil. The calerete
at the base of the profile is enriched in Ca relative to Mg, consistent with calcite
impregnation. On the other hand, the <2 mm fraction of the overlying horizon
consists largely of dolomite, as indicated by parallel trends in Ca and Mg abun-
338 A.R. Milnes
dance. The composition of hardened lithoclasts comprising the gravel fractions of

samples in this interval suggests that secondary dolomite has been involved in the
cementation processes, as noted in other profiles. In the overlying horizon, above
the boulder layer, the concentration of Ca increases upwards at the expense of Mg,
indicating that calcite is the more abundant carbonate phase. Lithoclasts separated
from the gravel fraction (>2 mm) are consistently enriched in Ca relative to Mg,
indicating that induration has been the result of impregnation by secondary calcite.
The surface soil, however, is comparatively uniform in composition with respect to
Ca and Mg, and the data are consistent with about equal concentrations of calcite
and dolomite.
The various processes involved in the formation of the profile are clearly illus
trated in a triangular composition diagram (Figure 13.16c). Chemical changes in the
<2 mm material indicate that there is a somewhat irregular but consistent increase
in calcite and acid-insoluble phases up-profile and that the abundance of dolomite
decreases. Exactly the same pattern is revealed in samples of the gravel fraction,
except that these samples are consistently enriched in carbonate. The increase in
calcite and acid-soluble phases is interpreted to be predominantly a property of the
primary sediment. Impregnation of the primary sediment by carbonate minerals to
produce nodules is one of the processes of calerete formation. In contrast with the
calcareous sand horizon, the surface soil has a restricted composition range and
contains high concentrations of acid-insoluble minerals and dolomite, indicating a
different provenance.
Mineralogical data
In terms of the mineralogy of the carbonate minerals, X-ray powder diffraction
analyses confirm the interpretations made from chemical data. Samples towards the
base of the profile in the bouldery horizon are dolomite-bearing and contain only
minor to trace amounts of calcite. Both dolomite and calcite are more abundant
in nodules or lithoclasts than in the unconsolidated materials and the levels of
quartz and layer-silicate clay minerals are consistently lower. In the upper calerete
pan, dolomite is a minor constituent compared with calcite and quartz, but in the
overlying soil, dolomite is relatively more abundant.
Layer silicate clay minerals identified in the <2 ìôç fraction of acid-treated sam
ples include kaolinite, illite and randomly interstratified material. Sepiolite occurs
in abundance in the lower part of the profile, below the boulder horizon.
Development of the profiles
The data reported for examples of calcretes in South Australia provides impor
tant clues to processes involved in their formation, although there are many details
which still require further scrutiny.
Calerete 339
Province of the Neogene coasts

Down-profile analysis of acid-soluble Ca and Mg concentrations in representa
tive, though heterogeneous, whole samples of calcretes and associated materials
reveals systematic changes both in the chemical composition of the carbonate frac
tion and in the proportions of carbonate and acid-insoluble components. Parent
materials in this province are eolian calcareous sands. Towards the base of the sec
tions, these are typically bioclastic sands formed of rounded fragments of shelly
marine fossils. Higher in the sequence, the sands are oolitic and finer grained. The
ooliths consist of finely laminated micrite coatings around fossil fragments, as well
as quartz and feldspar grains. These changes suggest that the landscapes became
gradually more remote from coastal marine environments which were the source of
bioclastic debris, through regression of the seas. Terrestrial processes, particularly
eolian activity interspersed with periods of pedogenesis, reworked the bioclastic
sands to produce the oolitic sand deposits. The physical environment was probably
like that described by Read (1974). Erosion on dune crests and progressive creep
of lithoclasts and soil materials down duneflanks,together with bioturbation, also
produced mixed sediments. Chemical and mineralogical data indicate a progressive
and systematic decrease in total carbonate content and an increase in the abundance
of acid-insoluble material (quartz, layer-silicate clays and feldspars) up-profile, thus
demonstrating the increasing terrestrial influence and the abundance of weathered
detritus in the landscapes.
Many processes have influenced the formation of the calcretes. In the lower parts
of the sections, zones of indurated bioclastic sands contain higher concentrations
of dolomite than the unconsolidated sands, indicating that dolomite is involved in
the cementation. In fact, SEM images show that intergranular pores are blocked
by subhedral microcrystals of calcite covered by thin, smooth cutans of dolomite
in which the crystals are too small to be resolved (Figure 13.10). The observations
suggest a combination of illuvial translocation of calcite and later precipitation of
dolomite from infiltrating solutions. By way of contrast, calcretes overlying these
dunes and capping oolitic sand deposits higher in the sequence are commonly hori
zons of mixed sedimentary and soil origin cemented by calcite. Investigations have
shown that the calcretes and the underlying sediment contain abundant evidence
of former soil environments in the form of rhizoliths and an extensive array of fos
silised micro-organisms including fungi and bacteria. Cementation of the horizon
is due to the accumulation of micritic calcite in pore spaces between ooliths. In
the same sense as proposed initially by Gilè et al. (1966) filling of the pore spaces
by micritic calcite progressively seals the horizon against water infiltration. Again,
it seems likely that at least some of the calcite has been illuviated from overlying
soil layers and some precipitated from solution. Thin section and SEM observations
indicate that residual pores are sites where fungi and other micro-organisms prefer
entially accumulate, presumably due to moist conditions. These are also sites where
the organic structures are calcified.
340 A.R. Milnes
As also pointed out by Gilè and his colleagues, the formation of laminar calerete
succeeds the eventual plugging of the horizons. However, in all of the calcretes stud
ied, there is clear evidence for dissolution events which have produced karst features
that are coated, draped or infilled by these laminated deposits of calcite. Some of
the dissolution features appear to have been initiated following physical breakup
of the calerete horizon by plant roots. Close examination of the laminations, which
were probably formed by evaporation of waterfilms,indicates a complex of associ
ated processes. For example,finesilt-size quartz grains throughout the laminae, and
sometimes concentrated in particular laminae, signify illuviation processes (Figure
13.17a). In addition, dark grey laminae are so colored because of the concentration
of fossil fungal filaments and other micro-organisms (Figure 13.17b, c), and their
abundance suggests a proliferation of organic activity during moist periods. These
processes are considered to have occurred on exposed surfaces in the manner pro
posed by Klappa (1979a) rather than beneath a soil mantle as suggested by Gilè et
al. (1966).
The source of the calcite for continued accession to the profile for induration and
the production of the laminated calerete crust was almost certainly eolian sediment.
In fact, the superposition of oolitic sands with other sediments and calcretes in the
sequences indicates an accreting sedimentary environment in which times of hiatus
corresponded with erosion and reworking of sediments, pedogenesis and calerete
formation. It is possible that the stage of laminar crust formation, on exposed
landsurfaces, constituted the final phase in a cycle which is reset by the deposition
of another layer of sediment. The contemporary soil mantle in the sites studied is
still comparatively rich in calcite (Figures 13.9 and 13.12). Dissolution of calcite and
illuviation offinesediment are therefore expected to be continuing processes during
rainfall and infiltration events. It is in this sort of soil environment that the oolith
structures that characterise the underlying sediments are thought to have formed,
specifically by evaporation of waterfilmson skeleton grains to produce thin calcitic
laminae. Some laminae preferentially contain concentrations of layer-silicate clays
and fine quartz particles (Figure 13.18) that are clearly of illuvial origin. Others
have accumulations of fungalfilamentsand other micro-organisms, like organic-rich
laminae in the laminar calcretes, and are grey colored as a consequence. These
features indicate the strong surface tension forces that have acted in the water films
as they retract around the skeleton grains due to evaporation.
The inland province

The situation in the inland areas is somewhat different in detail. The substrate
materials here are mainly non-calcareous and so the origin of what Phillips and
Milnes (1988) called the carbonate blanket has to be eolian. There is no evidence
for the lateral migration of carbonate in a catenary sense, as proposed by Ruellan
(1971). The source of the carbonate is less easy to specify, but a distant source in the
Neogene coastal regions is likely, as suggested initially by Crocker (1946) and var-
Calerete 341
Fig. 13.17. a. Photomicrograph of laminar calerete crust on sample ÌÐ297 of upper calerete pan,
Poochera quarry profile, showing alternation of light and dark colored (organic) laminae and scattered
silt-size quartz grains (white). Width of field 6 mm. Transmitted plane polarised light, b. SEM image of
broken surface of organic lamina in same sample, showing cross-sections of calcified fungal filaments
(arrows) embedded in micritic calcite matrix. Bar scale 10 ìÀç. e. SEM image of organic lamina showing
alveolar structure formed of calcitic rods embedded in micritic calcite matrix. Bar scale 4 ìðé.
342 A.R. Milnes
ious local sources in the form of continental playas containing dolomitic sediments
also contributed. In these examples, the deposited sediment was essentially loess.
Carbonate in the unconsolidated substrate sediment in some profiles in this re
gion is exclusively dolomite. Chemical and mineralogical data for these profiles
indicate that there is a progressive enrichment of dolomite and a depletion of
acid-soluble quartz and layer-silicate clay minerals up-profile (Figure 13.14). This
variation is due in part to the changing composition of the primary sediment. SEM
observations show that the sediment contains pelletai structures, like ooliths, that
are typical of the eolian sediments in the coastal province. SEM studies of mottled
zones, where dolomite has infiltrated non-calcareous substrate sediments (Figure
13.15), suggest that crystallites and crystal aggregates of dolomite were physically
translocated from overlying horizons. However, cementation within mottles seems
to have been effected by the precipitation of veryfinelycrystalline dolomite from so
lution. The chemical environment which favors the dissolution and re-precipitation
of dolomite in contrast to calcite is presently being investigated.
As in the coastal province, hardened nodules and calerete pans are invariably
enriched in calcite relative to dolomite and acid-insoluble phases. THangular chem
ical composition diagrams outline these distinctive patterns, which are the reverse
of trends obtained for profiles in the coastal province. Chemical data for the Klein
Point profile follow a complex pattern which reflects the progressive addition of
calcite and quartz plus layer-silicate clays through re-working of the sediment, over
printed by the influx of secondary calcite to produce indurated nodules and the
Fig. 13.18. X-ray scanning images showing the distribution of Si (a) and Al (b) reflecting the presence of
quartz and layer-silicate clay mineral particles in the laminar coatings around an oolith. Sample M138.
Bar scale 45 ìðé.
Calerete 343
uppermost calerete pan. The modem soil mantle is essentially carbonate free and
contrasts with the remainder of the profile. Other features, including the occur-
rence and distribution of rhizoliths and fossilised micro-organisms, the character of
pisoliths and larger clasts coated by laminated deposits of calcite, and the develop-
ment of laminar calerete zones, are also common to both provinces and allude to
geomorphic environments like that described by Read (1974).
Discussion
A great deal of work on calcretes throughout the world has provided the ba-
sis for an interpretation of processes in the broad sense. The suggestions of Gilè
and his colleagues (Gilè, 1961; Gilè et al., 1966), Crocker (1946) and Reeves
(1970) concerning illuviation processes, eolian source materials and aggrading con-
ditions in desertic landscapes, have received substantial support. Read (1974) and
Arakel (1982) perceived the intricacies of geomorphic and hydrologie environments;
Klappa (1979, 1980) and others (Philips et al., 1987) have clearly identified bi-
ological influences in calerete formation. The variation in calerete facies within
landscapes is widely accepted, if not fully understood, and there is now consider-
able information about the vertical chemical and mineralogical variation within cal-
erete profiles (Milnes and Hutton, 1983; Phillips and Milnes, 1988). Klappa (1983)
produced an enlightened model of calerete formation, drawing together facts and
possibilities from a wide-ranging review of the subject. He proposed the term pe-
dodiagenesis to illustrate the complex of pedogenic and diagenetic processes that
have acted to accumulate calcium carbonate in soils and to lithify the soil to form
various calerete facies. Mechanisms of calcite precipitation were reviewed and a
generalised model was proposed to account for the development of calcretes. Stages
in the model included preparation of the host material, soil formation, accumula-
tion of calcium carbonate, development of the profile, induration and reworking.
But, as Klappa (1983) admitted, there are many details yet to be investigated. Some
relevant questions are raised in the descriptions in this chapter of South Australian
calcretes.
The traditional concepts of pedogenic carbonate illuviation imply infiltration of
bicarbonate-rich solutions and precipitation of calcite at depth. However, there is
considerable evidence of illuviation of finely paniculate material, including quartz,
layer-silicate clays and carbonate minerals. The intimate association of calcified
micro-organisms with micritic carbonate minerals in calcretes records a significant
biological influence. The mechanisms of calcification of the biological structures
are unknown and relationships with the micritic cement, which accounts for the
progressive plugging of illuvial horizons, are not clear. The origin of dolomite and
the manner in which it infiltrates and cements substrate sediments, are of consider-
able interest in the context of the formation of calcretes in Australia. Rather than
344 A.R. Milnes
comparatively simple features in which secondary calcite has precipitated from in-
filtrating solutions and cemented a subsoil horizon, many calcretes are a complex
assemblage of reworked sediment and soil materials, impregnated by calcite and/or
dolomite, interrupted by karstic dissolution features and capped by laminar calcite
deposits. The time frame and pattern of landscape development reflected in these
materials is clearly extensive. Many cycles of change are implied, but our ability
to unravel the pattern through the available dating techniques is currently limited
and further investigations are warranted. There is controversy about the preser-
vation of the powdery calerete horizons beneath hardpans. For example, Reeves
(1970) suggested that comparatively rapid plugging and inhibition of infiltration
of near surface horizons has effectively preserved the unconsolidated character of
the parent sediment. Esteban and Klappa (1983) on the other hand, stated that
the so-called chalky horizon was maintained despite illuviation and precipitation of
secondary calcite because cementation was precluded by physical, chemical and bio-
logical disturbance. There is clearly further work to be undertaken in this context.
Attention to the changes in structure and fabric, chemistry, mineralogy, and the
identification of biological features in profiles in which calcretes occur now provides
considerable data with which to refine our understanding of processes. Field rela-
tionships are being interpreted in a reasonable fashion. Laboratory investigations
have progressed well. It now remains for us to embark on studies of detail using
techniques such as scanning electron microscopy and electron probe microanalysis
to map the associations of the secondary carbonate phases and interpret the pro-
cesses responsible. In addition, there is a clear requirement to design and utilise the
results of experimental studies of carbonate accumulation in soil systems.
References
Amundson, R.G., Chadwick, O.A., Sowers, J.M. and Doner, H.E., 1989. Soil evolution along an altitu-
dinal transect in the eastern Mojave Desert of Nevada, U.S.A. Geoderma, 43: 349-371.
Arakel, A.V., 1982. Genesis of calerete in Quaternary soil profiles, Hutt and Leeman Lagoons, Western
Australia. J. Sediment. Petrol., 52: 109-125.
Aristarain, L.F., 1970. Chemical analyses of caliche profiles from the High Plains, New Mexico. J. Geol.,
78: 201-212.
Bowler, J.M. and Polach, H.A., 1971. Radiocarbon analyses of soil carbonates: an evaluation from
paleosols in southeastern Australia. In: D.H. Yaalon (Editor), Paleopedology. Origin, Nature and
Dating of Paleosols. Isr. Soc. Soil Sci., Israel Universities Press, Jerusalem, pp. 97-108.
Brown, C.N., 1956. The origin of caliche on the northeastern Llano Estacado, Texas. J. Geol., 64: 1-15.
Callen, R.A., Wasson, R.J. and Gillespie, R., 1983. Reliability of radiocarbon dating of pedogenic
carbonate in the Australian arid zone. Sediment. Geol., 35: 1-14.
Callot, G., Guyon, A. and Mousain, D., 1985. Interrelations entre aiguilles de calcite et hyphes
mycéliens. Agronomie, 5: 209-216.
Churchward, H.M., 1963. Soil studies at Swan Hill, Victoria, Australia. III. Some aspects of soil devel
opment on eolian materials. Aust. J. Soil Res., 1: 117-128.
Calerete 345
Crawford, A.R., 1965. The geology of Yorke Peninsula. Geol. Surv. S. Aust. Bull., 39, 96 pp.
Crocker, R.L., 1946. Post-Miocene climatic and geologic history and its significance in relation to the
genesis of the major soil types of South Australia. CSIR Bull., 193, 56 pp.
Dan, J., 1977. The distribution and origin of nari and other lime crusts in Israel. Isr. J. Earth-Sci., 26:
68-83.
Dorn, R, 1988. A rock-varnish interpretation of alluvial-fan development in Death Valley, California.
Nati. Geogr. Res., 4: 56-73.
Esteban, M., 1974. Caliche textures and Microcodium. Soc. Geol. Ital. Bull., 92: 105-125.
Esteban, M., 1976. Vadose pisolite and caliche. Bull. Am. Assoc. Pet. Geol., 60: 2048-2057.
Esteban, M. and Klappa, CF., 1983. Subaerial exposure environment. In: P.A. Scholle, D.G. Bebout
and CH. Moore (Editors), Carbonate Depositional Environments, Am. Assoc. Pet. Geol. Mem., 33:
1-54.
Fairbridge, R.W. and Teichert, C, 1952. Soil horizons and marine bands in the coastal limestones of
Western Australia. J. Proc. R. Soc. N.S.W., 86: 68-87.
Gardner, L.R., 1972. Origin of the Mormon Mesa caliche, Clark County, Nevada. Geol. Soc. Am. Bull.,
83: 143-156.
Gilè, L.H., 1961. A classification of ca horizons in soils of a desert region, Dona Ana County, New
Mexico. Soil Sci. Soc. Am. Proc, 25: 52-61.
Gilè, L.H., Hawley, J.W. and Grossman, R.B., 1981. Soils and geomorphology in the Basin and Range
area of southern New Mexico — Guidebook to the Desert Project. New Mexico Bur. Min. Miner.
Res., Mem. 39, 222 pp.
Gilè, L.H., Peterson, F.F. and Grossman, R.B., 1965. The K horizon: a master soil horizon of carbonate
accumulation. Soil Sci., 99: 74-82.
Gilè, L.H., Peterson, F.F. and Grossman, R.B., 1966. Morphological and genetic sequences of carbonate
accumulation in desert soils. Soil Sci., 101: 347-360.
Gill, E., 1975. Calerete hardpans and rhizomorphs in western Victoria, Australia. Pac. Geol., 9: 1-16.
Goudie, A.S., 1972. On the definition of calerete deposits. Z. Geomorphol., 16: 464-468.
Goudie, A.., 1973. Duricrusts in Tropical and Subtropical Landscapes. Clarendon, Oxford, 174 pp.
Goudie, A.S., 1983. Calerete. In: A.S. Goudie and K. Pye (Editors), Chemical Sediments and Geomor
phology: precipitates and residua in the near-surface environment. Academic Press, London, pp.
93-131.
Hutton, J.T. and Dixon, J.C, 1981. The chemistry and mineralogy of some South Australian calcretes
and associated soft carbonates and their dolomitisation. J. Geol. Soc. Aust., 28: 71-79.
James, N.P., 1972. Holocene and Pleistocene calcareous crust (caliche) profiles: criteria for subaerial
exposure. J. Sediment. Petrol. 42: 817-836.
James, N.P. and Choquette, P.W., 1984. Diagenesis 9. Limestones — The meteoric diagenetic environ
ment. Geosci. Can., 11: 161-194.
Johnson, D.L., 1967. Caliche on the Channel Islands. Calif. Div. Min. Geol. Miner. Inf. Service, 20:
151-158.
Klappa, CF., 1979a. Lichen stromatolites: criterion for subaerial exposure and a mechanism for the
formation of laminar calcretes (caliche). J. Sediment. Petrol., 49: 387-400.
Klappa, CF., 1979b. Calcified filaments in Quaternary calcretes: organo-mineral interactions in the
subaerial vadose environment. J. Sediment. Petrol., 49: 955-968.
Klappa, CF., 1980. Rhizoliths in terrestrial carbonates: classification, recognition, genesis and signifi
cance. Sedimentology, 27: 613-629.
Klappa, CF., 1983. A process-response model for the formation of pedogenic calcretes. In: R.C.L.
Wilson (Editor), Residual Deposits: Surface Related Weathering Processes and Materials. Geol.
Soc. London, Spec. Pubi., Blackwell, Oxford, pp. 211-220.
Knox, G J., 1977. Caliche profile formation, Saldanha Bay (South Africa). Sedimentology, 24: 657-674.
346 A.R. Milnes
Ku, T.L., Bull, W.B., Freeman, S.T. and Knauss, K.G., 1979. Th 2 3 0 -U 2 3 4 dating of pedogenic carbonates
in gravelly desert soils of Vidal Valley, southeastern California. Geol. Soc. Am. Bull., 90: 1063-
1073.
Lamplugh, G.W., 1902. Calerete. Geol. Mag., 9: 575.
Lawrence, C.R., 1975. Murray Basin. In: J.G. Lawrence and J.A. Ferguson (Editors), Geology of Victo
ria. Geol. Soc. Aust., Spec. Pubi., 5: 276-288.
Milnes, A.R. and Hutton, J.T., 1983. Calcretes in Australia. In: Soils: An Australian Viewpoint. CSIRO,
Melbourne/Academic Press, London, pp. 119-162.
Milnes, A.R., Kimber, R.W. and Phillips, S.E., 1987. Studies in calcareous eolian landscapes of southern
Australia. In: Liu Tungsheng (Editor), Aspects of Loess Research. China Ocean Press, Beijing, pp.
130-139.
Milnes, A.R. and Ludbrook, N.H., 1986. Provenance of microfossils in eolian calcarenites and calcretes
in southern South Australia. Aust. J. Earth Sci., 33:145-159.
Netterberg, F., 1967. Some roadmaking properties of South African Calcretes. Proc. 4th Regional
Conf. Soil Mechanics and Foundation Engineering, Cape Town, South Africa, December 1967. A.A.
Balkema, Cape Town, pp. 77-81.
Netterberg, F., 1978. Dating and correlation of calcretes and other pedocretes. Trans Geol. Soc. S. Afr.,
81: 379-391.
Netterberg, F., 1969. The interpretation of some basic calerete types. S. Afr. Archaeol. Bull., 24: 117-
122.
Netterberg, F., 1980. Geology of southern African calcretes: 1. Terminology, description, macrofeatures,
and classification. Trans. Geol. Soc. S. Afr., 83: 255-283.
Northcote, K.H., Hubble, G.D., Isbell, R.F., Thompson, C.H. and Bettenay, E., 1975. A Description of
Australian Soils. CSIRO, Melbourne, Vic, 170 pp.
Phillips, S.E. and Milnes, A.R., 1988. The Pleistocene terrestrial carbonate mantle on the southeastern
margin of the St Vincent Basin, South Australia. Aust. J. Earth Sci., 35: 463-481.
Phillips, S.E., Milnes, A.R. and Foster, R.C., 1987. Calcified filaments: an example of biological influ
ences in the formation of calerete in South Australia. Aust. J. Soil Res., 25: 405-428.
Phillips, S.E. and Self, P.G., 1987. Morphology, crystallography and origin of needle-fibre calcite in
Quaternary pedogenic calcretes of South Australia. Aust. J. Soil Res., 25: 429-444.
Rabenhorst, M.C. and Wilding, L.P., 1986. Pedogenesis on the Edwards Plateau, Texas: III. New model
for the formation of petrocalcic horizons. Soil Sci. Soc. Am. J., 50: 693-699.
Read, J.F, 1974. Calerete deposits and Quaternary sediments, Edel Province, Shark Bay, Western
Australia. In: Evolution and Diagenesis of Quaternary Carbonate Sequences, Shark Bay, Western
Australia. Am. Assoc. Pet. Geol. Mem., 22: 250-282.
Reeves, C.C., 1970. Origin, classification, and geologic history of caliche on the southern High Plains,
Texas and eastern New Mexico. J. Geol., 78: 352-362.
Reeves, C.C., 1976. Caliche. Organisation, Classification, Morphology and Uses. Estacado Books, Lub-
bock, 233 pp.
Ruellan, A., 1971. Les sols à profil calcaire différencié des plaines de la basse Moulouya (Maroc
Oriental). Mém. ORSTOM, 54, 302 pp.
Rutter, N.W, Crawford, R.J. and Hamilton, R.D., 1979. Dating methods of Pleistocene deposits and
their problems, IV Amino acid racemization dating. Geosci. Can., 6: 122-128.
Seisser, WG., 1973. Diagenetically formed ooids and intraclasts in South African calcretes. Sedimentol-
ogy, 20: 539-551.
Sowers, J.M. (Editor), 1988. Geomorphology and Pedology on the Kyle Canyon Alluvial Fan, Southern
Nevada. In: D.L. Weide and M.L. Faber (Editors), Field Trip Guidebook for the 1988 Geological
Society of America Cordilleran Section Meeting in Las Vegas, Nevada. University of Nevada, Las
Vegas, Nev., pp. 137-157.
Calerete 347
Von der Borch, C.C., Lock, D.E. and Schwebel, D., 1985. Ground-water formation of dolomite in the
Coorong region of South Australia. Geology, 3: 283-285.
Warren, J.K., 1983. On pedogenic calerete as it occurs in the vadose zone of Quaternary calcareous
dunes in coastal South Australia. J. Sediment. Petrol., 53: 787-796.
Williams, G.E. and Polach, H.A., 1971. Radiocarbon dating of arid-zone calcareous paleosols. Geol.
Soc. Am. Bull., 82: 3069-3086.
Wilson, A.M., 1981. Calerete profiles near Bakara and Bow Hill in the western Murray Basin, South
Australia. Honours Thesis, Univ. Adelaide, Adelaide, S.A. (unpublished).
Yaalon, D.H. and Singer, S., 1974. Vertical variation in strength and porosity of calerete (nari) on chalk,
Shefela, Israel, and interpretation of itys origin. J. Sediment. Petrol., 44: 1016-1023.
Yair, A., Yaalon, D.H. and Singer, S., 1978. Thickness of calerete (nari) on chalk in relation to relief
factors, Shefela, Israel. Abstracts, 10th Int. Congr. on Sedimentology, Jerusalem, 1978, pp. 754-755.
349
Chapter 14
Silcretes
A.R. MILNES and M. THIRY
Introduction
"Silcretes" are rocks produced by the cementation of unconsolidated regolith

materials (sediments, saprolite, soils) by various forms of secondary silica, includ-
ing opal, cryptocrystalline quartz or well-crystallised quartz. The term "silcrete" is
widely used to refer to silicified materials of different structure and fabric, irrespec-
tive of their origin. However, supergene conditions are usually implied, particularly
processes above the phreatic zone or in environments associated with soil formation.
Silicification in both groundwater zones and soil environments can be regarded as
a form of diagenesis, according to the definition of Gary et al. (1973). Silcretes are
widely recognised features of landscapes in Australia, southern Africa, and in parts
of western Europe (Goudie, 1973; Summerfield, 1983a). They are usually classed
as duricrusts (Woolnough, 1927), along with materials that have been cemented by
relatively soluble cations deposited from solution as oxyhydroxides, carbonates and
sulfates in response to chemical reactions and physical conditions in near-surface
terrestrial environments (Goudie, 1973). Silcretes are also known in parts of North
and South America (Goudie, 1973; Summerfield, 1983a), northern Africa (Conrad,
1969) and the Arabian Gulf (Kahalf, 1988).
Silcretes have important associations with paleosurfaces (Tvidale et al., 1970;
Grandin and Thiry, 1983), although this has not always been appreciated. Contem-
porary materials resembling silcretes have not been recognised. Although there is
evidence of cementation of materials by silica in contemporary soil environments,
for example the duripans of North America (Flach et al., 1969; Chadwick et al.,
1987a, b), dorbanks of southern Africa (Ellis and Schloms, 1982) and red-brown
hardpans of Australia (Brewer et al., 1972; Chartres, 1985), these materials them-
selves usually occur on remnants of former landscapes. Therefore, by implication,
silcrete duricrusts are probably formed over time spans that are considerably greater
than the lifespan of individual soils, each of which record specific climatic, geomor-
phic, biological and hydrological environments. In addition, the development of
350 A.R. Milnes and Ai. Thiry
silcretes must require very stable geomorphic and geologic conditions to achieve the
extent of silicification and the observed relationship with paleosurfaces.
Studies of siliceous duricrusts over the years have referred to aspects such as
field distribution and petrology, geomorphological patterns and relationship to pale-
olandscapes, and mineralogical-geochemical characteristics. Silcretes display a wide
variety of forms; detailed investigations of these from the field scale to resolutions
ranging up to that available in scanning electron microscopes can provide clues to
their mode of origin. This approach has not been widely explored, for it is only in
the last decade that Thiry and his colleagues (Thiry, 1981) have made some sense
of the very complex textures and fabrics. The marriage of such studies of natural
materials (Arbey, 1980; Darragh et al., 1976) with data from experimental silica
systems (Her, 1979; Delmas et al., 1982; Williams et al., 1985; Williams and Crerar,
1985) has greatly facilitated interpretations of processes involved in the deposition
of secondary silica, and assessments of the environments in which such processes
were active. Current work of this type includes the studies of Thiry and Milnes (in
press) on silicification in paleoenvironments and the investigations of Chadwick et
al. (1987a, b) of contemporary pedogenic environments.
This chapter focusses on the most recent concepts relating to the formation of
silcretes, and is essentially a review of work in progress. It takes the view that silicifi-
cation is one of a complex of diagenetic changes that result in the conversion of soil
materials and sediments to rock. Relationships between processes involved in the
origin, transport and deposition of silica are discussed, environments of silicification
are suggested and research directions where further information should be sought
are identified.
The nature of silcretes
The discussion in this chapter deals specifically with silicified clastic materials.
Silicification of chemical sediments such as limestones is an important related phe-
nomenon observed in paleodrainage systems (Conrad, 1969; Khalaf, 1988; Arakel et
al., 1989), but is beyond the scope of this chapter.
An extensive literature contains very adequate reviews of observations and data
on the characteristics of silicified clastic sediments. In the main, studies appear to
date from the papers of Lamplugh (1902, 1907), who coined the term "silcrete"
to describe examples of silicified sediments in England and central Africa. South
African silcretes are well known from the work of Frankel and Kent (1937), Frankel
(1952), Mountain (1952), Smale (1973) and Summerfield (1982,1983a, b, c). French
examples have been described by (Thiiy, 1978; 1981; Thiry et al., 1983a, b). Lang-
ford-Smith (1978) summarised studies of Australian silcretes, and subsequent over-
views are given by Milnes and TWidale (1983), TWidale and Milnes (1983), Callen
(1983), Young (1985) and Wells and Callen (1986).
Silcretes 351
Descriptive terms like massive, columnar, nodular, bulbous, mammillated and

glaebular have been used for South African silcretes (Frankel, 1952; Summerfield,
1982, 1983b, c) and multiple silcretes have been recognised in some profiles (Sum-
merfield, 1983a). Australian examples have been described as massive, nodular,
ragged, vesicular, ropy, porcellanitic, concretionary, columnar, conglomeratic, platy,
brecciated and pillow-like (Stephens, 1971; Wopfner, 1978). In addition, there have
been observations on the relationships of the silicified materials with weathered pro-
files and recognition of the effects of tectonics on the disposition of silcrete horizons
(Stephens, 1971; Senior, 1978, Wopfner, 1978; Milnes and TWidale, 1983), as well as
studies attempting to tie silcretes into a stratigraphie framework and therefore into
ages and environmental conditions of formation (Jessup and Norris, 1971; Callen,
1983; Wells and Callen, 1986).
The chemical and mineralogical compositions and pétrographie character of sil-
cretes have been widely reported (Williamson, 1957; Hutton et al., 1972; Smale,
1973; Watts, 1977,1978; Summerfield and Whalley, 1980; Summerfield, 1982,1983b,
c; Milnes and TWidale, 1983; Isaac, 1983). Attempts to explain common features
such as the high concentrations of titania (anatase) and the distribution of various
silica polymorphs have provided additional input to hypotheses about the mecha-
nisms and environments of silicification. Although all of this work has produced data
to tantalise those of us working in the field to speculate further, it turns out with
hindsight that the manner in which most of us have made observations and collected
samples in the field were not sufficiently rigorous to provide definitive information.
Such information was not available until the detailed work of Thiry (1978,1981) on
silcretes in the Paris Basin.
The French experience
By careful and detailed sampling of sections through silicified zones, particularly

in quarries, Thiiy (1978, 1981) recognised the fact that silcretes which have a com-
plex macromorphological organisation nevertheless exhibit a systematic distribution
of micromorphological and mineralogical features. On the other hand, there are
silcretes such as those that occur as multiple lenses in the Fontainebleau Sand of
the Paris Basin that are comparatively simple in terms of macromorphology and
micromorphology.
Silcretes with complex structure andfabrìc

A typical example at Montigny-Lencoup (Figure 14.1) was formed by silicifica-
tion of Early Eocene sandy clays and is overlain by Late Eocene limestone (Thiry,
1978). The profile is 2.25 m thick and can be divided into several distinct horizons
(Figure 14.2).
352 A.R. Milnes and M. Thiry
\—f-\ Early Eocene [7ãã| silcrete fTTTI silcrete | ; ► ; | Late Eocene

i=^3 sandy clay ILLÜ columnar Ã° ° °- °l granular * ' ► ' ' limestone
Fig. 14.1. Schematic sketch of field relationships of silcrete in the Montigny-Lencoup quarry. The silcrete
forms lenses about 10 m in length at the top of the Early Eocene kaolinitic sandy clays.
Basal granular horizon

The lower part of this horizon is a facies in which small siliceous granules (2-5
mm diameter) occur throughout a yellow sandy clay. The subspherical to ovoid gran
ules are yellow-coloured, hard and internally structureless. They consist of quartz
grains with amoeboid and embayed borders (marked by concentrations of micro-
crystalline anatase), in a matrix of interpenetrating, amoeboid quartz crystals ( 5 -
10 ìðé) and microcrystalline quartz (1-3 /im) pigmented by anatase. Many of the
granules have oriented clay cutans. TEM and electron diffraction studies of the clay
matrix show that hexagonal-shaped kaolinites have corrosion embayments and are
coated and welded together by a silica gel.
Higher in the profile the granules become progressively larger (2-5 mm) and
are welded into centimetric nodules by brown coloured opal which has replaced
the clay matrix. The abundance of nodules ultimately defines a distinctive nodular
facies (Figure 14.2). The brown colour of the intergranular isotropie opal matrix
is due to pigmentation by very fine titania. A few anisotropie crystals distinguished
at high magnification are identified by electron probe microanalysis as silica, not
clay, indicating complete transformation of the original clay matrix. Micro-nodular
structures (100 ìôç diameter) and domains of very fine laminae are defined by
concentrations of titania pigment. In matrix zones rich in coarse quartz grains,
the laminated domains parallel the walls or bases of voids and indicate illuvial
deposition of opaline silica.
The uppermost part of the horizon is a massive quartzitic facies formed of tabular
blocks or slabs (Figure 14. 2), although the internal granular organisation is still ev
ident. However, the granules have acquired horizontally elongate form. In addition,
the intergranular brown opal matrix is replaced by microcrystalline quartz littered
with fine crystals of anatase. Skeletal quartz grains inherited, from the original
sediment, have embayed borders marked by accumulations of anatase and appear
to grade into the microcrystalline quartz matrix. Thin, discontinuous, horizontal
("wispy") laminae marked by yellow siliceous material are conspicuous, especially
Siìcretes 353
STRUCTURE AUTHIGENIC
HORIZONS
AND FABRIC SILICA
sand
pseudo-breccia
pseudo-nodular
columnar
prismatic
columnar
granular
silty clay i f—i silica

(section I I I coating
Fig. 14.2. Sketch of the macromorphological organisation of the Montigny-Lencoup silcrete profile and
distribution of secondary silica and dissolution structures.
near the tops of the blocks or slabs. They appear to have been formed through
the dislodgement and re-organisation of quartz grains and concentrations of micro-
crystalline titania in numerous irregular voids and channels. In general, there is no
evidence of the crystallisation of silica or the deposition of illuvial materials in the
voids or channels. However, deposits of opaline silica cap the blocks of this facies.
Prismatic columnar horizon

Vertically elongate blocks or columns formed by the intersection of vertical frac
tures and horizontal cracks characterise this horizon (Figures 14.2 and 14.3). The
columns tend to be coated with deposits of yellow-white opaline silica.
The interior of the columns is similar in many respects to the massive unit at
the top of the granular horizon. The matrix is formed of quartz grains with irregu
lar overgrowth apophyses which grade into titania-enriched microcrystalline quartz
(Figure 14.3). "Wispy" laminae are well developed at the base of the columns. They
are actually comprised of horizontally elongated voids and channels floored by thin
354 A.R. Milites and M. Thiry
Fig. 14.3. Pedogenic silcrete of Montigny-Lencoup (Seine-et-Marne, France), a. Columnar horizon.

The columns display a clear prismatic division. Their smooth upper part corresponds to the capping
structures, b. Void in the prismatic columnar horizon. Void is filled with nodules (7), thin particle-
size graded laminae (2) and, towards the top, thinly laminated opal (3). Ceiling of the void is lined
by titania-rich stalactites, c. Matrix of the prismatic columnar silcrete. Skeleton quartz grains have
irregular overgrowth apophyses which grade into the titania-enriched microcrystalline quartz matrix.
Crossed polars, d. Nodular aspect of the base of columns in the prismatic columnar horizon, e. Complex
nodular aspect of the pseudo-breccia horizon. The nodules are titania-rich and are rimmed by titania.
Scale bar is 250 ìðé.
laminated deposits of anatase alternating with fine quartz. The walls of the voids are
frequently lined with finely crystalline deposits of clear quartz. Vertical and horizon
tal fractures delimit prismatic structures within the columns and, where the fractures
are more prominent, for example near major fissures which define the margins of
the columns, pseudo-nodular structure is clearly evident. The pseudo-nodules are
Siìcretes 355
held in a matrix of fine silica. Associated voids havefloorsfilledby deposits of fine,

structureless opal. The horizontal joints display specific organisations. Laminae of
fine silica, alternating with layers of coarser grain size, are accumulated along the
base and are overlain by opal. Voids often remain above the opal and their ceilings
are covered by layered, mammillary deposits of titania similar in form to stalactites
(Figures 14.3 and 14.4). The deposits penetrate between quartz grains, replacing the
microcrystalline matrix. In places, the ceilings have collapsed and fragments of the
stalactites, together with skeleton quartz grains, have accumulated over thefloorsof
underlying voids.
At the base of the columns, a dense network of irregular cracks outlines nodu
lar bodies (Figure 14.3). The wider cracks (up to centimetric size) are filled with
coarse quartz and silicified debris of micro-nodular appearance, set in a matrix of
microcrystalline quartz. This material is size-graded in a vertical sense, and fines
upwards.
The columns themselves are covered by layered deposits consisting of thin lam
inae of isotropie silica alternating with accumulations of coarse skeleton quartz
grains in a matrix of microcrystalline quartz and titania. The basal layers in these
deposits show indications of fragmentation and micro-nodulation due to post-
Fig. 14.4. a. Photomicrograph of dark-coloured titania-rich lining the walls and roof of a former void in
the prismatic columnar horizon of the Montigny-Lencoup silcrete. The void filling is formed of layers
of particulate debris covered by deposits of thinly laminated opal. Residual void spaces are indicated
by arrows. Bar scale is 250 ìðé. b. The stalactite-like structure and associated coatings on the roof
of the former void are seen on the scanning X-ray image (Ti) to be composed of laminae containing
particulate titania.
depositional processes. The cappings are illuvial deposits which effectively seal the
columns and inhibit further infiltration of water.
The major vertical fissures in this horizon are also filled with illuvial material
comprised of thinly laminated deposits of opaline silica and titania alternating with
deposits of coarse debris. In places there is a succession of fissures, with new fissures
formed in material which has filled previous ones. All such features indicate a
complex regime of water percolation, involving some dissolution and removal of
materials, but primarily infilling and plugging of the horizon by various types of
material flushed down from above.
Pseudo-nodular columnar horizon

Macroscopic observations show that the columnar organisation in this horizon
clearly persists from below, but micromorphological analysis indicates that illuvial
deposits filling vertical fissures have been appreciably reorganised. In particular,
laminated deposits of opaline silica are fragmented and nodulated, and the resulting
nodules have developed rims of titania (especially at their base) concomitantly
with progressive conversion of the opal to microcrystalline quartz. Voids around the
nodules contain very little illuvial material, but are often lined by crystalline deposits
of clear automorphic quartz.
Pseudo-nodular structures within the columns tend to be centimetric in size and
subspherical in shape and are rimmed by deposits of beige silica. They are also
highly variable in terms of internal structure and the enclosing matrix is hetero
geneous, ranging from fine grained silica to material rich in quartz grains. Planar
cracks like those at the base of the prismatic columnar horizon are important fea
tures and contain accumulations of coarse clastic material and sporadic deposits of
opal. In the interior of nodules it is possible to recognise that some illuvial deposits
in voids do not conform to a geopetal orientation, showing that the nodules have
suffered displacement or rotation.
Laminated illuvial deposits of relatively coarse material, interspersed with thin
bands of microcrystalline titania with very little matrix, cover the tops of the columns
and large polygenetic nodules within the columns. At this level in the profile there
is a total absence of contemporary fine illuvial material. In fact, the numerous voids,
the nodulation and fragmentation of the matrix and of former laminated illuvial
structures, and the displacement of pseudo-nodules from an original geopetal orien
tation, indicate significant dissolution and removal of material.
Pseudo-breccia horizon
This relatively massive unit is dark grey in colour compared with the beige-yellow
colours of the underlying horizons. The fragmentation and nodulation of material is
very advanced and nodules are complex and polygenetic.
The largest elements are the major blocks or columns capped by very thick (up
to 5 cm) deposits of interstratified fine grained silica and coarser clastic materials
Silcretes 357
(with fragments up to 1 cm or so in size). In the interior of the columns, large

nodules contain smaller nodules formed of granular assemblages of coarse quartz
in a microcrystalline quartz matrix rich in titania; other nodules are fine grained
and almost opaque due to the high concentration of oxides (Figure 14.3). The inter-
nodular matrix is very heterogeneous. Larger nodules frequently have thin, well-
developed coatings of titania along their bases as well as caps of poorly organised,
coarse-grained illuvial material. Many nodules have been rotated or displaced since
the time of formation of their caps.
Irregular voids are abundant and well developed, especially in thefinestmaterial.
Voids are often developed around pseudo-nodules and probably account for dis
placement of the nodule from the orientation in which it formed. Voids are also well
developed along oblique cracks, at the base of large composite pseudo-nodules and
at the base of the major blocks or columns. Illuvial deposits on the floors of voids
are exceptional and wall linings of crystalline quartz are rare.
The features of this horizon indicate that it has suffered considerable losses of
material in an environment of dissolution and eluviation.
Sand horizon
Brown-ochre sand covers silcrete debris which marks the top of the pseudo-
breccia horizon (Figure 14.2). The debris is angular and highly variable in size.
Green plastic clays and sandy clays beneath the sands penetrate into thefissuresof
the pseudo-breccia horizon.
Development of the profile

Thiry (1978, 1981) and Thiry and Millot (1987) argued that the features of the
Montigny-Lencoup silcrete are typical of pedological environments. They recog
nised a specific distribution of secondary silica polymorphs within the profile (Fig
ure 13.2). In particular, opal occurs in the matrix of the lowermost granular horizon
and as finely laminated void and channel cutans in the middle columnar horizon,
but is absent from the upper parts of the profile. Microcrystalline quartz is present
as crystallaria in the granular horizon and is the major component of the matrix
cement and illuviation structures in the middle and upper horizons. Euhedral quartz
occupies voids in the uppermost horizons.
Based on these observations, Thiry and Millot (1987) suggested that the silcrete
profile evolved essentially as a consequence of the downward percolation of soil
water, in a succession of cycles of leaching and precipitation. The transformation
of layer-silicate clays to opal in the lower parts of the silcrete profile resulted from
dissolution of the clays in situ. However, infiltration and percolation of water from
the landsurface where it was able to dissolve quartz, to depths where progressively
higher concentrations of Si and other cations were achieved through evaporation,
produced various forms of secondary silica. Opal was precipitated in voids near the
base of the profile and quartz towards the top. Dissolution channels,fissuresand
voids are better developed towards the top of the profile (Figure 14.2). Dissolution
initiated physical release of skeleton quartz grains and various components of the
matrix, including microgranular titania. The eluviation of this clastic material and its
deposition at depth in association with precipitated silica tended to seal or plug the
lower parts of the profile against further downward infiltration.
At a different scale there is a systematic distribution of microquartz and granular
titania in the basal laminae of illuvial deposits, in contrast to overlying laminae
of biréfringent opal and uppermost laminae of pale brown, transparent opal. This
sequence was interpreted by Thiry and Millot (1987) to indicate recrystallisation of
early-formed opal during the formation of the deposits. This process may account
for the observation that opal is partly replaced by microquartz in the middle parts
of the profile. The preservation of opal in the lower horizons (Figure 14.2) may
indicate that it was effectively protected from percolating solutions which could
facilitate recrystallisation. Micro-stalactitic titania coatings on the ceilings of voids,
and rims of microcrystalline titania on micro-nodules and skeleton quartz grains are
conspicuous features of the silcrete profile. They were interpreted by Thiry (1978,
1981) to result from accumulation of titania through dissolution of silica matrix.
There is certainly good evidence for the development of microgranules of anatase
through transformation of titania-pigmented brown opal to microquartz.
Thiry and Millot (1987) suggested that, in order to explain successive dissolution
and precipitation, environmental conditions at the time of formation of the silcrete
profile were likely to have been dominated by alternating wet and dry periods.
Silcretes with simple structure and fabric
Distribution and form

Silicified zones in the Fontainebleau Sand (Early Oligocene) in the Paris Basin
represent excellent examples of this type of silcrete. As described by Thiry et al.
(1988), the silicified zones occur as flat-lying sandstone lenses (2-8 m thick) at
different levels within the formation. The uppermost silcrete forms a capping on
buttes, plateaux and ridges in the region north-west of Fontainebleau. The silcretes
are also exposed on the margins of plateaux and in valleys, but do not extend more
than a few hundred meters beneath the plateaux. Quarries in the Fontainebleau
Sand provide clear exposures of the nature and form of the silcretes, of which there
are one to four in a depth interval of 10-30 m (Figure 14.5). They are up to 3 m
thick, at variance with primary bedding in the sand, and consist of tightly cemented
quartz sand with a low residual porosity.
The lowest lens has a well-defined, almost planar upper surface, but a broadly-
curved, warped lower surface bordered by friable sandstone. It consists of detrital
quartz grains cemented by secondary quartz overgrowths. The detrital grains are
very clear and display almost no impurities. The overgrowths are subhedral or
sutured and have polygonal contacts. The bordering friable sandstone facies actually
Silcretes 359
LITHOLOGY PETROFABRIC CLAYS

5%
ß
Etampes limestone
contorted Sandstone
white sand with entirely

weathered flint pebbles } / · · ^ v v -^ · I P I
ondulated sandstone
with unweathered flint
pebbles
white sand with flint

N
pebbles showing kaolinite
a weathered rim
massive sandstone
with flint pebbles
\YLsmectite
yellow sand with
Y
un weathered flint 200 ju m illite
pebbles
I 1 void
Fig. 14.5. Schematic section through the Fontainebleau Sand of the Paris Basin showing the distribution
and micromorphological characteristics of silcrete lenses.
Fig. 14.6. Morphological features of the uppermost silcrete lens in the Fontainebleau Sand. Note the
contorted aspects of the silicified bodies (a). Successive rims on the upper silcrete lens: 1 = porous
border, 2 = glassy fringe, 3 = porous quartzite core.
has a variable degree of cementation by overgrowths, resulting in residual pores.

The upper silcrete is characterised by numerous concavities, with vertical struc
tures and pipes 0.1-2 m in diameter (Figure 14.6). In addition, this lens is sys
tematically rimmed by an outer zone of friable sandstone (1-5 cm thick) and an
Fig. 14.7. Petrofabric of the Fontainebleau Sandstone, a. Silcrete from the core of the lenses. Over
growths are subhedral, leaving polygonal voids. Polars crossed, b. Silcrete from the glassy fringe. The
residual pores between overgrowths are filled with microcrystalline quartz. Polars crossed, c. Outer
friable cortex. Overgrowths are restricted to residual bridges. Plane polarised light. Scale bar is 200 ìðé.
underlying zone (2-6 cm) of hard, glassy sandstone which surrounds a core of
porous sandstone. The core facies consists of detrital quartz grains with secondary
overgrowths, as in the lowest lens (Figure 14.7). In the outer friable sandstone, over
growths are corroded as a result of dissolution (Figure 14.7). The underlying glassy
zone generally consists of detrital quartz grains with polygonal overgrowths, but has
residual voids filled with microcrystalline quartz. Thiry and Millot (1987) suggested
that the microquartz was silica derived from dissolution of the outer margins of the
silcrete lenses.
Origin
Thiry et al. (1988) identified intense silica leaching in the upper levels of the
Fontainebleau Sand, as indicated by the dissolution of clay minerals, the corro
sion of detrital quartz grains and the disintegration of flint pebbles. This leaching
has to be related to acidic conditions resulting from oxidation of the primary sand
(Thiry et al., 1989). In addition, they suggested that the warped shape of the upper
most silcrete lenses was the result of dissolution by percolating water, leading to the
degradation of the sandstones into friable and ultimately unconsolidated sand. Thiry
and his colleagues suggested that the location of the silcrete horizons was controlled
by outcrop pattern and the position of water-table levels during the comparatively
recent evolution of the Paris Basin. The disposition of several silcrete horizons in
the scarps of plateaux and valleys indicates that silicification occurred at groundwa-
ter discharge zones which mark the level of successively lower water tables (Figure
14.8). As the water table dropped in relation to continued downcutting of the land
scape, former silcretes were stranded in the vadose zone where they were subjected
Sikretes 361
H^marl500m Î ™ 6
^ ^ Bsilcrete H ground water
Fig. 14.8. Schematic model of successive stages of groundwater silcrete formation in the Fontainebleau
Sand. 1. Dissection of the limestone cover and thefirstsilicification in the groundwater discharge zone.
2. Downcutting, with concomitant lowering of the water table. Former silcretes at the top of the sand
undergo dissolution and a lower level of silicification is established. 3. Repetition of the process, leading
to a thicker leached profile above the water table and a deeper silcrete level. Note that silcrete levels at
a particular level in the sand in different valleys or in the same valley were not formed during the same
episode of silicification. In any section, the upper silcrete is the oldest and the deeper lens the youngest.
to dissolution at their boundaries, releasing silica for deposition at the level of the
new water table.
Relationships of silcretes with paleosurfaces
The Tertiary formations of the Paris Basin display three main types of silcrete
which are linked to various paleosurfaces and geomorphological features (Figure
14.9; Thiry et al., 1983a). The first type, with complex pedogenic structures and
containing abundant titania, is associated with an Early Eocene paleosurface. The
second type, comprising silcretes with simple structure and fabric and of ground-
water origin, is restricted to thick sand formations. The third type, silcretes called
"meulières" with remnant silicified limestone structures, occur on structural lime
stone platforms eroded during the Quaternary. Clearly, a particular region can
have a complex arrangement of silcretes formed during successive episodes of land
scape evolution. The Paris Basin provides a particularly good case study because
the geological framework is well established through investigations of the numerous
quarries and drill-holes.
The silcretes with complex pedogenic structures mark out a wide paleo-pediplain
along the southern margin of the Paris Basin. The pediplain formed by reworking
groundwater silcrete pedogenic silcrete

in Fontainebleau Sand on Eocene paleosurface
Fig. 14.9. Distribution of different silcretes in the Tertiary and Quaternary formations of the Paris Basin.
of soils developed during Cretaceous and Early Tertiary times on basement and
sedimentary formations. Close to the basin, silicification of sandy clays produced
silcretes like that described from Montigny-Lencoup. On the southern edge of the
pediplain, silcrete developed in deposits of clay with flint pebbles, whereas on the
margin of the basin silicification affected conglomerates up to 10 m thick (Figure
14.10; Thiry, 1981). In grabens on the border of the basement, ferruginous kaolinitic
L I T H OL O G Y 1 STRUCTURE and FABRIC"
lacustrine limestone
with dissolution
features
massive glassy
only cappings
conglomerate without
silicification
chalk
|~oT-3 limpid flint pebble |ΓχΤ\] sandy matrix

b&saj illuviations and cappings of
ßüSa Ti enriched microcrystalline quartz
f l l l l l l microcrystalline quartz matrix
¼ï00'ï°ÃÉ microporous flint IIHIHHI void
Fig. 14.10. Sketch of the macromorphological and micromorphological organisation of the silcrete
profile on the Lower Eocene flint conglomerate on the border of the Paris Basin.
Siìcretes 363
paleosols on fault scarps and mottled horizons of graben deposits are silicified
(Thiry and Tbrland, 1985).
Groundwater silcretes occur in the Paris Basin in several Tertiary sand formations
that are interbedded with thick marine and lacustrine limestones. The silcretes crop
out in the scarps of plateaux, where outflow of groundwaters produced bleaching
and leaching of the sands and silicification, especially at near the limestone cover.
Meulières is a cavernous-weathering silcrete formed by weathering of silicified
limestones covering the Beauce and Brie Plateaux (Figure 14.9). It is from this ma
terial that the famous millstones of the Paris Basin were quarried. The meulières
occur as irregular bodies up to 6 m thick, characterised by irregular, silicified cores
dissected by planar and oblique joints (Figure 14.11; Menillet, 1987, 1988). At the
base of the profile, the limestones display karst dissolution features with infillings
of illuvial clays and silts. Higher in the profile, the clays exhibit hydromorphic
features and the residual limestone and flint masses are leached and weathered,
forming micro-porous zones and numerous voids. Abundant joints filled with pal-
lisadic quartz and chalcedony form a complex network ranging from millimeters to
centimeters in scale. At the top of the profile, clays and silica phases have been
leached to produce a characteristic cavernous facies.
soil and limon
cavernous meulière
massive meulière
brown hydromorphic
clays with
residual limestone
and flints
karst dissolutions
Fig. 14.11. Sketch of macromorphological organisation of the meulière profile developed in weathered
limestone materials on plateaux of the Paris Basin.
Australian examples
Southwestern part of the Eromanga Basin
Recent studies of silcretes in the precious opal mining areas of South Australia,
along the southwestern margin of the Eromanga (Great Australian) Basin, have
identified zones of both pedogenic and groundwater silicification in profiles through
the regolith. In a typical example in the Stuart Creek opalfield, near the northern
tip of Lake Tbrrens, deep bulldozer costeans have exposed a section through uncon-
solidated Cretaceous marine sediments and overlying Tfertiary fluviatile sands. Thiry
and Milnes (in press) described a composite profile comprising various groundwater
silcretes consisting of opalite and glassy sandstone at depth, and a nodular to colum-
nar pedogenic silcrete at the top of the section (Figure 14.12). All of the distinctive
features of the Fontainebleau sandstones and the Montigny-Lencoup silcrete of the
Paris Basin respectively are present, although there are variations.
LITHOLOGY MINERAL AUTHIGENIC SILICA
desert loam soil MICROCRYSTALLINE

QUARTZ
pedogenic columnar
silcrete
(OPAL)
variegated hematitic sands
EUHEDRAL
groundwater and
silicified sands MICROCRYSTALLINE
QUARTZ
amoeboid and branching — bg *\Lfl
silicified sand masses
groundwater OPAL
silicified sands*
opal-impregnated CHALCEDONY
Cretaceous shales
Cretaceous shales
Wï77] quartz | | clays Γ.·"-1 opal calcite FCT9°etn,te Ι aanatase

nai
ase
calcite L i l J hematite F (no scale)
Fig. 14.12. Schematic section through the regolith in the Stuart Creek opal field, South Australia,
showing the distribution of silcretes and associated features.
Sikretes 365
Zones of groundwater silicification

Macromorphological features
The Cretaceous sediments near the base of the section at about 10 m depth are
essentially grey shales with rounded quartzite dropstones and contain amoeboid-
shaped lenses of silicified shale (opalite). Similar grey to brown opalite lenses and
pods occur near the top of the Cretaceous sequence at about 9 m. The opalite
masses have distinctive desiccation fracture patterns of more or less concentric
cracks which cause them to easily break into small, blocky fragments.
The base of the Tertiary sequence at about 8.1 m is variably silicified to form a
bluish-coloured, lustrous sandstone. A massive sandstone lens between 6.9-8.0 m
depth consists of lustrous, glassy, vari-coloured grey, beige, red and purple patches
rimmed by porous sandstone (Figure 14.13). While the top of the sandstone lens is
generally planar, the base is broadly warped and mammillated. Bedding is preserved
in the sandstone.
Higher in the profile a massive lens of glassy sandstone between 4.7-5.3 m has
complex internal structures outlined by beige, pink, red and dark purple colours,
surrounded by zones of porous to friable sandstone. It has a highly irregular base,
with downward projections of stalactite-like sandstone structures. Unconsolidated
beige to ochre sand beneath contains numerous silicified tubiform masses, pendants
and nodules which display discordant surface layers of a gel-like form (Figure 14.13).
Micromorphological and mineralogical features

The opalite consists of detrital quartz grains in a clear to beige, or brown to
nearly opaque, opal matrix. Relatively high concentrations of various cations in the
opal matrix, including Al, indicate residual clay minerals. The darker coloured opal
has lower Al but a higher Ti content. The form of the matrix and the conservation
Fig. 14.13. Groundwater silicification at Stuart Creek, a. Irregular zone of purple glassy sandstone
surrounded by friable sandstone. Scale 50 mm. b. Silicified nodules of pendant-like form with discordant
surface gel-like layers.
of all primary sedimentary structures suggest in situ transformation of sedimentary

clays. Some voids in the opalite have a discoidal shape and represent dissolution
of former gypsum crystals. There is always a specific succession of secondary silica
deposits in voids, beginning with botryoidal cutans of clear micro-laminated opal
overlain by quartzine. Lutécite can fill residual voids. In some voids, lussatite forms
regular cutans and is commonly overlain by chalcedony which grades into euhedral
quartz in the centers of voids. The succession is consistent with a recrystallization
sequence.
The lower sandstones in the profile consist of rounded, detrital quartz grains ce
mented by four successive deposits of secondary silica. A variable amount of translu
cent, brown-coloured opal forms the matrix between quartz grains (Figure 14.14). It
can occupy all the intergranular space, or occur as rims and bridges between grains,
and is regarded as an alteration product of an original clay matrix. Void cutans
of lussatite encapsulate the brown opal matrix, and may be overlain by chalcedans
(Figure 14.14). Residual pore spaces can contain small quartz crystals. In the up
per silicified sand and associated tubiform masses and nodules, there is a gradation
from tightly cemented, glassy sandstone without residual pores, to porous sandstone
in which the secondary silica cement forms bridges between detrital quartz grains.
The cement does not contain opal but consists of quartz overgrowths, microquartz
and chalcedony. The matrix in the glassy zones is brown coloured microquartz which
is often intersected by cracksfilledwith micro-laminated cutans of fine quartz, pal-
lisadic quartz and ultimately euhedral quartz. Porous sandstone contains variable
amounts of brown microquartz matrix, in places preserved beneath syntaxial quartz
overgrowths. In some samples, quartz overgrowths have developed to a signifi
cant extent, leaving no residual porosity. In other samples, deposits of micro-zoned
quartz and drusy quartz occur in residual pores. The shapes of both the sandstone
Fig. 14.14. Petrofabric of groundwater silcretes at Stuart Creek, a. Detrital quartz grains (1) cemented
by brown opal (2; resulting from the alteration of original clay matrix), rimmed by a fringe of lussatite
(3); 4 — residual void. Plane polarised light, b. Detrital quartz grain (2) with sparse brown opal matrix
(2), fringe of lussatite (3), and chalcedony (4) in the residual space. Polars crossed. Scale bar is 200 ìðé.
Silcretes 367
lenses and the associated tubiform and nodular sandstone masses are essentially due
to dissolution of quartz overgrowth cements, and are hence regarded as more or less
residual.
Pedogenic silcrete
Macromorphological features
In the uppermost part of the section, between 0.5 and 3.0 m depth, yellow-white
sand is progressively silicified to form the typical nodular, massive to columnar, and
pseudo-breccia horizons (Figure 14.15) of a complex pedogenic silcrete (Thiry and
Milnes, in press).
At the base of the silcrete, white, friable, vertically-elongate nodules and pen
dants of sandstone occur in unconsolidated yellow-white to pink sands. An overlying
nodular horizon has interspersed zones of small centimetric nodules and larger nod
ules typically 0.10 m in diameter. The latter are rounded on top and have cuspate
bases formed of pendants projecting down into the sand. Internally, the nodules are
formed of small, hard, beige quartzitic nodules in a matrix of white, porous sand
stone. The small nodules have fine, laminated silica cappings and yellowish deposits
of silica and titania lining the base of voids.
The massive horizon above is disrupted by vertical fissures to produce a crude
columnar structure. The columns are from 0.1 to 0.3 m in diameter and consist of
beige to grey coloured, porous quartzite. They are capped byfinelylaminated films
1 HORIZONS I STRUCTURE [ MINERAL | TiQ2 1
^quartz ■ clays E 3 opa. ^ ca.cite H ü fë*h \ ™ ^
Fig. 14.15. Schematic section showing details of the macromorphological organisation and mineralogy
of the pedogenic silcrete at the top of the Stuart Creek section.
of yellowish silica and titania (Figure 14.16). In addition, they contain channels and
pipes up to 2 cm in diameter that are partly filled with deposits of lustrous, yellowish
silica and titania.
The uppermost part of the silcrete profile is a pseudo-breccia horizon formed
of angular quartzitic lumps of variable size up to 0.3 m. Most are tabular in shape
because of fracturing along planar joints. They are in part mottled with iron ox
ides and contain lustrous, densely silicified zones. Finely laminated illuvial deposits
of silica and titania are abundant. Some blocks are capped by deposits of similar
material, which include small-scale gravels and display micro-scale cross-bedding.
Burrow-like structures are conspicuous. Brown calcareous and gypseous soil mate
rial, silcrete debris and gypsum rhizoliths fill the fissures between the blocks.
Micromorphological features
Nodules in the basal horizon and similar structures in the underlying sand con
sist of detrital quartz grains cemented by deposits of clear brown opal or lussatite
in intergranular spaces. In zones with a lustrous appearance, chalcedony fills the
residual voids. The major nodules commonly contain smaller nodules bound by lam
inated illuvial deposits of size-graded quartz with variable but high concentrations
of microgranular titania.
At the base of the columnar horizon, quartz grains are cemented by brown opal,
whereas the cement near the top consists of dark coloured opal and microquartz.
The cement is usually restricted to bridges and menisci between grains, thus main
taining a high porosity. Deposits of microquartz, rich in granular anatase occur in
parts of rhizolith- or burrow-like structures. Voids in zones with a glassy appear
ance, contain cutans of brown opal overlain by chalcedony, pallisadic quartz and
ultimately euhedral quartz. These deposits post-date the opal and microquartz in
the matrix of the columnar silcrete.
Fig. 14.16. Pedogenic silcrete at Stuart Creek, a. Network of channels partly filled with illuviation
deposits of lustrous silica and titania. b. Laminated capping of silica and titania above a column. Coin
20 mm in diameter.
Silcretes 369
The pseudo-breccia horizon is essentially formed of quartz grains in a micro-

quartz matrix. Successive and variable illuviation and capping structures, which are
distinguished from the matrix by grain size and the nature and distribution of the
microquartz and titania cement, are distinctive features.
Formation of the silcretes

In contrast with the French examples, two types of silcrete occur in the one
section in the Stuart Creek opalfield. The particular pattern of distribution, namely
pedogenic silcrete near the landsurface and one or more horizons of groundwater
silicification at depth, is repeated throughout the opalfields on the southwestern
margin of the Eromanga Basin and extends into the Alice Springs region of central
Australia (Milnes et al., in press).
The pedogenic silcrete is considered to have formed under the direct influence
of processes that fashioned the landscape. Major re-structuring events have ensured
that original sedimentary structures and fabrics are largely obliterated. Dissolution
features, illuvial deposits of clastic materials and precipitated silica, and void and
channel cutans andfillingsof silica, imply cyclic conditions of infiltration and evapo
ration, as in a seasonally wet and dry climate.
The groundwater silcretes at depth, represented by opalite lenses in the Creta
ceous shales and sandstone pans in the Tfcrtiary sands, developed in environments
that were somewhat insulated from processes active at the landsurface. Primary
sedimentary structures and fabrics are conserved within these silcretes. Epigenesis
of the clay minerals in the matrix of the primary sediments to form opal in situ
probably occurred close to a phreatic surface which corresponded at the time to the
active weathering front. Acid conditions produced in part by oxidation of primary
sulfide minerals are thought to have resulted in the destruction of the clays with loss
of all cations except Si and Ti. Later partial dissolution of the opal matrix provided
circulating solutions with silica which was subsequently precipitated in voids and
channels.
The superposed zones of groundwater silicification in the regolith at Stuart Creek
opalfield and their counterparts in the regolith thoughout this region of Australia
are possibly consequences of the lowering of the water-table following landscape
dissection, as Thiry et al. (1988) discussed in the case of the Fontainebleau sand
stones. Under these conditions, the older silcretes were stranded in the unsaturated
zone where percolating waters caused recrystallisation and dissolution of the earlier
deposits of secondary silica, as well as the precipitation of new deposits. The pedo
genic silcrete may represent a first phase of silicification on the margins of mature
depositional basins in the continental landscape. In this case the complex of macro-
morphological and morphological features probably represents an integration of
processes over a long time period, even extending to the present day. The ground-
water silcretes are likely to be successively younger with depth, their formation
relating to the depression of the water table following dissection of the landscape.
However, there are examples in the opalfields of silcretes with pedogenic features
which appear to be superimposed on silicified materials of groundwater origin. Such
relationships imply that the two types of silcretes may not necessarily have formed
in the time sequence suggested above, and this is one area in which further detailed
investigations are required.
Time and environment

Recent studies of the isotopie composition of alunites in the weathered profiles in
which groundwater silcretes occur (Bird et al., 1989; in press) have produced data on
their age and environment of formation. In turn, this provides an indication of the
age and conditions under which the groundwater silcretes and associated deposits
of precious opal were formed. On the basis of hydrogen isotope composition, Bird
et al. (1989) concluded that alunite from various opalfield sections, including Stuart
Creek, formed in equilibrium with evaporated meteoric waters, probably late in
the weathering process and during the drying out of the profile. Sulfur isotope
values (£34S%o), on the other hand, were considered to be consistent with a major
input from cyclic marine sulfate. K/Ar data indicate that the alunites crystallised
at different times during the Miocene (Bird et al., in press) and, by association,
silicification processes were probably initiated in this time period.
More detailed investigations of the isotopie composition of alunite and secondary
silica phases (including precious opal) from weathered profiles in which silcretes
occur will be potentially very profitable.
Relationships of silcretes with paleogeography
The Beda Valley paleo-toposequence

Extensive dissection of landscapes in the opalfield regions around the margins
of the Eromanga Basin make it difficult to delineate any clear relationship between
silcrete horizons and geomorphic elements that can be assigned to specific parts of
former landscapes. However, studies in the Beda Valley at the southern end of Lake
Tbrrens have identified a general association between near-surface silcretes and a
paleopediment (TWidale et al., 1970; Hutton et al., 1972) which links the scarps of
Precambrian bedrock highlands with the Cenozoic sediments of the Lake Tbrrens
basin (Figure 14.17). More recent observations have shown that the silcretes exhibit
clear macromorphological and micromorphological indicators of a pedogenic origin,
and that they vary in facies depending on location on the paleosurface. In the
scarp-foot zone, for instance, slumped rock masses and boulders of Precambrian
sandstone have distinctive laminated coatings ("skins") of fine clastic debris and
titania-rich microcrystalline quartz on the walls of horizontal and vertical joints.
Downslope, coarsely laminated deposits of similar material form coatings on the
upper surfaces of sandstone boulders. In places, substantial capping structures in
excess of 0.3 m thick occur on very large columns which define a silcrete horizon
Siìcretes
ARKOONA LAKE
PLATEAU TORRENS
PLAIN
j:::::xj Arkoona quartzite [ I T U Woomera shale ßgj^| Tertiary deposits ÐÔÐÔÐÐ silcrete
Fig. 14.17. Schematic view of the Beda pediment showing the distribution of pedogenic silcrete facies.
more than 2 m thick. The lower parts of the pediment are characterised by a nodular
to columnar silcrete facies about 1 m thick.
Processes which controlled the distribution of silcretes on the Beda pediment
are not known; further studies are currently in progress. Detailed topographic in
formation and a more complete characterisation of the various facies may provide
explanations.
The Todd River Plain

Southeast of Alice Springs, in Central Australia, remnants of Tertiary basinal
sediments cover the syncline between the MacDonnell Ranges and the Ooraminna
Ranges (Figure 14.18). Pediments flanking the quartzite crests on the northern
margins of the former basin are formed on colluvial and alluvial sediments, whereas
lacustrine limestones to the south were deposited in the basin. Bedrock is exposed
at the landsurface or is covered by a thin veneer of Pleistocene-Holocene sediments.
The scarp foot zone in the north is covered by a thick armour of pedogenic silcrete
i!a:;:s,;a-i Tertiary sand groundwater gTertiary chalcedonic

fêi™ and silt pedogenic silcrete silcrete 11
º limestone
1 Heavitree Qz 2 Goyder Fm, 3 Mereenie Sandst. 4 Hermannsburg Sandst.
Proterozoic Silurian-Devonian Devonian
Fig. 14.18. Schematic section across the Todd River Plain, southeast of Alice Springs, central Australia,
showing the location of different silcrete fades.
which displays well developed columns and is enriched in titania. At the base of
the scarp foot, 10-30 m beneath the pedogenic silcrete, groundwater silcretes have
developed in the clastic sediments as lenses 10-30 m long and about 1 m thick.
Basinwards, groundwater silcrete occurs as caprocks on mesas and plateaux, and
formed in Devonian and Silurian sandstones. The Tertiary lacustrine limestones are
also silicified and contain irregular, both horizontally-elongated and cross-cutting
lenses of secondary silica.
The details of the processes of silicification in this landscape are not yet known
and investigations, which include detailed geomorphic and topographic mapping,
are continuing. One possibility is that the different silicification phenomena are re
lated to a single stage of landscape evolution. For example, meteoric water percolat
ing through clastic sediments on the basin margins might have become progressively
enriched in silica and produced groundwater outflows basinwards leading to the
silicification of Tertiary sands and limestones, as well as bedrock sandstones.
Discussion
The typical characteristics and pattern of distribution of silcretes in France and

Australia are repeated elsewhere, for example in southern Africa. There are, of
course, many variations. Thiry et al. (1983), Thiry and Türland (1985) and Thiry
and Millot (1987), for example, described a paleo-toposequence of silicified red
paleosols in the northern Massif Central of France. In addition, the Australian
red-brown hardpans (Brewer et al., 1972; Wright, 1983; Milnes et al., in press),
the dorbanks of southern Africa (Ellis and Schloms, 1982) and the duripans of
north America (Flach et al, 1969; Chadwick et al., 1987a, b) are sediments or
soils cemented by opaline silica deposited from percolating solutions. Studies of
the silicification processes responsible for the formation of these materials are still
largely in progress. In the case of pedogenic and groundwater silcretes it is possible
to make general comments about processes, but there are still many questions of
detail that remain unanswered.
Geochemistry of solutions and the sequence of silica deposits

TWo processes of formation of opaline silica are recognised in the silcretes de
scribed. On the one hand, opal has formed by the in situ transformation or replace
ment of sedimentary layer-silicate clays, with general conservation of structure and
fabric. This is characteristic of the groundwater silcretes and the deeper horizons of
the pedogenic silcretes. It implies an acidic environment in which Al and other ma
jor cations are more soluble than Si and are leached from the sediment while Si is
retained as opal. The physical process of transformation is essentially that described
by Roulin et al. (1986). Secondly, opal has been widely precipitated from percolating
Silcretes 373
solutions in voids and channels in the groundwater silcretes. The formation of these
deposits required levels of silica saturation in groundwaters up to that of amorphous
silica (120 ppm at ambient temperatures). Such concentrations of silica may have
been derived, in part, by dissolution of the relatively soluble matrix opal.
In addition to these processes, in the case of the pedogenic silcretes, meteoric
waters infiltrating the regolith appear to have dissolved quartz and other minerals
near the landsurface and carried suspended sediment down through cracks and
fissures to be deposited in channels and voids in the lower parts of the profile. The
initial solutions contained up to 6 ppm silica, the solubility of quartz at ambient
temperatures, but evaporation would have progressively concentrated the silica as
well as other cations. The preferential precipitation of opal in the lower horizons
of the pedogenic silcretes was favoured by highly concentrated silica solutions, high
rates of precipitation and the presence of impurities (Williams and Crerar, 1985).
Solutions with comparatively low concentrations of silica and low levels of impurities
can precipitate quartz (Delmas et al., 1982) and this appears to have been the case
in the upper horizons of the silcretes, nearer the land surface.
Water regime
Conspicuous fissure and void fillings in the pedogenic silcretes indicate to the
reworking andflushingof surface soil materials down into subsoil horizons. Coarse
material is deposited quickly asfissuresand channels narrow, or other barriers to
percolation are encountered. Fine materials travel further, but are eventually cap
tured by blockage of pore throats between framework grains. Upward size grading
in these deposits and the recurrence of interspersed, finely laminated zones of sil
ica and titania, indicate that intermittent but repeated infiltration events gradually
plugged the profile. Cementation due to precipitation of interstitial silica is likely to
have occurred between infiltration events, when saturated parts of the profile dried
out.
TUbiform, pendant and stalactite-like masses that are common in the Australian
sequences are also interpreted to have formed by evaporation of downward per
colating solutions which precipitated interstitial silica. Illuviation structures are not
present and so the solutions were free of suspended load material. However, many
of the structures exhibit dissolution features, suggesting that they are to some extent
residual.
Illuviation features are not present in the groundwater silcretes. Secondary silica
usually occurs in place of a former intergranular clay matrix, as void or channel
cutans of comparatively uniform thickness or as intergranular void fillings. Such
deposits are considered indicative of a saturated, groundwater-type environment.
Because of its comparatively low solubility, the precipitation of quartz relative to
opal in the groundwater silcretes may have required a high waterfluxand therefore
a substantial hydraulic gradient (Thiry et al., 1988). In the pedogenic environment,
however, evaporation of solutions containing about 6 ppm silica may have produced
a concentration factor between 3 and 5 sufficient to precipitate quartz. At the same
time, increasing concentrations of salts could have decreased the solubility of silica
(Thiry and Milnes, in press).
Time frameworks
Pedogenic and groundwater silcretes may or may not relate to one broad pe-
riod of silicification and there is an enormous scope for critical field studies in this
context. The superposition of silcretes in a particular section does not necessarily
indicate different periods of silcrete formation, but rather the relative positions of
the groundwater table at different times during the evolution of the landscape. Pe-
dogenic silcretes are the most critical indicators of environment and contain a wide
variety of features which register proximity to the landsurface. Conditions during the
formation of the pedogenic silcretes were clearly highly dynamic. Thiry and Milnes
(in press) suggested that the Australian pedogenic silcretes might represent the first
examples of silicification on the margins of mature continental depositional basins
of Neogene age. Groundwater silcretes are considered to be younger, having formed
after dissection of the landscapes at periods of hiatus during progressive lowering
of the water tables. Reconnaissance sampling of alunite, which occurs in associa-
tion with the groundwater silcretes and precious opal deposits in the Australian
opalfields, has yielded isotopie data that indicates that the weathering processes re-
sponsible for its formation were active in the Middle to Late Miocene (Bird et al.,
in press). However, more detailed studies will be needed to resolve age differences
between the various groundwater silcrete horizons. In addition, detailed geomor-
phic investigations are required to unravel the spatial and temporal relationships
between the pedogenic and groundwater silcretes.
References
Arakel, A.V., Jacobson, G., Salehi, M. and Hill, CM., 1989. Silicification of calerete in palaeodrainage
basins of the Australian arid zone. Aust. J. Earth Sci., 36: 73-89.
Arbey, F., 1980. Les formes de la silice et l'identification des évaporites dans les formations silicifiées.
Bull. Cent. Rech. Explor. Prod. Elf-Aquitaine, 4: 309-365.
Bird, M.I., Andrew, A.S., Chivas, A R. and Lock, D., 1989. An isotopie study of surficial alunite in
Australia: 1. Hydrogen and sulphur isotopes. Geochim. Cosmochim. Acta, 53: 3223-3237.
Bird, M.I., Chivas, A.R. and McDougall, I., in press. An isotopie study of surficiâl alunite in Australia
2: Potassi um-argon geochronology. Isot. Geosci.
Brewer, R., Bettenay, E and Churchward, H.M., 1972. Some aspects of the origin and development of
the red and brown hardpan soils of Bulloo Downs, Western Australia. Div. of Soils, CSIRO, Tech.
Pap., 13,13 pp.
Callen, R A., 1983. Late Tertiary "grey billy" and the age and origin of surficial silicifications (silcrete)
in South Australia. J. Geol. Soc. Aust., 30: 393-410.
Silcretes 375
Chadwick, O.A., Hendricks, D.M. and Nettleton, W.D., 1987a. Silica in Duric soils: I. A depositional
model. Soil Sci. Soc. Am. J., 51: 975-982.
Chadwick, O.A., Hendricks, D.M. and Nettleton, W.D., 1987b. Silica in Duric soils: II. Mineralogy. Soil
Sci. Soc. Am. J., 51: 982-985.
Chartres, C.J., 1985. A preliminary investigation of hardpan horizons in north-east New South Wales.
Aust. J. Soil Res., 23: 325-337.
Conrad, G., 1969. Dévolution continentale post-hercynienne du Sahar algérien (Saoura, Erg Chech-
Tanezrouft, Ahnet-Mouydir). Centre Rech. Zones Arides, Géologie, No 10, CNRS, Paris, 527 pp.
Darragh, PJ., Gaskin, A.J. and Sanders, J.V, 1976. Opals. Sei. Am., 234: 84-95.
Delmas, A.B., Garcia-Hernandez, J.F. and Pedro, G., 1982. Discussion sur les conditions et les méca
nismes de formation du quartz à 25°C en milieu ouvert. Analyse réactionelle par voie cinétique. Sci.
Géol. Bull., 35: 81-91.
Ellis, E and Schloms, B.H.A., 1982. A note on the Dorbanks (Duripans) of South Africa. In: J.A.
Coetzee and E.M. van Zinderen Bakker (Editors), Palaeoecology of Africa and the Surrounding
Islands. A.A. Balkema, Rotterdam, 15: 149-157.
Flach, K.W., Nettleton, W.D., Gilè, L.H. and Cady, J.G., 1969. Pedocementation: induration by silica,
carbonates, and sesquioxides in the Quaternary. Soil Sci., 107: 442-453.
Frankel, J.J., 1952. Silcrete near Albertinia. S. Afr. J. Sci., 49: 173-182.
Frankel, J.J. and Kent, L.E., 1937. Grahamstown surface quartzites (silcretes). Trans Geol. Soc. S. Afr.,
40: 1-42.
Gary, M., McAfee, R. and Wolf, C.L., 1973. Glossary of Geology. American Geological Institute,
Washington, D.C., 805 pp., plus Appendix, 52 pp.
Goudie, A.S., 1973. Duricrusts in Tropical and Subtropical Landscapes. Clarendon, Oxford, 174 pp.
Grandin, G. and Thiry, M., 1983. Les grande surfaces continentales Tertiaires des régions chaudes
succession des types d'altération. Can. ORSTOM, Sér. Géol., 13: 3-18.
Hutton, J.T, TWidale, C.R., Milnes, A.R. and Rosser, H., 1972. Composition and genesis of silcretes and
silcrete skins from the Beda Valley, southern Arcoona Plateau, South Australia. J. Geol. Soc. Aust.,
19: 31-39.
lier, R.K., 1979. Chemistry of Silica. Wiley-Interscience, New York, N. Y., 866 pp.
Isaac, K.P., 1983. Silica diagenesis of Palaeogene residual deposits in Devon, England. Proc. Geol.
Assoc. London, 94: 181-186.
Jessup, R.W and Norris, R.M., 1971. Cenozoic stratigraphy of the Lake Eyre basin and part of the arid
region to the south. J. Geol. Soc. Aust. 18: 303-331.
Kahalf, F.I., 1988. Petrography and diagenesis of silcrete from Kuwait, Arabian Gulf. J. Sediment.
Petrol., 58: 1014-1022.
Lamplugh, G.W., 1902. Calerete. Geol. Mag., 9: 575.
Lamplugh, G.W, 1907. The geology of the Zambesi Basin around the Batoka Gorge (Rhodesia). Q. J.
Geol. Soc. London, 63: 162-216.
Langford-Smith, T, 1978. A select review of silcrete research in Australia. In: T Langford-Smith (Edi
tor), Silcrete in Australia. University of New England, Annidale, N.S.W., pp. 1-11.
Menillet, F., 1987. Les meulières du Bassin de Paris (France) et les faciès associés. Réle des altérations
supergènes néogènes à Quaternaire ancien dans leur genèse. Thèse Sci. Univ. Strasbourg, 536 pp.
Menillet, F, 1988. Désilicifications et silicifications au Plioquaternaire dans le karsts de calcaires ter-
tiares du Bassin de Paris. Genèse et caractéristiques des meulières types. Bull. Inf. Géol. Bass. Paris,
25: 81-91.
Milnes, A.R. and TWidale, C.R., 1983. An overview of silicification in Cenozoic landscapes of arid central
and southern Australia. Aust. J. Soil Res., 21: 387-410.
Milnes, A.R., Thiry, M. and Wright, M.J., in press. Silica accumulations in saprolites and soils in South
Australia. In: B.J. Allen, W.D. Nettleton and C.T. Hallmark (Editors), Occurrence, Characteristics,
and Genesis of Carbonate, Gypsum, and Silica Accumulation in Soils. Spec. Pubi. Soil Sci. Soc. Am.
Mountain, E.D., 1952. The origin of silcrete. S. Afr. J. Sci., 48: 201-204.
Roulin, E, Boudeulle, M. and Truc, G., 1986. Transformations argile-opale dans les silcretes Eocenes du
bassin d'Apt (Vaucluse). Bull. Minéral., 109: 349-357.
Senior, B.R., 1978. Silcrete and chemically weathered sediments in southwest Queensland. In: T. Lang-
ford-Smith (Editor), Silcrete in Australia. University of New England, Annidale, N.S.W., pp. 4 1 -
50.
Smale, D., 1973. Silcretes and associated silica diagenesis in southern Africa and Australia. J. Sediment.
Petrol., 43: 1077-1089.
Stephens, CG., 1971. Laterite and silcrete in Australia: A study of the genetic relationships of laterite
and silcrete and their companion materials, and their collective significance in the formation of the
weathered mantle, soils, relief and drainage of the Australian continent. Geoderma, 5: 5-52.
Summerfield, M.A., 1982. Distribution, nature and probable genesis of silcrete in arid and semi-arid
southern Africa. In: D.H. Yaalon (Editor), Aridic Soils and Geomorphic Processes. Catena Suppl.,
1: 37-65.
Summerfield, M.A., 1983a. Silcrete. In: A.S. Goudie and K. Pye (Editors), Chemical Sediments and
Geomorphology. Academic Press, London, pp. 59-91.
Summerfield, M.A., 1983b. Petrography and diagenesis of silcrete from the Kalahari Basin and Cape
coastal zone, southern Africa. J. Sediment. Petrol., 53: 895-909.
Summerfield, M.A., 1983c. Geochemistry of weathering profile silcretes, southern Cape province, South
Africa. In: R.C.L. Wilson (Editor), Residual Deposits: Surface-related Weathering Processes and
Materials, Blackwell, Oxford, pp. 167-178.
Summerfield, M.A. and Whalley, W.B., 1980. Pétrographie investigation of sarsens (Cenozoic silcretes)
from southern England. Geol. Mijnbouw, 59: 145-153.
Thiry, M., 1978. Silicification des sédiments sablo-argileux de l'Yprésien du Sud-Est du Bassin de Paris.
Genèse et évolution des dalles quartzitiques et silcretes. Bull. Rech. Geol. Min., 1: 19-46.
Thiry, M., 1981. Sédimentation continentale et altérations associées: calcitisations, ferruginisations et
si li cificat ions les argile plastiques du Sparnacien du Bassin de Paris. Sci. Geol. Mem., 64,173 pp.
Thiry, M., Bertrand Ayrault, M. and Grisoni, J.-C, 1988. Ground-water silicification and leaching in
sands: Example of the Fontainebleau Sand (Oligocene) in the Paris Basin. Geol. Soc. Am. Bull., 100:
1283-1290.
Thiry, M., Delaunay, A., Dewolf, Y., Dupuis, C , Menillet, F., Pellerin, J. and Rasplus, L., 1983a. Les
périodes de silicification au Cénozoique dans le Bassin de Paris. Bull. Soc. Géol. Fr., 25: 31-40.
Thiry, M. and Millot, G., 1987. Mineralogical forms of silica and their sequence of formation in silcretes.
Thiry, M. and Milnes, A.R., in press. Pedogenic and groundwater silcretes at Stuart Creek opal field,
South Australia. J. Sediment. Petrol.
Thiry, M., Schmitt, J.M., Trauth, N., Cojean, R. and Turland, M., 1983b. Formations rouges "sidéroli-
thiques" et silicifications sur la bordure Nord du Massif Central. Rev. Géogr. Phys. Géol. Dyn., 24:
381-395.
Thiry, M. and Turland, M., 1985. Paléotoposéquences de sols ferrugineux et de cuirassements siliceux
dans le Sidérolithique du Nord du Massif Central (bassin de Montiucon-Domerai). Géol. Fr., 2:
175-192.
Thiry, M., Vinsot, A. and Bertrand-Ayrault, M., 1989. Faciès blancs et faciès sombres dans les Sables
de Fontainebleau (Stampien, Bassin de Paris). Miliex de dépôt et paléoaltérations. 2nd Congrès Fr.
Sédim., Paris, pp. 279-280.
Twidale, C.R. and Milnes, A.R., 1983. Aspects of the distribution and disintegration of siliceous duri-
crusts in arid Australia. Geol. Mijnbouw, 62: 373-382.
Silcretes 377
TWidale, C.R., Shephard, J.A. and Thompson, R.M., 1970. Geomorphology of the southern part of the
Arcoona Plateau and the Tent Hill region, west and north of Port Augusta, South Australia. Trans.
R. Soc. S. Aust., 94: 55-69.
Watts, S.H., 1977. Major element geochemistry of silcrete from a portion of inland Australia. Geochim.
Watts, S.H., 1978. A pétrographie study of silcrete from inland Australia. J. Sediment. Petrol., 48: 987-
494.
Wells, R.T. and Callen, R.A. (Editors), 1986. The Lake Eyre Basin — Cenozoic sediments, fossil
vertebrates and plants, landforms, silcretes and climatic implications. Australasian Sedimentologists
Group Field Guide Series No. 4,176 pp.
Williams, L.A. and Crerar, D.A., 1985. Silica diagenesis, II. General mechanisms. J. Sediment. Petrol.,
55: 312-321.
Williams, L.A., Parks, G.A. and Crerar, D.A., 1985. Silica diagenesis, I. Solubility controls. J. Sediment.
Petrol., 55: 301-311.
Williamson, W.O., 1957. Silicified sedimentary rocks in Australia. Am. J. Sci., 255: 23-42.
Woolnough, WG., 1927. Presidential address, Part I. The chemical criteria of peneplanation. Part II.
The duricrust of Australia. J. Proc. R. Soc. N.S.W, 61: 1-53.
Wopfner, H., 1978. Silcretes of northern South Australia and adjacent regions. In: T. Langford-Smith
(Editor), Silcrete in Australia. University of New England, Annidale, N.S.W, pp. 93-141.
Wright, M.J., 1983. Red-brown hardpans and associated soils in Australia. Trans. R. Soc. S. Aust., 107:
252-254.
Young, R.W, 1985. Silcrete distribution in eastern Australia. Z. Geomorphol., 29: 21-36.
379
Chapter 15
Diversity and terminology of lateritic profiles

Y. TARDY
Introduction: definition of latentes
This chapter will deal with the definition and the terminology of a related series
of materials ranging from latentes, lateritic bauxites, lateritic soils to all kinds of
intertropical weathering products, of detailed features characterizing the different
horizons, of horizons constituting a large variety of profiles, and of profiles within a
large intertropical climatic area.
The term latérite (from the Latin later, brick), is commonly attributed to Bu-
chanan (1807) who described, in Malabar, surficial natural hard materials used as
bricks. In some African dialects, surficial materials which are generally red are
called brick earth (Maignien, 1964) and the term latérite refers to blocs used in con-
struction (Prescott and Pendleton, 1952, in McFarlane, 1976). A controversy over
the definition of latérite has been going on for 150 years (Maignien, 1958; 1966;
McFarlane, 1976). Tbday, two positions emerge. First, many scientists have used
the term laterite to designate weathering products generally formed under tropical
conditions, rich in iron and aluminum, and either hard or subject to hardening upon
exposure to alternate wetting and drying (Pendleton, 1936; Kellog, 1949). Latérites
also include certain highly weathered material with sesquioxide-rich, humus-poor
nodules, that may be surrounded by earthy material that does not harden (Sivara-
jasingham et al., 1962). The term also includes all kinds of plinthites (from the
Greekplinthos, brick) which are latérites in the restricted sense (Mohr et al., 1972,
but excludes soft kaolinitic lithomarges, fine saprolites and non-indurated ferrallitic
soils. Second, for Maignien (1964) and Millot (1964) followed by Schellmann (1983,
1986) the word laterite is not restricted to indurated materials but largely includes
all kinds of tropical weathering products.
I recommend usage of the term laterite in its broadest sense, that is as products
of intense weathering made up of mineral assemblages that may include iron or
aluminum oxides, oxyhydroxides or hydroxides, kaolinite and quartz, and character-
ized by a ratio Si0 2 : (Al203+Fe 2 0 3 ) which does not exceed the value required to
380 Y Tardy
Fig. 15.1. The zone of rubéfaction (Pedro, 1968) coincides with the intertropical area of latente forma
tion, characterized by the development of kaolinite and hydroxides, oxi-hydroxides or oxides of iron and
aluminum. Latérites are limited northwards or southwards, under semiarid or arid climates, by belts
of silcretes, calcretes, vertisols or other smectitic weathering profiles. The two unbroken lines drawn
by Schwertmann (1988) delineate the two cold climatic domains characterized by absence of hematite
from the hot intertropical zone characterized by presence of hematite in almost all soils and weathering
profiles.
characterize quartz and kaolinite.

Thus the term latérite includes bauxites, ferricretes, iron or aluminum duricrusts,
mottled horizons, "carapaces", "cuirasses", plinthites, pisolite or nodule bearing
materials and is extended to the formations or horizons which are parts of red
or yellow ferrallitic soils, tropical ferruginous soils and other formations such as
kaolinitic lithomarges which are soft, cannot be hardened but therefore are, for
example, commonly associated with indurated ferricretes. I propose here that the
zone of the present-day latérite formation would correspond to the rubéfaction zone
of Pedro (1968) (Figure 15.1).
The three domains of common lateritic profiles
In typical lateritic profiles Bocquier et al. (1984) have distinguished three zones
or major horizons (Figure 15.2):
(a) zone of alteration, at the base;
(b) a glaebular zone, located in the middle part; and
(e) a soft zone, non indurated, located in the higher part of the profile.
Diversity and terminology of lateritic profiles 381
YELLOW (GOETHITE) DISMANTLING GRITTY

THE SOFT ZONE
CLAY (KAOLINITE)
SOFT HORIZON FERRICRETE
DISMANTLED CUIRASSE
HEMATITIC NODULES LITHORELICTUAL AND
PEDORELICTUAL
N 0 D U L ES
CARAPACE
RED (HEMATITE)
CLAY (KAOLINITE) THE GLAEBULAR ZONE
@1.: JMOTTLE ZONE
INDURATED HORIZON
PEDOSTRUCTURES
HEMATITIC NODULES
IN FORMATION
LITHOMARGE
lOR FINE SAPROLITE
SAPROLITE
LITHOSTRUCTURES
( QUARTZ THE ALTERATION ZONE
KAOLINITE,GOETHITE
AND HEMATITE )
ABSENCE OF NODULES
ARENE OR COARSE
SAPROLITE
HPARENT ROCK
Fig. 15.2. Schematic representations of a lateritic profile capped by a ferricrete and of the three major
zones of a nodular lateritic profile: alteration zone (A), glaebular zone (B) and soft surficial zone (C).
In the alteration zone, quartz, kaolinite, goethite and hematite form together. In the glaebular domain
hematitic nodules form in a red kaolinitic clay matrix. In the soft zone, hematitic nodules are dismantled
together with the development of a secondary goethite.
In the zone of alteration (coarse saprolite or arene, fine saprolite or lithomarge),

the parent rock volumes and structures are roughly conserved. This domain is es
sentially characterized by the incongruent dissolution of primary minerals and by
the leaching of most of the soluble materials; the least mobile elements (Al, Fé),
liberated by weathering, reorganize almost in situ with little or no transport.
The glaebular domain, generally shows indurated accumulations of iron or alumi
num, either continuous (ferricretes or bauxites) or discontinuous (nodules or piso
lites) resulting in a reorganization of the original material and in an absolute accu
mulation of iron and aluminum crystallized in various oxides, hydroxides oxyhydrox-
ides and also kaolinite.
The soft zone, non-indurated, is characterized by a relative accumulation of
primary minerals such as quartz, or secondary minerals such as kaolinite and oxyhy-
droxides, either resulting from dissolution, degradation and dismantling of glaebular
material or reworking from below by termite activity.
382 Y Tardy
Most of the lateritic profiles clearly show three typical domains but in several
situations one or two of these zones are absent either by incomplete formation or by
posterior erosion.
Saprolites and the so called alteration domain
The so called alteration domain, including different kinds of saprolites, is com-

mon to all lateritic profiles. These horizons are normally located below the ground
water table in the saturated zone, that is in permanently wet conditions. Here, the
mineralogy of the alteration domain is less sensitive to climatic conditions than
that of horizons located above the ground water table. Ttoo kinds of saprolites are
normally distinguished (Figure 15.2).
The coarse saprolite

At the bottom of lateritic profiles, is the unweathered parent rock. Immediately
above, lies the coarse saprolite, where abundant fragments of unweathered rock and
primary minerals are conserved, with original structures intact. On granitic rocks,
the coarse saprolite is traditionally called "arène" (from Latin arena, sand; Leneuf,
1959; Millot, 1964; Tkrdy, 1969). The limit between parent rock and coarse saprolite
is not generally a horizontal plane and the weathering fronts progress irregularly,
penetrating deeper down cracks, fractures and so on, and leaving unaltered boul-
ders, fragments of rocks and relicts of primary minerals sometimes far up the profile.
This gives a fairly diffuse weathering front. Under certain circumstances, the coarse
saprolite may be thick [as frequently observed on granitic rocks (Leneuf, 1959)], or
very thin [as generally observed on basic rocks (Delvigne, 1965)]. Furthermore, the
thickness of the saprolite is reduced in the humid tropics and relatively enlarged in
arid and in temperate regions (Tardy, 1969; Boulet, 1978).
In the early stages of weathering, the first steps of alteration of primary minerals
develop imperfectly closed systems (Tardy, 1969; Tïescases, 1973; Nahon et al., 1977;
Nahon and Bocquier, 1983), where the secondary minerals formed are site-specific.
Their nature depends on the rate of weathering of each primary mineral, and on the
rate at which the initial closed systems become progressively open to the circulating
solutions.
Weathering products in the coarse saprolite may be either a single mineral phase
such as vermiculite, smectite, kaolinite, gibbsite, imogolite, amorphous and hydrated
aluminosilicates, or assemblages such as vermiculite-kaolinite, smectite-kaolinite,
kaolinite-gibbsite (Leneuf, 1959; Tardy, 1969; Novikoff, 1974; Novikoff et al., 1972).
In a normal succession of minerals the most soluble phases (calcite, smectites)
appear in closed systems, while the less soluble ones (kaolinite, gibbsite) appear
in the most open systems, at the contact with circulating solutions. Consequently,
as demonstrated by Tardy et al. (1973), kaolinite and especially gibbsite occuring

in open systems are good indicators of the drainage conditions. On the contrary,
smectites found at the base of the lateritic profiles (in the coarse saprolites or
arènes) are controlled by the specificity of the primary mineral closed microsystems
and do not show any climatic significance.
Thefinesaprolite or lithomarge
Above the coarse saprolite is the fine saprolite (or lithomarge) in which the
structures of the parent rock and the original volumes are still preserved (Leneuf,
1959; Millot, 1964; Tkrdy, 1969). The progress of weathering is petrographically
expressed by an increase in porosity, a complete or partial transformation of most
of the parent rock minerals and a decrease in induration of the rock (Tardy, 1969;
Nahon 1986). Beside quartz, which dissolves slowly, and small pieces of resistant
primary minerals remaining partly undissolved, the dominant species are secondary
kaolinite and ferruginous hydroxides, oxyhydroxides and oxides (goethite, hematite
and amorphous phases).
Finally, in the fine saprolites, located below the ground water level, there is no
important loss or gain of aluminum or iron (Tkble 15.1), nor important migration
observable under the microscope (Tkrdy and Nahon 1985). Compared to the glae-
bular horizon located above, in which element transfers are important, iron and
aluminum, in lithomarges, are almost immobile elements.
Leaching or accumulation of iron and aluminum in lithomarges
There are three types of lithomarges, as related to movement of iron and alu
minum: leaching, lixiviation or accumulation (Figure 15.3, Table 15.1).
A lithomarge sensu stricto, or an idealized lithomarge called a C horizon, the
definition of which corresponds to the description given in the previous paragraph.
A leached lithomarge or lithomarge Ca2, in which a2 stands for an excess of leach
ing which corresponds to the definition of the pallid zone which includes: (a) the
removal of iron around voids, canalicules and channels from a fine saprolite which
locally becomes white, (b) out of these previously iron-poor areas, the removal of
kaolinite which is either dissolved or mechanically leached, and (c) the dissolution
or the mechanical transport of quartz whichfinallyleads to the formation of some
large cavities which may look karstic in origin (Tkrdy, 1969; Ambrosi and Nahon,
1986). Often, at the top of the fine saprolite and below the ferricrete, exists a hori
zon, rich in coarse grained quartz, depleted infinegrained quartz and kaolinite, and
showing some convergence of characters with stone lines or with the A2 horizon,
described by Leprun (1979) as the result of dismantling a lower ferricrete horizon.
A lithomarge Ca2b, in which the letter b stands for a secondary, either neoformed
or deposited kaolinite, accompanied by goethite, filling the small or large karstic
386
GRITTY LAYER
RING MADE OF G O E T H 1 T E
LITHOMARGE Ca ?b
CORE MADE OF
FERRICRETE
( C A O L I N I TE AND DISMANTLING
MACROVOID
FILLEO BY YELLOW BROUN HEMATITE
•CAOLINITE
PISOLITIC LAYER
LITHOSTRUCTURE
RICH IN KAOLIHITE
LITHOSTRUCTURE LITHOSTRUCTURE
RICH IN KAOLINITEl RICH IN QUARTZ
FINE SAPROLITE
LITHOMARGE C
SENSU STRICTO
LITHORELICTUAL AND
COARSE SAPROLITE PEDORELICTUAL NODULES
MASSIVE RED
FJMATRIX FORMING
NODULE
FORMING
Fig. 15.3. a. Alteration zone: coarse saprolite or arène, fine saprolite or lithomarge (C horizon sensu stricto), leached lithomarge (Ca2) and illuviated
lithomarge (Ca2b). b. Mottle zone and nodular horizon: channels, macrovoids, bleached domains, yellow-red soil matrix, litho- and pedo-relictual mottles
and nodules, c. Ferricrete zone and surficial horizon: a progressive nodulation towards the top of the profile and a secondary development of pisolite
close to the soil surface can be distinguished. Three types of nodules are distinguished: lithorelictual (white), pedorelictual (striped) and undifTerenciated
(black).
Y Tardy
Diversity and terminology of lateritic profiles
TABLE 15.1
Isovolumetric chemical balance (in g per 10 cm 3 of rock) of lithomarges and fine saprolites (horizon C), showing that some samples appear leached and
depleted in both Fe and Al (Ca 2 ) while others appear enriched in kaolinite, goethite, Fe and Al (Ca 2b )
Nature of rock Density Quartz Si02 A1 2 0 3 Fe203 Mn304 Ti02 CaO MgO Na20 ê2ï H20
Parent rock unaltered 2.60 70 188 35 6 0.08 0.5 4 1.5 10 12 1.0
Lithomarge C 1.40 60 93 28 5.5 0.03 0.8 0.2 0.2 0.1 0.2 12.0
(leached) Ca2 1.20 55 87 25 2.0 0.01 0.2 0.1 0.1 0.1 0.1 8.9
(illuviated) Ca2b 1.80 65 223 40 7.5 0.17 3.0 0.3 0.3 0.1 0.3 17.5
Lithomarge (calculated) 1.65 70 111 35 6 0.0 0.0 0.0 0.0 0.0 0.0 13.0
Mottle zone 1.49 54 93 33 8.0 0.03 0.9 0.3 0.3 0.2 0.1 13.5
Source: Tardy (1969).
387
386 Y Tardy
cavities characterizing the lithomarge Ca2. This accumulation, brown in color, con
trasts clearly with the pale colour of the void and channel surroundings which often
appear decolored (Tkrdy, 1969). When most of the initial porosity isfilled,a massive
saprolite is produced (Ambrosi and Nahon, 1986).
Lithomarges Ca2 and Ca2b both are located at the top of the fine saprolite zone,
underlining the mottle zone.
Problems of terminology
In my opinion, fine saprolite, lithomarge, pallid zone and variegated clays refer
to the same thing. They all designate the same quartz-kaolinite-iron oxyhydroxide
formation, previously described. However, the two last terms produce possibilities
of confusion and interpretive error.
The use of the term variegated clays (argiles bariolées) instead of lithomarge
or fine saprolite (Leprun, 1979; Ambrosi, 1984) introduce risks of confusion with
mottled clays (argiles tachetées) which are, genetically speaking, entirely different.
Variegation was used to describe differences in colour due to the original distri
bution of iron-rich (biotites, amphiboles) and iron-poor (quartz, feldspars) primary
minerals. Mottles are differences in colour due to a pedogenetic or a secondary
redistribution of iron, removed from some domains and reconcentrated in other
domains.
The distinction between mottle zone and pallid zone is due to Walther (1915).
Nevertheless, the term pallid zone, introduced by McLaren (1906) and Simpson
(1912), according to Prescott and Pendleton (1952) and McFarlane (1976), also cre
ates confusion. In fact, it has been understood by definition as a zone, pale-colored
due to leaching of iron, which is considered as the source of iron enrichment in
the overlying mottle zone ferricrete. McFarlane (1976) has presented a series of
arguments showing that this cannot be the case mostly because there are no rela
tions between the thickness of the pallid zone and the amounts of iron accumulated
above it, and also because some so called pallid zones do not show any iron release,
according to the observations described below. Consequently both terms variegated
clay or pallid zone should be avoided.
Smectitic saprolite and pistachio horizons
Under semi-humid or semi-arid conditions (rainfall < 1000 mm per year) within
the coarse and thefinesaprolites, a smectitic horizon called sometimes the pistachio
horizon because of its pale green color, may be developed. This horizon is generally
overlain by a smectitic soil in a dynamic equilibrium reflecting the local climate,
that is, rainfall 800-900 mm per year, temperature 25-30°C, for a typical Vertisol
(Bocquier, 1973). It may also be covered by a lateritic soil consisting of a lithomarge,
a mottle zone and in some cases a carapace and a ferricrete. The pistachio horizon
has been considered as secondarily developed, after the formation of the lateritic
profile, in peculiar cases, when the climate has been changing from humid (laterite)
to arid (Vertisol). The mechanism was observed in Senegal and in Burkina Riso by
Leprun (1979) and also described by Boulet (1978), Pion (1979), Ambrosi (1984)
and Ambrosi and Nahon (1986) (Figure 15.3).
In this case, the development of smectite, associated with kaolinite, invading a
part of the fine saprolite has a climatic significance according to Tkrdy et al. (1973).
This secondary development of underlying smectitic horizons is interpreted as one
of the most probable mechanisms of the secondary degradation of the basal hori
zons of the ferricretes previously formed during humid periods and then subjected
to semi-arid climates.
In summary, arene and lithomarge, equivalent to coarse and fine saprolites, are
basal weathering domains (C horizons) in which (a) volumes of the parent rocks are
preserved, (b) amounts of iron and aluminum and amounts of quartz are, conserved
within 10%, and (c) porosity is due to the leaching of the alkaline, alkaline-earth
elements, and silica in excess of that required to form kaolinite and to remain as pri
mary quartz. At the top of the lithomarge sensu stricto (C horizon) one may observe
additional leaching (Ca2) or secondary enrichment (Ca2b) of kaolinite and iron hy
droxides. However, there are neither short scale movements nor reconcentration
of iron and consequently no mottle formation. Both lithomarge Ca2 and Ca2b are
transitional with the overlying mottle zone, they are part of the alteration zone and
are not part of the glaebular zone which includes mottle and nodular horizons.
Fine saprolites or kaolinitic lithomarges can be very thick. Several tens of meters
are not exceptional (McFarlane, 1976) and an order of magnitude of 100 m is not
uncommon (Pham et al., 1988; Freyssinet, 1990).
Glaebular domains and the ferricrete profile
Under contrasted tropical climates (Γ = 25-30°C, rainfall = 1500 mm per year,

5 months of dry season, relative humidity of the air, HR9 = 65%) and above fine
saprolites and lithomarges, iron is generally redistributed and concentrated to char
acterize a glaebular zone, in which ferricrete may develop. A typical ferricrete
profile consists of three major horizons: mottle zone, ferricrete itself (carapace and
cuirasse) and, at the soil surface, a gritty layer and dismantling horizon.
Mottle zones and nodular horizons

The mottle zone is characterized by a contrast between bleached domains and
Fe-mottles which can be easily distinguished on a centimeter scale, in outcrops and
in samples, and on a micrometer scale, under the microscope (Figure 15.3).
Bleached domains consist mainly of quartz and kaolinite and exhibit a white or
gray colour due to the de-ferruginization of the previously associated kaolinite and
388 Y Tardy
FORMATION OF NODULES FORMATION OF MACROVOIDS

FROM MOTTLES FROM BLEACHED DOMAINS
Fig. 15.4. Formation of mottles and nodules, in originally kaolinite-rich domains, leading to lithore-
lictual features developed in primary lithostructures and pedorelictual features, developed in secondary
pedostructure. Formation of bleached domains, macrovoids and channels from originally quartz-rich do
mains. Secondary formation of pedorelictual mottles or argilomorphous nodules in previously formed
macrovoids.
iron oxyhydroxides. Original kaolinite aggregates, free of iron, can be dispersed and
kaolinite particles can migrate and even be leached out. These changes are accom
panied by a strong increase of porosity, leading subsequently to the formation of
macrovoids, such as tubules and alveoles. At that stage quartz also can precipitate in
the generalized eluvial migration (Figure 15.4; Nahon, 1986). Quartz and kaolinite
removal are probably due to termite activity (Eschenbrenner, 1987).
The de-ferruginization takes place in domains originally enriched in quartz and
poor in kaolinite. It begins at the top of the fine saprolite (lithomarge Ca2) and
becomes progressively more important towards the top of the mottle zone, where
channels and tubules are abundant (the mottle clay zone: "argile tachetée", which
was sometimes called channel clay zone: "argile à canaux").
Fe-mottleSy mostly of a brown red colour, are diffuse glaebules (Brewer, 1964)
and result in a concentration of iron which precipitates mainly as goethite and as
hematite together with kaolinite. Thus, domains of previous accumulation of kaoli

nite are natural hosts for secondary accumulation of iron in mottles and nodules
(Figure 15.4).
In the mottle zone, there are two situations in which kaolinite accumulation is
the precursor of nodule formation: lithorelicts and pedorelicts.
Lithorelicts and pedorelicts. Lithorelictual mottles or nodules are iron accumula
tions in which the original structure of the parent rock can be seen (Figure 15.4).
They are more accurately called alterorelicts by Faure (1985) and are mostly pri
mary assemblages of aluminum and iron-rich minerals which, by alteration, give
secondary stable associations of kaolinite and iron oxyhydroxides, in which porosity
is small in size. This is the case for schists, amphibolites and layers of migmatite
which are particularly rich in biotite, for example. Granitoid rocks including aplitic
and quartzo-feldspathic veins, initially poor in iron and consequently sensitive to
leaching of kaolinite, are not ideal parent rocks for the further accumulation of this
element.
Pedorelictual features (orpedorelicts) are due to secondary accumulation of kaolin
ite,fillingvoids previously created in the bleached domains. They are relicts of soil
forming processes. Kaolinite, derived from solution or translocation from overlying
upslope layers, followed by precipitation or deposition lower in the profile. These
secondary accumulations of kaolinite and associated quartz are generally accom
panied by a secondary accumulation of iron-generating brown-red colored clays,
providing a contrast with the white-gray decolored areas, located around the tubules
and channels (Nahon, 1986). Higher in the ferricrete, these mottles evolve into
nodules as accumulation progresses. They are what Nahon (1976) has called the
argilomorphous nodules, which are old voids, filled by kaolinite and iron oxyhy-
droxide. Mottles and nodules of this kind are traces of the secondary pedogenetic
activity occuring in the unsaturated zone. In the nodular zone, lithorelicts are more
abundant on schists and amphibolites while pedorelicts dominate on sandstones and
granitic rocks rich in quartz (Figure 15.4).
Towards the top of the profile, the mottle zone evolves towards either a nodular
horizon (Ambrosi et al., 1986) or a ferricrete (Figure 15.3; Nahon, 1976).
The ferricrete formation

The ferricrete profile has been described by Nahon (1976), Nahon et al. (1979)
and Leprun (1979). Its mechanism of development has been studied and interpreted
by Didier et al. (1983a, b, 1985), Tardy and Nahon (1985), Nahon (1986), Müller and
Bocquier (1986) and Ambrosi et al. (1986).
Above the mottle zone and the non-indurated nodular horizon, onefindssucces
sively, from the bottom to the top, variously indurated horizons (Figure 15.3).
Carapace is intermediate between mottle zone located below and cuirasse located
above. It corresponds to a progressive accumulation of iron and as a consequence,
390 Y Tardy
to a progressive development of hematitic iron nodules, either pedorelictual or

lithorelictual. The bleached zone is progressively reduced in size, so that the yellow-
white colored domains decrease in size while the purple-red indurated domains
enlarge and develop.
Cuirasse or ferricrete or iron duricrust sensu stricto is indurated, purple-red in
color, considerably enriched in iron, particularly in hematite. In ferricrete sensu
stricto, the bleached zones are almost absent. Furthermore channels may be reduced
in size but always survive in abundance. The edges of the channels often appear
lined by iron accumulations in which goethite dominates. The channels are com
monly empty but may be filled by kaolinite and fine quartz. They are progressively
transformed into argilomorphous, pedorelictual nodules in which iron accumulates
as hematite (Nahon, 1976).
Towards the soil surface, and from the carapace to the cuirasse, in the discontin
uous iron accumulation horizon, hematitic nodules grow and join each other in a
way recognized by McFarlane (1976) as the major process of ferricrete formation.
However, there are domains in which the clay matrix surrounding the early formed
nodules is itself invaded by iron (also as hematite) so that in certain circumstances
the resulting ferricretes form an almost continuous accumulation of iron in a kaoli
nite plasma. This results in a massive structure which may or may not be crossed
by large voids and channels. This facies of ferricrete is called for that reason mas
sive and vacuolar, and generally is the most evolved, that is the richest in iron, in
hematite, in kaolinite and the poorest in quartz, if present in the original parent
rock. The massive ferricrete is equivalent to the faciès gréseux simple defined by
Nahon (1976) on quartz-rich sandstones.
The ferricrete dismantling horizon. Towards the top, a new system of secondary
voids is developed. The hematite-kaolinite nodules are rehydrated and corroded
at their edges. Kaolinite is dissolved and Al-hematite is transformed into Al-
goethite. A goethitic cortex (concentric yellow brown) develops at the periphery
of the purple-red hematitic nodules. The ferricrete facies becomes either pseudo-
conglomeratic if blocs are present, sometimes recemented by goethite, or pisolitic if
single nodules are individualized (Nahon, 1976).
In summary, a ferricrete is a glaebular domain made of several successive hori
zons: a mottle zone, a nodular non indurated horizon, a carapace and a cuirasse,
massive, pseudonodular, nodular and pisolitic. All these horizons are parts of a
single destruction-formation system.
The gritty layer of surface dismantling
A gritty horizon is developed at the top of the profile, made of the products of the
dismantling of the pseudoconglomeratic or the pisolitic-underlined ferricrete. At the
soil surface, over the indurated horizon, two kinds of soft materials are accumulated
locally: (a) surficial sandy or silty layer made of corroded quartz liberated by the
dissolution of the ferricrete, and mixed with (b) pebbly layer which develops at the
expense of the pisolitic iron crust and which comes from the early formed hematitic
nodules. The size of the pebbles diminishes with time while goethite develops at
the expense of hematite. Iron, aluminum and silica leached from the surface in the
dismantling upper horizon precipitate in a deeper horizon and thus reconstitute the
ferricrete below. The formation of gritty layers at the surface is a component of
ferricrete metabolism and is an essential phase of its reconstitution.
The A2 leached horizon and sub-surf ace ferricrete dismantling
Below the ferricrete, the top of the lithomarge and the lowest part of the mottle
zone often appear decolored and leached, so that a stone-line (labelled A2) made
of large quartz grains, underlies the carapace and the ferricrete and sometimes re
places entirely the mottle zone. The impoverishment of all of the fine fractions in a
coarse material characterized by a high permeability over a lithomarge richer in clay
and less permeable, produces a perched water table. Here, water circulates laterally
and creates a reduced zone where iron is solubilised. This leads to the destabi-
lization of iron-kaolinite aggregates, which induces the leaching of the kaolinite
particles (Chauvel, 1977) and the dismantling of the overlying ferricrete. In humid
regions, the A2 horizon may be attributed to termite activity inducing a selective
upward and downward motion of fine material (Eschenbrenner, 1987). In less hu
mid regions, close to the Sahelian fringes, this phenomenon, described by Leprun
(1979) is general. In dry climate areas, the pistachio smectitic horizon, sometimes
well developed, may also induce a lateral circulation of ground water high in the
profile, favorable to a sub-surface dismantling of ferricrete previously formed under
more humid climates (Leprun, 1979).
Terminology
Mottle zones, nodular horizons, nodular ferricretes (carapace or cuirasse) and

pisolitic horizons are all part of the glaebular zone. Glaebules are three dimensional
units within the s-matrix, recognized in this case by a greater concentration of iron,
the edges of which are either sharp (nodules) or diffuse (mottles) (Brewer, 1964).
Clearly, the mottle zone is a part of the glaebular zone. Chatelin (1972) has intro
duced a distinction between alterite and alloterite. Thefirststands for saprolites and
was extended to mottle zones in which the parent rock structures are distinguishable
and the original volume is conserved. They are clearly isovolumetric alterations in
the sense of Millot and Bonifas (1955). Alloterite, on the contrary, stands for hori
zons in which the original primary architecture of parent rocks is destroyed. Thus,
in most cases, carapaces, cuirasses and more generally bauxites and ferricretes can
be considered as alloterites (Boulangé, 1984). Herbillon and Nahon (1988) have dis
tinguished two major zones: a relative accumulation zone including most saprolites
392 Y Tardy
and in some cases, mottle clays, and an absolute accumulation horizon including all
the ferricrete layers. Muller et al. (1981) have described a mottled alterite distinct
from a nodular zone, so that a certain ambiguity is introduced, concerning a pos
sible difference between mechanism of formation of mottles in mottle zones and
nodules in nodular horizons. Mottles and nodules form in response to an identical
filiation process. Both diffuse mottles and sharp edged nodules are glaebules, so that
mottled zone and ferricrete must be included in the glaebular zone. Despite this,
the so-called mottled clay horizon is generally isalteritic, though the results of the
beginning of the mobilization and the absolute concentration of iron can be seen.
An essential distinction is proposed between (a) coarse and fine saprolites, per
manently located in wet conditions, below the ground water table and in which little
if any movement of iron is detected, and (b) mottle zones, nodular horizons and
ferricretes, located in the unsaturated domain, seasonally hydrated and dryed, in
which mass transfer of iron is currently observed.
Conclusion: ferricrete metabolism

An iron crust is generally built at the top of the lateritic mantle by a combination
of successive small-scale migration of iron, leaching or dissolution of kaolinite and
quartz grains, formation of voids, secondary accumulation of kaolinite together with
small quartz crystals, and ferruginization of these accumulations (Nahon, 1986).
Several remarks can be added.
(a) Despite some lateral and upward vertical movements, particularly observed
downslopes, where the water table is close to the soil surface, the transfers of iron
and aluminum are supposed to be essentially vertical and downwards.
(b) Kaolinite-rich domains are selective hosts for hematitic iron accumulations.
They evolve towards lithorelictual nodules if kaolinite accumulation comes directly
from the original parent rock (lenses of schists, mica-rich layers in migmatites,
aluminous-rich domains in basic rocks) or towards argilomorphous pedorelictual
nodules if kaolinite comes from a secondary accumulation either by neoformation or
by mechanical deposition in the macrovoids. They are small sized porous domains,
in which thermodynamic activity of water is commonly low (Didier et al., 1983a, b;
Tfordy and Nahon, 1985).
(c) By contrast, primary quartz-rich domains appear as selective sites for decol
oration, deferruginization and destabilization of the iron-clay aggregates. Conse
quently, they are also privileged sites for chemical or mechanical leaching of quartz
grains and finally for creation of channels, glosses and zones of water circulation.
They appear generally to have large pores and are seasonally filled by water of high
thermodynamic activity, close to the conditions of saturation (Tkrdy and Novikoff,
1988).
(d) Finally these two types of facies taking over from one to another in space, are
the two privileged sites for the two simultaneous stages of ferricrete metabolism:
(1) a destructive or dismantling stage in which dissolution and degradation takes

place close to the soil surface, in pores of large size temporarily hydrated, together
with (2) a formative or a reconstitutive stage in which precipitation concretion and
nodule formation takes place in clay-rich and small sized pore domains of deeper
horizons temporarily dehydrated. In the first case (1) of the process, goethite, a
hydrated mineral, prevails, while in the second case (2), hematite, a dehydrated
mineral, dominates. The first stage (1) takes place during the wet season and is
dominated by circulation in pores of large diameter, leaching and reactions occuring
in saturated domains and high water activity. The second stage (2) takes place
during the dry season and is dominated by imbibition and impregnation of pores of
small size, occuring in unsaturated domains and low water activity.
The soft zone
In tropical areas a large number of glaebular soils are capped by a soft horizon
(Figure 15.2), the origin of which is diverse or subject to different interpretations.
Three possible orìgins

In some regions, such as Amazonia for example (Lucas, 1989), the soft hori-
zon is very thick (several meters) and is regularly and widely distributed. This so
called "argilla de Belterra", 2-10 m thick, has been alternately considered either
as a sedimentary cover (Sombroek, 1966; TKcart, 1978; Putzer, 1984; Truckenbrodt
and Kotschoubey, 1981) or as an in situ alteration of the sedimentary parent rock
(Chauvel et al., 1983; Irion, 1984). The question is not yet settled.
In Africa, soft horizons cover large tropical or subtropical areas, overlying in most
cases, a glaebular zone or a ferricrete. In the latter situation the soft horizon or the
surficial sandy layer, non indurated and characterized by a relative accumulation
of primary minerals such as quartz, or secondary minerals such as kaolinite and
goethite, generally shows some geochemical relationship to the underlying ferricrete
and has been considered as an in situ dismantling product (Nahon, 1976; Leprun,
1979; Müller et al., 1981; Rosello et al., 1982; Chauvel et al., 1983; Fritsch, 1984;
Bocquier et al, 1983). However, despite a surficial degradation of the underlying
ferricrete, the soft horizon as a whole cannot entirely derive from the ferricrete.
The soft horizon may be due to upward transfer of material coming from the mottle
zone, by termite activity.
Concretion and pisolite formation in Ultisols or Oxisols
Concretions and pisolites are widespread in non-indurated latentes (Maignien,

1958; 1964; Martin, 1967; Stoops, 1968; McFarlane, 1976) and nodular soils are
widely distributed in South America and in Africa (d'Hoore, 1963).
394 Y Tardy
In western Africa, they appear distributed in two major climatic zones (Maig-
nien, 1958). The first zone corresponds to the highly contrasted tropical climates
under which Ultisols (sols ferrugineux tropicaux) show a B horizon enriched in
kaolinite, in which mottles, ferruginous concretions and nodules are currently de
veloped (Maignien, 1958; Fauck, 1973; Leveque, 1975; Boulet, 1978; Leprun, 1979).
The second zone corresponds to humid tropical climates under which Oxisols (sols
ferrallitiques) also show a nodular or pisolitic horizon (Martin, 1967).
These kinds of soft but nodular Oxisols or ferralitic soils have been extensively
described by Muller et al. (1981) in Cameroon, Gabon and Congo, and by Eschen
brenner (1987) and Eschenbrenner and Badarello (1987) in Ivory Coast, all regions
subjected to humid tropical climate.
KA HE GO Gì QU
Fig. 15.5. Nodular Oxisol profile in Cameroon (from Müller, 1987). Distinguishable are the alteration
zone (A), the glaebular zone (B) and the soft horizon (C) and within them: (1) saprolite with inherited
rock structure and texture; (2) red and compact matrix; (3) yellow and friable matrix; (4) ferruginous
lithorelict; (5) argilomorphous nodules; (6) yellow and compact matrix; (7) organic matter accumulation
(from Müller and Bocquier, 1986). The mineralogical distribution of kaolinite (KA), hematite (HE),
goethite (GO), gibbsite (GI) and quartz (QU) (from Müller, 1987).
Müller (1987) described in detail the soil-sequence of Goyoum in the Eastern

part of Central Cameroon. Above the fine saprolite, the glaebular domain com-
prises a red hematitic clay in which dark red nodules are individualized. The soft red
clay is, diffusely impregnated by hematite and goethite (hematite > goethite), asso-
ciated with kaohnite in microstructures. Within the soft red clay matrix two types of
nodules are developed: (a) argilomorphous, with a structure similar to that of the
clay matrix, or (b) lithorelictual, showing a structure similar to the pétrographie or-
ganization of the parent rock. Nodules are in general made up of an assemblage of
kaohnite and hematite. Towards the top of the profile, the red clay matrix becomes
progressively yellow and goethite dominates over hematite. The hematitic nodules
are surrounded by a goethitic cortex the thickness of which also increases towards
the top of the profile. Close to the soil surface, the red nodular horizon is dismantled
and replaced by a soft clayey horizon, made of an assemblage of goethite, kaolinite
and quartz.
In Ultisols nodules or pisolites can be of two different origins: (a) either (Figure
15.5) they are relatively recent secondary nodules growing within the clay-rich soil
matrix, en route to the development of a ferricrete, and following a similar pro-
cess of accumulation of iron as hematite; or (b) they are primary hematitic nodules
generated by a secondary disagregation of older ferricretes, today covered by a soft
goethite-kaolinite horizon and en route to being dismantled through the forma-
tion of a nodular stone-line. The first process is generally accepted and nodular
oxisols are often considered as the precursors of the ferricrete (McFarlane, 1976,
1983). The second was recently discovered with the realization that the equatorial
rain forest was previously underlain by ferricretes which are now being dismantled
(Novikoff, 1974; Tkrdy et al., 1988; Martin and Volkoff, 1989; Nahon et al., 1989).
Oxisols with no glaebular development
Mohr et al. (1972) state: "typical oxisols without plinthite are deep, friable soils.
Horizon differentiation is indistinct and there is generally no clay movement. These
soils are porous and are rapidly drained. The oxic horizon does not harden upon
exposure to air; there is no segregation of iron; its distribution is very homogeneous
and concretions are absent or near absent. If the solum is very deep, it is sometimes
underlain by a mottle clay horizon. Otherwise the solum lies directly on the weath-
ered rock", that is on the lithomarge or the fine saprolite. These red ferralitic soils,
also known in Brazil as Latosols, are developed on various parent materials under
humid tropical or equatorial climates in regions covered by rain forest and particu-
larly in the well drained upslopes of landscapes (Chauvel, 1977; Volkoff, 1985); they
typically show a micronodular structure made up of an assemblage of kaolinite and
quartz cemented by iron oxides (Figure 15.6; Bennema et al, 1970; Pedro et al.,
1976; Chauvel and Pedro, 1978a, b).
396 Y Tardy
Fig. 15.6. Typical Oxisols do not generally show clay movements, colour gradients and subsequent hori
zon differentiations. Because of their micronodular structures made up of an assemblage of hematite,
kaolinite and quartz, they are porous and characterized by rapid drainage (from Chauvel, 1977)
There are few studies devoted to the composition of iron minerals in these kinds
of soils. As suggested by the red colour they are mixtures of ferrihydrite, goethite
and hematite (Schwertmann, 1988). Gibbsite is also common and sometimes abun-
dant in Oxisols or Latosols (Sieffermann, 1973; Bourgeat, 1972). The mineral dis-
tribution within the profile is a function of depth: hematite increases towards the
top of the soil-surface while kaolinite or gibbsite may either increase or decrease
(Leneuf, 1959; Schellmann, 1964; Sieffermann, 1973; Delvigne, 1965; Mohr et al,
1972). Furthermore Volkofï (1985) pointed out that the mineralogical composition
may depend on the nature of the parent rock: kaolinite, hematite, gibbsite on basic
rocks; quartz, kaolinite, goethite on sandstones.
Lateritic bauxites
A bauxite is a thick accumulation of aluminum resulting from long term lateritic

weathering under humid tropical or equatorial climates (Millot, 1964; Valeton, 1972;
Bardossy, 1982; Boulangé, 1984; Lucas, 1989).
The rainfall required for the separation between a ferricrete and a bauxite is
probably about 1700 mm a year. Above this limit, ferricrete dismantling takes place
while gibbsite forms, develops and accumulates. The humidity of the air and tem-
perature are also determinant factors. Ferricrete and iron accumulation are favored
Diversity and terminology oflateritic profiles 397
by a relatively low thermodynamic activity of water and relative humidity of the air
(HR = 60%, annual average) and high temperature (T = 28°C) while bauxite and
aluminum enrichment can stand lower temperature (T > 22°C) and are favored by
a higher thermodynamic activity of water and a higher relative humidity of the air
(HR > 80%; Tkrdy and Novikoff, 1988; Tkrdy et al., 1988a, b).
Some of these lateritic bauxites are very old and have been evolving under
various tropical climates, since the Jurassic, the Cretaceous, the Paleocene or
the Eocene (Michel, 1973). Others are younger and may have formed since the
Miocene, the Pliocene and even the Quaternary (Bardossy, 1979). Lateritic bauxites
are widespread all around the world and appear under different latitudes.
The old bauxitic profiles which have been submitted to climaticfluctuationsshow
in general a large variety of petrographical facies and a high complexity of mineral
distribution.
Distribution ofgibbsite
Several lateritic bauxite profiles, representing quite a large variety of situations,
were recently described by Boulangé (1984) and by Lucas (1989). Most profiles
show two different gibbsitic layers. The first one occurs in the coarse saprolite
at the bottom of the profile and close to the contact with the parent rocks. The
second one, bauxitic, sensu stricto is located at the top of the profile close to the soil
surface. In between, a kaolinite-rich horizon, from which gibbsite may be absent,
appears similar to the fine saprolite or lithomarge already described. However other
situations are also described. In some cases only kaolinite is found in the coarse
saprofite, while in others, gibbsite may occur all through the profile. In the latter
situation, kaolinite, always present in the fine saprolite, disappears in the bauxite
(Figure 15.6).
The bauxite zone itself occurs in different major facies. The most classical
bauxitic profiles show the development, more than 10 to 15 m deep, of a massive
gibbsitic accumulation. Towards the top, vacuoles develop within a goethite-gibbsite
yellow plasma while nodules red in color, made of an assemblage of hematite-
gibbsite, are individualized (Boulangé, 1984). The formation of pisolites is not ob
served.
The second type remarkably, shows a secondary development of boehmite to
gether with a large formation of pisolites (Tardy et al, 1988a, b).
Secondary formation ofpisolites and boehmite

Boehmite forms in some bauxitic profiles, particularly those located close to the
edges of a plateau and in an horizon close to the soil surface. When this occurs,
the gibbsite-hematite original plasma is reorganized into pisolites and a strong
segregation of hematite is observed.
398 Y Tardy
a. Iron accumulation b, Aluminum accumulation
o o o o
Hematite o o o o Boehmite + Hematite
o o o o
DOD DOD a Da DDD
ODD D D D a a Da a a
d o o ODD D a a D a a
DOD D D D Bauxite a a a D a a Gibbsite + Hematite + Goethite
Gibbsite D D DD D D
D D DODD
D a D a a a
a a DD a a
n a n D DO
a D a ODD
DOD D D a
ρβΙΙΕβΙ
DOD D D D Ferricrete |yi;iοiοWi;iοiii'^;oY;;-ijjj!l Hematite -f Goethite + Kaolinite
Kaolinite Lithomarge Kaolinite
Fig. 15.7. Distribution of iron and aluminum in bauxites: iron generally tends to accumulate close to the
top of gibbsitic profiles (a), while it tends to concentrate at the bottom of boehmitic profiles (b).
These pisolites rich in boehmite are progressively transformed towards the soil
surface and secondarily surrounded by cortex, septarias and a cristalliplasma of
gibbsite. In an intermediate horizon together with the formation of nodules of
boehmite and hematite, dehydrated minerals form in abundance. This layer is in-
terbedded between two horizons in which gibbsite, a hydrated mineral, is formed
(Figure 15.7).
Distribution of hematite and goethite

In both the coarse saprolite and thefinesaprolite, goethite is more abundant than
hematite, while in the bauxitic layer hematite dominates goethite. Furthermore,
when boehmite develops, goethite disappears and hematite is the single iron-rich
mineral present. This distribution, frequently observed in West African alteritic
bauxites (Boulangé, 1984) is similar to that frequently described by Combes (1969)
and by Bardossy (1982) in karst bauxites of the South of France, Hungary and
Greece.
Conclusion
Aluminum and iron are distributed from the bottom to the top of the profiles, in
two mineral sequences:
gibbsite (1) - kaolinite - gibbsite (2) - boehmite - gibbsite (3)

goethite (1) - hematite - goethite (2)
These are ideal sequences and not every layer is always present. However, in both
cases, hematite and boehmite which are dehydrated or relatively dehydrated miner-
als are generally located between more hydrated phases, (goethite and gibbsite).
The gibbsitic profiles developed under humid tropical or equatorial climates are
often capped by an iron rich horizon (Grandin, 1976). By contrast, gibbsitic horizons
which were later transformed into boehmitic horizons under seasonally contrasted
tropical climates are dominantly poor in iron. As a consequence of this secondary
change, iron previously accumulated close to the surface moves down below and a
ferricrete develops at the bottom of the profile (Figure 15.7).
Conclusion
Lateritic profiles show a kaolinitic lithomarge or in some cases a bauxite at

their base. Above the Hthomarge either a bauxite or a ferricrete or a lateritic soil
develops.
Eight different lateritic soils can be described (Figure 15.8).
(a) Pale colored Oxisols, with a thick surficial horizon, poor in kaolinite and rich
in quartz. These soils are developed on sandstones and quartz-rich parent rocks.
They may present some characteristics of tropical podzols.
(b) Yellow Oxisols, primarily rich in kaolinite and goethite, with quartz and
gibbsite as accessories, are located on quartz-poor rocks particularly on downslope
setting of the landscape.
(c) Red Oxisols primarily rich in kaolinite and hematite, with quartz and gibbsite
as accessories, are found in the upper slopes of a landscape.
(d) They may or may not have a glaebular horizon, associated in the latter case
only with a mottle zone.
(e) Ferricrete profiles may be covered by sandy surficial horizons originating
from termite activity leading to the subsequent dismantling of hematitic duricrusts,
the product of which remains within the soft horizon as a relictual nodular stone
line.
(f) Ferricrete profiles may show the development of sandy, quartz-rich, bleached
horizons forming stone lines at the top of mottle zones and leading subsequently to
sub-surface dismantling of iron duricrusts.
400 Y Tardy
È ®
QUARTZ GOETHITE
+ HEMATITE
KAOLINITE
+ NODULES
GIBBSITE
MOTTLES
KAOLINITE
LITHOMARGÊ GOETHITE
+
HEMATITE
+
KAOLINITE
Γ+τττττΙ
PARENT
ROCK
t.t.t + -
EQUATORIAL YELLOW RED NODULAR

PODZOL OXISOL OXISOL OXISOL
GRITTY HORIZON
0 JlLsOFT
HORIZON
®^
STONE FERRICRETI LEACHED
© Ö ® © — LINE HORIZON
NODULAR
V····
HORIZON "2
HORIZOr HORIZON^ • · ·
QUARTZ
ii: ''L.Ul-2evi JMOTTLE_^ MOTTLE • · ·
ZONE "ZONE
KAOLINITE KAOLINITE
LITHOMAR iSAPROLITE
GE
SMECTITE
SAPROLITE
::::::::::::|PARENT — ROCK —
tt-t-tt + + + +
NODULAR FERRICRETE FERRICRETE FERRUGINOUS

OXISOL WITH UNDERNEATH ULTISOL
A 2 LEACHED WITH SMECTITE
HORIZON HORIZON
Diversity and terminology of laterìtic profiles 401
(h) Beige Ultisols show the development of surficial leached horizons (A2)
overlying horizons (B) in which kaolinite accumulates and in which mottles and
hematitic nodules subsequently develop.
All these soils are latérites, characteristic of a large variety of present-day climate,
paleoclimates, parent rocks, topographic positions and hydrologie regimes.
References
Ambrosi, J.P., 1984. Petrologie et géochimie d'une séquence de profils latéritiques cuirassés ferrugineux
de la région de Diouga, Burkina Faso. Thèse, Univ. Poitiers, 223pp.
Ambrosi, J.P. and Nahon, D., 1986. Petrological and geochemical différenciation of lateritic iron crust
profiles. Chem. Geol., 57: 371-393.
Ambrosi, J.P., Nahon, D. and Herbillon, A.J., 1986. The epigenetic replacement of kaolinite by hematite
in laterite. Pétrographie evidence and the mechanisms involved. Geoderma, 37: 283-294.
Bardossy, G., 1979. The role of tectonism in the formation of bauxite deposits. In: Travaux de l'ISCOBA,
Zagreb, 15, pp. 15-34.
Bardossy, G., 1982. Karst Bauxites. Developments in Economy Geology 14. Elsevier, Amsterdam, 441
pp.
Beaudet, G., 1978. Essai sur la zonation et la signification géomorphologique des cuirasses ferrugineuses
en Afrique Occidentale. Trav. Doc. CEGET, Bordeaux, 33: 35-52.
Bennema, J., Jongenus, A. and Lemos, R., 1970. Micromorphology of some oxic and argillic horizons
in south Brazil in relation to weathering sequences. Geoderma, 4: 335-355.
Bocquier, G., 1973. Genèse et évolution de deux toposéquences de sols tropicaux du Tchad. Interpréta
tion biogéodynamique. Thèse Sci. Strasbourg, Mém. ORSTOM, 62, 350 pp.
Bocquier, G., Boulangé, B., Ildefonse, P., Nahon, D. and Muller, D., 1983. Transfers, accumulation
modes, mineralogical transformations and complexity of historical development in lateritic profiles.
In: A. J. Melfi and A. Carvalho (Editors), 2nd Int. Seminar on Lateritisation Processes, Säo Paulo,
pp. 331-343.
Bocquier, G., Müller, J.P. and Boulangé, B., 1984. Les latérites. Connaissances et perspectives actuelles
sur les mécanismes deleur différenciation. Livre Jubilaire du Cinquantenaire de l'AFES, Paris, pp.
123-138.
Boulangé, B., 1984. Les formations bauxitiques latéritiques de Côte d'Ivoire. Trav. Doc. ORSTOM,
Paris, 175, 341 pp.
Boulet, R., 1978. Existence de systèmes à forte différenciation latérale en milieu ferrallitique guyanais:
un nouvel exemple de couvertures pédologiques en déséquilibre. Sci. Sol, 2: 75-82.
Bourgeat, E, 1972. Sols sur socle ancien à Madagascar. Mém. ORSTOM Paris, 57, 335 pp.
Brewer, R., 1964. Fabric and Mineral Analysis of Soils. Wiley, New York, N.Y., 470 pp.
Fig. 15.8. Sequence of lateritic soil profiles from Oxisols (A, B, C, D) to iron duricrust (£, F, G) and
Ultisols (//). In the oxisol group and above the lithomarge and fine saprolite, ferrallitic soil is either
pale and made of quartz (A), yellow and made of quartz, kaolinite, gibbsite and goethite (B) or red and
made of quartz, kaolinite, gibbsite and hematite (C and £>). In the ferricrete group some of the profiles
are covered by a sandy horizon which may be originated from termite activity (E) and some others
present a leached horizon below the ferricrete (G). Deep saprolites located below Ultisols may present
a pistachio colored smectitic horizon (//).
402 Y Tardy
Buchanan, F., 1807. A journey from Madras through the countries of Mysore, Kanara and Malabar.
East India Company, London, 2, pp. 440-441.
Chatelin, Y., 1972. Les sols ferrallitiques: historique, développement des connaissances et formation
des concepts actuels. Init. Doc. Tech., ORSTOM Paris, 20, 98 pp.
Chatelin, Y., 1974. Les sols ferrallitiques. Laltération. Init. Doc. Tech., ORSTOM Paris, 24,144 pp.
Chauvel, A., 1977. Recherches sur la transformation des sols ferrallitiques dans la zone tropicale à
saisons contrastées. Trav. Doc. ORSTOM, 62, 532 pp.
Chauvel, A., Boulet, R., Join, P. and Bocquier, G., 1983. Aluminum and iron oxyhydroxides segregation
in nodules of latosols developed on a Tertiary sediment (Barreiras Group), near Manaus, Amazon
Basin, Brasil.In: A. J. Melfi and A. Carvalho (Editors), 2nd Int. Seminar on Lateritisation Processes,
Säo Paulo, pp. 508-526.
Chauvel, A. and Pedro, G., 1978a. Genèse de sols beiges (ferrugineux tropicaux lessivés) de Casamance
(Sénégal). Cah. ORSTOM, Sér. Pédol., 16: 231-249.
Chauvel, A. and Pedro, G., 1978b. Sur l'importance de l'extrême dessication des sols (ultra-dessiccation)
dans l'évolution pédologique des zones tropicales à saisons contrastées. C. R. Acad. Sci. Paris, Sér.
D, 286: 101-113.
Combes, J.P., 1969. Recherches sur la genèse des bauxites dans le Nord-Est de l'Espagne, le Languedoc
et l'Ariège (France). Mém. CRGH, Montpellier, 432 pp.
D'Hoore, J., 1963. Carte des sols d'Afrique à 1:5.000.000. C.C.T.A./S.P.I., Lagos.
D'Hoore, J., 1964. La carte des sols d'Afrique à 1:5.000.000. Lovanium. Congo Leop. Afr. Soils, 9:
55-64.
Delvigne, J., 1965. Pédogenèse en zone tropicale. La formation des minéraux secondaires en milieu
ferrallitique. Mém. ORSTOM, Paris, 13,177 pp.
Didier, Ph., Nahon, D., Fritz, B. and Tardy, Y., 1983a. Activity of water as geochemical controlling
factor in ferricretes. A thermodynamic model in the system: kaolinite Fe-Al-oxyhydroxides. Sci.
Géol. Mém., 71: 35-44.
Didier, Ph., Fritz, B., Nahon, D. and Tardy, Y., 1983b. Fe 3+ -kaolinites, Al-goethites and Al-hematites
in tropical ferricretes. In: D. Nahon and Y. Noacks (Editors), Petrology of weathering and soils. Sci.
Geol. Mém., 71, pp. 35-44.
Didier, Ph., Perret, D., Tardy, Y. and Nahon, D., 1985. Equilibres entre kaolinites ferriferes, goethites
alumineuses et hématites alumineuses dans les systèmes cuirassés. Rôle de l'activité de l'eau et de
la taille des pores. Sci. Géol. Bull., Strasbourg, 38(4): 383-397.
Eschenbrenner, V, 1986. Contribution des termites à la microagrégation des sols tropicaux. Cah.
ORSTOM, Sér. Pédol., 22: 397-408.
Eschenbrenner, V, 1987. Les glébules des sols de Côte d'Ivoire. Thèse Sci., Université de Bourgogne,
498 pp.
Eschenbrenner, V. and Badarello L., 1987. Etude pédologique de la région d'Odienné (Côte d'Ivoire).
Carte des paysages morpho-pédologiques. Feuille Odienné à 1200.000. ORSTOM Paris, Not. Expl.,
74,123 pp. -I- 1 carte 1:200.000 + 7 cartes 1:50.000.
Fauck, A., 1973. Les sols rouges sur sable et sur grès d'Afrique Occidentale. Mém. ORSTOM, Paris,
257 pp.
Faure, P., 1985. Les sols de la Kara, Nord-Est Togo. Relations avec l'environnement. Trav. Doc.
ORSTOM, Paris, 183, 281 pp.
Freyssinet, Ph., 1990. Géochimie et minéralogie des latérites du Sud-Mali. Evolution des paysages.
Prospection géochimique de l'or. Thèse de l'Université Louis Pasteur, Strasbourg (in prep.).
Fritsch E., 1984. Les transformations d'une couverture ferrallitique en Guyane française. Thèse 3ème
Cycle, Paris, ORSTOM, 190 pp.
Grandin G., 1976. Aplanissements cuirassés et enrichissement des gisements de manganèse dans
quelques régions d'Afrique de l'Ouest. Mém. ORSTOM, Paris, 82, 268 pp.
Herbillon, AJ. and Nahon D., 1988. Latentes and laterization process. In: J.W. Stucki, A.B. Goodman
and U. Schwertmann (Editors), Iron and Soils and Clay Minerals. Reidel, Dordrecht, pp. 779-797.
Irion, G., 1984. Sedimentation and sediments of Amazonian rivers and evolution of the amazonian
landscape since Pliocene times. In: The Amazon: Limnology and Landscape Ecology of a Mighty
Tropical River and Its Basin. Monogr. Biol,. 56: 201-214.
Kellog, CE., 1949. Preliminary suggestions for the classification and nomenclature of great soil groups
in tropical and equatorial regions. Commonw. Bur. Soil Sei. Tech. Commun., 46: 76-85.
Leneuf, N., 1959. Ealtération des granites calco-alcalins et des granodiorites en Côte d'Ivoire forestière
Leprun, J.C., 1979. Les cuirasses ferrugineuses des pays cristallins de l'Afrique Occidentale sèche.
Genèse — Transformation — Dégradation. Sci. Géol. Mém. 58, Strasbourg, 224 pp.
Leveque, A., 1975. Pédogenèse sur le socle granito-gneissique du Togo. Différenciation des sols et
remaniements superficiels. Thèse, Univ. Louis Pasteur, Strasbourg, 224 pp.
Lucas, Y., 1989. Systèmes pédologiques en Amazonie brésilienne. Equilibres, déséquilibres et transfor
mations. Thèse, Univ. Poitiers, 157 pp.
Maignien, R., 1958. Le cuirassement des sols en Guinée, Afrique Occidentale. Mém. Serv. Carte Géol.
Alsace Lorraine, 16, 239 pp.
Maignien, R., 1964. Compte rendu des recherches sur les latérites. UNESCO NS/HT/125, WS/0264.32-
NA, 266 pp.
Maignien, R., 1966. Compte rendu de recherches sur les latériies. Col. Rech. Res. Nat. 4, UNESCO,
Paris, 155 pp.
Martin, D., 1967. Géomorphologie et sols ferrallitiques dans le Centre Cameroun. Can. ORSTOM, Sér.
Pédol., 5: 189-216.
Martin, D. and Volkoff, B., 1989. Signification paléoclimatique des cuirasses et des nappes de nodules
ferrugineux dans les sols d'Afrique Centrale (Rive droite du Zaire). Schwartz.
McFarlane, M.J., 1976. Laterite and Landscape. Academic Press, London, 151 pp.
McFarlane, M.J., 1983. Latérites. In: A.S. Goudie and K. Pye (Editors), Chemical Sediments and
McLaren, M., 1906. On the origin of certain latérites. Geol. Mag. 43: 536-547.
Michel, P., 1973. Les bassins des fleuves Sénégal et Gambie. Etude géomorphologique. Thèse Sci.
Strasbourg, Mém. ORSTOM, 63, 3 vols., 752 pp.
Millot, G. and Bonifas M., 1955. Transformations isovolumétriques dans les phénomènes de lateritisa-
tion et de bauxitisation. Bull. Serv. Carte Géol. Alsace Lorraine, 8: 3-10.
Mohr, E.C.J., Van Baren, F.A. and Van Schuylenborgh, J., 1972. Tropical soils. A comprehensive study
of their genesis. Mouton, Ichtiar Bam, VanHoeve, Paris, Djakarta, The Hague, 481 pp.
Muller, J.P., 1987. Analyse pétrologique d'une formation latéritique meuble du Cameroun. Thèse, Univ.
Paris VII, 173 pp.
Müller, J.P., Bocquier G., Nahon D. and Paquet H., 1981. Analyse des différenciations minéralogiques
et structurales d'un solferrallitique à horizons nodulaires du Congo. Can. ORSTOM, Sér. Pédol., 18:
87-109.
Müller, J.P. and Bocquier, G., 1986. Dissolution of kaolinites and accumulation of iron oxides in lateritic-
ferruginous nodules: mineralogical and microstructural transformations. Geoderma, 37:113-136.
Nahon, D., 1976. Cuirasses ferrugineuses et encroûtements calcaires au Sénégal Occidental et en Mau
ritanie, systèmes évolutifs: géochimie, structures, relais et coexistence. Sci. Géol. Mém. 44, 232 pp.
Nahon, D., 1986. Evolution of iron crusts in tropical landscapes. In: S.M. Colman and D.P. Dethier
191.
Nahon, D., Janot, C , Karpoff, A.M., Paquet, H. and Tardy, Y., 1977. Mineralogy, petrography and struc-
404 Y Tardy
tures of iron-crusts (ferricretes) developed on sandstones in the western part of Senegal. Geoderma,
19: 263-277.
Nahon, D., Janot, C, Paquet, H., Parron, C. and Millot, G., 1979. Epigénie du quartz et de la kaolinite
dans les accumulations et cuirasses ferrugineuses superficielles. La signification des goethites et
hématites alumineuses. Sci. Géol., Bull., 32:165-180.
Nahon, D. and Bocquier, G., 1983. Petrology of element transfers in weathering and soil systems. Sci.
Géol. Mém., 72:111-119.
Nahon, D., Melfi, A. and Conte, C.N., 1989. Présence d'un vieux système de cuirasses ferrugineuses
latéritiques en Amazonie du Sud. Sa transformation in situ en latosols sous la forêt equatoriale
actuelle. C.R. Acad. Sci. Paris, Sér. II, 308: 755-760.
Novikoff, A., 1974. Ealtération des roches dans le Massif du Chaillu (République Populaire du Congo).
Thèse Doct. Sci., Université Louis Pasteur, Strasbourg, 298 pp.
Novikoff, A., Tsawlassou, G., Gac, J.Y., Bourgeat, F., and Tardy, Y., 1972. Altération des biotites dans
les arènes des pays tempérés, tropicaux et équatoriaux. Sci. Géol. Bull., 25: 287-305.
Pedro, G., 1968. Distribution des principaux types d'altération chimique à la surface du globe. Présenta
tion d'une esquisse géographique. Rev. Géogr. Phys. Géol. Dyn., 2,10, 5: 457-470.
Pedro, G., Chauvel, A and Melfi, A.J., 1976. Recherches sur la constitution et la genèse des terra roxa
estructurada du Brésil. Ann. Agron. 27: 265-294.
Pendleton, R.L., 1936. On the use of the term laterite. Am. Soil Surv. Bull., 17: 102-B.
Pham, V.N., Boyer, D., Freyssinet, Ph. and Tardy, Y., 1988. Cartographie magnéto-tellurique des profils
d'altération épais. Relations entre la nature et la structure du substratum profond et la morphologie
du paysage latéritique au Sud Mali. C. R. Acad. Sci., Paris, Sér. II, 307:1355-1362.
Pion, J.C., 1979. Altération des massifs cristallins basiques en zone tropicale sèche. Etude de quelques
toposéquences en Haute-Volta. Thèse Sci., Strasbourg, Mém Sci. Géol., 37, 220 pp.
Prescott, S.A. and Pendleton, R.L., 1952. Laterite and lateritic soils. Commonw. Bur. Soil Sci. Tech.
Commun., 47, Farnham Royal Bucks., 45pp.
Putzer, H., 1984. The geological evolution of the Amazon basin and its mineral resources. In: The
Amazon: limnology and landscape ecology of a mighty tropical river and its basin, Monogr. Biol.,
56: 15-46.
Rosello, V, Muller, J.P., Ildefonse, P. and Bocquier, G., 1982. Contrôle de la solubilité du fer et de
l'aluminum en milieu ferrallitique. Geochim. Cosmochim. Acta., 46: 1267-1279.
Schellmann, W, 1964. Zur lateritischen Verwitterung von Serpentinit. Geol. Jahrb., 81: 645-678.
Schellmann, W., 1983. Geochemical Principles of lateritic Ni ore formation. In: A. J. Melfi and A.
Carvalho (Editors), 2nd Int. Seminar on Lateritisation Processes, Sâo Paulo, pp. 119-136.
Schellmann, W, 1986. A new definition of laterite. Geol. Surv. India, Mem., 120: 1-7.
Schwertmann, U., 1988. Occurrence and formation of iron oxides in various pedoenvironments. In: J.W.
Stucki, B.A. Goodman and U. Schwertmann (Editors), Iron and Soils and Clay Minerals. Reidel,
Dordrecht, pp. 267-306.
Sieffermann, G., 1973. Les sols de quelques régions volcaniques du Cameroun. Variations pédologiques
et minéralogiques du milieu equatorial au milieu tropical. Mém. ORSTOM, 66, 183 pp.
Simpson, E.S., 1912. Notes on laterite in western Australia. Geol. Mag., 49: 399-406.
Sivarajasingham, S., Alexander, L.T., Cady, J.G. and Cline, M.G., 1962. Laterite. Soil Sci., 121: 1-60.
Sombroek, WG., 1966. Amazon Soils. A Reconnaissance of the Soils of the Brasilien Amazon Region.
PUDOC, Wageningen, 300 pp.
Stoops, G., 1968. Micromorphology of some characteristic soils of the lower Congo (Kinshasa). Pédolo
gie, 18: 110-149.
Tardy, Y., 1969. Géochimie des altérations. Etude des arènes et des eaux des massifs cristallins d'Europe
et d'Afrique. Mém. Serv. Carte Géol. Alsace Lorraine, 31, 199 pp.
Tardy, Y., Bardossy, G. and Nahon, D., 1988. Fluctuations de l'activité de l'eau et successions de
Diversity and terminology of letterine profiles 405
minéraux hydratés et déshydratés au sein des profils latéritiques ferrugineux et bauxitiques. C.R.
Acad. Sci., Paris, Sér. II, 307: 753-759.
Tardy, Y., Bocquier, G., Paquet, H. and Millot, G., 1973. Formation of clay from granite and its distri
bution in relation to climate and topography. Geoderma, 10: 271-284.
Tardy, Y., Melfi, A.J. and Valeton, I., 1988. Climats et paleoclimats périatlantiques. Rôle des facteurs
climatiques et thermodynamiques: température et activité de l'eau, sur la répartition et la composi
tion minéralogique des bauxites et des cuirasses ferrugineuses au Brésil et en Afrique. C. R. Acad.
Sci., Paris, Sér. II, 306: 289-295.
Tardy, Y. and Nahon, D., 1985. Geochemistry of latérites, stability of Al-goethite, Al-hematite and Fe-
kaolinite in bauxites and ferricretes: an approach to the mechanism of concretion formation. Am. J.
Sci., 285: 865-903.
Tardy, Y. and Novikoff, A., 1988. Activité de l'eau et déplacement des équilibres gibbsite-kaolinite dans
les profils latéritiques. C. R. Acad. Sci., Paris, Sér. II, 306: 39-44.
Trescases, J.J., 1973. Involution géochimique supergène des roches ultrabasiques en zone tropicale.
Formation des gisements nickelifères de Nouvelle-Calédonie. Mém. ORSTOM, 78, 259 pp.
Tricart, L.F., 1978. Ecologie et développement: l'exemple amazonien. Ann. Geogr., 481: 257-291.
Thickenbrodt, W. and Kotschoubey, B., 1981. Argila de Bel terra. Cobertura terciara das bauxitas ama-
zonicas. Rev. Bras. Geosci., 11: 203-208.
Valeton, L, 1972. Bauxites. Development in Soil Science. Elsevier, Amsterdam, 226 pp.
Volkoff, B., 1985. Organisations régionales de la couverture pédologique du Brésil. Chronologiedes
différenciations. Can. ORSTOM, Sér. Pédol., 21: 225-236.
Walther, J., 1915. Laterit in Westaustralia. Z. Dtsch. Geol. Ges. 67 B: 113-140.
407
Chapter 16
Geochemistry and evolution of lateritic

landscapes
Y. TARDY and C. ROQUIN
Introduction
Latentes are widespread in Australia, India, Burma, Brazil and in intertropical

Africa. In these regions, bedrocks are generally weathered to depths of over 10
to 150 m, depending on age of laterization, regional tectonic activity, climate and
nature of the parent materials. Despite a certain homogeneity in mineralogical and
chemical composition, they offer a large variety of colors, textures and pétrographie
facies which may not obviously reflect the parent rock from which they derive. Over
most of the areas of the Brazilian and the African shields, a very thick lateritic
mantle has been continuously formed over more than 100 Myr (King, 1957, 1962;
Michel, 1973). Latérites are mostly old formations which are not necessarily fos-
sils but have been evolving and continue to evolve under various tropical climates,
changing or fluctuating from equatorial, humid tropical to semi-arid. As a conse-
quence of the complexity of climatic history, formation and destruction mechanisms
of lateritic profiles and landscapes are generally somewhat difficult to decipher.
Despite a considerable amount of work and a large bibliography (d'Hoore, 1954;
Maignien, 1958, 1966; Leneuf, 1959; Millot, 1964; Tkrdy, 1969; Mohr et al., 1972;
McEarlane, 1976, 1983; Tkrdy and Nahon, 1985), several questions remain unre-
solved, new observations are being established, new data are being produced and
new ideas and theories are now emerging.
The purpose of this chapter is to present some of the recent results of research
into the evolution of profiles and landscapes, of latérites, associated bauxites, ferri-
cretes or soft kaolinitic nodular horizons. Their geochemistry and paleoclimatology
will be stressed.
Autochthon^ allochthony and lithodependence of ferricretes
In western and central Africa, the lateritic weathering mantle is very thick,
widespread, and, in many cases, capped by old and widely developed ferricretes
408 Y Tardy and C. Roquin
(King, 1948; Maignien, 1958; Michel, 1973; Nahon, 1976; McFarlane, 1976; Leprun,
1979).
Ferricretes consist of relative accumulations of poorly mobile major elements
such as iron, aluminum and silicon, minor or trace elements such as vanadium,
chromium, phosphorus, and insoluble minerals such as quartz, ilmenite, tourmaline,
zircon. These elements and minerals may concentrate in situ or may be transported
over short distances either in solution in the percolating waters or mechanically in
bioturbation products. Most of them are autochthonous. The most soluble minerals
are dissolved, and moderately or highly mobile elements are leached in solution and
removed from lateritic landscapes. Beside the resistant primary minerals (mainly
quartz), ferricretes are generally composed of a fairly constant secondary assem
blage made of hematite, goethite, kaolinite and sometimes gibbsite.
Despite their apparent homogeneity, ferricretes show a large diversity of miner-
alogical and chemical compositions, related to their age, their degree of evolution
and the nature of the parent rock from which they derive. Some very immature
and recent ferricretes retain the mineralogical and the chemical signature of their
parent rock. They are called "lithodependent". In others, older and more mature,
mineralogical and chemical compositions are homogenized and the composition of
the parent rock is almost totally modified. These can be called "lithoindependent".
The terminology used here is explained in detail in Tardy (this book).
Mineralogical and geochemical diversity of the lateritic patchworks
At different scales of observation, from regional to local, latérites outcrop as

patchworks of a high degree of mineralogical and chemical diversity.
Age of lateritic surfaces

On a continental scale as shown by King (1957,1962), Michel (1973), McFarlane
(1973), Boulet (1978) and Grandin (1976) in western and central Africa, units of
latérites, different in their degree of evolution, as well as in their mineralogical and
chemical composition, cover surfaces of different ages. Michel (1973) has recog
nized three bauxitic surfaces (Jurassic, Cretaceous, Eocene) followed by four levels
capped by ferricretes at different stages of evolution: relief intermédiaire (Pliocene),
haut glacis (early Quaternary) and moyen glacis (middle Quaternary; Figure 16.1).
This regional sequence of peneplanation surfaces has been interpreted as resulting
from the alteration of humid and arid periods which supposedly succeeded since the
Jurassic. Ages of the Jurassic, Cretaceous and Eocene bauxitic surfaces are prob
ably correct while the ages of ferricretes associated with what is called haut glacis
and moyen glacis are certainly older than Quarternary. In fact, calculated or evalu
ated ages of profiles (Leneuf, 1959; Tkrdy, 1969; Gac, 1980; Gac and Tardy, 1980;
Fritz, 1975; Nahon and Lappartient, 1977; Nahon, 1986) suggest that ferricretes a
few metres thick, overlying lithomarges of several tens of metres in thickness need
Geochemistry and evolution oflateritic landscapes 409
Fig. 16.1. The seven peneplanation surfaces of Michel (1973) capped by latérites of different kinds
(bauxites or ferricretes) and of different degree of evolution, observable from the Fouta Djalon in
Guinea to the Niger alluvial plain in Mali.
several million years to form. In other words it is proposed here that, in Western
Africa, from the Jurassic to the present time, latérites were continuously formed.
Bauxites have been covered since the Oligocene by ferricretes. Ages of ferricrete
overlying morphological surfaces may range from about 1 to 40 million years and
consequently, at a continental scale, degrees of evolution of outcropping ferricretes
may be very diverse.
At a regional scale (20 km) a ferricrete lateritic landscape consists generally of
a myriad of plateaus capped by iron duricrusts (carapaces, cuirasses, gritty layers)
and separated by depressions or thalwegs, in which there are outcrops of the lower
horizons such as mottled zones, lithomarges, saprolites and sometimes parent rocks
(Figure 16.2). In a given system, for example the haut glacis of Michel (1973), all
plateaus were supposed to lie on the same morphological level and were supposed
FERRICRETE I ( OLD )
A
»J FERRICRETE PROFILE
THE BED-ROCK IS DEEP LITHOMARGE

LITHOMARGE
Fig. 16.2. Lateral distribution of lateritic facies in a tropical landscape shaped by erosion and subsequent
alteration.
410 Y. Tardy and C. Roquin
to exhibit the same characteristics. In fact, they do not show the same mineralogical
and chemical composition and consequently the same degree of evolution (Mazal
tarim, 1989; Boeglin, 1990).
The same kind of heterogeneity is also observable on a local scale in a set of
ferricrete samples collected at the surface of a given morphological unit of a plateau
system, covered by a continuous ferricrete (Figure 16.3; Mazaltarim, 1989; Roquin
et al, 1989; Boeglin, 1990; Freyssinet, 1990).
At each scale of observation, the heterogeneity of outcropping facies has received
different interpretations (Figure 16.4). Some authors invoke climatic differences
(Michel, 1973). Others claim that tectonics plays an important part in differen
tiating the altitudes of peneplanation (McFarlane, 1973). Yet others have looked
at the heterogeneity as being due to variations in chemical compositions of par
ent rock (Leprun, 1979), or to lateral variations of facies due to lateral migration
of elements (Fe, Al, Si) within profiles (Nahon, 1976). A systematic mapping of
ferricrete plateaus located in Mali (Kangaba) in Burkina Faso (Gaoua) and Cen
tral Africa (Zemio) (Mazaltarim, 1989; Beauvais, 1990; Boeglin, 1990; Freyssinet,
1990) shows that the surficial heterogeneity can be explained by a differential sur
face erosion process. The erosion leads in different places and at different times to
an exposure of subsurface horizons younger and less evolved than other ones, lo
cated close-by and not eroded. After erosion, all the outcropping young and old fa
cies are submitted to continuous weathering under similar conditions (Figures 16.2
and 16.3).
A continuous ferricrete covering a given morphological surface appears in fact
as a patchwork of petrographical, mineralogical and chemical facies differing in age
and degree of evolution. This can be regarded as a consequence of the differential
erosion whichfinallyinvolves all the ferricrete surfaces.
Allochthony or autochthony offerricretes

The diversity of the pétrographie facies or of the mineralogical and chemical
compositions of the outcropping ferricrete samples make the interpretation of the
surficial geochemistry difficult. For many years the question of the allochthony or
autochthony of ferricretes has been debated. Are they the result of lateral trans
portation of iron in solution precipitating in a lithodependent host saprolite (Maig-
nien, 1958; Michel, 1973; McFarlane, 1976) or are they formed from long distance
mechanically transported materials, so that they are entirely independent of the sur
rounding rocks (Lamotte and Rougerie, 1962; Vogt, 1967; Seret, 1978)? Should we
consider the lateritic mantles as masks without any relationship with the underly
ing rocks or as skins developed from the underlying parent materials (Millot et al.,
1976)?
During these last ten years, evidence of ferricrete autochthony has accumulated
(Wackermann, 1975; Blot et al, 1976, 1978; Boulet, 1978; Pion, 1979; Schellmann,
km
Fig. 16.3. Patchworks of different petrographical, mineralogical and geochemical facies of femcrete, due
to differential erosion over a given morphological surface. The soil surface was permanently submitted to
a contrasted tropical climate and outcropping ferricretes are of different ages and degrees of evolution
(Mazaltarim, 1989; Boeglin, 1990; Freyssinet, 1990).
a MICHEL (1973):
PENEPLANATION SURFACES
AN UNIQUE BUT DISSECTED PENEPLANATION SURFACE
LEPRUN (1979) :
AN UNIQUE PENEPLANATION SURFACE AND TWO DIFFERENT FERRICRETES OF THE SAME AGE
Fig. 16.4. Five interpretations for lateral differentiations of lateritic landscapes: a. surfaces of different
ages formed under different climatic conditions (Michel, 1973); b. a single surface dissected at different
topographical levels (Beaudet, 1978); c. a single surface of ferricretes differing because of their parent-
rock composition (Leprun, 1979); d. a lateral differential evolution (Nahon, 1976); e. a differential
erosion which induces the outcropping of ferricrete fades of different ages and degrees of evolution.
1986; Leprun, 1979). In most cases, quartz and resistant minerals as well as iron and
some trace elements reflect the composition of the underlying rocks. The lithode-
pendence of ferricretes is now generally admitted.
Lithodependence ofyoung duricrusts and chemical homogenization of old ferncretes

Ferricretes of different degrees of evolution and age, overlying parent rocks
of different compositions are clearly lithodependent in young and poorly evolved
systems while they are lithoindependent in old and highly evolved facies, which
are chemically homogenized. In Table 16.1 several analyses of ferricretes collected
in the region of Gaoua (Burkina Faso) include parent rocks (acid PRA or basic
PRB) and their associated ferricretes classified as a function of their iron content
(Fe203 %) (FYA: young on granitic rocks; FMA and FMB: mid-evolved on acid
and basic rocks, respectively; FOA and FOB: highly evolved on acid and basic rocks,
respectively). Some evolutionary trends are shown in Figure 16.5.
0.20
0.18 1.0.
0.16 0.9 H
J5 0.14
S 0.8.
0.12 0.7.
TT 0.10 0.6.
O 0.5.
ro 0.08
o
2 0.06 0.3 J
0.04.
—i 1 1 1 1 1 1 1 1 Fe O 1 1 1 1 1 1 1 1 1 FP O
0.02. 20 25 30 35 40 45 50 55 60 2 3 20 25 30 35 40 45 50 55 60 ^ 3
PR FERRICRETES PR FERRICRETES
60
50
40_|
9r 30J
25 30 35 40 45 50
FeA 30 35 40 45 50
PR FERRICRETES FERRICRETES
Fig. 16.5. An example of the mineralogical and chemical composition of ferricretes of Gaoua (Burkina
Faso), located on different basic or granitic parent rocks (PR). For low Fe2<>3 contents, compositions
are distinct: ferricretes are lithodependent. For high Fe2C>3 contents, compositions are homogenized:
ferricretes seem to be lithoindependent. (From Tardy et al., 1988.)
TABLE 16.1
Average mineralogical and chemical composition of granitic rocks (PRA) and basic rocks (PRB) and of
their associated ferricretes (37 samples): moderately evolved (FYA, on granities), midi-evolved (FMA,
on granites and FMB, on basic rocks) and highly evolved (FOA, on granites and FOB, on basic rocks;
MEN: average FOA and FOB)
PRA PRB FYA FMA FMB FOA FOB MEN
(" = 6) a (* = ?) ("=2) (" - 18) (" = 10) (" = *) (*=2) (" = ?)
Si02 70.75 55.76 52.20 30.81 25.81 16.49 17.55 16.80
AI2O3 14.56 15.12 10.52 14.87 18.70 14.53 12.94 14.07
Fe203 2.01 9.17 27.59 42.67 42.90 57.54 56.27 57.18
MgO 0.68 4.95 0.09 0.10 0.11 0.04 0.11 0.06
CaO 2.39 7.60 0.20 0.21 0.21 0.20 0.20 0.20
Na20 4.88 3.19 0.05 0.07 0.09 0.08 0.09 0.08
K2O 2.14 0.64 0.26 0.11 0.20 0.06 0.15 0.09
Mn304 0.03 0.17 0.04 0.20 0.15 0.10 0.15 0.12
P2O5 0.11 0.16 0.15 0.27 0.23 0.27 0.31 0.28
Ti02 0.27 0.67 0.53 0.83 0.75 0.89 0.73 0.84
H20 1.27 3.17 7.23 9.45 10.79 9.32 11.48 9.94
Sum 99.17 100.66 98.68 99~45 99.80 99.31 99^89 99.47
Sr 449 317 12 14 20 5 22 10
Ba 656 314 46 206 101 34 120 59
V 34 191 387 846 794 1194 1073 1160
Ni 9 55 48 54 49 69 72 70
Co 8 35 21 44 61 26 48 32
Cr 11 187 160 358 485 416 540 451
Zn 45 123 23 57 67 84 219 122
Cu 13 57 61 109 131 132 162 141
Se 4 25 12 28 37 57 32 50
Y 5 15 11 11 11 11 18 13
Zr 127 75 224 214 210 147 119 139
La 26 9 8 14 11 8 11 9
Ce 44 27 36 80 62 33 43 37
Eu 1.1 1.0 1.0 1.0 1.0 1.0 1.0 1.0
Yb 0.3 1.2 1.4 0.8 1.2 1.4 1.7 1.5
Lu 0.3 0.3 1.0 0.3 0.6 0.7 0.6 0.6
Nb 4 3 16 12 12 13 11 13
Qtz 28 8 43 17 9 3 5 4
Kaol 23 31 41 32 29 30
Goet 27 32 33 29 52 35
Hema 6 17 16 37 15 31
Gib 0 2 2 0 0 0
RPHG b 18 35 32 56 23 47
Note: there are no samples in the group FYB, the moderately evolved ferricretes on basic rocks.
a
n = number of samples
b
RPHG = 100 x hematite/(goethite + hematite).
Quartz and P, Ba, Cr, Se, Ni, Zn, Cu, Ce and La are characteristic of the nature
of the parent rock. In carapaces and in poorly or mid-evolved ferricretes (FYA,
FMA, FMB), which are relatively poor in iron (30-40%, Fe2C>3), the nature of
Geochemistry and evolution of lateritic landscapes 415
the parent rock still can be detected, while in ferricretes rich in iron (55-60%)
mineralogical and chemical compositions are almost totally homogenized. In old
and highly evolved ferricretes, the nature of parent rocks is not recognizable and
there is no way to separate those which are formed either on granites, gneiss, schists,
amphibolites, gabbros, sedimentary or volcanic rocks. Independently of the nature
of parent rocks from which they derive, the final composition is the same: poor
in quartz (5%) and mostly made of kaolinite (30%), goethite (35%) and hematite
(30%). When aging, ferricretes become richer in iron and in hematite relatively
to goethite and poorer in quartz. Kaolinite content remains almost constant. Tïace
elements which are originally located in heavy and resistant primary minerals such
as zircon, ilmenite, cassiterite for example, are progressively redistributed: they are
either leached out, incorporated into or adsorbed on to, goethite, hematite and
kaolinite (Tardy et al, 1988a; Mazaltarim, 1989; Boeglin, 1990; Freyssinet, 1990).
Conclusion
Most of the ferricretes develop through a downward vertical movement of iron

and resistant minerals (such as quartz and zircon) accompanying the lowering of
the landscape surface. They occur as lithodependent autochthonous relicts; at a
landscape scale they are autochthonous and do not result from a large lateral trans-
port. Therefore, the lithodependence principle refers only to the column of parent
rock previously located above, which is in most cases similar to the unweathered
material located below. The degree of lithodependence diminishes with age and
degree of evolution. Along the progress of alteration a chemical homogenization
takes place and all the ferricretes of different origin become similar with time. The
homogenization can also be partly due to lateral movements.
Remobilization of surficial material by termite activity
Erhardt (1951) was probably the first to suggest that termites play an impor-
tant role in ferricrete formation. The subject was discussed at length by McFarlane
(1976). Some researchers such as Russell (1962) believe that what are called vesic-
ular, vacuolar or tubular structures are formed by termite activity, while others like
Griffith (1953), de Swardt (1964) and Millot (1964) deny that termites could be
responsible for lateritic structures and iron accumulation.
After the work of Watson (1972), Weber (1966), Yakushev (1968), Lee and Wood
(1971), Boyer (1973), Lepage et al. (1974), Wood and Sands (1978), Cloud et al.
(1980), Lavelle (1983), Grasse (1984,1986) and others, Eschenbrenner (1986,1987)
has shown how the structure of lateritic profiles and particularly the structure of
glaebular horizons depend on termite activity. The observations can be summarized
as follows (Figure 16.6).
FERRICRETE
CHEMICAL LATERAL TRANSPOR
DISMANTLING IN SOLUTION
LATERAL AND MECHANICAL

TRANSPORT BY GRAVTY AFTER
EROSION
FERRICRETE
MECHANICAL
DISMANTLING
TERMITE ACTIVITY LATERAL TRANSPORT

C O L L U V I U M AND
ALLUVIUM FORMATION
UPWARDS MOTION OF CLAY AND S I L T
FROM MOTTLE ZONES REINCORPORATION

UNDERNEATH DISMANTLING IN SURFICIAL
OF FERRICRETE HORIZONS
Fig. 16.6. Lateral movements of material in landscapes: a. lateral migration of iron in solution; b. lateral
mechanical movements of disaggregated blocs of ferricretes; c. lateral transport of material reworked
by termite activity.
(a) Tèrmite mounds are commonly very abundant on ferricrete plateaus. Termite
activity consists mainly of mechanical remobilization within the whole profile and
within the mottle zone in particular.
(b) Tèrmite activity also results in an upward motion of clay, silt or even sand
particles, which are deposited on the soil surface and then submitted to mechanical
erosion. The depth of uptake [up to 12 or 13 m, according to Boyer, (1973)] is
important in determining what materials are transported. Commonly the uptake
source corresponds mainly to the mottle zone or at least to the top of the lithomarge
of the typical lateritic profile.
For this reason, the surficial silty materials accumulated in lateritic landscapes
show a chemical and mineralogical composition similar to that of the mottle zone.
For instance, the shape and the surface of the gold grains that are commonly found,
Geochemistry and évolution oflateritic landscapes 417
in silty surficial materials, are almost identical to those of the mottle zone but differ
from those of both ferricrete or lithomarge (Figure 16.6; Freyssinet, 1990).
(e) Tèrmite activity may also lead to downward movements of paniculate materi-
als in the vacuolar voids and channels of the profiles. Normal weathering leads to a
decrease in quartz, zircon, tourmaline, cassiterite and other resistant minerals from
the lithomarge to the top of the massive ferricrete horizons. By contrast, termite ac-
tivity and incorporation of surficial material into deeper horizons, yield an increase
in resistant minerals from the bottom to the top and from the massive ferricrete
to the surficial gritty horizon. Consequently most of the chemical properties of the
surface horizons are modified by bioturbation (Figure 16.6).
(d) Because of lateral transport of clay and silt material by mechanical erosion
of the soil surface, termite activity is also responsible for lateral movements of lat-
eritic material along the slopes. Consequently, it may be responsible for the lateral
dispersion of elements and the formation of aureoles and halos well identified in
geochemical prospection.
(e) Tèrmite activity is also most probably responsible for the formation of the soft
clay-silt horizon which in many situations overlies either a bauxite or a ferricrete
or a nodular horizon or a stone line. The soft horizon can even be very thick and,
being clay-rich, can temporarily retain water and induce ferricrete dismantling and
the formation of a stone line (Tkrdy, this book).
Vertical lowering of lateritic landscapes
Despite lateral movements of iron in solution (Maignien, 1964; Millot, 1964;

McFarlane, 1973) and despite lateral transport of fine materials reworked by termite
activity, one can consider that at a landscape scale and at a regional scale most of
the movements are vertical (Tkrdy et al., 1988b). Landscapes evolve with time by a
vertical lowering of soil surface and of weathering fronts.
Gold distribution in latérites

Gold is one of the most valuable of metals which concentrate in latérites. Fur-
thermore native gold, reputed to be very stable and almost insoluble, appears as
a significant marker of the vertical lowering of lateritic landscapes (Freyssinet et
al., 1989). It accumulates in ferricretes together with quartz and other resistant
minerals. In mineralized districts, lateritic weathering profiles can be very thick (50
to 100 m) and quartz veins or fine stockworks bearing the gold mineralization,
may intersect the whole profiles; they remain almost unaltered in the saprolites
and lithomarges. Quartz grains become disaggregated and corroded in the mottled
zone and quartz veins are, as a rule, entirely dismantled when crossing the ferri-
crete. Quartz particles may sometimes remain at the surface in the gritty horizon.
Fig. 16.7. Formation of geochemical halos around mineralized quartz veins. Vertical lowering of the
soil surface, deepening of the weathering front and apparent lateral motion of material upslope or
downslope. (Adapted from Leprun, 1979.)
The gold mineralization remains relatively concentrated around the quartz veins
in saprolites and lithomarges, but becomes concentrated and scattered in the mot
tled zone and highly dispersed in the ferricrete. Ferricrete develops together with a
relative accumulation of gold and resistant minerals and with apparent mechanical
dispersion of these materials. The apparent mechanical dispersions around quartz
veins are evidences for a vertical lowering of the lateritic landscapes through time
(weathering fronts and soil surface; Figure 16.7).
The mechanical dispersion is also associated with a chemical remobilization.
Despite its reputation as an immobile element, several authors have shown that
gold may be mobilised in supergene conditions (Krauskopf, 1951; Cloke and Kelly,
1964; Koroleva and Pogregnyak, 1984; Webster, 1986; Machairas, 1970; Evans,
1981; Herail, 1984; Mann, 1984; Webster and Mann, 1984; Wilson, 1984; Stofïre-
gen, 1986). Morphological and geochemical evidence of chemical migrations and
neoformations of gold in latérites has been found in the mineralized district of
Kangaba, South Mali (Freyssinet et al., 1987). In lateritic profiles, the primary gold
flakes associated with quartz structures are mechanically liberated, chemically dis
persed and concentrated in the residual ferricrete. The chemical dispersion and
concentration include phases of corrosion, dissolution of primary gold and precip
itation of secondary gold. In the weathering profile and towards the top, the gold
grains are progressively corroded, their size decreases and shapes characteristic of
chemical rounding appear. At the same time, secondary gold can be formed in
pores of weathered materials, in etch-pits of goldflakesor at the surface of goethite
and hematite. Microspherulites of flexuous membranes (0.25 ìôç wide), including
weathering ferruginous matrix and dendrites, also develop. Secondary gold can also
crystallize around the primary etched nuggets and is easily distinguished from pri-
mary gold by its fineness and chemical purity. In fact, primary gold yields between
5 to 9% silver, as well as traces of copper (900 ppm), nickel (300 ppm) and sul
fur (200 ppm), while secondary gold is very pure but may include traces of iron,
probably inlayed during the secondary crystallization. The chemical dispersion by
dissolution-precipitation is evidence that ferricretes require a long time to form.
Rates of chemical erosion and weathering and of landscape lowering

By taking into account the composition of waters and of hydrological balances,
the weathering rates, that is, the lengths of time required to form, for example, 1
m of kaolinite lithomarge and all the associated horizons can be estimated. Results
given in Tkble 16.2 show that 1000 mm of water percolating each year through a
profile allows the formation of 1 m of kaolinite lithomarge and consequently allows
a lowering of the weathering fronts of 10 m per Myr or 1000 m per 100 Myr. Thus
lowering of the landscape since the Cretaceous (100 Myr) may be of the order of
several hundreds of metres and at least 1 km (Tardy, 1969, in press; Gac, 1980; Gac
and Tkrdy, 1980).
Schematic evolution of a ferricrete and bauxite covered landscape

At a landscape scale or at a regional scale, iron or aluminum accumulations as
well as soil surfaces and weathering fronts move downwards and mostly vertically.
Ferricretes and bauxites are thus lithodependent. However, the lithodependence is
not on the rocks presently located below but on the rocks previously located above.
Ferricretes and bauxites as well as most lateritic profiles, are relicts or residuals of
large amounts of parent rock which has been altered and dissolved over large thick
ness. Some undissolved resistant minerals (gold or quartz, for example) remaining
at the soil surface and later reincorporated in profiles can produce geochemical ha
los with apparent lateral dispersion of the same order of magnitude as the vertical
lowering of the landscape (Figure 16.8).
TABLE 16.2
Rates of chemical weathering of granitic rocks under various climates
HC03 S1O2 Rainfall Drainage T Weathering rates

(mmoles/kg) H2Q (mm/yr) (mm/yr) fC) (m/Myr) (km/100 Myr) (yr/m)
Norway 0.08 0.05 1200 600 5 3 0.3 350,000

France 0.13 0.15 1200 600 10 8 0.8 126,000
Chad 0.89 1.42 700 0 32 1 0.1 1,000,000
Centrafica 0.62 0.33 1200 200 26 5 0.5 190,000
Ivory Coast 0.09 0.33 1500 550 25 14 1.4 70,000
Malagasy 0.10 0.16 2500 1500 28 33 3.3 30,000
Source: Tardy (1969; in press).
422
PRESENT DAY SURFACE III
Y. Tardy and C. Roquin

Ä=Ä^^
Successions of hydra ted and dehydrated minerals in lateritic profiles
Dehydration reactions in Al, Fé and Mn-systems
Lateritic profiles and particularly bauxites, ferricretes and kaolinitic mantles,

show successions of horizons more or less indurated and sequences of minerals
which are in different hydration states. In the unsaturated zone located above the
ground water table, the less hydrated phases are interstratified between two hori-
zons containing more hydrated minerals. This general distribution is explained by
seasonal fluctuations of the thermodynamic activity of water within the profiles.
Such fluctuations also explains how duricrusts such as ferricretes for example, are
formed, conserved or destroyed (Tkrdy and Novikoff, 1988; Tkrdy et al., 1988a, b).
In lateritic profiles several sequences of hydrated-dehydrated-hydrated min-
erals have been observed: gibbsite-kaolinite-gibbsite in Oxisols (Bourgeat, 1972,
Delvigne, 1965); goethite-hematite-goethite in ferricretes (Nahon, 1976; Tardy
and Nahon, 1985); gibbsite-boehmite-gibbsite in lateritic bauxites (Valeton, 1972;
Boulangé, 1984); or even kaolinite—diaspore (boehmite)-gibbsite in karst bauxites
(Combes, 1969; Bardossy, 1982). These mineral sequences are not always complete,
as given (for example in bauxites). However, the maxima of boehmite, diaspore
and sometimes corundum (Medvedev, 1971) always correspond to the maxima of
hematite and to the optimal development of ooids and pisoids (Bardossy, 1982). A
similar process occurs in manganese duricrusts in which the sequence manganite-
pyrolusite-cryptomelane is observed (Weber et al., 1979; Nahon et al., 1984). In fer-
ricretes in which the mechanism of hydration-dehydration-rehydration was demon-
strated (Nahon, 1976; Tkrdy and Nahon, 1985), the optimal induration and accu-
mulation of iron correspond to the zone of nodule formation and to the maximum
value of the ratio hematite/goethite. The secondary goethite which appears at the
top of the profiles corresponds to a rehydration process which is thought to be re-
sponsible for the destruction of the original nodules and for ferricrete dismantling
(Nahon, 1976).
Our designation of the label hydrated or dehydrated to a mineral is based on the
reactions:
2 FeOOH (goethite) = Fe 2 0 3 (hematite) + H 2 0 (1)
MnOOH (manganite) 4- 0.25 0 2 = Mn0 2 (pyrolusite) + 0.5 H 2 0 (2)
Al(OH)3 (gibbsite) = AlOOH (boehmite) + H 2 0 (3)
Al(OH)3 (gibbsite) + H4Si04(aq) = 0.5 Al 2 Si 2 0 5 (OH) 4 (kaolinite) + 2.5 H 2 0 (4)
Goethite, manganite and gibbsite are hydrated minerals while hematite, pyrolusite,
boehmite and kaolinite are dehydrated minerals. Hydrated minerals appear at the
bottoms of profiles close to the ground water table or downslope where the unsatu-
rated domain is reduced in size. They also appear again close to the surface of thick
profiles, at the edges of large channels, temporarily hydrated during the wet season.
Dehydrated minerals appear in the intermediate part of the unsaturated zone (per
manently dryer). They form concretions, almost always accompany induration, and
contribute to duricrust-formation. The dehydrated minerals are typically minerals of
latérites sensu stricto (subject to hardening).
Gibbsite and /caolinite successions in lateritic profiles
Both weathering of primary minerals and secondary mineral transformation oc-

curing in soils, commonly lead to the exchange of H2O molecules. For the gibbsite-
kaolinite reaction (4) the expression of the equilibrium constant is the following:
log[H4Si04] - 2.5 log[H20] = log* R (5)
Few studies have dealt with the importance of the thermodynamic activity of
water in low P-T reactions (Chesworth, 1972,1975; Bourrié and Pedro, 1979; Mat-
tigod and Kittrick, 1980; Tkrdy, 1982; Tïolard et al., 1986; Tïolard and Tardy, 1987,
in press), and normally it is taken to be equal to 1. This is equivalent to consid
ering H 2 0 to be pure liquid water in its standard state. Thus in equation (5) log
[H 2 0] = 0.
In fact both the thermodynamic activities of silica and of water are important
in defining equilibrium in reaction (4). When the activity of silica, in solution, de
creases, gibbsite may be stabilised relative to kaolinite (Garrels and Christ, 1965;
Gardner, 1970). If the activity of water decreases, the stability field of kaolin
ite enlarges while the gibbsite field is reduced, (Tardy, 1982; Tkrdy and Novikofï,
1988).
The value which has to be chosen for the equilibrium constant depends on several
factors, particularly on the crystal size and the cristallinity of both gibbsite and
kaolinite. Fritz and Tkrdy (1973) chose log [H4S1O4] = -4.5, at 25°C, in order to
take into consideration that in most African soils gibbsite currently forms on basic
bedrocks, from which quartz is absent, and rarely develops on granitic or sandstone
bedrocks, in which quartz is abundant.
The distribution of gibbsite and kaolinite in lateritic profiles is in fact more com
plex than usually described in textbooks. Gibbsite, for example, not only forms in
the near surface horizon, located above the kaolinitic layer, or in the high parts
of the landscapes, but also in the coarse saprolite at the contact with the parent
rock, located below a soft kaolinitic layer, and also in the mid-slopes of the land
scape relief, downslope from kaolinitic profiles (Figure 16.9). Bauxitic deposits of
Amazonia, constantly submitted to an equatorial climate and overlying quartz-rich
sandstones, clearly show that gibbsite and quartz can be closely associated without
Fig. 16.9. Distribution of gibbsite and kaolinite in lateritic profiles of the Ivory Coast, showing: a. only
gibbsite; b. a succession of gibbsite-kaolinite-gibbsite; and c. a succession of kaolinite-gibbsite from
top to bottom of the profiles and from upslope to downslope on the landscapes (after Delvigne, 1965).
forming kaolinite (Lucas, 1989). From this, we suppose that, under permanently hu-
mid conditions, gibbsite and quartz may be more stable than kaolinite, while under
a dryer climate kaolinite is more stable than the gibbsite-quartz association. If log
[H4S1O4] - 2.5 log [H 2 0] = -3.952, at 25°C, and if the activity of water [H 2 0] = 1,
quartz and gibbsite are stable together. When the activity of water decreases and
becomes lower than [H 2 0] < 0.8, kaolinite is more stable than gibbsite and quartz.
Climatic and hydrostatic factors governing the thermodynamic activity of

water in lateritic profiles
The activity of water, written as aw or [H 2 0], is a dimensionless number which

express how much a given state of capillary water (aw < 1) is different from the
standard state of the pure liquid and free water (aw = [H 2 0] = 1). By convention
water in its standard state is a pure water at the pressure of 1 bar.
The difference between energies of water located in a pore either behind a planar
interface (P = 1 bar in both sides, in air and in water; aw = 1) or a curved interface
(P = 1 bar, in air; P < 1 bar, in water; aw < 1) is also equal to the difference
of the chemical potentials characterizing water in the two situations (at P = 1 bar
and at P < 1 bar). Furthermore, the difference of the two chemical potentials is,
by a definition, a function of the activity of water; the activity of water is also equal
to p/p°y the ratio of the partial water pressures in the dry air and in the saturated
air, respectively (Hillel, 1971). This energy difference of water trapped behind a
meniscus, the radius of which is equal to r, is also equal to that required to elevate
one mole of water to a capillary height /ic, so that:
V
ì„ - ì \ = RTÌnav, = RTlnp/p° = VAP = -2ó— = -pgVhc
in which V stands for the molar volume of water, ó for the interfacial tension, p for
the water density and g for the gravitational field.
In a porous medium in which hydration and dehydration processes take place,
the activity of water is controlled by capillarity. At the equilibrium between humid
air and capillary water: the lower the ground water level, the higher the capillary
height, the smaller the radius of the menisci, the smaller the radius of pores in
which water is trapped, the smaller the relative humidity of the air, the smaller the
chemical potential of water, the smaller the water activity and the more the capillary
water environment leads to dehydration of the minerals.
Hydrological factors play an important role in determining mineralogical se
quences in the deep and thick lateritic profiles (Lelong and Lemoine, 1968; Degal-
lier, 1975; Tbulemont, 1984; Vallès, 1986). In the unsaturated zone of such profiles,
water activityfluctuatesduring the year as a function of depth, between the ground
water level ([H2O] = 1) and the soil surface where activity is most of the time
lower than one ([H2O] < 1). Remarkably, a minimum in water activity at about 1
to 5 m depth, is possible, where prevailing conditions are drier than in the lower
horizons (close to the water table) throughout the year, and also drier than in the
surface horizon during the wet season (Vachaud et al., 1978; Vauclin and Vachaud,
1981; Imbernon, 1981; Vauclin, 1982; Figure 16.10). During the dry season dehy
dration takes place at the surface and in this hydric hollow. During the wet season
the quasi-saturation of the surface horizons acts in favour of rehydration of the
minerals previously dehydrated during the dry season. This rehydration takes place
only in pores and voids of large size where the thermodynamic activity of water
was previously high. In the small-size pores the activity of water remains small. The
hydric hollow region in which dehydration takes place almost permanently is the
most probable location for induration and nodule formation and therefore for the
production of ferricretes or other duricusts. Its depth and its thickness is primarily
dependant on the dryness of the climate and on the depth of the ground water
table.
The thermodynamic activity of water is closely correlated with climatic and hy
drostatic factors which both characterize the physical state of water in the unsat
urated zone located above the ground water table. At the piezometric level, water
activity is by definition equal to one ([H2O] = 1). Above the piezometric water level,
water activity is smaller than one ([H 2 0] < 1) and dehydration reactions take place
Geochemistry and evolution oflateritic landscapes
x% RELATIVE HUMIDITY OF AIR 100 %

^ _ DRY SEASON ACTIVITY OF WATER WET SEASON — ^
0 a w *ry season ^ a w e t season 1
H i-
LOW WATER CONTENT HIGH
Fig. 16.10. Seasonal distribution of water activity across a weathering profile, in a non
situation, above a deep ground water level (10 m) (after Tardy et al., 1988a).
(Hillel, 1971; Tkrdy and Nahon, 1985; Tïolard and Tkrdy, 1987; Tkrdy and Novikoiï,
1988; Tardy et al, 1988a).
Chemical composition and climatic distribution of goethite and hematite in

soil profiles and landscapes
Schwertmann (1988) analyzed occurence and formation of iron oxides in var-

ious pedoenvironments. In weathering profiles, the ratio of hematite to hematite
plus goethite [Hm/(Hm + Go)] is controlled by temperature, humidity of the soil
atmosphere, pH, organic matter content, altitude and latitude.
Temperature
The ratio Hm/(Hm + Go) formed from ferrihydrite increases with temperature
(from 4° to 25°C) at which the system is aging. Besides temperature, pH also has a
strong effect (Schwertmann, 1988). In Oxisols, Ultisols and Inceptisols of Brazil, the
proportion of hematite (Hm) compared to goethite increases with increasing mean
annual air temperature (Kämpf and Schwertmann, 1983). In the northern hemi-
sphere, a line drawn at approximately 40°N latitude separates soils without hematite
(North) from soils with hematite (South) (Figure 16.1; Schwertmann, 1988).
Relative humidity of the soil atmosphere
The relative humidity of the soil atmosphere at the contact of which hematite
and goethite form is controlled by soil moisture at a profile scale or at a landscape
scale, and by the difference between rainfall and evaporation. The effect of rela-
tive humidity (RH) on the crystallization of hematite and goethite from synthetic
ferrihydrite is shown by Tbrrent et al. (1982). Amounts of hematite and goethite
both decrease, but Hm/(Hm + Go) increases when the relative humidity of the air
decreases.
On a landscape scale, soils have generally less hematite, yield more goethite
and become yellower and less red with increasing wetness downslope (Curi and
Franzmeier, 1984; Adams and Kassim, 1984; Schwertmann, 1988). Red ferrallitic
soils, rich in hematite and formed of kaolinite-hematite micronodules are gener-
ally located upslope, in well drained conditions, while yellow ferrallitic soils rich
in goethite are located downslope where the water table is higher in the profile
(Chauvel, 1977; Chauvel and Pedro, 1978; Müller, 1987).
On a regional scale, in Southern Brazil for example, goethite is more abundant in
soils subjected to an excess of rainfall over évapotranspiration (Kämpf and Schwert-
mann, 1983).
pH and organic carbon content

With regards to hematite-goethite formation, pH and organic carbon content
of soils are typical kinetic factors. A decrease of pH below 8, and an increase of
organic carbon both act in favour of goethite instead of hematite (Schwertmann,
1965; 1988).
Correlating the ratio Hm/(Hm + Go) with such factors as temperature, relative
humidity, latitude, elevation, pH and organic matter content yields interesting re
sults. Some of these factors, for example altitude (elevation) or latitude, include
the effect of temperature and moisture. Furthermore, on a regional scale, some pa
rameters such as soil pH and organic matter content are closely related to rainfall.
Therefore, it is difficult to separate their specific influence. Soil temperature is reg
ulated by climatic factors such as latitude, altitude and continentality. The relative
humidity of the soil atmosphere, that is, the thermodynamic activity of water, is
controlled by rainfall, evaporation, permeability, pore size, and depth to the ground
water table.
If pH and organic matter are typically kinetic parameters, temperature and ac
tivity of water are both thermodynamic factors which have to be taken into account
as well. In fact, the mass action law applied to the equilibrium hematite-goethite,
yields:
FeOi.5(hematite) + 0.5 H 2 0 (water) = FeOOH (goethite)
_ [FeOOH]
K{T)
" [Fe0 15 ] [H2O]0 5
The mass action law constant, K(T)> is a function of temperature, and the equi
librium conditions are chiefly dependent on the activity of water (*zw = [H20]).
However, hematite-goethite equilibria are also regulated by the activity of the
two solid phases: [FeOOH] for goethite and [FeOi 5] for hematite. Solid phase
activities are determined by three conditions which have to be taken into con
sideration separately: purity, chemical composition (aluminum content) and grain
size. Thus solid phase activities are dependent on both thermodynamic and kinetic
parameters.
Berner (1969), reviewing the stability conditions of hematite in water, has pointed
out that although hematite is more stable than goethite, kinetic phenomena may be
the reason for the non-spontaneous formation of hematite. Langmuir (1969) is in
agreement with Berner's view (1969) thatfine-grainedgoethite is thermodynami-
cally unstable relative to well crystallized hematite. Later, Schwertmann and Taylor
(1977) and Kämpf and Schwertmann (1983) also indicated that kinetics is the major
factor that governs the rate of dissolution or precipitation of the iron oxides and
oxyhydroxides. However, natural observations made on ferricretes suggest that be
side the kinetic interpretation, the distribution of goethite and hematite and their
Al-content variations can be understood in terms of thermodynamically metastable
equilibria, in which not only the size of the grains but also the temperature and the
chemical potential of water are involved.
Distribution ofAl-hematite and Al-goethite in latérites
Tkrdy and Nahon (1985) and Ttolard and Tkrdy (1987, in press) have reviewed
the abundant littérature devoted to Al-goethite, Al-hematite and Fe3+-kaolinite
distribution in latérites.
The aluminum contents of goethite and hematite should be regarded as a func-
tion of the nature of other minerals with which they are associated. In goethites
of red soils, the substitution rates diminish when the climate becomes less humid
and percolating solutions are richer in silica (Tkrdy, 1971). This is probably due to
the presence of kaolinite which controls the chemical composition of the solution
(Ttolard and Tkrdy, in press).
Compositions range from 2 to 20 mole percent of AIO(OH) in goethites of
ferricretes and other soils in which kaolinite is present. Compositions range from
18 to 27 mole percent in goethites associated with gibbsite in bauxites. Thus, the
substitution rate of AIO(OH) into FeO(OH) is lower in ferricretes than in bauxites,
as generally admitted by Nahon (1976), Nahon et al. (1977) and Leprun (1979), in
latérites of Senegal and Ivory Coast. Hematite generally has less substitution than
goethite and both are less aluminous when kaolinite and quartz are present than
when gibbsite forms.
Obviously, the aluminum contents of goethites and hematites are controlled by
the activity of aluminum in solution which in turn is determined by the solubility of
aluminous minerals, that is gibbsite and kaolinite. It turns out that the relative sta-
bility fields of goethite and hematite may depend on their degree of Al-substitution
which is dependent on the stability of the other aluminous minerals with which they
are associated. The conditions of formation of goethite and hematite may be ulti-
mately determined by the activity of silica in solution and by the activity of water in
pores in which these minerals form.
Thermodynamic stability field of Al-goethite and Al-hematite
Stability fields of Al-goethite and Al-hematite, gibbsite, boehmite and kaolinite

can be considered to be a function of particle size, temperature and the thermody-
namic activity of water (TVolard, 1988; Tïolard and Tkrdy, 1987,1989).
In ferricretes or in synthetic samples, the particle sizes of Al-goethite crystals
commonly range from 100 to 400 Â, while those of Al-hematite range from 40
to 400 Â (Schwertmann, 1965; Langmuir, 1969, 1971; Langmuir and Whittemore,
1971; Amouric et al., 1986; Koch et al, 1986).
Stability diagrams
For a given temperature, Ã, and according to the equilibrium conditions ex

pressed earlier, stability diagrams are constructed as a function of the aluminum
content and water activity. Five simplified diagrams are given in Figure 16.11 for 5°,
15°, 40°, 55° and 80°C. When temperature decreases, several modifications are ob
served. The stabilityfieldof the association of Al-hematite and boehmite is reduced.
The stability field of the association of the Al-goethite and gibbsite becomes larger.
For a given composition and a given water activity, the Al-substitution in Al-goethite
associated with gibbsite decreases. When temperature increases the contrary is ob
served. The stability field of Al-goethite is reduced, gibbsite disappears near 40°C
and is replaced by boehmite; the stability field of Al-hematite is enhanced; at about
75°C, Al-goethite disappears.
For a constant water activity ([H2O] = 1, for example) and for a given composi
tion [Al203/(Fe2C>3 + AI2O3)] the degree of substitution of diaspore in Al-goethite,
when Al-goethite is associated with gibbsite, increases with temperature. The degree
of substitution of corundum in Al-hematite also increases with temperature. These
variations are represented in Figure 16.12.
T=5eC T=15°C
T=25°C
Mole fraction : AIA

(Fe203+Al203>
1/3
8»
Caption:
0.9 mmm-
·?:·:►:►:·:·:::·:¥:::
Goe AI
mmm
b
Goe A , +Gib
mmm
mmm
':£&$$$K-
GoeA| +Boe
0.0 0.5
Moles of Al203 per 1 mole of Fe203
OB
03
►g GoeAI +HemA,
HemAi
;
0 1/a
Fig. 16.11. Stability diagrams for Al-goethite, Al-hematite, gibbsite and boehmite at 25°C and at different
temperatures as a function of water activity (from Trolard and Tardy, 1987).
[H20]= 0.9 [H 2 0]=0.8

QJ
►
[H20] =1.0
m
50 ]
Mole fraction : AIA ^Éß
(Fe 2 0 3 MI 2 0 3 )
100 -I
1/3
~^^1
Ö
3
ES
(0
k.
50 Jjl 1
Ö
Goe+Boe
looJHiik—-I
Q.
E Al 100
1/2
HemA1+Boe Caption:
Al
Goe Al
GoeA, +Gib
0.0 0.5 GoeAi +Boe
GoeA) + HemA,
Moles of ALCL per 1 mole of Fe O HemAi
Hem A, +Boe
0 1/2
Fig. 16.12. Stability diagrams for Al-goethite, Al-hematite, gibbsite and boehmite at [H 2 0] = 1 and at
different activity of water, as function of temperature (from Trolard and Tardy, 1987).
Figures 16.11 and 16.12 show that an increase in temperature and a decrease in
water activity, have qualitatively similar effects on the stabilityfields.This is because
the entropy and the heat capacity of water in gibbsite (Al203.3 H2O), goethite
(Fe 2 0 3 .H 2 0) and boehmite (A1203.H20) are smaller than those characterizing
liquid water.
Conclusions
It is clear that the chemical potential of water plays a major role in the transfor
mation of gibbsite into boehmite, as already pointed out by Chesworth (1972), for
the transformation of boehmite into corundum at low temperature (Bardossy, 1982)
and for the transformation of goethite into hematite (Tardy and Nahon, 1985).
Equilibrium diagrams show the conditions of formation of Al-goethites, Al-
hematites, gibbsite and boehmite in latérites, bauxites and ferricretes. The stability
fields of the various associations are presented in terms of the thermodynamic ac
tivity of water, temperature and particle size. The diagrams allow for substitutions
of Al for Fe in goethite and hematite and we have regarded these solid solutions as
ideal.
The fact that an ideal solid solution must extend continuously from one member
to the other is masked in the diagram by the appearance of other stability fields
which partly overlap the solid solutionfieldsof AlOOH-FeOOH and Al 2 03-Fe 2 03.
The assumptions of ideal solid solutions leads to the prediction of composition limits
for Al-goethites (<30%, mole fraction of diaspore) and Al-hematites (<15%, mole
fraction of corundum) which agree very well with Al-contents of natural associations
as reviewed by Tardy and Nahon (1985).
In summary, the stability diagrams based on activity of water and temperature
permit the following conclusions.
(a) Dehydration transformations of the type: goethite —▲ hematite; gibbsite —>
boehmite —▲ corundum; or gibbsite —▲ kaolinite, can be promoted by a decrease in
the thermodynamic activity of water (at constant Ã), by an increase in temperature
(at constant aw) or by aw decreasing and T increasing simultaneously.
(b) A decrease in water activity has a similar effect to an increase in temperature.
The successions of mineral phases and parageneses are similar in both cases: (1) Al-
goethite -*· Al-goethite + Al-hematite —▲ Al-hematite, for systems of low aluminum
content, and (2) Al-goethite + gibbsite —▲ Al-goethite + boehmite —▲ Al-hematite
-* (Fe-corundum), for systems of high aluminum content.
(c) When minerals are in association such as (Al-goethite + gibbsite) or (Al-
goethite + Al-hematite), the Al-goethite or Al-hematite compositions do not de
pend on the composition of the system but only on water activity or temperature.
The aluminum contents increase as water activity decreases or temperature in
creases.
(d) When Al-goethite and boehmite are stable together, the Al-goethite compo
sition does not depend on the composition of the system nor on temperature and
water activity. It is remarkable that, in this case, the aluminum content of goethite
is almostfixed[for example, (Alo.25 Feo.75s)OOH at 25°C]. The same considerations
are valid for the association of hematite and corundum.
(e) When only Al-goethite or Al-hematite are stable, their aluminum content
depends neither on water activity nor on temperature. The Al-goethite and Al-
hematite compositions depend only on the ratio of Á12è3/(Á12è3 + Fe2C>3) in the
system.
(f) The stability field of kaolinite in the presence of gibbsite or boehmite is
dependent both on activity of water and activity of silica. At 25°C and 100 kPa total
pressure, if the silica activity is lower than 10"4·57, kaolinite is not stable whatever
the activity of water. In the dissolved silica activity interval [10~4 57 to 10~4 50]
gibbsite, then kaolinite and finally boehmite can be stabilized, successively, when
the water activity decreases. In the interval [10~4 5 to 10~4 °] only kaolinite and
boehmite are stable. For a decreasing water activity, the stability field of boehmite
progressively overlaps the kaolinite domain (Tïolard and Tkrdy, 1989).
(g) The presence of kaolinite instead of gibbsite and boehmite in soils or weather
ing profiles should promote changes in the Al-content of aluminous goethite, as the
equilibrium with kaolinite decreases when water activity or silica activity increases.
This may explain why, in general, goethites are poorer in aluminum downslope, in
lowlands and in hydromorphic soils compared to Ultisols or Oxisols located at the
top of profiles or in high landscape situations where the aqueous pore solutions are
depleted in silica.
(h) The iron content in kaolinite is shown to be dependent on the water activity
and on dissolved silica activity. An increase of water activity at a fixed silica activity
or an increase of silica activity at a fixed water activity should induce an increase of
iron relative to aluminum in the ferruginous kaolinite. However, the dominant fac-
tor which probably controls the iron content in ferruginous kaolinite is its solubility
related to its ciystallinity which generally decreases from the bottom to the top of
the profiles: the lower the crystallinity, the higher the solubility and consequently
the higher the iron-content in ferruginous kaolinite.
Clearly the chemical composition of goethite, hematite and kaolinite are con-
trolled by equilibrium with other phases present in the system, by silica activity
regulated by the rate of percolating waters, by temperature and by the activity of
water which is a function of climatic and regional drainage conditions.
Periatlantic climates and paleoclimates; the role of temperature and water

activity in the distribution and mineralogical composition of bauxites and
ferricretes
Since the Jurassic, parts of the Brazilian and the African shields have been
subjected to either equatorial, humid tropical, or arid tropical climates. Both are
covered by a thick and widely distributed lateritic mantle, the age, the nature and
the mineralogical composition of which varies from one region to another. Bauxites
form under hot and humid equatorial climates (25°C and rainfall > 1800 mm per
year), while ferricretes develop under seasonally contrasted tropical climates (30°C
and rainfall between 1300 and 1700 mm per year). Since the Jurassic and the open-
ing of the Atlantic ocean, Africa and America have been moving northwards and
equator and tropics have been shifted southwards. Consequently the surface of the
periatlantic continents were covered by latérites of different kinds and of different
ages (Figures 16.13 and 16.14).
In the southern hemisphere, and particularly in Brazil and in South- and South-
East Africa, ferricretes are not widely distributed, but bauxites are abundant and
widely developed.
Ferricretes are not widely distributed, either because they were previously formed
under contrasted tropical climates and later dismantled under equatorial climates
or because paleoclimates were never sufficiently hot and contrasted or because
contrasted paleoclimates did not operate during a long enough time (Tardy et al,
1988b). In Brazil for example, scattered hematitic ferricretes were described in the
0 1000 km 0 1000 km
Fig. 16.13. Paleoclimatic potential distribution of bauxites (dark shadowed) in equatorial domains of
ferricretes (light shadowed) in contrasted tropical zones of Africa (from Parrish et al., 1982, and after
Kobilsek, 1989).
Nordeste, Goias and Matto Grosso and seem to be preserved in these semi-humid
regions. Ferricretes also outcrop in several parts of Amazonia but appear to have
been dismantled under the rain forest equatorial climate (Melfi et al., 1979; Tardy et
al., 1988b; Nahon et al, 1989).
Ferricretes are also present in South Africa (Milnes et al., 1987; Milnes and
Fitzpatrick, 1987; Fitzpatrick, 1988). However, they have not the same extent and
the same development as in the northern hemisphere. Particularly in western and
central Africa, hematitic nodular or pisolitic ferricretes are widely developed (Maig-
nien, 1958; Michel, 1973; Nahon, 1976; McFarlane, 1976; Leprun, 1979; Petit,
Fig. 16.14. Paleoclimatic potential distribution of bauxites (dark shadowed) in equatorial domains of
ferricretes (light shadowed) in contrasted tropical zones of South-America (from Parrish et al., 1982,
and after Kobilsek, 1989).
1985). Bauxites are widespread in Guinea, Ivory Coast, Nigeria and Cameroon.
Some bauxites are very old and seem to have been formed since the Jurassic and
through the Cretaceous and the Eocene in near equatorial conditions (Michel,
1973; Boulangé et al., 1973; Hieronymus, 1973; Boulangé, 1984; Valeton and Beiss-
ner, 1986; Figures 16.13 and 16.14). In South America bauxites may be very old
(Jurassic) or younger (Miocene) (Weber, 1959; Dennen and Norton, 1977; Aleva,
1981; TYuckenbrodt and Kotschoubey, 1981; Kronberg et al., 1982; Groke et al.,
1982; Melfi and Carvalho, 1983; Lemos and Villas, 1983; TYescases and Melfi, 1985);
in South-East Africa, they are reported to be essentially Eocene (Grubb, 1973,1979;
Geochemistry and evolution oflaterìtic landscapes 435
Parrish and Curtis, 1982; Parrish et al., 1982; Fitzpatrick, 1983, 1988; Patterson et
al., 1986). In both cases, for a million years climates have been continuously humid
and paleoclima tes have been evolving from arid to humid (Figure 16.15; Tardy et al.,
1988b). In both cases, bauxites are dominantly gibbsitic and generally show massive
structures with no pisolite development nor boehmite formation.
The remaining old bauxitic latérites which are abundant in Mali and in Burkina
Faso located at the edges of Sahara yield mainly gibbsite but also some boehmite
(Balkay and Bardossy, 1967), formed later together with a secondary development
of pisolites. In these regions, probably at the end of the Eocene or since the
Miocene, under contrasted tropical climates, gibbsitic bauxites ceased to form and
hematitic and kaolinitic ferricretes began to develop. The secondary formation of
boehmite and pisolites in bauxite is probably contemporaneous with the extension
of the ferricrete, occuring when equatorial climates were replaced by tropical, as the
Sahara progressively moves southwards (Tkrdy et al., 1988b). Thus on a continental
scale, regions in which gibbsitic bauxites were formed continuously or over a long
period of time do not show a wide development of ferricretes, while those which
exhibit boehmitic and pisolitic bauxites are covered by a large mantle of surficial
iron accumulation. This is particularly the case in western and central Africa, where
ferricrete may have been formed since the end of the Cretaceous (McFarlane, 1976;
Tardy et al, 1988).
In the present-day equatorial regions covered by rain forest, gibbsitic and goethit-
ic soils are now forming. However, ferricretes that are being dismantled also out
crop (Novikoff, 1974; Volkoff, 1985). It is clear that some, if not all, of the hematitic
nodular red soils or nodular stone lines may result from a secondary transforma
tion of ancient ferricretes formed earlier under seasonally contrasted climates and
presently submitted to humid equatorial conditions. This is clearly the case in a
large part of the Congo and the Amazon basins (Nahon et al, 1989).
Earlier it was demonstrated that the relative stability of goethite compared to
hematite, and gibbsite compared to boehmite, is controlled by two thermodynamic
and climatic factors, that is, temperature (Ã) and water activity (aw). Thermodynam
ic activity of water is a parameter equivalent to the relative humidity of the air [aw =
p/p° = 100HR (%)] which characterizes the dryness or the humidity of the climates.
Both an increase of temperature and a decrease of the water activity, can lead to
the formation in latérites, of dehydrated minerals instead of gibbsite [Al(OH)3]
(Tardy and Nahon, 1985; Tïolard and Tardy, 1987). Consequently, tropical climates
such as those prevailing south of the Sahara, and characterized by a mean annual
temperature of about 30°C and by a dry season of several months during which the
relative humidity of the air becomes lower than 70%, encourage the formation of
hematite and kaolinite in iron accumulations and of secondary boehmite in gibbsitic
bauxites, previously formed.
During the last 100 Myr and due to the slow drift of continents, climatic zones
have progressively shifted. During the Jurassic, the equator was located in what is
4^
0\
MA: MANAUS; BE: BELEM; RE: RECIFE; BEL: BELO HORIZONTE; SP: SAO PAULO KE: IM KEBDENE; BA: BAMAKO; KO: KORHOGO; YA : YAOUNDE; Z E : ZEMIO
BR: BRASILIA; CUIABA KI : KINSHASA ; DA : DAR ES SALAM
Fig. 16.15. Paleoclimates and succession of potential lateritic domains in Brazil and in Africa. 1: Domain submitted to contrasted tropical climate and ^
probably continuously covered by ferricretes since Cretaceous (100 Myr) or Eocene (50 Myr). 2: Exclusively bauxites: equatorial domain covered by H
bauxites since Cretaceous (100 Myr). 3: Bauxite transformation and ferricrete formation domain previously submitted to equatorial climate and later 2.
exposed to contrasted tropical climates; previously covered by bauxites and later by ferricretes. 4: Ferricrete dismantling and bauxite formation, in ö
domains previously covered by ferricretes and currently evolving under the rainforest equatorial climates. I!.
n
s*
now Mauretania and Northern Venezuela. Then Africa rotated and the two con-
tinents were displaced northwards, shifting the equator southwards (Parrish et al.,
1982). Both Brazil and southeast Africa, initially more arid, have become progres-
sively more humid. By contrast, west and central Africa, more humid to begin with,
later became dryer and less humid. Western and central Africa on one hand, and
southeast Africa and central Brazil on the other, during the Cretaceous and the Ter-
tiary, were subjected to climates which, independently of global fluctuations, have
been changing in an opposite way. Consequently, some climatic contrasts, which
today separate the first from the second group of continental regions, might have
also operated in the past
Finally, the spatial and temporal distribution as well as the observed mineralog-
ical composition of bauxites and ferricretes, in Brazil and in Africa, not only result
from the diversity of the present-day climates but also reflect the paleoclimatic
evolution which took place since the latest Jurassic and the opening of the Atlantic.
References
Adams, W.A. and Kassim, J.K., 1984. Iron oxyhydroxides in soils developed from Lower Paleozoic
secondary rocks in mid Wales and implication for some pedogenic processes. J. Soil Sci., 35: 117-
126.
Aleva, GJ.J., 1981. Essential differences between the bauxite deposits along the southern and northern
edges of the Guiana shield, South America. Econ. Geol., 76:1142-1152.
Amouric, D., Baronnet, A., Nahon, D. and Didier, P., 1986. Electron microscopic investigations of iron
oxyhydroxides and accompanying phases in lateritic iron-crust pisolites. Clays Clay Miner., 34: 45-
52.
Balkay, B. and Bardossy, G., 1967. Etude des processus élémentaires de la laterisation sur latérites
guinéennes. Földt. Közl. Bull. Soc. Géol. Hongr., 1: 91-110.
Bardossy, G., 1982. Karst Bauxites. Developments in Economy Geology, 14. Elsevier, Amsterdam, 441
pp.
Beaudet, G., 1978. Essai sur la zonation et la signification géomorphologique des cuirasses ferrugineuses
en Afrique Occidentale. Trav. Doc. CEGET, Bordeaux, 33: 35-63.
Beauvais, A., 1990. Thèse, Univ. Louis Pasteur, Strasbourg (in prep.).
Berner, R.A., 1969. Goethite stability and the origin of red beds. Geochim. Cosmochim. Acta, 33: 267-
273.
Blot, A. and Leprun, J.C., 1973. Influence de deux roches mères de composition voisine sur les
altérations et les sols. Un exemple sur le socle cristallin au Sénégal Oriental. Can. ORSTOM Sér.
Géol., 5: 45-57.
Blot, A., Leprun, J.C. and Pion, J.C, 1976. Originalité de l'altération et du cuirassement des dykes
basiques dans le massif de Saraya (Sénégal oriental). Bull. Soc. Géol. Fr., 18: 45-49.
Blot, A., Leprun, J.C. and Pion, J.C, 1978. Corrélations géochimiques entre les cuirasses ferrug
ineuses et les roches du socle cristallin au Sénégal et en Haute-Volta. Lithodépendance et héritage
géochimique. CR. Acad. Sci. Paris, Sér. D, 286:1331-1334.
Boeglin, J.L., 1990. Thèse, Univ. Louis Pasteur, Strasbourg (in prep.).
Boulangé, B., 1984. Les formations bauxitiques latéritiques de Côte d'Ivoire. Trav. Doc. ORSTOM,
Paris, 175, 341 pp.
Boulangé, B., Delvigne, J. and Eschenbrenner, V, 1973. Descriptions morphoscopiques, géochimiques

et minéralogiques des faciès cuirassés des principaux niveaux géomorphologiques de Côte d'Ivoire.
Can. ORSTOM, Sér. Géol., V: 59-81.
Boulet, R., 1978. Existence de systèmes à forte différenciation latérale en milieu ferrallitique guyanais:
un nouvel exemple de couvertures pédologiques en déséquilibre. Sci. Sol, 2: 75-82.
Bourgeat, E, 1972. Sols sur socle ancien à Madagascar. Mém. ORSTOM Paris, 57, 335 pp.
Bourrié, G. and Pedro, G., 1979. La notion de pF, sa signification physicochimique et ses implications
pédogénétiques. I — Signification physicochimique. Relation entre pF et activité de l'eau. Sci. Sol.,
4: 313-322.
Boyer, Ph., 1973. Action de certains termites constructeurs sur l'évolution des sols tropicaux. Première
partie: les termites et le sol. Ann. Sci. Nat. Zool., Paris, 15: 329-498.
Chauvel, A., 1977. Recherches sur la transformation des sols ferraliitiques dans la zone tropicale à
saisons contrastées. Trav. Doc. ORSTOM, 62, 532 pp.
Chauvel, A. and Pedro, G., 1978. Genèse de sols beiges (ferrugineux tropicaux lessivés) de Casamance
(Sénégal). Can. ORSTOM, Sér. Pédol., 16: 231-249.
Chesworth, W, 1972. The stability of gibbsite and boehmite at the surface of the earth. Clays Clay
Miner., 20: 369-374.
Chesworth, W, 1975. Soil minerals in the system AI2O3-S1O2-H2O: phase equilibrium model. Clays
Clay Miner., 23: 55-60.
Cloke, P.L. and Kelly, W.C., 1964. Solubility of gold under inorganic supergene conditions. Econ. Geol.,
59: 259-270.
Cloud, L., Gustafson, L.B. and Watson, J.A.L., 1980. The works of living social insects as pseudofossils
and the age of the oldest known metazoa. Science, 210: 1013-1015.
Combes, J.P., 1969. Recherches sur la genèse des bauxites dans le Nord-Est de l'Espagne, le Languedoc,
et l'Ariège (France). Mém. CRGH, Montpellier, 432 pp.
Curi, N. and Franzmeier, D.P., 1984. Toposequence of oxisols from the Central Plateau of Brazil. Soil
Sci. Soc. Am. J., 48: 341-346.
D'Hoore, J., 1954. âccumulation des sesquioxydes libres dans les sols tropicaux. Pubi. Ineac, Sér. Sci.
62, Bruxelles, 132 pp.
Degallier, R., 1975. Interprétation des variations naturelles du niveau des nappes souterraines. Appli
cation aux données provenant du bassin versant de Korhogo en Côte d'Ivoire. Thèse Sci., Univ.
Montpellier, 231 pp.
Delvigne, J., 1965. Pédogenèse en zone tropicale. La formation des minéraux secondaires en milieu
Dennen, W.H. and Norton, D.A., 1977. Geology and geochemistry of bauxite deposits in the Lower
Amazon Basin. Econ. Geol., 72: 82-89.
Erhart, H., 1951. Sur l'importance des phénomènes biologiques dans la formation des cuirasses ferrug
ineuses en zone tropicale. C. R. Acad. Sci. Paris, 233: 804-806.
Eschenbrenner, V, 1986. Contribution des termites à la microagrégation des sols tropicaux. Can.
ORSTOM, Sér. Pédol., 22: 397-408.
Eschenbrenner, V, 1987. Les glébules des sols de Côte d'Ivoire. Thèse Sci., Univ. Bourgogne, 498 pp.
Evans, D.L.C., 1981. Lateritisation as possible contribution to gold placers. Eng. Min. J., 182: 86-91.
Fitzpatrick, R.W, 1983. Occurence and mineralogy of ferruginous bauxite along the eastern seaboard of
South Africa. In: R. de Villiers and P.A. Cawthorn (Editors), Proceedings First Int. Congr. Applied
Mineralogy (ICAM 87), Spec. Pubi. 7, Geol. Soc. S. Afr., Johannesburg, pp. 87-96.
Fitzpatrick, R.W, 1988. Iron compounds as indicators of pedogenic processes: examples from the south
ern hemisphere. In: J.W. Stucki, G.A. Goodman and U. Schwertmann (Editors), Iron in Soils and
Clay Minerals. Reidel, Dordrecht, pp. 351-396.
Freyssinet, Ph., 1990. Géochimie et minéralogie des latérites du Sud-Mali. Evolution des paysages.
Prospection géochimique de l'or. Thèse, Univ. Louis Pasteur, Strasbourg (in prep.).
Freyssinet, Ph., Zeegers, H. and Tardy, Y., 1989. Morphology and geochemistry of gold grains in latente
profiles of southern Mali. In: S.E. Jenness (Editor), Geochemical Exploration 1987. J. Geochem.
Explor., 32: 17-31.
Freyssinet, Ph., Zeegers, H. and Tardy, Y., 1987. Néoformation d'or dans les cuirasses latéritiques:
dissolution, migration, précipitation. C. R. Acad. Sci., Paris, Sér. II, 305: 867-874.
Fritz, B., 1975. Etude thermodynamique et simulation des réactions entre minéraux et solutions. Appli
cation à la géochimie des altérations et des eaux continentales. Sci. Géol. Mém., Strasbourg, 41,152
pp.
Fritz, B. and Tardy Y., 1973. Etude thermodynamique du système gibbsite, quartz, kaolinite, gaz car
bonique. Application à la genèse des podzols et des bauxites. Sci. Géol. Bull., 26: 339-367.
Gac, J.Y., 1980. Géochimie du Bassin du Lac Tchad. Bilan de l'altération, de l'érosion et de la
sédimentation. Trav. Doc. ORSTOM, Paris, 123, 251 pp.
Gac, J.Y. and Tardy, Y., 1980. Géochimie d'un paysage tropical. Le bassin du lac Tchad. In: Y. Tardy
(Editor), Géochimie des interactions entre les eaux, les minéraux et les roches. SARL, Elements,
Tarbes, pp. 199-239.
Gardner, L.R., 1970. A chemical model for the origin of gibbsite from kaolinite. Am. Minéral., 55:
1380-1389.
Garrels, R.M. and Christ, C.L., 1965. Solutions, Minerals and Equilibria. Harper and Row, New York,
N.Y., 450 pp.
Grandin, G., 1976. Aplanissements cuirassés et enrichissement des gisements de manganèse dans
quelques régions d'Afrique de l'Ouest. Mém. ORSTOM Paris, 82, 268 pp.
Grasse, PP., 1984. Termitologia. 2: Fondation des sociétés. Construction. Singer-Polignac, Masson, Paris,
613 pp.
Grasse, PP., 1986. Termitologia. 3: Comportement, Sociabilité, Écologie, Évolution, Systématique.
Singer-Polignac, Masson, Paris, 715 pp.
Griffith, G., 1953. Vesicular laterite. Nature (London), 171: 530.
Groke, M.C.T., Melfi, AJ. and Carvalho, A., 1982. Bauxitic alteration on basic and alkaline rocks in
the state of Säo Paulo, Brazil, under tropical humid climate. Proc. 2nd Int. Seminar on Laterisation
Processes, Säo Paulo, pp. 237-250.
Grubb, PL., 1973. A preliminary account of bauxite in Rhodesia, Mozambique and Malawi. 3rd IC-
SOBA International Symposium, Nice, pp. 223-232.
Grubb, PL., 1979. Genesis of bauxite deposits in the Lower Amazon Basin and Guianas Coastal Plain.
Econ. Geol., 74: 735-750.
Herail, G., 1984. Géomorphologie et gitologie de l'or détritique. Piedmonts et bassins intramontagneux
du NW de l'Espagne. Editions du CNRS, 456 pp.
Hieronymus, B., 1973. Etude minéralogique et géochimique des formations bauxitiques de l'Ouest du
Cameroun. Can. ORSTOM, Sér. Géol., 5: 97-110.
Hillel, D., 1971. Soil and Water. Academic Press, London, 288 pp.
Imbernon, J., 1981. Variabilité spatiale des caractéristiques hydrodynamiques d'un sol du Sénégal.
Application au calcul d'un bilan sous culture. Thèse, Univ. et Inst. Nati. Polytechnique de Grenoble,
152 pp.
Kämpf, N. and Schwertmann, U., 1983. Goethite and hematite in a climosequence in southern Brazil
and their application in classification of kaolinitic soils. Geoderma, 29: 27-39.
King, L.C., 1948. On the ages of the African land surfaces. Q. J. Geol. Soc., 104: 439-459.
King, L.C., 1957. The geomorphology of Africa, I. Erosion surfaces and their mode of origin. Sci. Prog.
London, 45: 672-681.
King, L.C., 1962. The Morphology of the Earth. Olivier and Boyd, London, 699 pp., 2nd ed., 1967.
Kobilsek, B., 1989. Etude pétrographique et dynamique d'un gisement bauxitique d'Amazonie brési-
Henne. Distribution des cuirasses bauxitiques et ferrugineuses en fonction des paléoclimats dans
l'hémisphère sud. Thèse Univ. Louis Pasteur, Strasbourg, in preparation.
Koch, G, Madsen M., Morup, S., Christiansen, G., Gerward, L. and Villadsen, J., 1986. Effect of heating
on microcrystalline synthetic goethite. Clays Clay Miner., 34: 17-24.
Koroleva, G.P. and Pogregnyak, Y.F., 1984. Experimental study of the dissolving and transfer of gold by
some acqueous solutions. Sov. Geol. Geophys., 25: 77-81.
Krauskopf, K.B., 1951. The solubility of gold. Econ. Geol., 46: 858-878.
Kronberg, B., Fyfe, W.S., Mckinnon, B.J., Couston, J.EG, Stilianidi, B. and Nash, R.A., 1982. Model
for bauxite formation (Paragominas, Brazil). Chem. Geol., 35: 311-320.
Lamotte, M. and Rougerie, G., 1962. Les apports allochtones dans la genèse des cuirasses ferrugineuses.
Rev. Géomorphol. Dyn., 10(13): 145-160.
Langmuir, D., 1969. The Gibbs free energies of substances in the system Fe203-CC>2 at 25°G U.S.
Geol. Surv., Prof. Pap., 650 B: 180-184.
Langmuir, D., 1971. Particle size effect on the reaction goethite = hematite + water. Am. J. Sci., 271:
147-156.
Langmuir, D. and Whittemore, D., 1971. Variations in the stability of precipitated ferric oxyhydroxides.
Adv. Chem. Ser., 106: 209-234.
Lavelle, P., 1983. The soil fauna of tropical savannas, II. The earthworms. In: F. Bourbière (Editor),
Tropical Savannas. Elsevier, Amsterdam, pp. 485-504.
Lee, K.E. and Wood, T.G., 1971. Physical and chemical effects on soils of some Australian termites and
their pedological significance. Pedobiologia, 11: 376-409.
Lee, K.E. and Wood, T.G., 1971. Termites and Soils. Academic Press, London, 251 pp.
Lelong, F and Lemoine, J., 1968. Les nappes phréatiques des arènes et des altérations argileuses. Leur
importance en zone intertropicale. Les difficultés de leur exploitation. Bull. BRGM, 3(2): 41-52.
Lemos, VP. and Villas, R.N., 1983. Alteraçao supergenica das rochas basicas do grupo Grao-Para.
Implicacoes sobre de a genese do deposito de bauxita de N5, Serra dos Carajas. Rev. Bras. Geoci.,
Säo Paulo, 13, 3:165-177.
Leneuf, N., 1959. Laltération des granites calco-alcalins et des granodiorites en Côte d'Ivoire forestière
Lepage, M., Morel, G. and Resplendido, CI., 1974. Découverte de galeries de termites atteignant la
nappe phréatique profonde dans le nord du Sénégal. GR. Acad. Sci., Sér. D, Paris, 278:1855-1858.
Leprun, J.G, 1979. Les cuirasses ferrugineuses des pays cristallins de l'Afrique Occidentale sèche.
Génèse-Transformation-Dégradation. Sci. Géol. Mém., 58, Strasbourg, 224 pp.
Lucas, Y., 1989. Systèmes pédologiques en Amazonie brésilienne. Equilibres, déséquilibres et transfor
mations. Thèse, Univ. Poitiers, 157 pp.
Machairas, G., 1970. Contribution à l'étude minéralogique et métallogénique de l'or. Bull. BRGM, 2:
86-91.
Maignien, R., 1958. Le cuirassement des sols en Guinée, Afrique Occidentale. Mém. Serv. Carte Géol.
Alsace Lorraine, 16, 239 pp.
Maignien, R., 1964. Compte rendu des recherches sur les latérites. UNESCO NS/HT/125, WS/0264.32-
NA, 266 pp.
Maignien, R., 1966. Compte rendu de recherches sur les latérites. Col. Rech. Res. Nat., 4, UNESCO,
Paris, 155 pp.
Mann, A.W, 1984. Mobility of gold and silver in lateritic weathering profiles: some observations from
western Australia. Econ. Geol., 79: 38-49.
Mattigod, S.V. and Kittrick, J.A., 1980. Temperature and water activity as variables in soils mineral
activity diagrams. Soil Sci. Soc. Am. J., 44: 149-154.
Mazaltarim, D., 1989. Geochimie des cuirasses ferrugineuses et bauxitiques de l'Afrique Occidentale et
Centrale. Thèse, Univ. Louis Pasteur, Strasbourg, 263 pp.
McFarlane, M.J., 1973. Laterite and topography in Buganda. Uganda J., 36: 9-23.
McFarlane, M.J., 1976. Laterite and Landscape. Academic Press, London, 151 pp.
McFarlane, M.J., 1983. Latentes. In: A.S. Goudie and K. Pye (Editors), Chemical Sediments and
Medvedev, Y.T., 1971. Some mineralogical and structural-textural peculiarities of bauxites in the Angara
Riverine. In: Platform Bauxites of the USSR. Izd. Nauka, Moscow, pp. 263-275.
Melfi, A.J. and Carvalho, A., 1983 (Editors). Proc. 2nd Int. Seminar on Laterisation Processes, Säo
Paulo.
Melfi, A.J., Trescases, J.J., Barros de Oliveira, S.M., 1979. Les "latérites" nickelifères du Brésil. Can.
ORSTOM, Sér. Géol., XI: 15-42.
Michel, P., 1973. Les bassins des fleuves Sénégal et Gambie. Etude géomorphologique. Thèse Sci.,
Strasbourg, Mém. ORSTOM, 63, 752 pp. (3 Vol.).
Millot, G., Bocquier G. and Paquet H., 1976. Géochimie des paysages tropicaux. La Recherche, Paris,
7: 236-244.
Milnes, A.R. and Fitzpatrick, R.W., 1987. Titanium and zirconium minerals. In: J.B. Dixon and S.B.
Weed (Editors), Minerals in Soil Environments. Soil Sei. Soc. Am.
Milnes, A.R., Bourman, R.P. and Fitzpatrick, R.W., 1987. Petrology and mineralogy of latérites in
southern and eastern Australia and southern Africa. Chem. Geol., 60: 237-250.
Mohr, E.C.J., Van Baren, FA. and Van Schuylenborgh, J., 1972. Tropical Soils. A Comprehensive Study
of their Genesis. Mouton, Ichtiar Baru, VanHoeve, Paris, Djakarta, The Hague, 481 pp.
Müller, J.P., 1987. Analyse pétrologique d'une formation latéritique meuble du Cameroun. Thèse, Univ.
Paris VII, 173 pp.
Nahon, D., 1976. Cuirasses ferrugineuses et encroûtements calcaires au Sénégal Occidental et en Mau
ritanie, systèmes évolutifs: géochimie, structures, relais et coexistence. Sci. Géol. Mém., 44, 232
pp.
191.
Nahon, D. and Lappartient, J.R., 1977. Time factor and geochemistry in iron crust genesis. Catena, 4:
249-254.
Nahon, D., Beauvais, A., Nziengui-Mapangou, P. and Ducloux, J., 1984. Chemical weathering of Mn-
garnets under lateritic conditions in northwest Ivory Coast (West Africa). Chem. Geol., 45: 5 3 -
71.
Nahon, D., Janot, C , Karpoff, A.M., Paquet, H. and Tardy, Y., 1977. Mineralogy, petrography and struc
tures of iron-crusts (ferricretes) developed on sandstones in the western part of Senegal. Geoderma,
19: 263-277.
Nahon, D., Melfi, A. and Conte, C.N., 1989. Présence d'un vieux système de cuirasses ferrugineuses
latéritiques en Amazonie du Sud. Sa transformation in situ en latosols sous la forêt equatoriale
actuelle. C.R. Acad. Sci. Paris, Sér. II, 308: 755-760.
Novikoff, A., 1974. Laltération des roches dans le Massif du Chaillu (République Populaire du Congo).
Thèse Sci., Univ. Louis Pasteur, Strasbourg, 298 pp.
Parrish, J.T. and Curtis, R.L., 1982. Atmospheric circulation, upwelling and organic rich rocks in the
Mesozoic and Cenozoic Eras. Palaeogeogr., Palaeoclimatol., Palaeoecol., 40: 31-66.
Parrish, J.T, Zirgler, A.M. and Scotese, C.R., 1982. Rainfall patterns and the distribution of coals and
evaporites in the Mesozoic and Cenozoic Eras. Palaeogeogr., Palaeoclimatol., Palaeoecol., 40: 67-
101.
Patterson, S.H., Kurtz, H.F., Olson, J.C. and Neeley, C.L., 1986. World bauxite resources. Geology and
resources of aluminum. U.S. Geol. Surv., Prof. Pap., 1076B: 151.
Petit, M., 1985. A provisional world map of duricrust. In: Douglas and Spencer (Editors), Environment
Change and Tropical Morphology, I. pp. 269-279.
Pion, J.C., 1979. Altération des massifs cristallins basiques en zone tropicale sèche. Etude de quelques
toposéquences en Haute-Volta. Thèse Sci., Strasbourg, Mém Sci. Géol., 37, 220 pp.
Roquin, C, Dandjinou, T., Freyssinet, Ph. and Pion, J.C, 1989. The correlation between geochemical
data and SPOT satellite imagery of lateritic terrain in southern Mali. In: S.E. Jenness (Editor),
Geochemical Exploration 1987. J. Geochem. Explor., 32:149-168.
Russell, E.W., 1962. Soil Conditions and Plant Growth. Longmans, London, 9th ed., 688 pp.
Schellmann, W, 1986. A new definition of laterite. Geol. Surv. India, Mem., 120: 1-7.
Schwertmann, U., 1965. Zur Goethit und Hämatibibildung aus amorphen Eisen. Ill — Hydroxide. Z.
Pflanzenernaehr. Dueng. Bodenk., 108: 37-45.
Schwertmann, U., 1988. Goethite and hematite formation in the presence of clay minerals and gibbsite
at 25°C. Soil Sci. Soc. Am. J., 52: 288-291.
Schwertmann, U. and Taylor, R.M., 1977. Iron oxides. In: J.P. Dixon (Editor), Minerals in Soil Environ
ments. Soil Soc. Am., pp. 145-180.
Seret, G., 1978. La notion de pénéplaine à la lumière d'une stratigraphie paléoclimatique du Quater
naire intertropical africain et des cuirasses latéritiques. Acta Géol. Hisp., 13: 145-147.
Stoffregen, R., 1986. Observations on the behavior of gold during supergene oxidation at Summitville,
Colorado, U.S.A., and implications for electrum stability in the weathering environment. Appi.
Geochem., 1: 549-558.
Swardt, A.M.J. de, 1964. Laterisation and landscape development in parts of equatorial Africa. Z.
Geomorphol., N.S., 8: 313-333.
Tardy, Y., 1969. Géochimie des altérations. Etude des arènes et des eaux des massifs cristallins d'Europe
et d'Afrique. Mém. Serv. Carte Géol. Alsace Lorraine, 31, 199 pp.
Tardy, Y., 1971. Characterization of the principal weathering types by the geochemistry of the waters
from some European and African crystalline massifs. Chem. Geol., 7: 253-271.
Tardy, Y., 1982. Kaolinite and smectite stability in weathering conditions. Estud. Geol., 38: 295-312.
Tardy, Y., in press. LErosion. Encyclopedia Universalis.
Tardy, Y., Bardossy, G. and Nahon, D., 1988a. Fluctuations de l'activité de l'eau et successions de
minéraux hydratés et déshydratés au sein des profils latéritiques ferrugineux et bauxitiques. C. R.
Acad. Sci., Paris, Ser. II, 307: 753-759.
Tardy, Y., Lesniak, P., Duplay, J. and Prost, R., 1980. Energies d'hydration des argiles. Application à
l'hectorite. Bull. Miner., 103: 217-223.
Tardy, Y., Mazaltarim, D., Boeglin, J.L., Roquin, C , Pion, J.C, Paquet, H. and Millot, G., 1988b.
Lithodépendance et homogénéisation de la composition minéralogique et chimique des cuirasses
ferrugineuses latéritiques. C. R. Acad. Sci., Paris, Sér. II, 307: 1765-1772.
Tardy, Y. and Nahon, D., 1985. Geochemistry of latérites stability of Al-goethite, Al-hematite and Fe-
kaolinite in bauxites and ferricretes: an approach to the mechanism of concretion formation. Am. J.
Sci., 285: 865-903.
les profils latéritiques. C. R. Acad. Sci., Paris, Sér. II, 306: 39-44.
Torrent, J., Parrà, M.A. and Guzman, R., 1980. The influence of relative humidity on the transformation
of ferrihydrite to hematite. Abstracts, 4th Meet. Eur. Clay Group, Freising RFA, pp. 155-156.
Toulemont, J., 1984. Modèle physicochimique d'évaporation. Transferts d'eau et de sels dans les sols par
capillarité et diffusion. Thèse, Univ. Paul Sabatier, Toulouse, 108 pp.
Trescases, J.J. and Melfi, A.J., 1985. Les gisements latéritiques du Brésil. Pangea, 5: 7-16.
Trolard, F, 1988. Physicochimie des cuirasses latéritiques. Domaines de stabilité des oxydes et hydroxy-
des de fer et d'aluminum. Thèse, Univ. Louis Pasteur, Strasbourg, 213 pp.
Trolard, F., Vallès, V, Partyka, S. and Tardy, Y., 1986. Propriétés thermodynamique de l'eau adsorbée à
la surface du quartz. Bull. Miner., 109: 199-218.
Trolard, F. and Tardy, Y., 1987. The stabilities of gibbsite, boehmite, aluminous goethites and aluminous
haematites in bauxites, ferricretes and latentes as function of water activity, temperature and particle
size. Geochim. Cosmochim. Acta, 51: 945-957.
Trolard, F and Tardy, Y., 1989. A model of Fe3+-kaolinite, Al3+-goethite, Al3+-hematite equilibria in
latérites. Clay Miner., 24: 1-21.
Thickenbrodt, W. and Kotschoubey, B., 1981. Argila de Belterra. Cobertura terciara das bauxitas ama-
zonicas. Rev. Bras. Geoci., 11: 203-208.
Vachaud, G., Dancette, C, Sonko, S. and Thony, J.L., 1978. Méthodes de caractérisation hydrody
namique in situ d'un sol non saturé. Application à deux types de sols au Sénégal en vue de la
détermination des termes du bilan hydrique. Ann. Agron., 29:1-36.
Valeton, I., 1972. Bauxites. Developments in Soil Science. Elsevier, Amsterdam, 226 pp.
Valeton, I. and Beissner, H., 1986. Geochemistry and mineralogy of the lower tertiary in situ latérites
on the Jos Plateau, Nigeria. J. Afr. Earth Sci., 5(5): 535-550.
Vallès, V, 1986. Etude et modélisation des transferts d'eau et de sels dans un sol argileux. Application
au calcul des doses d'irrigation. Sci. Géol. Mém., Strasbourg, 79,148 pp.
Vauclin, M., 1982. Infiltration in unsaturated soils. NATO advanced lecture. Newark, Del., 50 pp.
Vauclin, M. and Vachaud, G., 1981. Bilan hydrique dans le Sud-Tunisien. II. Modélisation numérique et
prévision des transferts hydriques en sol stratifié. J. Hydrol., 49: 53-73.
Vogt, J., 1967. Première note sur un remarquable remblaiement inactuel des têtes de vallées des pays
Gouin, Karaboro et Lobi (Haute-Volta). VP Congr. Panafricain de Préhistoire, Dakar, pp. 442-450.
Volkoff, B., 1985. Organisations régionales de la couverture pédologique du Brésil. Chronologie des
différenciations. Can. ORSTOM, Sér. Pédol., 21: 225-236.
Wackermann, J.M., 1975. âlteration des massifs cristallins basiques en zone tropicale semi-aride. Etude
minéralogique et géochimique des arènes du Sénégal oriental. Conséquences pour la cartographie
et la prospection. Thèse Sci., Strasbourg, 373 pp.
Watson, J.P., 1962. The soil below a termite mound. J. Soil Sci., 13: 46-51.
Watson, J.P., 1972. The distribution of gold in termite mounds and soils at a gold anomaly in Kalahari
sand. Soil Sci., 113: 317-321.
Weber, B.N., 1959. Bauxitizaçao no distrito de Poços de Caldas, Minas Gérais, Brazil. Bol. Soc. Bras.
Geol., 8: 17-30
Weber, N.A., 1966. Fungus-growing ants and soil nutrition. Colloquium Latino-Americano dos Suelas,
Unesco, Montevideo, 1: 221-256.
Weber, F, Ledere, J. and Millot, G., 1979. Epigénies manganésifères successives dans le gisement de
Moanda (Gabon). Sci. Géol. Bull., 32: 147-164.
Webster, J.G., 1986. The solubility of gold and silver in the system Au-Ag-S-0 2 -H 2 0 at 25°C and 1
atm. Geochim. Cosmochim. Acta, 50:1837-1845.
Webster, J.G. and Mann, A.W., 1984. The influence of climate, geomorphology and primary geology on
the supergene migration of gold and silver. J. Geochem. Explor., 22: 321-335.
Wilson, AF, 1984. Origin of quartz free gold nuggets and supergene gold found in latérites and soils.
A review of some new observations. Aust. J. Earth Sci., 31: 303-316.
Wood, T.G. and Sands, W.A, 1978. The role of termites in ecosystems. In: M.V Brian (Editor), Produc
tion Ecology of Ants and Termites, Cambridge University Press, Cambridge, pp. 245-292.
Yakushev V.M., 1968. Influence of termite activity on the development of laterite soil. Soviet Soil Sci.,
1:109-111.
445
Chapter 17
Metallogeny of weathering: an introduction

D.B. NAHON, B. BOULANGÉ and F. COLIN
Introduction
Soils and weathering mantle while developing at the expense of certain parent
rocks concentrate or simply refine useful substances or metals. The latter, according
to their mobility under given weathering conditions, can concentrate at different
levels and in different horizons of weathering profiles.
The most spectacular examples belong to tropical regions where weathering is,
or has been very intense. Such is the case of lateritic weathering which covers at
present a third of all emerged lands.
The purpose of this article is not to present an exhaustive series of examples
but to illustrate by means of a few selected case-histories how such concentrations
develop and may evolve into economic deposits. However, it is at first necessary to
define the main horizons differentiated within lateritic weathering.
Lateritic weathering profiles
The lateritic weathering mantle developed in the tropical zone is a real patchwork
in which older profiles coexist with younger ones. As yet, no technique is available
for the absolute dating of these various profiles whose thickness, distribution, and
nature are controlled by several factors: composition of parent rock, type of climate
and paleoclimates, and geomorphology, all of which are briefly discussed below.
The mineralogical composition of the parent rock as well as its geological and
structural history plays a fundamental role in the circulation of meteoric fluids and
their interaction with minerals, in the differential rates of dissolution, and finally in
the generation of weathering products (Millot, 1964; Berner, 1978; Nahon, in press).
The nature of the climate to which the rocks are subjected is expressed by the
amount of water percolating through the profiles and by the physico-chemical prop-
erties of these fluids (such as pH, pC0 2 , p 0 2 and activity).
446 D.B. Nation, B. Boulangé and F. Colin
The nature of various paleoclima tes involved in the development of the profile, is
more difficult to grasp.
Current and past geomorphological location plays or has played, an important
role in the vertical and lateral dynamics of solutions, namely in the leaching or
confinement of elements transported in the state of simple ions or complexes, or in
the state of particles (Tardy, 1969).
Five main types of horizons vertically or horizontally differentiate from a given
parent rock: saprolites, massive crusts, aggregated or nodular unconsolidated clay
horizons, brecciated, conglomeratic or pisolitic crusts, and deeply leached unconsol
idated horizons. These various types of horizons are discussed below.
Saprolites are horizons with the original structure of the parent rock preserved
through is volumetric geochemical weathering. These very porous horizons consist
of weathering phyllosilicates (mainly 1/1 but with 2/1 type at the base), associated
with oxihydroxides or hydroxides. These saprolites can form almost entire profiles
with a variable number of subdivisions (Nahon, 1986, in press).
Massive crusts are indurated horizons in the upper part of the profile, where
the original structure of the parent rock is still preserved. These crusts consist of
strongly cemented oxides or oxihydroxides.
Unconsolidated, aggregated clay horizons in which kaolinite is intimately associ
ated with oxihydroxides in microaggregates, or unconsolidated clay horizons with
kaolinite and in which oxides are differentiated as spots and nodules, develop at the
expense of the saprofite or even of the massive crust. According to the intensity of
the wet season, these horizons are called oxic, mottled clays, nodular horizons and
soon.
Brecciated, conglomeratic or pisolitic crusts develop in situ at the expense of mas
sive crusts. Original structures of the parent rock are no longer visible. These hori
zons consists of oxides, hydroxides and oxihydroxides.
Deeply leached horizons develop at the expense of all other types of horizons,
and particularly from unconsolidated clay horizons (such as saprolites and micro
aggregated horizons). These horizons are very porous, impoverished in clay, and
consist often of a residual skeleton of quartz grains or of fragments of crusts.
These five main types of horizons can become differentiated within the same
profile or within the same weathering sequence. Certain profiles may show only the
development of saprofite overlain by another horizon.
During the differentiation of one or several of these horizons, supergene metal
lic concentrations are generated or redistributed. These concentrations occur as
secondary carbonates, silicates or oxides precipitated from solutions, or as detri-
tal residual particles accumulated during the geochemical lowering of weathering
profiles. Several examples can illustrate these different situations, such as lateritic
concentrations of gold, bauxite, manganese, nickel and copper.
Metallogeny of weathering: an introduction 447
Concentrations of gold
Gold-bearing lateritic profiles of Gabon
Lateritic weathering of gold-bearing sulfide, carbonate or quartz parent rocks

leads to the redistribution of gold in weathering horizons. This redistribution can be
controlled by geochemical, biogeochemical or simple mechanical processes.
Since the theoretical works of Krauskopf (1951) and Cloke and Kelly (1964), the
dissolution of primary gold particles and the precipitation of secondary gold have
been documented by field studies and/or by pétrographie and geochemical investi-
gations in Brazil (Freise, 1931; Vasconcelos and Kyle, 1989), in the United States
(Evans, 1981; Stoffregen, 1986), in Australia (Webster and Mann, 1984; Mann,
1984; Wilson, 1984; Davy and Ansary, 1986; Webster, 1986; Lawrance, 1988), in
New Guinea (Webster and Mann, 1984), in India (Nair et al, 1987) and in tropical
Africa (Colin et al., 1989; Freyssinet et al., 1989).
Although these investigations show that the behaviour of gold is related to the
development of lateritic profiles, some of these profiles are subjected today to more
arid climates. Such is the case of the major part of Western Australia.
All these studies reveal strong similarities in the evolution of the "nuggets" of
primary gold during lateritic weathering. In Gabon, where lateritic weathering is
still very active (equatorial climate with average annual rainfall of 2000 mm), the
localities of Mébaga, Dondo and Ovaia (Colin et al., 1989) show the same type of
evolution. The parent rock is a metabasite intersected by quartz veins and enclosed
within a granite-gneiss (Figure 17.1). The "nuggets" of primary gold occur in the
quartz veins and consist of an alloy of gold and silver (5-9% Ag). As in most of
the deposits described in the literature, the maximum content in gold of the quartz
veins in the profiles of Gabon is 100 g t" 1 .
The dissolution of primary gold is recognizable only from the upper part of the
saprolite and continues to the surface. Ttoo simultaneous evolutionary trends are
observable: a vertical and a lateral one.
The vertical evolutionary trend shows: (1) a gradual roundness leading to blunt
gold particles looking like alluvial "nuggets"; (2) a decrease of the average size of
the "nuggets" (1000 μτη in the protore to 500 μπι at the top of the mineralized
zone); and (3) a decrease of the average content in silver, this depletion occurring
at the margins of the "nuggets" (5 to 1% Ag at the top; Figure 17.2).
The lateral evolutionary trend is shown by the differentiation of a dispersion halo
in the ferruginous nodular horizons and the surficial sandy-clayey horizons (Figure
17.1). Thus from the centre of the halo toward its rim, gold particles are (1) dimin
ished (from 500 μτη to a few microns), and (2) increasingly rounded. Whereas in
the centre of the halo (mineralized zone), the weathering which increases upward
raises the gold content of the periphery of the "nuggets", along the margins of the
448 D.B. Nahon, B. Boulangé and E Colin
,,^\
cCv^h
\ Ml a
5Q/t
—m
—IV
WEATHERING \ *
FRONT
20 m
'l00g/t
20m
20m
I I I SANDY-CLAYER LAYER
| H | FERRUGINOUS NODULAR LAYER
ΓπΠ SAPROLITE
pβH BARREN FRESH ROCK
\WsA METABASITE BODY

Π Π Au CONTENT OF QUARTZ VEINS
Fig. 17.1. Geological setting of the studied areas. Surface dispersion halo of gold contents (g f 1 ) .
Ag
%I CENTRAL PART OF THE HALO
RIM OF THE HALO
10
l
®
& /->. O SANDY CLAYEY LAYER I
8
1
5
^ ty y,
® m FERRUGINOUS NODULAR LAYER Π
^L <fr SAPROLITE IH
6-
o
1 <&<&►► o
► ► ► o
4-
2-
o 1 i i 1 1 1 ~r*~ i— i i
►dV
r
90 92 94 96 98 100 96 98 /JßéË
Fig. 17.2. Ag % vs Au % in increasing weathering.
Metallogeny ofweatherìng: an introduction 449
halo (beyond the mineralized vein), the contents in gold and in silver (0.3 to 3%)
decrease with depth (Figure 17.2).
The profiles of Gabon show an increasing geochemical weathering immediately
above the gold-bearing vein combined with a mechanical dispersion as a surficial
halo, with its own increasing weathering. The inversion of the contents in gold and
silver along the margins of the halo results from a downward translocation of the
most weathered particles by regolith reduction and by biological activities.
Reprecipitation of gold in latentes
Reprecipitations of secondary gold within surficial dispersion patterns are limited

and very localized in the examples studied in Gabon, which are mostly characterized
by dissolution processes. However, supergene reprecipitation of gold in lateritic en
vironments can be an important factor in the secondary enrichment in gold (Webster
and Mann, 1984). Many proofs of precipitation of secondary gold were observed,
they concern geometrical, chemical (fineness) and crystalline characteristics, and
they vary according to conditions of weathering. These proofs are briefly discussed
below.
The geometrical distribution of certain concentrations, as particles or crystals,
argues in favour of their secondary origin. Indeed, gold can concentrate as follows.
(a) At the base of the oxidized zone, at the level of the water table of the
mineralized zones or within rocks which originally were barren. Such is the case
at Wau in Papua, New Guinea, where a thin horizon of gold occurs in lateritized
ultrabasic rocks, at 40 meters, at the present level of the water table (Webster and
Mann, 1984).
(b) In fissures, cracks and cavities which intersect iron oxides, as at Westonia,
Western Australia (Webster and Mann, 1984).
(c) At the surface of lateritic ferruginous nodules, or of concentrations of man
ganese oxides, or of weathering clays, or interstratified with iron oxides, as shown
by the examples of Kangaba, in Mali (Freyssinet et al., 1989), of Dondo in Gabon
(Colin et al., 1989) and of Western Australia (Wilson, 1984; Lawrance, 1988).
The chemical composition of gold distributed in the ways just mentioned is dif
ferent from that of primary gold. Indeed, this gold is depleted in silver compared to
the primary alloy Au(i_JC) Ag^). This depletion depends on weathering conditions.
In extreme cases, as at Kalgoorlie, Western Australia, high purity gold particles or
crystals (Au = 99.91%) have been found (Wilson, 1984). Here, weathering acts as a
natural agent for refining gold.
The crystalline forms of the gold depleted in silver also indicate its secondary
origin. This type of gold occurs in various forms such as arborescent dendrites, 0.01
to 4 mm in size, (Webster and Mann, 1984; Freyssinet et al., 1989), "flatmakes"
consisting of minute cubes (<2 mm) or of delicate wires (Webster and Mann,
1984), thinfilmsand coatings called "paint gold" (Wilson, 1984), geode crystals with
perfect forms of a size of about 20 /im (Mann, 1984) and crystals associated in
botryoidal masses (Colin et al., 1989).
Mechanisms oflateritic weathering of gold

The occurrence of dissolution features in the primary gold grains and crystals of
secondary gold raises the question of the mechanisms of dissolution, of transport
and of reprecipitation of gold under lateritic conditions.
Dissolution of gold occurs when the weathering of enclosing rocks is very ad
vanced and when oxidizing solutions percolate through the environment. Indeed,
dissolution features appear only in the upper part of the saprolite and become more
intense in the horizons of mottled clay, of ferruginous nodules, or with ferruginous
or bauxitic crusts, where oxides and oxihydroxides of iron accumulate.
Supergene gold crystalline shapes occur at different levels in the profile: in the
crust horizon, in the mottled clay horizon, or very low in the profile, at the level
of the water table, at the base of the saprolite. This implies different conditions of
transfer either at very short distance or very long distance, before reprecipitation of
authigenic secondary gold.
The coupled dissolution-precipitation of gold must therefore operate either
within the same environment or microenvironment, or be completely separate, de
pending on the conditions of weathering. Therefore, particular attention should be
given to the "nuggets" of primary gold which survive in the upper part of lateritic
profiles.
These "nuggets" display two types of contacts according to the observations of
Mann (1984), either with the clay ferruginous matrix, or with the iron oxide matrix
(Figure 17.3).
The contact with iron oxides suggests a replacement of the "nugget" of primary
gold by iron oxides with an almost simultaneous mobilization of gold and silver
of the primary alloy (Mann, 1984). However, a slight discontinuity is visible, at
the contact, between the contents in Au and Ag, the latter showing a preferential
depletion. This could explain why gold relics in the iron oxide attached to the nodule
are impoverished in gold. The portion of the "nuggets" which is not in contact with
iron oxides shows, on the other hand, a preferential depletion of silver. Thus, a
ring enriched in gold develops, and its boundary with the primary alloy is sharp,
although irregular (Colin et al., 1989; Freyssinet et al., 1989). The ring is sometimes
discontinuous and has a thickness that does not exceed 15-45 ìðé.
This natural refining of primary gold (Au(i_A:) Ag^) where 0.04 < x < 0.1) may
thus be attributed either to a preferential depletion of silver by atomic diffusion or
to a dissolution of the alloy, loss of silver and reprecipitation in situ of gold.
Careful observation of the electron microprobe scan across the silver depleted
outer rim of the "nugget" (Mann, 1984, fig. 7, p. 6) shows in reality that the cen-
tripedal silver depletion occurs in "regular" steps of nearly 15 ìçé in width (Figure
Metallogeny of weathering: an introduction
higher purity gold rim périphérie iron oxide containing pieces of gold
(Ag. depleted outer rim) void / (with<0.2%Ag)
iron oxide
external
surface
Fig. 17.3. Sketch of the two types of evolution of primary gold "nuggets" (from Mann, 1984).
453
452 D.B. Nahon, B. Boulangé and F. Colin
17.3). This fact suggests that depletion of silver occurs by successive diffusionary
steps.
Furthermore, the thickness of the Ag-depleted outer rim of the primary gold
particles remains constant (15-45 /xm) while the gold particles size is reduced with
increasing weathering from the base to the top of the profiles. Obviously, the pref
erential dissolution of silver by diffusion is followed by a congruent dissolution of
gold. Once the outer rim is formed, the dissolution fronts of silver and gold advance,
simultaneously and centripedally at the expense of the primary alloy.
The dissolution of gold is probably promoted by a cryptoporosity generated by
the silver depletion. However, the replacement of primary alloys by iron oxides
(Mann, 1984) indicates the simultaneity of the dissolution of primary alloy, and
precipitation of iron oxide. In such a case, silver is removed as rapidly as gold in
the solution of the microenvironment in which the replacement occurs. Obviously,
dissolution of gold is increased where iron oxides precipitate.
Most workers refer to the studies of Krauskopf (1951) for an explanation of the
behaviour of gold in the lateritic environment. In particular, it is clear that in acid
solutions rich in chloride ion, silver and gold form chloride complexes, the most
important of which are AgCl° and AuCl^", considering silver in its monovalent form
and gold under its trivalent form. In such a case, the oxidation reactions are (Mann,
1984):
4Ag + 4 C r + 0 2 + 4H + -* 4AgCl° + 2 H 2 0 (1)
4Au° + 16CF + 3 0 2 + 12H + -> 4AuClJ + 6H 2 0 (2)
These reactions imply well oxygenated and chloride-rich waters. Such conditions
occur in the profiles of Western Australia where climate is semi-arid and where
chlorine contents are regularly provided by rain coming from the coast.
Complexed gold in this form can migrate downwards and reprecipitate as sec
ondary crystals of very high purity. The reducing agents of AuCl^~ could be the
ions Mn2+ or Fe2+ inducing a coprecipitation of gold and oxihydroxides of iron or
manganese according to the following equations (Mann, 1984):
AuCl^ + 3 Fe2+ + 6 H 2 0 -> Au° + 3 FeOOH + 4 CP + 9 H + (3)
AuCÇ + 3 Mn2+ + 6 H 2 0 -» Au° + 3 MnOOH + 4 CP + 9 H + (4)
However, these conditions do not prevail under a more humid climate where
concentrations of chlorine in water are lowered by dilution. Such is the case of
tropical zones with contrasted seasons where lateritic profiles are covered by thick
ferruginous crusts, or in very humid tropical (equatorial) zones as in Gabon. Reac
tions should be written as a function of these lower concentrations. Furthermore,
secondary gold found in this type of climate seems concentrated in ferruginous hori
zons (crust, ferruginous nodules), not co-precipitated but crystallized infissuresand
cavities which intersect iron oxides. The example of replacement of primary alloys
by iron oxides (Mann, 1984) indicates that gold is dissolved, and not precipitated
simultaneously with iron oxides, as suggested by equation (3). The dissolution of the
primary alloy requires the presence of a ligand and of an oxidizing agent consider-
ably stronger than H + alone (Krauskopf, 1951, p. 853). In equatorial forest areas,
the ligand could be provided by organic acids or by chloride ion in much smaller
amount than in semi-arid lands, because of dilution by water.
Organic acids (humic, fulvic), originating from bacterial activity in the decay
products of the vegetal cover, can play an important role (Baker, 1978). The chlo-
ride ion could still play a critical role here by associating with gold in the form of
AuCl(OH)" according to Colin and Vieillard (oral commun.). The ferric ion Fe 3+ ,
although a weaker oxidizing agent than oxygen, reaches a high concentration at the
top of lateritic profiles and thus could play the role of oxidizing agent. The following
equations (5) and (6):
Au(s) + Fe(aq) + Cr + OH" - AuCl(OH)- q) + Fefa+q) (5)
2 Fefa+q) + 1 02(aq) + 2 H 2 0 - Fe203(hem) + 4 H + (6)

could explain the simultaneous dissolution of gold and precipitation of the iron
oxide replacing it by assuming, from the pétrographie observations of Mann (1984),
that the rates of the two reactions (5) and (6) toward the right are of the same order
of magnitude.
Bauxitic concentrations
Lateritic bauxitization
Lateritic bauxites have been the subject of a recent review (Bardossy, 1982; Tardy,
this book).
Concentrations of alumina in the humid tropical zone generate lateritic bauxites.
They derive from two types of processes (Millot, 1964). The first is direct bauxiti-
zation, which is a direct weathering (or with transitory intermediate products) of
parent silicates into alumina hydroxides. This type of weathering occurs in leached
and well-drained environments. The second is indirect bauxitization which produces
first an important kaolinitic weathering, which subsequently evolves toward alu-
minum hydroxides. These two processes generate different pétrographie structures
and textures (Boulangé, 1984). In direct bauxitization, the original structure of par-
ent rocks is preserved up into the massive bauxitic crust, whereas such is not the
case for indirect bauxitization.
Examples of direct bauxitization allow, by means of the preserved structures and
textures of the parent rock, an estimation of isovolumetric mass balances. This has
454 D.B. Nation, B. Boulangé and E Colin
been done on nepheline syenites of the Islands of Los, Guinea (Lacroix, 1913; Millot
and Bonifas, 1955), in Arkansas (Gordon and Ttacey, 1952), in Brazil at Poços de
Caldas (Almeida, 1977) and at Passa Quatro (Sigolo and Boulangé, 1987), and also
on trachytes in Cameroon (Hieronymus, 1985) and on granites in the Ivory Coast
(Boulangé, 1984).
Bauxites of Lakota, Ivory Coast
The deposit of Lakota (700,000 tons) is located at the top of a hill (Mount Tato)
which is 380 m high. It reaches a thickness of about 20 m, and consists of a saprolite,
about 8 m thick, overlain by a thick massive crust whose uppermost meter is strongly
dismantled.
In this case of direct bauxitization, the observed mineral sequences show that
the main parent minerals are weathered, starting with thefirstfew centimetres of
the saprolite, into gibbsite with or without intermediate phases (gels, interlayered,
kaolinite). Although plagioclases and orthoclase are more rapidly weathered than
microcline, all feldspars are pseudomorphed, as early as theirfirststage of weather
ing, into amorphous alumino-silicate products (gels). Similar examples are reported
in younger weathering systems in the Ivory Coast (Delvigne, 1965). The evolution of
these gels toward gibbsite is general, with sometimes another intermediate weather
ing phase: metahalloysite, halloysite, or even kaolinite. All these transitional phases
between feldspars and gibbsite are ephemeral and develop only over a few centime
tres at the base of the lower saprolite (Figure 17.4). Biotite is also pseudomorphed
by an interlayered biotite-vermiculite which itself evolves into goethite + kaolinite
Bcehmite - Hematite + Gibbsite - Al-Gthite
N Gibbsite
Gibbsite + Al-Gthite
Kaolinite
Goethite
metahalloysite
Al - Si - Gels Al - Goethite
Biot. - Verm.
+ Gibbsite
inter layer
ORTHOCLASE"! | Na PLAGIOCLASES Ca || MICROCLINE | | ~ BIOTITE

E I
| | H O R N B L E N D E | | MUSCOVITE
Fig. 17.4. Mineral sequences of weathering during direct bauxitization (example of Lakota).
before the appearance of pseudomorphic gibbsite which is accompanied by alumi

nous goethite. This evolution of biotite toward gibbsite starts at the very base of the
lower saprolite.
Hornblende does not occur at Lakota, but to the NNE, at Orumbo Bocca. It
weathers directly into septa of aluminous goethite with gibbsite crystals in the cen
tre.
Muscovite remains unweathered in the saprolite as shown by electron microprobe
studies (Boulangé, 1984). Its weathering into gibbsite only occurs at the base of
massive bauxitic crust when quartz is completely dissolved.
Thus, with the exception of muscovite and quartz, all parent minerals of the gran
ite evolve into gibbsite or into gibbsite + aluminous goethite from the first stages
of weathering at the base of the saprolite. These hydroxides and oxihydroxides are
pseudomorphs after minerals from which they derive. This spatial distribution al
lows microscopic and macroscopic recognition of the texture and structure of the
parent rock.
In the massive bauxitic crust, the original structure of the parent rock is still
preserved. The mineral assemblage consists of gibbsite + hematite and aluminous
goethite locally associated with boehmite.
Horizons of brecciated, conglomeratic or pisolitic crust, in the region of Lakota,
develop in weathering profiles from granite only over morphologies located below
Mount Tato. In this case, the saprolite consists only of kaolinite and is overlain by
a massive crust of gibbsite + aluminous goethite and hematite. In the horizons of
brecciated, conglomeratic crust, boehmite is associated with gibbsite and hematite.
In the pisolitic crust, boehmite and hematite are dominant minerals.
At Mount Tkto, pseudomorphic weathering of parent minerals into gibbsite and
aluminous goethite generates a very high microporosity. Then the porosity is rapidly
filled by coatings of gibbsite and hematite precipitated from circulating solutions.
These transfers of materials can be very important and clog most of the micro
porosity estimated at about 40% of the volume of the rock in the lower part of the
TABLE 17.1
Isovolumetric estimate of relative gains in AI2O3 and Fe203 in weathering horizons with preserved
structure of the bauxitic profile of Mount Tkto.
Layer Thickness AI2O3 Fe2Ü3

(m) (%) (%)
Massive dismantled crust 1 130 290
Upper massive bauxitic crust 0.5-1 140 80
Lower massive bauxitic crust 10-15 120 200
Transition horizon 0.2-0.5 60 250
Upper saprolite 3-5 60 250
Lower saprolite >2 20 80
Parent granite 44 g per 100 cm3 4.5 g per 100 cm3
saprolite. Reprecipitated gibbsite and hematite in the pores of the saprolite give a
very high degree of cohesion to the rock which is then called a massive bauxitic
crust.
Since original volumes are still recognizable, isovolumetric mass balance calcula
tions allows estimation of the gains and losses in alumina and iron with respect to
the parental granite (Tkble 17.1; Millot and Bonifas, 1955).
Mechanisms of bauxitization
Since the works of Garrels and Christ (1965), numerous thermodynamic ap
proaches have been proposed to explain the mechanism of bauxitization by means
of activity-activity or Eh-pH diagrams, or a combination of both.
The basic nature of bauxitization can be understood by means of the relations
between kaolinite and gibbsite, without neglecting the relations with the other ox
ides or hydroxides of aluminum (diaspore, corundum, boehmite, amorphous hy
droxides), and the oxides and hydroxides of iron (Al-hematite, Al-goethite) which
appear as solid solutions corundum-hematite and diaspore-goethite (Nahon, 1976;
Tkrdy and Nahon, 1985).
Construction of solubility curves and of stability field diagrams depends mostly
on the selection of values of the standard Gibbs free energy of formation (AG°f) of
the solid and dissolved species. Pétrographie analysis helps to decipher the chrono
logical orders of precipitation or of dissolution of weathering minerals. Considering
the activity of water equal to 1, the values of AG°f reported by Zen (1972, set 2)
give the following order of decreasing stability: boehmite, gibbsite, diaspore, corun
dum; this is in agreement with data from Hem and Roberson (1967) and Norton
(1973).
Stabilities of the various aluminum hydroxides, of solid solutions goethite-dia-
spore or hematite-corundum, and of kaolinite, are in fact closely related to vari
ations of activity of ions in percolating waters. Furthermore, the activity of water
itself, related to the size of the pores containing it, plays a fundamental role.
The activity of silica [H4S1O4] and the activity ratio [Al3+]/[H+] of the solutions
control the stability of kaolinite, of gibbsite, and the rate of aluminum substitution
of iron oxides and hydroxides (Didier, 1983; Tardy and Nahon, 1985). In the pres
ence of quartz only, the rate of aluminum substitution is low, it is higher in the
presence of kaolinite and even higher in the presence of gibbsite. This explains why
in bauxites, the rate of aluminum substitution of iron hydroxides is higher than in
ferruginous crusts and also why goethites are more substituted in leached and acid
soils than in soils less acid and richer in silica.
Similarly, kaolinites are very often Fe-substituted. The rate of substitution varies
from 1 to 4 mole percent Si2Fe205(OH)4. The rate increases with increasing iron
content of the environment where kaolinites form. Crystallinity of kaolinites be
comes poorer and their crystal size smaller with increasing iron substitution (Canti-
nolle et al., 1984; Didier, 1983). In very iron-rich environments, kaolinite is desta-
bilized during the precipitation of oxides which releases large amounts of protons
(Ambrosi and Nahon, 1986). Therefore, stability of kaolinite for constant activities
in [Al3+] and in [H4S1O4] of the solutions can depend on the content of iron in the
environment.
Similarly, for the same activity ratio [Al3+]/[H+] of the solutions, the equilibrium
gibbsite-kaolinite can be strongly displaced when the activity of water decreases
(Mattigod and Kittrick, 1981). Furthermore, Tkrdy and Novikoff (1988) showed that
the activity of water plays an important role in the equilibrium gibbsite-boehmite.
This idea had been previously suggested under another form by Chesworth (1972b)
in terms of the chemical potential of water.
Finally, precipitations of aluminous goethite, of aluminous hematite and of their
reciprocal association or with gibbsite, are, for a given [Al3+]/[H+] activity ratio,
dependent on the activity of water. The activity of water can be related to the size
of the pores containing it (Bourrié and Pedro, 1979). In such a case, it is easy
to understand how the bauxitic evolution, by the endlessly repeated combination
of dissolutions and precipitations, namely opening and closing of porosities, can
influence through self-organization the evolution of mineral parageneses.
Concentrations of oxides and hydroxides of manganese
Role of the parent rock

It is difficult to present a typical weathering profile of Mn-accumulation because
it depends on the nature of the original parent rock and, particularly, on its struc-
ture. The profiles we have studied in detail developed from sedimentary rocks with
subhorizontal bedding at Moanda (Gabon), with oblique bedding at Azul (Brazil),
and from metamorphic rocks at Ziemougoula (Ivory Coast) and at Conseilhero
Lafaiette (Brazil).
Weathering of manganiferous parent rocks consists essentially of the oxidation of
silicate or carbonate parent phases of Mn2+ into oxihydroxides with Mn 3+ , and then
into oxihydroxides and oxides with Mn 4+ .
Parent minerals do not weather at the same rate (Goldich, 1938) and this allows,
in horizons located at the base of weathering profiles, the preservation of relics
of parent minerals. These relics, because of their late weathering, produce new
oxides through the elements they liberate, and sometimes real retromorphosis of
oxihydroxides and oxides formed during the first stages of weathering.
Mineral sequences of weathering

Weathering of parent rocks with rhodochrosite (MnCC>3) were studied at Nsuta,
in Ghana (Perseil and Grandin, 1978; Roy, 1981), at Moanda, in Gabon (Bricker,
1965; Weber, 1968; Weber et al., 1979; Nahon et al, 1983) and at Azul, in Brazil
(Valarelli et al., 1978; Beauvais et al., 1987).
Weathering of parent rocks consisting essentially of manganiferous silicates were
studied at Ziemougoula, in the Ivory Coast (Nahon et al., 1984, 1985), at Blafo-
Gheto, in the Ivory Coast (Grandin and Perseil, 1983; Perseil and Grandin, 1985),
at Conseilhero in Brazil (Horen, 1953; Boeglin, 1981) and at Dongri-Buzurg and at
Magnitogorsk, in the Soviet Union (Chukhrov et al., 1983).
Mineral sequences of weathering generated by these two types of parent rocks
(Figures 17.5 and 17.6) were abstracted by Nahon and Pare (1989) and Pare (1989).
RAMSDELLITE PYROLUSITE
NSUTITE
CRYPTOMELANE
NSUTITE
A—A A
LITHIOPHORITE
RAMSDELLITE
BIRNESSITE
iGROUTITEl IMANGANITEl
iPYROXMANGp [TEPHROITE] | Ì Ð - Á Ì Ñ Ç É Â 1 ISPESSART IRHOPONITÊI
Fig. 17.5. Mineral sequences in the weathering of Mn-bearing silicates (heavy arrow indicates retromor-
phosis of oxides).
CRYPTOMELANE
| PYROLUSITE 1
RAMSDELLITE
NSUTITE
CRYPTOMELANE
PYROLUSITE
ð
BIRNESSITE llTODOROKITE
MANGANITE
HAUSMANNITE
RHODOCHROSITE
BEARING PARENT ROCK
Fig. 17.6. Mineral sequences in the weathering of Mn-bearing carbonates (heavy arrows indicate retro-
morphosis of oxides).
It is apparent that different parent minerals can generate the same oxidized
weathering phase and that different minerals may generate in the same horizon
different weathering minerals. In thefirstcase, thefluxof solutions controls the for
mation of the secondary phase; in the second case the crystallo-chemical structure
of the parent mineral seems the controlling factor of the weathering product. Only
the control by weathering solutions was studied by thermodynamic simulation (Pare
et al., 1989). By estimating the free energies of formation of weathering minerals
from field observations, Pare et al. (1989) presented log [Mn 2+ ]/[H + ] 2 as a function
of l°g/o 2 showing the relative stability of the minerals as a function of the degree
of oxidation (Figure 17.7a and b). These diagrams also show, with dashed lines,
for increasing pressures of CO2, the stability field of rhodochrosite. It can thus be
understood why rhodochrosite weathers in the investigated deposits into different
oxihydroxides.
For instance, under high partial pressures of C0 2 (log/co2 > - 2 ) rhodochrosite
weathers directly into birnessite (Mn5Mn2Oi3. 5H 2 0) or into cryptomelane
(KosMny 5Ìçï.5èé6) if potassium is available in the weathering environment, par
ticularly during simultaneous weathering of potassic feldspars, muscovites or illites.
Weathering of rhodochrosite into manganite, often observed in thefield,occurs only
for values of log / C o 2 = - 2 .
Oxihydroxides thus formed, subsequently weather into oxides with higher va
lences of manganese higher in the profile during the opening of the environment.
Nevertheless, it is common to observe in profiles a very oxidized manganiferous
phase grading into a less oxidized phase (Figures 17.5 and 17.6). Such a retromor-
-26 -22 -18 -14 -10 -6 log f 0 2 -26 -22 -18 -14 -10 -6 log f 0 2
Fig. 17.7. a. Stability field of nsutite, cryptomelane and manganite at 25°C, 1 atm. The tie-lines represent
different limits of the rhodochrosite stability field as a function of log/co 2 - b. Stability field of nsutite,
birnessite and manganite at 25°C, 1 atm. The tie-lines represent different limits of the rhodochrosite
stability field as a function of log/co 2 -
phosis can be explained by late weathering of primary minerals in the profiles. Thus,
potassium released late by weathering of micaceous minerals allows the retromor-
phosis of nsutite (Mn^Mny 02-y (OH)^) and even of pyrolusite (Mn 4+ 0 2 ) into
cryptomelane. Similarly lithiophorite (Mn3 Al2 0 9 3H 2 0), which strongly develops
toward the top of lateritic weathering profiles, shows modifications of valence of
manganese due to the incorporation in its structure of transition elements, particu-
larly Ni2+ (Manceau et al., 1987).
Aluminum required for the formation of lithiophorite originates either from the
parent mineral (alumino-manganiferous garnets), or from the destruction in the up-
per part of the profile of weathering kaolinite. This last case produces lithiophorite
in large amounts. Nahon et al. (1989) showed that the relative concentration of
manganese oxides can locally generate very acidic environments, allowing the re-
lease of aluminum by dissociation of kaolinite and its simultaneous incorporation
into the structure of the precipitating lithiophorite.
This enrichment in aluminum at the top of lateritic weathering profiles can gen-
erate concentrations of gibbsite by dissolution of lithiophorite, to the extent of form-
ing thick horizons rich in aluminum hydroxides (Beauvais and Nahon, 1985). This
situation confirms that in lateritic environments, manganese remains more mobile
than aluminum.
The stability field of lithiophorite is, therefore, intermediate between that of
kaolinite, that of gibbsite, and that of manganese oxides devoid of alumina. The
nature of the latter varies for a given activity of Al 3+ , as a function of the partial
pressure of oxygen, that is, as a function of the opening (aeration) of the environ-
ment (Pare et al., 1989).
Nickeliferous concentrations
Nickeliferous silicate and oxide horizons
Weathering profiles developed from ultrabasic rocks almost always show com-
mercial grade enrichments in nickel. These profiles show at the base greenish sili-
cate horizons overlain by thick oxidized horizons where goethite remains the main
mineral. Concentrations of nickel can occur both in the silicate and oxidized hori-
zons. Such is the case of the latérites of New Caledonia (Lacroix, 1943; Tïescases,
1975), of Indonesia (Kühnel et al, 1978), of Western North America (Hotz, 1964)
and of Brazil (Melfi et al., 1980). Other nickeliferous latérites were described as
oxidized deposits only, as in Cuba (de Vletter, 1955), in the Philippines (Frasche,
1941; Santos Ynigo, 1964), in Venezuela (Jurkovic, 1953), in Australia (Tbrner, 1968;
Zeissink, 1968) and in South Africa (de Waal, 1971).
In the silicate and oxide deposits, nickel is associated with minerals of "gar-
nierite" type in horizons at the base of the profile, and with goethite toward the
top of the profile. Most commonly, garnierite consists of a mixture of talc, serpen
tine, smectite and quartz (Brindley and Hang, 1973). In the oxide deposits, nickel is
associated with goethite or with amorphous iron oxihydroxides.

Several mineral sequences of weathering were studied in Brazil (Colin, 1985;
Colin et al., 1985), in West Africa (Paquet et al, 1982; Nahon et al, 1982) and in
New Caledonia (TVescases, 1975; Figure 17.8).
The investigated ultrabasic parent rocks contain in variable proportions olivine
(most commonly forsterite-rich), pyroxene (enstatite-rich orthopyroxenes and diop-
side-rich clinopyroxenes) and serpentine (predominantly lizardite). Nickel-bearing
parent minerals are essentially olivine and serpentine which appears most often as
the hydrothermal alteration product of olivine.
Differential weathering of the minerals (Goldich, 1938) of the ultrabasic rocks
results in a much faster weathering of olivine than that of orthopyroxene which in
turn is faster than that of clinopyroxene. Serpentine weathers at the slowest rate,
and is therefore found higher in the weathering profiles. This is why olivine can
release nickel to weathering solutions. Thus, when zones enriched in olivine are
located upslope from zones enriched in pyroxenes, as in the deposit of Jacuba in
Brazil, the weathering products of pyroxenes are also nickeliferous, because they
result from the interaction with a weathering solution previously enriched in nickel
(Colin et al, 1985).
Whether the location is on a well-drained plateau (or upslope), or in a poorly-
Ni - Gœthite Ni - Gœthite ; Al - Goethite Ni - Goethite ; Al - Goethite
Ni - Tale ; Talc + Quartz
Ni - Fé Smectites Al - Fé Smectites
Ni - Fé Smectites
Ni - Serpentine
Ni - Mg Smectites
+ Silico - ferric
Mg Carbonates Gel
Silico-ferric
Ni - Gel
OLIVINE PYROXENES SERPENTINE
Fig. 17.8. Mineral sequences in the weathering of nickeliferous ultrabasic rocks.

drained and confined gulley (or downslope), the weathering of minerals can gener
ate different products. Thus, weathering of olivine in confined environments (bot
tom of profile, bottom of slope) generates nickeliferous silico-ferric gels (sometimes
called iddingsite), or nickeliferous and magnesian smectites associated with mag-
nesian carbonates. In well-drained environments, Ni-Fe smectites (Ni-Fe saponite,
then Ni-hisingerite according to Paquet et al., 1981,1983) are formedfirst,although
this stage can be reached later and higher in the profiles of poorly-drained environ
ments.
Pyroxenes can weather at the bottom of profiles into amorphous ferruginous
products. However, in late weathering within better-drained horizons, they are pseu-
domorphed by of Al-Fe or Ni-Al-Fe smectites according to the composition of the
pyroxenes themselves, and whether weathering solutions are enriched in nickel or
not. Most often, weathering of pyroxenes generates Ni talc (kerolite-pimelite series)
accompanied by neoformed quartz. Serpentine (lizardite), without weathering into
new minerals, can rapidly become enriched in nickel which substitutes for magne
sium, which is evacuated by the solutions. This enrichment forms a Ni-end member
serpentine called nepouite (Brindley and Wan, 1975).
In the most oxidized portion of the profiles, all the weathering minerals of olivine,
pyroxenes and serpentine further weather into goethite whose nickel content de
pends on that of the mineral from which it originates. Furthermore, aluminous
pyroxenes can, in the upper part of profiles, generate aluminous goethites in their
final stage of evolution (Nahon and Colin, 1982).
Distribution of nickel
With respect to nickel content, nickeliferous latérites contain 2-5% NiO in sili
cate horizons and 1-3% on the average in oxidized zones with goethite.
The distribution of nickel in silicate phases was studied in the phases considered
the richest, namely nickeliferous talcs of the kerolite-pimelite series, nickeliferous
serpentines of the lizardite-nepouite series (Maksimovic, 1966, 1973; Brindley and
Wan, 1975; Brindley et al, 1979) and nickeliferous smectites (Paquet et al., 1981).
However, various spectroscopic techniques succeeded in locating nickel with
more precision than standard X-ray diffraction techniques and chemical micro-
analyses. Indeed, the crystallographic parameters of thefirstcoordination shell were
precisely established by diffuse reflectance and K-edge absorption spectroscopy,
whereas local ordering at the scale of several angstroms was determined by EXAFS
(Extended X-ray absorption fine structure) analysis as reported by Manceau and
Calas (1985) and Decarreau et al. (1987).
The results obtained by these authors are as follows:
(a) Structural configuration of Ni in the minerals of the kerolite-pimelite series:
two types can be considered, either Ni and Mg are clustered inside octahedral
sheets, or several Ni-Fe sheets are irregularly distributed within the phyllosilicate
Metaïlogeny of weathering: an introduction 463
structure. This would be a mixture of kerolite and pimelite sheets at the scale of
several sheets.
(b) Structural configuration of Ni in the serpentines of the lizardite-nepouite
series: nickel enrichment occurs by substitution of Mg in small domains of a size
smaller than 300Â, without any structural reorganization.
(c) Structural configuration of Ni in nickeliferous smectites: these smectites capa
ble of containing up to 22% Ni, would consist of an assemblage, with variable spatial
distribution of nontronite-like dioctahedral domains, rich in iron, and of pimelite-
like trioctahedral domains, rich in Mg and Ni. The latter, where all the nickel would
be concentrated, would represent about a third of the sample.
In summary, in silicate phases nickel would be concentrated in the structure of
serpentines (nepouite) by substitution of Mg according to a transformation in the
sense of Millot (1964), or in the structure of talcs (pimelite) during the precipitation
of these minerals. Substitution in serpentine could occur preferentially, and nicke
liferous serpentine could be slightly more stable than nickeliferous talc, as shown by
the thermodynamic diagram (Figure 17.9) established by Boeglin et al. (1983). This
diagram clearly shows that Ni-phyllosilicates are more stable than Ni-oxides and
hydroxides, which would explain why nickel concentrates preferentially in phyllosil-
icates. However, little is yet known about nickel in oxidized zones: is it substituted
in the structure of goethite, or is it in the form of nickel oxides and hydroxides
intimately mixed with goethite?
og(Ni2 + ) + 2pH
11 quartz I Si02
am.
Ni(OH)2
bunsenite
0-
-,
H º
I
1
öJ —. 1 -4
1
H 1— -
log(H4Si04)
Fig. 17.9. Solubility diagram of nickeliferous minerals (from Boeglin et al., 1983).
Cupriferous concentrations
Character of weathering cupriferous concentrations
Supergene enrichment in copper of cupriferous sulfides has been extensively

studied (Sillitoe and Clark, 1969; Sinclair and Gasparini, 1980; Thornber, 1975,
1983; Nickel, 1984).
The behaviour of copper in the weathering zone itself has been investigated
by several authors (Blain and Andrew, 1977; Parisot and Melfi, 1983; Mosser and
Zeegers, 1987; Toledo Groke et al., 1987), but the works of Koud (1988) and Creach
(1988) are the most complete and are used extensively in this section.
In this case also, it is difficult to present a typical weathering profile because the
nature of the parent rock is variable. The original cupriferous sulfide mineraliza-
tion can be massive or disseminated in very different host rocks and according to a
variable tectonic scenario. Furthermore, copper being a relatively mobile element,
is rapidly enriched in secondary cupriferous minerals in the weathering horizons
developed from cupriferous rocks. Thus, if original sulfides generate in situ sec-
ondary minerals, most of the latter precipitate in cavities or replace other minerals
which are not originally copper-bearing. Consequently, horizons of concentration of
copper do not occur necessarily above parent rocks, such as bauxites, but in hori-
zons developed laterally or downslope, particularly with respect to concentrations of
copper carbonates and silicates. Copper oxides are then generated in situ from the
latter.
In summary, all situations are possible: vertically with concentration at the top,
vertically with concentration at the bottom or laterally with concentration downs-
lope. The state of fracture and the structure of the host rock enclosing the sulfide
body are of critical importance.
Starting from parent sulfide minerals, the order of appearance of minerals with
increasing weathering is discussed from pétrographie observations made on the
deposits of Djenguile and Mindouli in Congo (Koud, 1988), and of Santa Blandina
(Creach, 1988) and Salobo (Toledo Groke et al., 1987) in Brazil, which display
well-developed lateritic weathering profiles (Figure 17.10).
In the investigated profiles, the source of copper consists of sulfides either in
massive lenses or in veins, or in small lenses disseminated in the surrounding host
rocks. Probably, the evolution of chalcopyrite toward secondary sulfides, such as
bornite, digenite, chalcosite, covellite and pyrite seems related to alterations in
the zone of secondary enrichment (Peterson, 1954), in the presence of meteoric
waters. In general, sulfates are poorly developed, if not completely absent. The
most common situation is the evolution of sulfides toward carbonate phases, in
Tenorite CuO + gihite

+hematite
(/} A k
LLI Cuprite Cu 2 0 + gthite 1
Ϋ
X + hematite |
O A A
Chrysocolla + gthite 1
Cu8Si802o(OH)8 8H 2 0 1
il 1 A
co r Chrysocolla + smectite + gthite
LU
Planchιite + gthite
Cu 7 Si 8 0 2 2 (OH) 2 11 i1
< L
o i " ii A1
CO
Dioptase + gthite
Cu Si 0 3 H 2 0
i1 ii Si - Cu Hydrated ferrie gel
if)
LU
Malachite + gthite i t
| Cu2 C0 3 (OH)2 1 Cu - Fe Hydrated siliceous gel
o * i A 1
1 i1 A
Azurite + gthite 11
< Cu3 (C0 3 ) 2 (OH)2 ϋ
o
ω
LU
Chalcancite or
tBrocantite
1
< CuS0 4 5 H 2 0 Cu 4 (OH) 6 S0 4
LL
CO
i 1 1
CHALCOCITE + COVELLITE + pyrite
Cu2S Cu S
i 1 +
CO
LU
1 DIGENITE + pyrite
Q 1 CU9S5
LL
A
3
CO
BORNITE Cu 5 FeS 4
A 1
CHALCOPYRITE
Cu FeS2 ____J
Fig. 17.10. Sequences of weathering minerals of cupriferous sulfides.
particular the most stable of them malachite which appears as a replacement of

sulfides, or precipitated in cavities of the host rocks (Chesworth, 1972a). Evolution
toward copper silicates, such as dioptase, plancheite, chrysocolla, appears higher up
in the profiles, where quartz and silicates of the host rocks are dissolved, releasing
most of their silica to weathering solutions. Replacement of quartz by dioptase and
plancheite is typical of the deposits of Mindouli in Congo, whereas replacement
of quartz by chrysocolla is characteristic of the profile of Santa Blandina in Brazil.
Replacement of carbonates by silicates is also common, but according to the types
of deposits and the ease of weathering of silicates in the surrounding host rocks,
sulfide parent rocks may weather directly to copper silicates.
In the surrounding host rocks located beneath or downslope from the lenses of
copper sulfides, weathering of non-copper bearing minerals (pyroxenes, garnets)
occurs in non-crystallized cupriferous weathering products, such as hydrated silico-
ferric gels in which the contents in copper and iron increase with the progress of
weathering. These weathering products may lead to mixtures of smectite, kaolinite
and goethite whose content in copper (established by means of a phase separation
in X-ray diffraction and microprobe analysis) seems related to smectite. Analyses by
infra-red spectrometry and EXAFS spectrometry (Creach, 1988) show that chryso-
colla forms particulated concentrations intimately associated with smectite and that
chrysocolla contains in its structure most of the copper in the form of Cu 2+ .
Finally, in the oxidized zone, copper carbonates and sulfates weather into cuprite
and, particularly, into tenorite, associated with iron oxides and oxihydroxides.
The high mobility of copper allows this element to enrich percolating solutions
and to precipitate in the weathering products of non-copper-bearing surrounding
host rocks. This particular feature is to be stressed in lateritic weathering profiles
where concentrations of cupriferous carbonates and silicates do not necessarily
immediately overlie parent sulfides.
Conclusions
Lateritic weathering is an efficient natural agent for refining ores. The variable
mobility of elements during lateritic weathering generates enrichments whose pét-
rographie and geochemical characteristics vary from one case to another. Specific
elements, such as copper, are so mobile that they concentrate laterally with respect
to the original protore. Others, such as gold build surficial dispersion halos. Finally,
it is obvious, especially in the case of nickel and copper that the silicate phases con-
taining them can only be detected by means of sophisticated techniques. However,
little is known, with any precision about oxidized phases.
In the future a special effort should be made in two directions.
(a) We need better to understand the partitioning and crystal chemistry of metal
cations in the structures of silicates and oxides under lateritic conditions; analytical,
theoretical and experimental approaches would be desirable.
(b) We need to place more emphasis on the geochemical significance of spatial
properties of lateritic ore bodies, such as the shape of the body and the textures
of the authigenic accumulations. Equilibrium approaches and models are in this

regard of little use, because the equilibrium condition is inherently blind to space.
Approaches and equations that include time and space are needed (Merino, 1987,
in press).
References
Almeida, E.B. de, 1977. Geology of the bauxite deposits of the Poços de Caldas alkaline district, State
of Minas Gérais, Brazil. Ph.D. Thesis, Stanford Univ., 265 pp. (unpublished).
Ambrosi, J.P. and Nahon, D., 1986. Petrological and geochemical differentiation of lateritic iron crust
profiles. Chem. Geol., 57: 371-393.
Baker, W.E., 1978. The role of humic acid in the transport of gold. Geochem. Cosmochim. Acta, 42:
645-649.
Bardossy, G.Y., 1982. Karst Bauxite. Developments in Economic Geology 14. Elsevier, Amsterdam, 441
pp.
Beauvais, A. and Nahon, D., 1985. Nodules et pisolites de dégradation des profils d'altération mangané-
sifères sous conditions latéritiques, exemples de Côte d'Ivoire et du Gabon. Sci. Geol. Bull., 38:
359-381.
Beauvais, A., Melfi, A , Nahon, D. and Trescases, J.J., 1987. Petrologie du gisement latéritique mangané-
sifère d'Azul (Brésil). Miner. Deposita, 22: 124-134.
Berner, R.A., 1978. Rate control of mineral dissolution under earth surface conditions. Am. J. Sci., 218:
1235-1252
Blain, C.E and Andrew, R.L., 1977. Sulphide weathering and the evaluation of gossans in mineral
exploration. Min. Sci. Eng., 9: 119-150
Boeglin, J.L., 1981. Minéralogie et géochimie des gisements de manganèse de Conselheiro Lafaiete au
Brésil et de Moanda au Gabon. Thèse 3ème Cycle, Univ. Toulouse, 155 pp. (unpublished).
Boeglin, J.L., Melfi, A., Nahon, D., Paquet, H. and Tardy, Y., 1983. Early formed mineral parageneses
in the deep zones of supergene ores in lateritic weathering. In: A.J. Melfi and A. Carvalho (Editors),
Lateritisation Processes, II Int. Sem. Proc, Säo Paulo, 1982, pp. 71-88.
Boulangé, B., 1984. Les formations bauxitiques latéritiques de Côte d'Ivoire. Les faciès, leur transfor
mation, leur distribution et l'évolution du modelé. Trav. Doc. ORSTOM, Paris, 341 pp.
Bourrié, G. and Pedro, G., 1979. La notion de pF, sa signification physico-chimique et ses implications
pedogenétiques. I. Signification physico-chimique. Relations entre pF et activité del'eau. Sci. Sol, 4:
313-322.
Bricker, O., 1965. Some stability relations in the system Ì ç - è 2 - Ç 2 è at 25°C and one atmosphere total
pressure. Am. Mineral., 50:1296-1354.
Brindley, G.W. and Hang, P.T., 1973. The nature of garnierites. I: structures, chemical compositions and
colour characteristics. Clays Clay Min., 22: 27-40.
Brindley, G.W. and Wan, H.N., 1975. Compositions, structures and thermal behaviour of nickel contain
ing minerals in the lizardite-nepouite series. Am. Mineral., 60: 833-871.
Brindley, G.W, Bish, D.L. and Wan, H.M., 1979. Composition, structures and properties of nickel-
containing minerals in kerolite-pimelite series. Am. Mineral., 64: 615-625.
Cantinone, P., Didier, P., Meunier, J.D., Parron, C, Guendon, J.L., Bocquier, G. and Nahon, D., 1984.
Kaolinites ferrifères et oxyhydroxydes de fer et d'alumine dans les bauxites des Canonettes (SE de
la France). Clay Miner., 19: 125-135.
Chesworth, W, 1972a. Thermodynamic study of the relative stability of malachite and azurite in soils.
Soil Sci., 113: 303-307.
Chesworth, W, 1972b. The stability of gibbsite and boehmite at the surface of the earth. Clays Clay
Miner., 20: 369-374.
Chukhrov, F.V., Gorshov, A.L., Sivtsov, A.V. and Beresovkaya, V.V., 1983. Mineralogy of manganese in
products of lateritic weathering. Int. Geol. Rev., 25: 719-730.
Cloke, P.L. and Kelly W.C., 1964. Solubility of gold under inorganic supergene conditions. Econ. Geol.,
59: 259-270.
Colin, E, 1985. Etude pétrologique des altérations de pyroxénite du gisement nickélifère de Niquelandia
(Brésil). Thèse, Univ. Paris VII, TDM ORSTOM, F17,135 pp.
Colin, E, Noack, Y., Trescases, J.J. and Nahon D., 1985. âlteration latéritique débutante des pyroxé-
nites de Jacuba, Niquelandia, Brésil. Clay Miner., 20: 93-113.
Colin, E, Edou Minko, A. et Nahon, D., 1989. Eor particulaire dans les profils latéritiques: altération
géochimique et dispersion superficielles en conditions équatoriales. C.R. Acad. Sci., Paris, Sér. II,
309: 553-560.
Creach, M., 1988. Accumulation supergène de cuivre en milieu latéritique: étude pétrologique, cristal-
lochimique et géochimique de l'altération du skarn de Santa Blandina (Itapeva, Brésil). Thèse, Univ.
Poitiers, 123 pp. (unpublished).
Davy, R. and El Ansary, M., 1986. Geochemical patterns in the laterite profile at the Boddington Gold
deposit, Western Australia. J. Geochem. Explor., 26: 119-124.
Decarreau A., Colin, E, Herbillon, A., Manceau, A., Nahon, D., Paquet, H., Trauth-Badaud, D. and
Trescases, J.J., 1987. Domain segregation in Ni-Fe-Mg smectites. Clays Clay Miner., 35: 1-10.
Delvigne, J., 1965., Pedogenèse en zone tropicale. La formation des minéraux secondaires en milieu
Didier, PH., 1983. Paragénèses à oxydes et hydroxydes de fer dans les bauxites et les cuirasses ferrug
ineuses. Thèse 3ème Cycle, Univ. Poitiers, 150 pp. (unpublished).
Evans, D.L.C., 1981. Lateritization as a possible contribution to gold placers. Eng. Min. J., 182: 86-91.
Frasche, D.E, 1941. Origin of the Surigao iron ores. Econ. Geol., 36: 280-305.
Freise, EW., 1931. The transportation of gold by organic underground solutions. Econ. Geol., 26: 421-
431.
Freyssinet, P., Zeegers, H. and Tardy, Y., 1989. Morphology and geochemistry of gold grains in lateritic
profiles of Southern Mali. J. Geochem. Explor., 32:17-31.
Garrels, R.M. and Christ, C.L., 1965. Solutions, Minerals and Equilibria. Harper and Row, New York,
N.Y., 450 pp.
Goldich, S.S., 1938. A study in rock-weathering. J. Geol., 46: 17-58.
Gordon, M. and Tracey, J.I., 1952. Origin of the Arkansas bauxite deposits. AIMME Symp., Problems
of Clay and Laterite Genesis, pp. 12-34.
Grandin, G. et Perseil, E.A., 1983. Les minéralisations manganésifères volcano-sédimentaires du Blafo-
Gueto (Côte d'Ivoire): paragénèses, altération climatique. Miner. Deposita, 18: 99-111.
Hem, J.D. and Roberson, CE., 1967. Form and stability of aluminum hydroxide complexes in dilute
solution. U.S. Geol. Surv., Water Supply Pap., 1827 A, 55 pp.
Hieronymus, B., 1985. Etude de l'altération des roches eruptives de l'Ouest du Cameroun. Thèse Doct.,
Univ. Paris VI, 273 pp. (unpublished).
Hören, A., 1953. The manganese mineralization at the Merid mine, Minas Gérais, Brazil. Ph.D. Thesis,
Harvard Univ., 224 pp. (unpublished).
Hotz, P.E., 1964. Nickeliferous latérites in Southwestern Oregon and Northwestern California. Econ.
Geol., 59: 355-396.
Jurkovic, I., 1953. Some geochemical aspects about the genesis of the nickel deposit Lorna de Hierro
(Venezuela). Geol. Vjesn., 17: 103-112.
Koud, J.M., 1988. Genèse et évolution des silicates de cuivre dans le gisement de Mindouli au Congo.
Sci. Geol. Bull, 41: 279-288.
Krauskopf, K.B., 1951. The solubility of gold. Econ. Geol., 46: 858-878.
Kuhnel, R.A., Roorda, HJ. and Steensma, U.S., 1978. Distribution and partitioning of elements in
nickeliferous latérites. Bur. Rech. Geol. Min. Bull., II: 191-206.
Lacroix, A., 1913. Les latérites de la Guinée et les produits d'altération qui leur sont associés. Nouv.
Arch. Mus., 5: 255-356.
Lacroix, A., 1943. Les péridotites de la Nouvelle Calédonie, leurs serpentines et leurs gites de Nickel et
de Cobalt. Les gabbros qui les accompagnent. Mém. Acad. Sci., Paris, 66: 1-143.
Lawrance, L.M., 1988. Behaviour of gold within the weathering profile in the Yilgarn Block, Western
Australia. Pubi. Univ. Western Aust., 12: 335-351.
Maksimovic, Z., 1966., B-kerolite-pimelite series from Goles Mountains, Yugoslavia. Proc. Int. Clay
Conf., Jerusalem, 1: 97-105.
Maksimovic, Z., 1973. Lizardite-nepouite isomorphic series. Zap. Miner. Obslich, 102: 143-149.
Manceau, A. and Calas, G., 1985. Heterogeneous distribution of nickel in hydrous silicates from New
Caledonia ore deposit. Am. Mineral., 70: 549-558.
Manceau, A., Llorca, S. and Calas, G., 1987. Crystal chemistry of cobalt and nickel in lithiophorite and
asbolane from New Caledonia. Geochim. Cosmochim. Acta, 51: 105-114.
Mann, A.W., 1984. Mobility of gold and silver in lateritic weathering profiles: some observations from
western Australia. Econ. Geol., 79: 38-49.
Mattigod, S.V. and Kittrick, J.A., 1980. Temperature and water activity as variables in soil mineral
activity diagrams. Soil Sci. Soc. Am. J., 44:149-154.
Melfi, A., Trescases, J.J. and Barros de Oliveira, S.M., 1980. Les latérites nickélifères du Brésil. Can.
ORSTOM, Sér. Géol., 11:15-42.
Merino, E., 1987. Textures of low temperature self-organization. Proc, Int. Meet. Geochem. Earth Surf.
and Proc. Min. Form., I, Grenada, 1986, pp. 597-610.
Merino, E., in press. The geochemistry of habits and textures of authigenic quartz. Int. Meet. Geochem.
Earth Surf, and Proc. Min. Form., II, Aix en Provence, Chem. Geol. Spec. Pubi, (abstract).
Millot, G., 1964. Géologie des argiles. Masson , Paris, 499 pp.
Millot, G., and Bonifas, M., 1955. Iransformations isovolumétriques dans les phénomènes de latéritisa-
tion et de bauxitisation. Bull. Serv. Carte Géol. Alsace Lorraine, 8: 3-10.
Mosser, C. and Zeegers, H., 1987. Copper in minerals formed by weathering of a granodiorite with
chalcopyrite (Burkina Faso). Proc. Int. Meet. Geochem. Earth Surf. Proc. Min. Form., Grenada,
1986, pp. 245-257.
Nahon, D., 1976. Cuirasses ferrugineuses et encroûtements calcaires du Sénégal Oriental et en Mauri
tanie. Systèmes évolutifs: géochimie, structures, relais et coexistence. Mém. Sci. Géol., Strasbourg,
44: 232 pp.
Nahon, D., in press. Introduction to the Petrology of Soils and Chemical Weathering. Wiley, New York,
N.Y.
191.
Nahon, D., Paquet, H. and Delvigne, J., 1982. Lateritic weathering of ultramafic rocks and the concen
tration of nickel in the Western Ivory Coast. Econ. Geol., 77: 1159-1175.
Nahon, D. and Colin, F, 1982. Chemical weathering of orthopyroxenes under lateritic conditions. Am.
J. Sci., 282: 1232-1243.
Nahon, D., Beauvais, A., Boeglin, J.L., Ducloux, J. and N'ziengui Mapangou, P., 1983. Manganite
formation in thefirststage of the lateritic manganese ores in Africa. Chem. Geol., 40: 25-42.
Nahon, D., Beauvais, A., N'ziengui Mapangou, P. and Ducloux, J., 1984. Chemical weathering of Mn-
garnets under lateritic conditions in northwest Ivory Coast (West Africa). Chem. Geol., 45: 53-
71.
Nahon, D., Beauvais, A. and Trescases, J.J., 1985. Manganese concentration through chemical weath
ering of metamorphic rocks under lateritic conditions. In: J.I. Drever (Editor), The Chemistry of
Weathering. Reidel, Dordrecht, pp. 277-291.
Nahon, D. and Pare, S., 1989. Lateritic concentration of manganese oxihydroxides and oxides. Geol.
Rundsch., 79.
Nahon, D., Herbillon, A. and Beauvais, A., 1989. The epigenetic replacement of kaolinite by lithio-
phorite in a manganese-lateritic profile, Brazil. Geoderma, 44: 247-259.
Nair, N.G.K., Santosh, M. and Mahadevan, R., 1987. Lateritization as a possible contributor to gold
placers in Nilaubur Valley, SW India. Chem. Geol., 60: 309-315.
Nickel, R.E., 1984. The mineralogy and geochemistry of the weathering profile of the Teutonic Bore
Cu-Pb-Zn-Ag sulphide deposit. J. Geochem. Explor., 22: 239-264.
Norton, S.A., 1973. Laterite and bauxite formation. Econ. Geol., 68: 353-361.
Paquet, H., Duplay, J. and Nahon, D., 1982. Variations in the composition of phyllosilicates monopar
ticules in a weathering profiles of ultrabasic rocks. Proc. 7th Int. Clay Conf., pp. 595-603.
Paquet, H., Duplay, J., Nahon, D., Tardy, Y. and Millot, G., 1983. Analyses chimiques de particules
isolées dans les populations de minéraux argileux. C. R. Acad. Sci. Paris, 296: 699-704
Parc, S., 1989. Contribution à l'étude cristallochimique et thermodynamique des oxy-hydroxydes de
manganèse dans l'altération latéritique. Thèse, Univ. Aix Marseille III, 128 pp. (unpublished).
Parc, S., Nahon, D., Tardy, Y. and Vieillard, P., 1989. Estimated solubility products and fields of sta
bility for cryptomélane, nsutite, birnessite, and lithiophorite based on natural lateritic weathering
sequences. Am. Mineral., 74: 466-475.
Parisot, J.C. and Melfi, A., 1983. Lateritic alteration and metasedimentary rocks with copper sulfide
mineralization in Central Brazil. In: A.J. Melfi and A. Carvalho (Editors), Lateritisation Processes.
Proc. II Int. Sem., Säo Paulo, pp. 185-195.
Perseil, E.A. and Grandin, G., 1978. Evolution minéralogique du manganèse dans trois gisements
d'Afrique Occidentale: Mokta, Tambao, Nsuta. Miner. Deposita, 3: 295-321.
Perseil, E.A. and Grandin, G., 1985. Altération supergène des protores à grenats manganésifères dans
quelques gisements d'Afrique de l'Ouest. Miner. Deposita, 20: 211-219.
Peterson , Í.Ñ., 1954. Copper cities copper deposit, Globe Miami district, Arizona. Econ. Geol., 49:
362-372.
Roy, S., 1981. Manganese Deposits. Academic Press, London, 458 pp.
Santos-Ynigo, L., 1964. Distribution of iron, alumina and silica in the Pujada laterite of Mati, Davao
province, Mindanao island (Philippines). Proc. 22nd Int. Geol. Congr., New Dehli, 14: 126-141.
Sigolo, J.B. and Boulangé, B., 1987. Caracterizâo das fades de alteraçao de urna toposeqiiencia no
Maciço Alcalino de Passa Quatro, MG, Brasil. Rev. Brasil. Geoci., 17: 269-275.
Sillitoe, R.H. and Clark, A.H., 1969. Copper and copper iron sulfides as the initial products of supergene
oxidation, Copiago mining district, Northern Chile. Am. Mineral., 54: 1684-1710.
Sinclair, I.G. and Gasparini, E., 1980. Textural features and age of supergene mineralization in the
Detour copper-zinc-silver deposit, Quebec. Econ. Geol., 75: 470-477.
Stoflregen, R., 1986. Observation on the behavior of gold during supergene oxidation at Summitville,
Colorado, U.S.A., and implications for electrum stability in the weathering environment. Appi.
Geochem., 1: 549-558.
Tardy, Y., 1969, Géochimie des altérations. Etude des arènes et des eaux de massifs cristallins d'Europe
et d'Afrique. Mém. Serv. Carte Géol., Alsace Lorraine, 31,199 pp.
Tardy, Y. and Nahon, D., 1985. Geochemistry of latérites, stability of Al-gœthite, Al-hematite, and
Fe 3+ -kaolin ite in bauxites and ferricretes: an approach to the mechanism of concretion formation.
Am. J. Sci., 285: 865-903.
les profils latéritiques. C.R. Acad. Sci., Paris, Sér. II, 306: 39-44.
Metallogeny of weathering: an introduction All
Thornber, M.R., 1975. Supergene alteration of sulphides. I. A chemical model based on massive nickel
sulphide deposits at Kambalda, Western Australia. Chem. Geol., 15: 1-14.
Thornber, M.R., 1983. Weathering of sulphide ore bodies. The chemical processes of gossan formation.
In: R.E. Smith (Editor), Geochemical exploration in deeply weathered terrains. CSIRO, Wembley,
W.A., pp. 67-72.
Toledo-Groke, M.C., Melfi, A.J. and Parisot, J.C., 1987. Comportamento do cobre durante o intem-
perismo das rochas xistosas cupriferas do Salobo 3A, Serra dos Carajas. Geodi. Bras., 1: 187-200.
Trescases, J.J., 1975. dévolution géochimique supergène des roches ultrabasiques en zone tropicale.
Formation des gisements nickélifères de Nouvelle Calédonie. Mém. ORSTOM, 78, 259 pp.
Turner, A.R., 1968. The distribution and association of nickel in the ferruginous zones of latérites of the
Giles complex, Australia. Amdel. Bull., 5: 76-93.
Valarelli, J.V., Bemardelli, A. and Beisigel, W.R., 1978. Aspectos genéticos do minério de manganes do
Azul. Proc. XXX Congr. Bras. Geol., Recife, 4: 1670-1679.
Vasconselos, P. and Kyle, J.R., 1989. Supergene geochemistry and crystal morphology of gold in a semi-
arid weathering environment: application to gold exploration. XIII Int. Geochem. Explor. Symp.,
Rio, p. 22 (abstract).
Vletter, R. de, 1955. How Cuba nickel ore was formed. A lesson in laterite genesis. Eng. Min. J., 156:
84-87.
Waal, S.A. de, 1971. South African nickeliferous serpentinites. Miner. Sei. Eng., Pretoria, 3: 32-45.
Weber, F, 1968. Une série précambrienne du Gabon: le Francevillien, sedimentologie, géochimie, rela
tions avec les gites minéraux associés. Mém. Serv. Carte Géol. Alsace Lorraine, Strasbourg, 28, 326
pp.
Weber, F, Leclerc, J. and Millot, G., 1979. Epigénies manganésifères successives dans le gisement de
Moanda (Gabon). Sci. Géol. Bull., 32: 147-164.
Webster, J.G., and Mann, A.W., 1984. The influence of climate, geomorphology and primary geology on
the supergene migration of gold and silver. J. Geochem. Explor., 22: 321-335.
Webster, J.G., 1986. The solubility of gold and silver in the system Au-Ag-S-0 2 -H 2 0 at 25°C and 1
atm. Geochim. Cosmochim. Acta, 50: 1837-1845.
Wilson, A.F., 1984. Origin of quartz free gold nuggets and supergene gold found in latérites and soils.
A review and some new observations. Aust. J. Earth Sci., 31: 303-316.
Zeissink, H.E., 1968. The mineralogy and geochemistry of a nickeliferous laterite profile (Greenvale,
Queensland, Australia). Miner. Deposita, 4: 132-152.
Zen, E.A., 1972, Gibbs free energy, enthalpy and entropy of ten rock forming minerals: calculations,
discrepancies, implications. Am. Mineral., 57: 524-553.
475
Chapter 18
Paleopedology: stratigraphie relationships and

empirical models
VR WRIGHT
Introduction
A paleosol is a soil formed on a landscape of the past (Ruhe, 1956; Working

Group on the Nature and Origin of Paleosols, 1971; Valentine and Dalrymple,
1975). Such ancient soils may be buried beneath younger materials (buried pale-
osol), may be still within the solum representing soils formed during an earlier
phase of landscape evolution but persisting into a period of changed environmental
conditions (relict paleosol), or may represent a buried paleosol exposed by erosion
(exhumed paleosol).
Paleosols have been the subject of considerable attention by pedologists, Qua-
ternary scientists and geologists. They can reveal evidence on past environments
as regards climate, vegetation cover, geomorphology, length of pedogenesis and
sedimentation rates. They have been used for paleoenvironmental reconstruction
in archaeological research (Macphail, 1986), in Quaternary geology (Yaalon, 1971;
Mahaney, 1978; Boardman, 1985) and, increasingly, in the pre-Quaternary. Much
literature is concerned with the glacial and interglacial paleosols of northern Europe
and northern North America. In other parts of the world prolonged weathering,
even back to the Mesozoic or Palaeozoic, has created complex landscapes with
residual soils where study involves a different conceptual approach (Beckmann,
1983).
Arguably the major challenge in paleopedology is to tackle the enormous amount
of material in the pre-Quaternary geological record. Many ancient alluvial and
deltaic sequences each contain hundreds of paleosols, which may provide the means
for a major new initiative in interpreting ancient environments. An analogy can be
drawn with the revolution which occurred in geology in the 1960's when sedimentary
structures were used to interpret ancient flow regimes and depositional settings.
This chapter aims to provide a review of the stratigraphie concepts of paleope-
dology. Tb achieve this both Quaternary and pre-Quaternary examples are used, and
are integrated to provide a framework for use in understanding paleosol sequences,
476 VP. Wright
especially in the thick alluvial sequences so common in the geological record. In

addition soil formation associated with major periods of landscape evolution (such
as at major unconformities) will also be briefly discussed.
Paleosol recognition
In the past geologists working on pre-Quaternary paleosols relied heavily on

features such as fossil rootlets to identify paleosols. Now a more process-based ap-
proach is used with paleosols being recognized on the basis of exhibiting features
reliably attributed to pedogenic processes such as illuviation, eluviation, gleying
and so on, which operated before the material was buried. There are many fea-
tures produced by pedogenic processes which are, or so it appears at the present
time, unique to soils and so are diagnostic of pedogenesis. Other features, such as
destratification, rubifaction, duripan formation and mottling, while being charac-
teristic of soil formation, are not diagnostic. One particular problem here is that
early hydromorphism, caused by rising groundwaters as the sediment pile aggrades,
can create colour changes, including mottling, and even some forms of duripans
(Wright, 1990). Much remains to be done to clarify the reliability of many soil-like
features in the older paleosols, especially those in the pre-mid Paleozoic when the
biological activity in soils was radically different from the present time.
A constructive approach to this problem came from Valentine and Dalrymple
(1975) who have argued that a crucial test of whether a paleosol really is a true soil
is that it can be related to a paleocatena. Such topographic relationships would be
unlikely to be produced by diagenetic processes. This criterion is not an easy one
to apply for two reasons:firstly,most geological outcrops are small and recognizing
lateral relationships is very difficult if not impossible; secondly, the majority of pre-
Quaternary paleosols so far documented weredeveloped on extensive alluvial plains
with negligible relief. Paleocatenas have been recognized in many pre-Quaternary
sequences but they are not common. This partly reflects the fact that many Qua-
ternary paleosols have been found in areas where topographic factors have been
essential in burying the soils, by colluvial processes (Butler, 1967; Rose et al., 1985),
and the possibilities for recognizing catenas are thus greater in such settings.
Stratigraphy terminology and concepts
If proof of the former activities of diagnostic pedogenic processes can be estab-

lished (Retallack, 1988; Wright, 1990), it may be desirable to designate a formal
name for it, for stratigraphie purposes. Several terms have appeared for this pur-
pose such as pedoderm, geosol or paleosol, and much debate has taken place on
Paleopedology: stratigraphie relationships and empirical models All
their usage (Brewer et al., 1970; Catt, 1988, 1989; Morrison, 1978; North Ameri
can Commission on Stratigraphie Nomenclature, 1983; Parsons 1981; Walker et al.,
1984).
Stratigraphie relations in paleosols are complex on several levels. It has been
argued that all soils evolve and are thus, by their very nature, polygenetic (Johnson
and Watson-Stegner, 1987). Pedogenic processes are depth-related and aggradation,
degradation or deepening of the soil profile creates complex overprinting within
single soil profiles.
In sequences of paleosols these overprinting relationships can be highly complex
and Morrison (1978) has provided simple terms to describe the typical relation
ships encountered in such sequences (Figure 18.1). Composite paleosols are profiles
which overlap one another, a consequence of the burial of one profile by an incre
ment of sediment which is thinner than the soil profile which subsequently develops
in it. A compound profile consists of non-overlapping profiles separated by sedi
ment. Bos and Sevink (1975) introduced other terms to describe these relationships.
Soils typically have complex lateral stratigraphie relationships, especially in areas of
slope development or even fault movement (Figure 18.1).
In both compound and composite profiles, sedimentation has been episodic with
long periods between each phase allowing horizonation to develop. If the rate of
sedimentation is high and continuous, the residence time of the material in the soil
may be so low that no horizonation may occur and thick, weakly pedified soils may
occur. If the rate of sedimentation is low, well developed, but thin composite profiles
may develop. Figure 18.2 illustrates some of these relationships for a hypothetical
calcrete-bearing sequence.
Thefirstpre-Quaternary paleosols to be widely documented were probably those
associated with coals, especially from the Upper Carboniferous. These serve to
illustrate a particular problem surprisingly common in many paleosols, mainly the
occurrence of polygenetic profiles, which adds to the difficulties in interpretation.
These coal-bearing paleosols developed in extensive, low relief delta plains and
hydromorphic groundwater gley soils were very common but podsolic soils also
developed on better drained areas showing prominent albic horizons (Besley and
ÃÀin i n
composite paleosols
active fault compound paleosols
Fig. 18.1. Schematic diagram showing soil stratigraphie relationships for an active fault zone. Compound
and composite relationships can occur in any stacked sequence (after Machette, 1978).
478 VR Wrìght
Sedimentation
Soil Rate
Development low high
Fig. 18.2. Relationships between rates of sedimentation, soil maturity, and time. Soils 1 and 2 are
compound paleosols; 3 and 6 are weakly developed cummulate paleosols with few carbonate nodules;
4 and 5 are mature composite paleosols. In this case calerete bearing paleosols have been used. The
numbers I-IV refer to stages of maturity defined by Machette, 1985. Density of vertical lines indicates
maturity of calerete.
Fielding, 1989; Fielding et al., 1988; Percival, 1986; Gardner et al., 1988). However,
these latter paleosols commonly occur beneath in situ coals (originally peats). The
podsolic paleosols reflect free drainage while the coal horizons indicate prolonged
hydromorphism and thick peat formation. Thus the paleosols are polygenetic.
These compound paleosols are so common that they have been the subject of
much discussion and several explanations have been offered for this apparent con
tradiction (Gardner et al, 1988).
(a) The albic horizon [termed gannisters or underclays (Percival, 1986)] resulted
from leaching (causing eluviation, loss of chlorite and kaolinite formation) which
occurred before peat formation, which represents a later accumulation of organic
matter under hydromorphic conditions.
(b) The original peat formation and leaching were synchronous and the differ
ences reflect local changes in pH, rates of bacterial decomposition.
Paleopedology: stratigraphie relationships and empirical models 479
(c) Leaching occurred after peat formation by acid waters (perhaps by phreatic
flow?)
While the second (b) mechanism seems unrealistic, the third may explain some
modern examples of kaolinite-enrichment beneath swamp peats (Staub and Cohen,
1978), but is not applicable because eluvial horizons apparently do not occur. The
consensus is for the first model and, in the case of some Pennsylvanian leached
soils from the U.S., Gardner et al. (1988) were able to show that the leached pa-
leosol horizon developed up to 5m above the contemporaneous water-table, and
certainly not synchronously with peat formation. However, the model requires an
initial phase of lowered water-table followed by a rise to create hydromorphic con-
ditions. This rise may reflect compaction, subsidence or a transgression, and many
gannister-coal sequences are overlain by marine shales representing major sea-level
rises (Percival, 1986).
Not only are vertical relationships complex, but so also are lateral relationships.
Finkl (1980) has provided a general review of the problems of soil and paleosol
stratigraphy. Pawluk (1978) has also discussed aspects of paleosols and stratigraphy.
Paleosol-landscape relationships
T\vo broad groupings of paleosols can be recognized based on their relationships

to landscapes. Firstly there are relict soils or paleosols associated with very old
landsurfaces and deep weathering profiles such as Western Australia or much of
era tonic Africa. Typically, these represent exposure periods of millions of years. The
other grouping consists of paleosols formed in essentially depositional landscapes.
Their occurrence reflects landscape stability but their preservation requires that
deposition has occurred.
The following sections review aspects of these two types relevant to pre-Quater-
nary paleosols and concentrate on the stratigraphie implications of such paleosols.
Paleosols at major unconformities
The bulk of the literature on Quaternary paleopedology has concentrated on

paleosols in the northern hemisphere. Pleistocene glaciations removed most of the
earlier weathered zones and most Quaternary paleosols are hosted in glacial or
interglacial deposits such as till, alluvium or loess. However, in the southern hemi-
sphere some landscapes are as old as the Cambrian (Beckmann, 1983). The concep-
tual framework for studying such areas is radically different to that used in studies
of aggraded Quaternary covers, but is critical for those studying paleosols at major
unconformities.
The landscape evolution of Australia provides a possible analogue with which to
interpret unconformity paleosols. Here a series of weathering profiles with latentes
480 VR Wrìght
and silcretes have developed on landsurfaces persisting from the late Paleozoic
onwards, many of which have been preserved in surface or near surface settings.
These remarkable weathering profiles have been dated using a variety of tech
niques including paleomagnetism and oxygen isotopes (Bird and Chivas, 1988).
Oilier et al. (1988) have provided a useful case study of the geomorphic and pe-
dogenic history of the Kalgoorlie area of Western Australia which covers a time
span or over 200 million years, based upon stratigraphie studies of a variety of
saprolites, surficial materials and duricrusts. The stratigraphie sequences record the
interplay of large scale tectonic and climatic changes, eustacy, prolonged periods of
landscape stability and weathering, phases of landscape inversion and cut-and-fill
sequences, and importantfluctuationsin water-table positions. Theirfindingscon
firm the model of complex polygenetic development for cratonic areas offered by
Fairbridge and Finkl (1980).
Very few studies of major unconformity paleosols have been carried out to date.
One such example is the extensive sub-late Cretaceous "lateritic" profiles of the
Middle East and the eastern Mediterranean (Nicholas and Bildgen, 1979). Such
lateritic profiles require periods of 106-107 years to form (McFarlane, 1983; Idnurm
and Senior, 1978) and develop within complex, frequently tiered landscapes. The re
worked latentes of the Jurassic of Israel (Goldbery, 1982), incorporated into other
paleosol-bearing sequences, illustrate how the evolution of ancient landscapes can
be reconstructed in certain circumstances. A variety of sedimentological and pedo-
logical techniques may be required to decipher the pedostratigraphic sequences
(Finkl, 1984), which then have to be related into spatial and temporal scenarios.
Such paleosols and related materials represent a major challenge especially for re
searchers familiar only with northern hemisphere soils and geomorpho-stratigraphic
concepts. The Quaternary soils and soil sequences of the northern hemisphere do
not provide the only basis for interpreting paleosols, and will be unlikely to apply to
most situations in the geological record.
Paleosols in ancient alluvial sequences: geomorphic and stratigraphie relationships

Most pre-Quaternary paleosols are to be found in ancient alluvial sequences.
They typically occur in large numbers; for example the Lower Old Red Sandstone
of southern Britain contains an estimated 600 calcrete-bearing paleosols within its
3 km thickness (Allen, 1986), while the Eocene Willwood Formation of Wyoming
contains 500-1200 superposed paleosols (Kraus and Bown, 1986).
The abundance of paleosols in ancient alluvial sequences requires some explana
tion. The degree of pedogenic modification that an alluvial sediment will undergo
depends on the residence time of material within the zone of active soil forma
tion, and the rate of pedogenesis. The residence time depends on the sedimentation
rate and the thickness of that zone. If we take afloodplain,which has an average
sedimentation rate of 2-3 mm per year (Bridge, 1984), and assume the zone of
significant pedogenesis is 2 m thick, then the averagefloodplainsediment will have

a residence time of up to 103 years in the zone of soil formation before it is pro
gressively buried and isolated. This is enough time to radically alter some sediment
properties, depending on soil conditions. By this argument most floodplain sedi
ments should have received some pedogenic modification resulting in simple soil
profiles, but this time period may not be enough for well developed profiles to form.
However, the abundance of well developed paleosols in ancient alluvial se
quences is evidence of the punctuated nature of sedimentation (Kraus and Bown,
1986). Thus the rates of sedimentation in alluvial settings could not have been uni
form and varied both in time and space. The rate varies across the depositional
surface with respect to the source of sediment. In addition the sedimentation rate
varies in time both as a consequence of intrinsic factors and because of changes
in the overall rate of sediment supply (extrinsic controls). In the case of the latter,
tectonic or eustatic changes in base-level and climatic change influence the supply
rate. A major intrinsic control arises from the fact that the depositional rate appears
to decline through time as thefloodplainsurface approaches a hypothetical regime
level (Nanson, 1980; Hayward, 1985).
The absence of pedogenic overprinting in floodplain deposits may reflect high
rates of sedimentation, low rates of pedogenesis, the loss of evidence by diage-
netic overprinting, or simply that the evidence has been missed by the observer or
misinterpreted as a burial diagenetic effect. The opposite case is where floodplain
deposits contain evidence of pedogenic processes but which represent reworked soil
material. The widespread occurrence of particles identified as former soil aggre
gates in Australian river sands are a striking example of this (Nanson et al., 1986;
Rust and Nanson, 1989).
Understanding either the lateral or sequential relationships of paleosols requires
an appreciation of geomorphic factors. A soil is intimately linked to the landscape
in which it forms, and sequences of paleosols provide us with a picture of land
scape change. We can then evaluate, using pedostratigraphic models, whether in
trinsic or extrinsic factors have affected pedogenesis. To illustrate this approach,
soil-landscape and pedostratigraphic models will be examined from both floodplain
and alluvial fan settings.
Floodplain soils: geomorphology and pedostratigraphy

Several simple soil/landscape relationships can be recognized on floodplains
which, when identified in ancient sequences, can be used to reconstruct past land
scapes. Figure 18.3 illustrates these catenary, pedofacies and chronosequence rela
tionships. Reviews offloodplainsoils have been given by Gerrard (1981,1987).
(a) Catenary relationships. Catenas (or toposequences) are groups of soils with simi
lar parent materials, developed under similar climates, but with different character
istics related to variations in relief and drainage.
482 VE Wright
Most studies on catenas have looked at hill slopes, but they also occur on flood-
plains, reflecting the topographic differences between alluvial ridges and the flood-
basins (backswamps). The ridges are areas adjacent to the channel and are slightly
above the level of the floodplain and are generally above the water-table. Soils
(Entisols) developed here are commonly better drained than adjacent lower-relief
areas, although this is partly due to the coarser nature of the ridge sediments
(Figure 18.3). The topographically lower areas on thefloodplain,characterized by
finer-grained sediments, are either frequently flooded or have their surfaces close
to, or even below, the water-table. As a result the drainage is typically poor. The
types of soil to develop will depend on the climate, and in humid areas gley or
pseudo-gley soil conditions may develop, whereas in areas with a prominent dry
season (and in alluvium rich in swelling clays) Vertisols may develop. Prominent
Vertisol formation requires that the local soil moisture regime enhances the climatic
contrast, and typically these soils are best developed in areas of impeded drainage
(Duchaufour, 1982).
Areas removed from the zone offlooding,such as relict alluvial ridges and terrace
surfaces, will have better drained conditions.
As the floodplain aggrades, the water-table may rise more slowly, resulting in
improved drainage (Hayward, 1985). This will be the case also as alluvial ridges
become established following avulsion or by ridge migration. As a result poorly
drained soils may be overlain by better drained ones (Farrell, 1987). At what scales
of channel size these differences become well developed is unclear.
DISTANCE FROM CHANNEL
-Alluvial Ridge 1 l· -Floodplain- H Terrace
Channel,
mature soils
locally poorly drained moderately mature
soils - climatic soils.
controlled. Immature -
mature.
CATENARY RELATIONSHIP -
PEDOFACIES RELATIONSHIP "
coarser sediment,
well-drained, immature CHRONOSEQUENCE
soils. RELATIONSHIP
Fig. 18.3. Soil landscape relationships on a floodplain.

Païeopedology: stratigraphie relationships and empirical models 483
(b) Pedofacies relationships. The pedofacies concept was introduced by Bown and
Kraus (1987), from their studies of the lower Eocene Willwood Formation in
Wyoming. The term denotes: "laterally contiguous bodies of sedimentary rock that
differ in their contained laterally contiguous paleosols as a result of their distance
(during formation) from areas of relatively high sediment accumulation" (Bown and
Kraus 1987, p. 599).
Pedofacies develop because the rate of sedimentation across thefloodplainde
creases sharply with distance from the channel (source) area (Figure 18.3; Bridge
and Leeder, 1979). The degree (maturity) of soil development will reflect the re
lationship between the rates of sedimentation and pedogenesis. In areas where the
residence time of the sediment in the soil is great (low rate of sedimentation), a
relatively more mature profile may develop; but where the reverse is the case, no
significant soil development, or a thick, weakly developed, overlapping (composite)
set of profiles may occur (Leeder, 1975). Thus areas near the alluvial ridge will tend
to have thick, immature profiles while positions more distal on thefloodplainshould
exhibit more mature profiles.
Kraus (1987) and Kraus and Bown (1988) applied this concept to understand
ing various scales of alluvial sequences. They noted packages of paleosols which
systematically changed up-sequence (Figure 18.4). Such sequences were interpreted
as reflecting maturity differences due to changing positions on thefloodplainin
relation to the avulsing channel belt. As the channel belt avulsed away from a
point on thefloodplain,the profiles became more mature, and vice versa. Such a
model is a very useful tool to use on thick, aggraded sequences but the model en
visages simple avulsion and continuous sedimentation and aggradation. The model
produces overlapping, composite profiles across the pedofacies gradient, and the
Willwood Formation does consist of such superimposed profiles (Bown and Kraus,
Fig. 18.4. Maturity changes in a soil sequence reflecting pedofacies relationships during successive
avulsion events. 1-3 represent increasing degrees of soil maturity.
484 VF Wright
1987). However, many ancient paleosol sequences are characterized by discrete

(compound), mature paleosols, typically separated by non-pedified or very lightly
pedified intervals. Such sequences will not be generated by the gradualistic pedofa
cies model, but require episodic (punctuated) deposition (Kraus and Bown, 1986)
and they reflect the interaction of both intrinsic and extrinsic factors. For example
the Permo-Tïiassic alluvial Violette Horizonte paleosols of Germany are strongly
horizonated with prominent iiluvial horizons and are traceable over an outcrop of
several hundred kilometres in extent (Ortlam, 1970, 1971, 1974) and must reflect
formation during major regional pauses of sediment accumulation. Isolating the in
dividual controlling factors from the rock record can be difficult, especially where
both local tectonic and marine-influenced base-level changes were known to have
occurred (Wright and Robinson, 1988).
The pedofacies model developed by Bown and Kraus (1987) raises a problem
when reconsidering rates of avulsion. Relatively little is known of the actual rates
of avulsion on floodplains and a figure of once every 103 years is often quoted
(Allen, 1974; Bridge, 1984). The number of avulsions separating successive channel
deposits in a vertical sequence should range between one and ten (Bridge, 1984).
This means that eachfloodplaininterval should represent 103-104 years. Such time
periods could allow up to 20-30 m of uncompacted sediment to form, but is not long
enough for several mature paleosols to have formed. Bown and Kraus (1987) record
Spodosols and Alfisols in their avulsion-controlled pedofacies sequences from the
Willwood Formation. However, Birkeland (1984) notes that Alfisols require 104
years or more to develop. This means that just one profile would constitute all the
time available in a sequence, leaving no time for other profiles developed during the
full avulsion cycle.
(c) Terracing: landscape dissection. Many ancientfloodplainsequences exhibit promi

nent mature carbonate horizons of Alfisols and Aridisols (Allen, 1986; Lehman,
1989; McPherson, 1979). Such horizons require long periods of time to form, based
on studies of Quaternary soils (Machette, 1985). They would have been unlikely
to have formed in an avulsion-controlled situation as currently envisaged. Indeed
their modern counterparts are generally to be found on old (pre-Holocene) terraces
(Gilè and Grossman, 1979).
Many, if not most, modern alluvial systems exhibit multiple, stepped surfaces.
These may be purely erosional or the remnants of earlier depositional surfaces
isolated by base-level changes. It would be naive to imagine ancient landscapes as
lacking such features. A geomorphic surface such as a terrace, isolated from the
main channel belt, will be subjected to lower sedimentation rates than the active
floodplain (eolian input and erosion need also to be considered). It is on such
surfaces that mature soils develop. Hall (1983) has provided an introduction to
terrace/soil stratigraphies.
From outcrop or subsurface data, it is very difficult to recognize evidence for
terracing in ancient sequences. Rare examples of gullying are known (Kraus and
Middleton, 1987), but major terraced sequences must be inferred rather than actu
ally seen. Behrensmeyer and Tkuxe (1982) were able to correlate mature paleosols
in the Miocene Siwalik Group of Pakistan with phases of valley cutting and terrac
ing. Retallack (1986b), in a study of Eocene-Oligocene paleosol sequences from the
Badlands National Park, South Dakota, recognized mature soil development linked
to phases of downcutting during periods of drier climates and reduced vegetation.
With limited exposure and a lack of detailed stratigraphie control, proving land
scape dissection can be very difficult. Terracing must be inferred rather than demon
strated. Figure 18.5 illustrates the difficulties in interpreting ancient alluvial se
quences. The three hypothetical pedo-sequences, A, B and C differ radically from
one another yet represent the same time interval and are the same thickness. There
is no means of reliably correlating between sections. The mature soil in sequence
A represents three phases of soil development and two separate phases of erosion
elsewhere. Tb form such a mature soil has required isolation on a terrace surface.
The whole lithosome, on a large scale, must be complex, a result of phases of valley
incision andfill.This sort of stratigraphie model must apply to many ancient coastal
alluvial sequences, reflecting small scale sea-level changes (20 and 40,000 year or
bital cycles). One possible example of such a lithosome has been described by Allen
and Williams (1982) from the lower Devonian of Dyfed. In this case they were able
to use ash-fall tuffs as markers, to reconstruct the architecture of a 15-30 m thick al
luvial suite. In their stratigraphie reconstructions the paleosol intervals are laterally
discontinuous and in some cases as many as five paleosol intervals are truncated by
B C
__
7
[ÃÔ
7
TT — 7
rm
4-6
TTTT
3
TT ^ ^ ^ ^5-6 TT \
2 6 TT
TT
1
TT 1 ΓΓΠ 1 ΓΓΠ
Fig. 18.5. Schematic diagram showing complex soil stratigraphies generated by simple phases of valley
incision andfilling.1-7 are geomorphic surfaces; A-C are sections. The density of short vertical lines is
a guide to the length of pedogenesis.
486 VR Wright
a single channel base. It could be argued that the presence of mature profiles and
complex stratigraphies is the signature for terracing.
Finally it should be stressed that any suite of alluvial paleosols will, in general,
exhibit pedofacies, catenary and terracing-related relationships. The alluvial soils of
the Hadejia River in Nigeria clearly exhibit the types of relationships shown in Fig
ure 18.3, with Entisols developed on the levees, Vertisols developed on backswamp
clays, and Alfisols developed on nearby terraces. Smith (1990) has recently reported
pedofacies relationships in Permian alluvial sequences from South Africa, but in this
case catenary factors (hydromorphism) were the dominant controls in influencing
the types of soils which formed.
Arid alluvial fans: geomorphology and pedostratigraphy

Paleosols have been widely documented from arid-type alluvial fan deposits.
Their pedostratigraphies can be used to evaluate the influences of tectonic, climatic
and intrinsic controls on fan deposition. While most fan deposits encountered in the
geological record are of the semi-arid or arid types, humid fan deposits have also
been recognized (Besley and Tbrner, 1983; Besley and Fielding, 1989).
Sedimentation is not uniform on alluvial fans and this affects the distribution and
nature of soil development (McCraw, 1968; Wright and Alonso-Zarzo, 1989). Soil
formation will take place on stable fan surfaces which form either because erosion
and deposition has stopped over the whole fan, as a result of climatic change, or
because the locus of deposition/erosion is moved to some other part of the fan as
a consequence of fan entrenchment and avulsion or dissection. Sediment is trans
ported onto the fan typically by a fan head trench cut into the upper (proximal) fan
(Figure 18.6). The presence of this trench is critical in determining the pattern of
soil development on the fan (McCraw, 1968). Fans which lack fan-head entrench
ment have the least developed soils in the fan head region where aggradation rates
are highest (Figure 18.7a), as sediment is deposited near the apex. On entrenched
fans the trench may cut through only part of the fan and intersect the fan surface
at the intersection point (Figure 18.6), below which deposition occurs as braided
channel or sheet flow dominated lobes. Areas adjacent to this trench are starved
of sediment and soils develop. Each active lobe covers only a part of the lower fan
Intersection
point Fan surface
Basin Floor (playa)
or axial drainage
Apron
Fig. 18.6. Schematic diagram showing major features of alluvial fan systems.
Fig. 18.7. Diagrammatic cross-sections of idealised soil stratigraphies, a. Non-trenched fan. b. Proximally
trenched fan. c. Through trenched fan. d. Complex radial dissected fan. e. Nested fan. f. Climatically
controlled soil stratigraphy based on Talbot and Williams, 1979. This latter example is a very case but
climatically controlled development also occurs on other fan types climatic change inducing changes in
trenching and dissection. In such cases much more complex pedostratigraphies develop.
surface, and as a consequence of avulsion (periodic switching) of the trenches, the

fan surface below the intersection point will alternate between phases of erosion/
deposition and pedogenesis (Collinson, 1986). Thus, areas below the intersection
point will have both incomplete soil cover and the soils which develop will be less
mature than those in the upper fan (Figure 18.7b; Talbot and Williams, 1979). In
situations where through-fan trenching has occurred the whole fan surface can
undergo pedogenesis (Figure 18.7c). Such continuous trenching may result from
head-cut erosion and fan head entrenchment (Harvey, 1987).
488 VP. Wright
The non-entrenched case develops in young fans and as a consequence of fan

ageing, as well as climatic, tectonic and related base-level changes, fans become
entrenched and finally dissected (Christenson and Purcell, 1985). Dissected fans
have complex soil surface patterns, depending on whether they are incised radially
(Figure 18.7d) or concentrically nested (Figure 18.7e).
The distribution of paleosols, in conjunction with sedimentary features (evidence
of deep channelling and dissection) can be used to infer the geomorphology of
ancient alluvial fans. For example Wright and Alonso-Zarza (1990) were able to
characterize non-trenched fans from the Miocene of the Madrid Basin based largely
on their paleo-pedostratigraphic patterns.
Beyond the lobes a broad, finer grained apron commonly occurs, passing down-
slope into either a basin plain with or without playas, or an axial drainage system
with associated channel and floodplain deposits. In the case of the former, saline
conditions may be present, or even Vertisols. A broad catenary relationship exists
between the fan and the playa basin, with well drained soils on the former and often
less well drained, even saline hydromorphic soils on the latter, especially near the
toe of the fan where springs emerge. However catenary relationships on arid slopes
are complex and dependent on hydrological factors such as the amount of leaching
caused by run-off-derived waters and "reverse catenas" can form [compare studies
in the Negev by Wieder and Yaalon (1985) and Wieder et al. (1985)].
In summary, pedofacies relationships exist on alluvial fans and crude catenary
relationships can be recognised between the fan and the adjacent playa basin. How
ever, these pedofacies models do not explain the pedostratigraphic successions seen
in some Quaternary fans, which appear to have climatically-controlled stratigra
phies. In their study of fan sediments in Janjari (Niger), Tklbot and Williams (1979)
found that this intrinsic (threshold-controlled) model of fan sedimentation did not
explain the distribution and nature of the soil stratigraphy. Figure 18.7b shows their
predicted stratigraphy for a threshold-controlled fan with a moderate degree of en
trenchment. The most mature soils are restricted to the upper fan while the lower
fan, below the intersection point, has less well developed soils which are in more
complex lateral relationships. The Janjari fan soil stratigraphy is shown schemati
cally in Figure 18.7f. All the soils are strongly developed and cover or covered the
whole fan during each phase of stability. Although this pattern resembles that of a
fully entrenched fan, the entrenched channels in the Jinjari fan exhibited evidence
of prolonged abandonment.
The Basin and Range region of southern New Mexico is a geomorphologically
complex area which has been intensively studied for its soil-landscape relationships
(Gilè and Grossman, 1979; Gilè et al., 1981). It provides a possible analogue to the
Jinjari situation, with evidence of climate as the major control on pedostratigraphy.
Arid-type fans are well developed in the area, and four major geomorphic areas can
be defined (Figure 18.8): Rio Grande flood plain, valley borders, piedmont slopes
and basin floor. The Rio Grande flood plain is an axial drainage system which has
V A L L E Y BORDER PIEDMONT S L O P E
Fig. 18.8. Block diagram showing the major landforms in part of southern New Mexico. Note the highly
dissected terrains of the valley border area and the axial drainage along the Rio Granderift.(After Gile
et al., 1981.)
had a profound effect on the surrounding piedmont fan areas (valley borders). The
piedmont slopes and the highly dissected borders are sites of fan deposition, while
the closed basinfloodof the Jornada del Muerto Basin is the site of dune and playa
development.
The area has been volcanically and tectonically active in the recent past. Local
tectonic displacement of lower Pleistocene deposits has exceeded 90 m, and upper
Quaternary sediments have been displaced by as much as 9 m. Local geological
factors, such as basement topography, have also profoundly influenced the geo-
morphic development. However, short-term, regional climaticfluctuationhas been
the main factor controlling the deposition in the individual basin and river-valley
zones. Despite the complex geomorphology, many of the major buried and relict
soils can be correlated over the whole region, analogous to the Janjari system but
here on a much larger scale. These soil phases correspond to climatically controlled
stabilisation periods. Colder phases led to increased river discharge, entrenchment
of the major valleys, and flooding of the basin floors to form perennial lakes. The
vegetation cover became more extensive and large areas of the piedmont and valley-
border slopes were stabilized. During the glacial-interglacial transition, environ-
490 VF Wright
mental change took place, particularly increased aridity. This led to both changes in
the vegetation cover and in the rainfall pattern to an increased frequency of higher
intensity events. The landscapes became unstable and both widespread erosion and
deposition took place. This is comparable to the Janjari example where fan sur
face stability, and pedogenesis occurred during humid phases when the vegetation
cover was dense. During more arid phases aggradation or degradation occurred
dependant, it seems, on the incidence of intense rainstorms (Tklbot and Williams,
1979).
The Rio Grande Rift and Janjari fans are not the only Quaternary fans where
climatic change is the major cause of the pedostratigraphic sequence, with alter
nating phases of instability (erosion/deposition) and stability (pedogenesis). Similar
climatic controls have been suggested for fans in the south west United States (Blair
et al, 1990; Dorn et al., 1987; Lettis, 1985, Ponti, 1985). A key factor in stabil
ising these fans has been the development of a vegetation cover, in more humid
periods. In pre-Late Palaeozoic times with weakly vegetated soils, fans may have
responded quite differently to climate change and it would be instructive to look at
pedostratigraphies in such fan deposits.
Climatically controlled fan stratigraphy should be characterised by widespread
soil development. This seems to be the case in the intermontane, closed-drainage
region to the east of the Rio Grande floodplain, especially the basin floor areas.
Here the topography is smooth, with gentle slopes and relatively continuous soils
occur. In the valley border area the fans are highly dissected with terracing resulting
from the base level changes associated with the axial drainage system, possibly
related to the major glacial-interglacial cycles. However much of this complexity
is a result of the effects of lithologie and structural factors. The pedostratigraphic
model for such areas would be more complex than that shown in Figure 18.7.
Tectonic activity is an important factor in controlling the location and larger scale
cyclicity in fans (Harvey, 1989) but it is not a major control on the smaller scale
stratigraphies of fans and even tectonic movements in the upper fan may not affect
deposition on the medial and lower fan zones (Harvey, 1984).
One problem in using such fan pedostratigraphies to interpret ancient fan de
posits is that the local controls, such as the degree of entrenchment, may operate on
a shorter time scale than the climatic controls, and may themselves be ultimately a
response to extrinsic factors. Further work is needed to clarify these factors.
While calerete horizon-bearing soils (especially Aridisols) are commonly associ
ated with both present day and ancient arid fan systems, they are not the only soil
type to occur. Many arid fans are veneered by reg soils, also classifiable as Aridis
ols (usually Orthids) or Entisols (Torrents). However records of ancient forms are
very scarce. In studies of alluvial fans in the Negev, Israel, a four stage chronose-
quence has been recognised which could provide a useful tool for assessing relative
sedimentation rates in ancient sequences (Amit and Gerson, 1986; Gerson and
Amit, 1987). The most mature profiles in the Negev (older than 100,000 years) have
prominent, loess-derived matrix-rich layers in their upper parts, up to 50 cm thick.

One is left wondering if such units might not be dismissed simply as inter-channel
fines in ancient sequences.
Fine-grained sediments accumulate in the nearby basinfloorareas, typically asso
ciated with playa lakes. The paleosols in comparable ancient deposits have received
relatively little attention. Recently Rust and Nanson (1989) have suggested, by anal
ogy with similar Quaternary deposits from the Lake Eyre Basin, Australia, that
much of thefine-grainedsediment in ancient alluvial red mudstones was derived
as bed-load sand-sized pelleted mud, later compacted into structureless mudstones.
In the Lake Eyre basin much of the fluvial sand is composed of sand-sized mud
aggregates (parna) produced pedogenically in smectite-rich soils (Vertisols). Rust
and Nanson (1989) offered a similar origin for two Mesozoic and one Carboniferous
red mudstone sequences of eastern Canada.
Distal alluvial fan-playa red mudstones occur in the Permo-TViassic of Britain.
One of these units, the late Tïiassic Mercia Mudstone of southwest Britain, pos
sesses vertic paleosols which exhibit pedofacies relationships. The unit had previ
ously been interpreted as a subaqueous lacustrine deposit (Tücker, 1978). While
lacustrine deposits are locally developed, the majoi part consists of an apparently
monotonous, structureless silty red mudstone. Clear evidence of clay pellets has not
been found so a direct analogy with parna deposits is not yet possible. However,
paleosols are common in thefinergrained lithologies, and distinctive types of cyclic
gypsic Vertisols have been identified (Wright et al., 1988). Their differences may re-
PROXIMAL INTERMEDIATE
y v vi fc>S·.·'/.·!^— halite
(B)C
Fig. 18.9. Pedofacies variation across the late Triassic Mercia Mudstone "playa" basins of the Bristol
Channel Basin, southwest England (from Wright et al., 1988).
492 VP. Wrìght
fleet pedofacies variations related to deposition rate and proximity to alluvial source
areas (Figure 18.9).
In the distal settings, receiving only silt-grade sediment or finer, the paleosol
profiles are up to 2 m thick with three horizons (Figure 18.8): (a) a basal nodu
lar orange-red/green mottled horizon with decimetre-sized gypsum nodules (Ces)
overlain by nodular mudstones with well ordered dolomite (Cca); (b) a dolomitic
orange-red/green mottled mudstone with decimetre-scale pseudo-anticlines defined
byfibrousgypsum-filled veins (structural (B)C horizon); and (c) an upper orange-
red dolomitic mudstone (structural B horizon) with prominent columnar structures
a few centimetres wide and up to 60 cm long, separated and locally sharply overlain
by a red mudstone. Packages of up to six profiles have been found and in such
sequences coarser siliciclastic deposits are volumetrically minor. They have been
interpreted as Vertisols.
In the intermediate settings, the red mudstones contain thin intercalations of
coarse sheet sands. The paleosols are up to 2.5 m thick and consist of three units
(Figure 18.9; Curtis, 1982): (a) at the base is a nodular gypsum horizon (Ces hori
zon) similar to those in the distal cycles; (b) the overlying mottled dolomitic mud
stone has tapering gypsum-filled polygonal desiccation cracks, a few centimetres
wide and up to a metre deep as well as gypsum-defined pseudo-anticlines; and (c)
the cycles are capped by laminated green siltstones and clays up to 30 cm thick with
wave ripples and halite pseudomorphs, which gradationally pass up into the overly
ing nodular horizon of the next cycle. These might also be classified as Vertisols.
In proximal settings the mudstones are thinner and are intercalated with fining-
upwards sandstones and siltstones which contain current-ripples and trough cross-
stratification. The paleosols (Figure 18.9) contain a basal horizon of red siltstone, up
to a metre thick, with irregular calcite-filled nodules (up to 10 cm in diameter) rep
resenting replaced gypsum-anhydrite (Tucker, 1976). Overlying this is up to a metre
of red mudstone, often silty or sandy, containing decimetre-scale pseudo-anticlines,
and root structures. Small polygonal desiccation cracks occur at the top of the hori
zon. The cycles are frequently incomplete, pedogenesis having been terminated by
sheetflood events. Such soils constitute Entisols.
The increase in grain-size and proportion of coarser lithofacies in the three se
quences reflects increased proximity to source, in this case to alluvial fan systems
bordering Paleozoic uplands (Tücker, 1977). Vertic features such as deep desicca
tion cracks and pseudo-anticlinal slickensided slip planes form relatively rapidly in
soils; the most complete and mature profiles with columnar structures may have
formed distally to the source areas where depositional rates were lowest. Gypsum is
absent from the most proximal facies where periods of pedogenesis were short lived
and interrupted by relatively frequent sheetfloodevents. Thus the differences in soil
profiles reflect pedofacies variations in the sense of Bown and Kraus (1987). How
ever, under the arid conditions which the Mercia Mudstone was deposited, saline
Vertisols were only weakly developed even in distal settings.
Vertisols are a common soil type in playas such as those in the intermontane
basins in southern New Mexico (Gilè et al, 1981). They require phases of wetting
and drying to form the distinctive deformation features. The desiccation phases are
prolonged and are interrupted byflooding,on average once every few years, forming
ephemeral lakes.
All the evaporites seen in the unit in the Avon-Somerset area are associated
with soil profiles, although much of it was remobilised into fractures during Tertiary
uplift and rehydration. The position of the gypsum in the lower parts of the profiles
suggests it was derived from downward moving soil waters and not from sahne
groundwater, and so it is not analogous to sabkha evaporites.
The recognition of such cycles in other deposits may aid facies analysis and even
allow the assessment of proximity to coarser clastic facies.
Modelling of alluvial pedostratigraphic sequences

Simple computer models have proved very useful in interpreting cyclic facies
sequences in the geological record (Goldhammer et al., 1987; Read et al., 1986).
Multiple paleosol sequences in ancient alluvial deposits lend themselves to this ap
proach and computer simulated sequences can be compared against actual paleosol
sequences to ascertain which combinations of intrinsic and extrinsic factors operated
during their accumulation.
At this stage two distinct models are worthy of discussion: intrinsic pedofacies
model and orbital-forcing. The pedofacies sequences of Kraus (1987) are primarily
avulsion driven and the model provides a simple intrinsically controlled system
where subsidence is uniform and continuous and the paleosol stratigraphies are
controlled by the one major process. Such a model is capable of generating both
paleosol sequences that become less or more mature upwards (Figure 18.4), but
Kraus was able to detect variations in this pattern reflecting external factors.
Orbitally-forced (Milankovitch-type) cyclicity is currently becoming increasingly
widely accepted as a means of generating many different types of shallow marine
cyclic facies (Goldhammer et al., 1987; Read et al., 1986). Such cycles of sea-level
(base-level) change typically display rapid transgressive phases (base-level rise),
followed by slow regressive phases (base-level fall) during which time sediment
aggradation/progradation overtakes the falling base-level and, as a result, many
shallow marine sequences are capped by prolonged exposure surfaces.
Recently Olsen (1988) has suggested that Milankovitch 20,000 and 100,000 year
cyclicity can be detected in the late Devonian Upper Old Red Sandstone alluvial
sequences of Greenland, based on cyclic changes in the dimensions and postulated
discharges of channel sandstones, reflecting climatic change. Modelling climatic
change is extremely difficult for the relationships between such changes and sedi
mentation rates and rates of pedogenesis are far from clear and quantifiable. The
changes Olsen noted were gradational but whether a drainage system responds in
494 VP. Wrìght
this way to climatic changes is questionable. It seems more likely that as geomorphic
thresholds are crossed more abrupt changes influvialbehaviour would occur. Many
ancient alluvial sequences were deposited on low relief, coastal plains; the Siluro-
Devonian Old Red Sandstone of Europe and America being such an example.
In such situations orbitally-forced base-level changes may have occurred but such
cyclicity could not create the regular symmetrical patterns (Figure 18.4) envisaged
in the avulsion-controlled model but could generate cyclicity with very mature pale-
osols, corresponding to major base-level falls. Further comparisons must await the
appearance of computer simulations where subsidence factors will be particularly
crucial.
Summary
Paleosols not only occur in Quaternary sequences but are relatively common in
pre-Quaternary non-marine and marginal marine sequences, especially alluvial de-
posits. They exhibit highly complex stratigraphie relationships, compared to purely
depositional lithosomes, especially those developed on major cratonic unconformi-
ties.
Ancient alluvial sequences can contain hundreds of vertically stacked profiles
which can be used to evaluate the controlling factors operating during their deposi-
tion, such as subsidence, sediment supply and climate. Differences in soil type seen
in paleosol sequences can reflect catenary or pedofacies relationships, or both. Cli-
matic changes and, possibly related, base-level changes also create variations within
paleosol sequences. Pedofacies relationships seen in vertical successions result from
the step-wise avulsion of channel belts and can create relatively orderly sequences of
increasing and decreasing profile maturities. Phases of terrace development result
in poor correlation between individual paleosols or sequences of paleosols. In allu-
vial fans two models currently exist: the threshold-controlled (intrinsic) model and
the climatic model. The former generally creates marked variations within the fan
sequence in terms of paleosol continuity and maturity. The climatically-controlled
sequence is characterized by fan-wide, uniformly mature paleosols reflecting major
phases of fan stability. Quaternary fan stratigraphies in the Jinjari Fan (Niger) and
in the Rio Grande Rift zone appear to fit this second model.
Part of this paper has attempted to summarise the current knowledge of paleosol
stratigraphy (pedostratigraphy) stressing its application to the pre-Quaternary pale-
osol record, for it is here where the greatest scope lies for paleopedologists. Alluvial
paleosol sequences lend themselves to such studies because of their remarkable
abundance. However, alluvial soils have attracted less attention than other soil types
and much remains to be done to understand the dynamics of sedimentation on
floodplains. Even at this early stage of study it is clear that the pedofacies relation-
ships revealed in ancient sequences suggest that some views held by sedimentolo-
gists on the rates of floodplain processes such as avulsion require rethinking.
The late Quaternary sedimentary record is too short to allow the assessment of
long-term controls on sedimentation. Sequences of paleosols, seen as indicators of
geomorphic processes, can allow a new perspective on alluvial basin filling. Forward
modelling of alluvial basin fills using computers (Bridge and Leeder, 1979) should
attempt to generate pedostratigraphies which can then be tested against actual
sequences to evaluate, for each example, the likely specific controls operating. In
this respect paleosols will play a key role in improving our understanding of alluvial
architecture with obvious implications for hydrocarbon exploration.
References
Amit, R. and Gerson, R., 1986. The evolution of Holocene reg (gravelly) soils in deserts — an example
from the Dead Sea region. Catena, 13: 59-79.
Allen, J.R.L., 1974. Studies in fluviatile sedimentation: implications of pedogenic carbonate units, Lower
Old Red Sandstone, Anglo-Welsh outcrop. Geol. J., 9: 181-208.
Allen, J.R.L., 1986. Pedogenic calcretes in the Old Red Sandstone facies (Late Silurian-early Car
boniferous) of the Anglo-Welsh area, southern Britain. In: VP. Wright (Editor), Paleosols: Their
Recognition and Interpretation. Blackwell, Oxford, pp. 58-86.
Allen, J.R.L. and Williams, B.P.J., 1982. The architecture of an alluvial suite: the rocks between the
Townsend Tuff and Pickard Bay Tuff Beds (early Devonian), southwest Wales. Philos. Trans. R. Soc.
London, Ser. B, 297: 51-89.
Beckmann, G.G., 1983. Development of old landscapes and soils. In: Soils: an Australian viewpoint.
Division of Soils, CSIRO, Melbourne/Academic Press, London, pp. 51-72.
Behrensmeyer, AK. and Tauxe, L., 1982. Isochronous fluvial systems in Miocene deposits of northern
Pakistan. Sedimentology, 29: 331-352.
Besley, B. and Turner, P., 1983. Origin of red beds in a moist tropical climate (Etruria Formation, Upper
Carboniferous, UK). In: R.C.L. Wilson (Editor), Residual Deposits: Surface Related Weathering
Processes and Materials. Geol. Soc. London, Spec. Pubi., 11: 131-147.
Besley, B.M. and Fielding, C.R., in press. Paleosols in Westphalian coal-bearing and red bed sequences.
Central and Northern England. Palaeogeogr., Palaeoclimatol., Palaeoecol.
Bird, M.I. and Chivas, A.R., 1988. Oxygen isotope dating of the Australian regolith. Nature, 331: 513-
516.
Birkeland, P.W., 1984. Soils and Geomorphology. Oxford University Press, New York, N.Y.
Blair, T.C., Clark, J.S. and Wells, S.G., 1990. Quaternary continental stratigraphy, landscape evolution,
and application to archeology: Jarilla piedmont and Tularosa graben floor, White Sands Missile
Range, New Mexico. Geol. Soc. Am. Bull., pp. 749-759.
Boardman, J. (Editor), 1985. Soils and Quaternary Landscape Evolution. Wiley, Chichester, 391 pp.
Bos, R.H.G. and Sevink, J., 1975. Introduction of gradational and pedomorphic features in descriptions
of soils. J. Soil Sci., 26: 223-233.
Bown, T.M. and Kraus, M.J., 1987. Integration of channel and floodplain scutes in aggrading alluvial sys
tems. I. Developmental sequence and lateral relations of Lower Eocene alluvial paleosols, Willwood
Formation, Bighorn Basin, Wyoming. J. Sediment. Petrol., 57: 587-601.
Brewer, R., Crook, K.A.W and Speight, J.G., 1970. Proposal for soil-stratigraphic units in the Australian
stratigraphie code. J. Geol. Soc. Aust., 17: 103-109.
Bridge, J.S., 1984. Large scale facies sequences in alluvial overbank environments. J. Sediment. Petrol.,
54: 583-588.
496 VP. Wrìght
Bridge, J.S. and Leeder, M.R., 1979. A simulation model of alluvial stratigraphy. Sedimentology, 26:
617-644.
Butler, B.E., 1967. Soil periodicity in relation to landform development in southeastern Australia.
In: J.N. Jennings and J.A. Mabbutt (Editors), Landform Studies from Australia and New Guinea.
Australian National University Press, Canberra, pp. 231-255.
Catt, J.A., 1986. Soils and Quaternary Geology: A Handbook for Field Scientists. Clarendon Press, 267
pp.
Catt, J.A., 1988. Quaternary Geology for Scientists and Engineers. Ellis Horwood, 340 pp.
Christenson, G.E. and Purcell, C, 1985. Correlation of Quaternary fan sequences, Basin and Range
province, southwest United States. Geol. Soc. Am., Spec. Pap., 203:115-122.
Collinson, J.D., 1986. Alluvial Sediments. In: H.G. Reading (Editor), Sedimentary Environments and
Facies. Blackwell, Oxford, pp. 20-62.
Curtis, M.T., 1982. Playa cycles in the Mercia Mudstone (Keuper Marl) of Aust. Cliff Avon. Proc. Bristol
Nats. Soc., 42:13-22.
Dorn, R.I., DeNiro, M.J. and Ajie, H.O., 1987. Isotopie evidence for climatic influence on alluvial-fan
development in Death Valley, California. Geology. 15:108-110.
Duchaufour, P., 1982. Pedology. Allen and Unwin, London.
Esu, I.E., 1989, A pedological characterization of soils of the Hadejia alluvial complex in the semi-arid
region of Nigeria. Pedologie, 39:171-190.
Fairbridge, F.W. and Finkl, C.W., 1980. Cratonic erosional unconformities and peneplains. J. Geol., 88:
69-86.
Farrell, K.M., 1987. Sedimentology and fades architecture of overbank deposits of the Mississippi River,
False River Region, Louisiana. In: FG. Etheridge, R.M. Flores and M.H. Harvey (Editors), Recent
Developments in Fluvial Sedimentology, Soc Econ. Paleontol. Mineral., Spec. Pubi., 39:112-120.
Fielding, C.R., Al-Rubaii, M. and Walton, E.K., 1988. Deltaic sedimentation in an unstable tectonic
environment — the Lower Limestone Group (Lower Carboniferous) of East Fife, Scotland. Geol.
Mag., 125: 241-255.
Finkl, C.W., 1980. Stratigraphie principles and practices as related to soil mantles. Catena, 7:169-194.
Finkl, C.W., 1984. Chronology of weathered materials and soil age determination in pedostratigraphic
sequences. Chem. Geol., 44: 311-335.
Gardner, T.W., Williams, E.G. and Holbrook, P.W., 1988. Pedogenesis of some Pennsylvanian under-
clays: ground-water, topographic, and tectonic controls. In: J. Reinhardt and WR. Sigleo (Editors),
Paleosols and Weathering Through Geologic Time: Principles and Applications. Geol. Soc. Am.,
Spec. Pap., 216: 81-101.
Gerrard, A.J., 1981. Soils and Landforms: an Integration of Geomorphology and Pedology. Allen and
Unwin, London, 219 pp.
Gerrard, A.J., 1987. Alluvial Soils. Van Norstrand Reinhold Soil Science Ser., Hutchinson Ross, New
York, N.Y.
Gerson, R. and Amit, R., 1987. Rates and modes of dust accretion and deposition in an arid region
— the Negev, Israel. In: L. Frostick and I. Reid (Editors), Desert Sediments: Ancient and Modern.
Geol. Soc. London, Spec. Pubi., 35:157-169.
Gilè, L.H. and Grossman, R.B., 1979. The Desert Project Monograph. Soil Conservation Service, U.S.
Dep. of Agriculture, 984 pp.
Gilè, L.H., Hawley, J.W and Grossman, R.B., 1981. Soils and geomorphology in the Basin and Range
area of southern New Mexico — Guidebook to the Desert Project. N.M. Bur. Mines Miner. Resour.,
Mem., 39, 222 pp.
Goldbery, R., 1982. Paleosols in the Lower Jurassic Mishor and Ardon Formations ("laterite-derivative
facies"), Makhtesh Ramon, Israel. Sedimentology, 29: 669-690.
Goldhammer, R.K., Dunn, P.A. and Hardie, L.A., 1987. High frequency glacio-eustatic sea level oscil-
Païeopedology: stratigraphie relationships and empirical models 497
lation with Milankovitch characteristics recorded in Middle Triassic platform carbonates in northern
Italy. Am. J. Sci., 287: 853-892.
Hall, G.F., 1983. Pedology and geomorphology. In: L.P. Wilding, N.E. Smeck and G.F. Hall (Editors),
Pedogenesis and Soil Taxonomy. 1. Concepts and Interactions. Elsevier, Amsterdam, pp. 117-140.
Harvey, A.M., 1984. Aggradation and dissection sequences on Spanish alluvial fans: influence on mor
phological development. Catena, 11, 289-304.
Harvey, A.M., 1987. Alluvial fan dissection: relationships between morphology and sedimentation. Geol.
Soc. London, Spec. Pubi., 35: 87-103.
Harvey, A.M., 1989. The occurrence and role of arid zone alluvial fans. In: D.S.G. Thomas (Editor),
Arid Zone Geomorphology. Belhaven Press, London, pp. 136-158.
Hayward, M., 1985. Soil development in Flandrian floodplains: River Severn case-study. In: J. Boardman
(Editor), Soils and Quaternary Landscape Evolution. Wiley, Chichester, pp. 281-299.
Idnurm, M. and Senior, B.R., 1978. Paleomagnetic ages of Late Cretaceous and Tertiary weathered
profiles in the Eromanga Basin, Queensland. Palaeogeogr., Palaeoclimatol., Palaeoecol., 24: 263-
277.
Johnson, D.L. and Watson-Stegner, D., 1987. Evolution model of pedogenesis. Soil Sci., 143: 349-366.
Kraus, M.J., 1987. Integration of channel and floodplain suites in aggrading alluvial systems, II. Vertical
relations of Lower Eocene paleosols, Willwood Formation, Bighorn Basin, Wyoming. J. Sediment.
Petrol., 57: 602-612.
Kraus, M.J. and Bown, T.M., 1986. Paleosols and time resolution in alluvial stratigraphy. In: V.P. Wright
(Editor), Paleosols: Their Recognition and Interpretation. Blackwell, Oxford, pp. 180-207.
Kraus, M.J. and Bown, T.M., 1988. Pedofacies analysis: a new approach to reconstructing ancient fluvial
sequences. Geol. Soc. Am., Spec. Pap., 216:143-152.
Kraus, M.J. and Middleton, L.T., 1987. Dissected paleotopography and base-level changes in a Thassic
fluvial sequence. Geology 15: 18-21.
Leeder, M.R., 1975. Pedogenic carbonates and flood plain accretion rates: a quantitative model for
alluvial arid-zone lithofacies. Geol. Mag. 112: 257-270.
Lehman, T.M., 1989. Upper Cretaceous (Maastrichtian) paleosols in Trans Pecos Texas. Geol. Soc. Am.
Bull., 101: 188-203.
Lettis, W.R., 1985. Late Cenozoic stratigraphy and structure of the west margin of the central San
Joaquim Valley, California. Geol. Soc. Am., Spec. Pap., 203: 97-114.
McCraw, J.D., 1968. The soil pattern of some New Zealand alluvial fans. 9th Int. Congr. Soil Sci. Trans.,
4: 631-640.
McFarlane, M.J., 1983. Latérites. In: A.S. Goudie and K. Pye (Editors), Chemical Sediments and
McPherson, J.G., 1979. Calerete (caliche) paleosols in fluvial redbeds of the Aztec Siltstone (Upper
Devonian), southern Victoria Land, Antarctica. Sedimentol. Geol., 22: 267-285.
Machette, M.N., 1978. Dating Quaternary faults in the southwestern United States by using buried
calcic paleosols. U.S. Geol. Surv., J. Res., 6: 369-381.
Machette, M.N., 1985. Calcic soils of the southwestern United States. In: D.L. Wiede (Editor), Soils
and Quaternary Geology of the Southwestern United States. Geol. Soc. Am., Spec. Pap., 203:1-21.
Macphail, R.I., 1986. Paleosols in archaeology: their role in understanding Flandrian pedogenesis. In:
V.P. Wright (Editor), Paleosols: their Recognition and Interpretation. Blackwell, Oxford, pp. 263-
290.
Mahaney, W.C. (Editor), 1978. Quaternary Soils. Geoabstracts, Norwich, 250 pp.
Morrison, R.B., 1978. Quaternary soil stratigraphy — concepts, methods and problems. In: W.C. Ma
haney (Editor), Quaternary Soils. Geoabstracts, Norwich, pp. 77-108.
Nanson, G.C., 1980. Point bar and floodplain formation of the meandering Beatton River, north east
British Columbia, Canada. Sedimentology, 27: 3-29.
498 VP. Wrìght
Nanson, G.C., Rust, B.R. and Taylor, G., 1986. Coexistent mud braids and anastomosing channels in an
arid-zone riven Cooper Creek, central Australia. Geology, 14:175-178.
Nicholas, J. and Bildgen, P., 1979. Relations between the location of karst bauxites in the North Hemi
sphere, the global tectonics and the climatic variations during geological time. Palaeogeogr., Palaeo-
climatol., Palaeoecol., 29: 205-239.
North American Commission on Stratigraphie Nomenclature, 1983. North American stratigraphie code.
Bull. Am. Assoc. Pet. Geol., 67: 841-875.
Oilier, CD., Chan, R.A., Craig, M.A. and Gibson, D.L., 1988. Aspects of landscape history and regolith
in the Kalgoorlie region, Western Australia. BMR J. Aust. Geol. Geophys., 10: 309-321.
Olsen, H., 1988. Autoprocesses and climatic control on a meandering river — Devonian, East Green
land. Abstracts, Brit. Sediment. Res. Group Annu. Meet., Cambridge 1988, British Antarctic Surv.,
Cambridge.
Ortlam, D., 1970. Die Randfazies des germanishchen Buntsandstein im südlichen Schwarzwald. Geol.
Jahrb. 89: 135-168.
Ortlam, D., 1971. Paleosols and their significance in stratigraphy and applied geology in the Permian
and Triassic of southern Germany. In: D.H. Yaalon (Editor), Paleopedology. International Society
of Soil Science and Israel Universities Press, Jerusalem, pp. 321-327.
Ortlam, D., 1974. Inhalt und Bedeutung fossilen Bodenkomplexe im Perm und Trias von Mitteleuropa.
Geol. Rundsch., 63: 850-884.
Parsons, R.B., 1981. Proposed soil-stratigraphie guide. Paleopedol. Newslett. for 1981, pp. 6-13.
Pawluk, S., 1978. The pedogenic profile in the stratigraphie section. In: W.C. Mahaney (Editor), Qua
ternary Soils. Geoabsracts, Norwich, pp. 61-75.
Percival, C.J., 1986. Paleosols containing an albic horizon: examples from the Upper Carboniferous
of northern England. In: VP. Wright (Editor), Paleosols: Their Recognition and Interpretation.
Blackwell, Oxford, pp. 87-111.
Ponti, DJ., 1985. The Quaternary alluvial sequence of the Antelope Valley, California. Geol. Soc. Am.,
Spec. Pap., 203: 79-96.
Read, J.F., Grotzinger, J.P., Bova, J.A. and Koerschner, W.F., 1986. Models for the generation of car
bonate cycles. Geology, 14: 107-110.
Retallack, G.J., 1986a. The fossil record of soils. In: VP. Wright (Editor), Paleosols, their Recognition
and Interpretation. Blackwell, Oxford, pp. 1-57.
Retallack, G.J., 1986b. Fossil soils as grounds for interpreting long-term controls on ancient rivers. J.
Sediment. Petrol., 56: 1-18.
Retallack, G.J., 1988. Field recognition of paleosols. Geol. Soc. Am., Spec. Pap., 216,1-20.
Rose, J., Boardman, J., Kemp, R.A and Whiteman, C.A., 1985. Paleosols and their interpretation of
British Quaternary History. In: K.S. Richards, R.R. Arnett and S. Ellis (Editors), Geomorphology
and Soils. Allen and Unwin, London, pp. 348-375.
Ruhe, R.V., 1956. Geomorphic surfaces and the nature of soils. Soil Sci., 82: 441-455.
Rust, B.R. and Nanson, G.C., 1989. Bedload transport of mud as pedogenic aggregates in modern and
ancient rivers. Sedimentology, 36: 291-306.
Smith, R.M.H., 1990. Alluvial paleosols and pedofacies sequences in the Permian Lower Beaufort of
the southwestern Karoo Basin, South Africa. J. Sediment. Petrol., 60: 258-276.
Staub, J.R. and Cohen, A.D., 1978. Kaolinite-enrichment beneath coals: a modern analogue, Snuggedy
Swamp, South Carolina. J. Sediment. Petrol., 48, 203-210.
Talbot, M.R. and Williams, M.A.J., 1979. Cyclic alluvial fan sedimentation and the flanks of fixed dunes,
Janjari, central Niger. Catena, 6: 43-62.
Tlicker, M.E., 1976. Quartz replaced anhydrite nodules (Bristol Diamonds) form the Triassic of the
Bristol District. Geol. Mag., 113: 569-574.
Tucker, M.E., 1977. The marginal Triassic deposits of South Wales: continental facies and palaeogeog-
raphy. Geol. J., 12: 169-188.
Tucker, M.E., 1978. Triassic lacustrine sediments from South Wales: shore zone elastics, evaporites and
carbonates. In: A. Matter and M.E. Tucker (Editors), Modern and Ancient Lake Sediments. Int.
Assoc. Sedimentol. Spec. Pubi., 2: 205-222.
Valentine, K.W.G. and Dalrymple, J.B., 1975. The identification, lateral variation and chronology of two
buried paleocatena s at Woodhall Spa and West Runton, England. Quat. Res., 4: 551-590.
Valentine, K.W.G. and Dalrymple, J.B., 1976. Quaternary buried palaeosols: a critical review. Quat.
Res., 6: 209-222.
Walker, P.H., Beckmann, G.G. and Brewer, R., 1984. Definition and use of the term "pedoderm". J. Soil
Sci., 35: 505-510.
Wieder, M. and Yaalon, D.H., 1985. Catenary soil differentiation on opposite-facing slopes related to
erosion-deposition and restricted leaching processes, northern Negev, Israel. J. Arid Environ., 9:
119-136.
Wieder, M., Yair, A. and Arzi, A., 1985. Catenary soil relationships on arid hillslopes. Catena Suppl., 6:
41-57.
Working Group on the Origin and Nature of Paleosols, 1971. Criteria for the recognition and classi
fication of paleosols. In: D.H. Yaalon (Editor), Paleopedology. Int. Soc. Soil Sci., Israel University
Press, Jerusalem, pp. 153-158.
Wright, VP, 1990. Early diagenesis in terrestrial settings: pedogenesis and paleosol recognition. In: G.V.
Chilingarian and K.H. Wolf (Editors), Diagenesis III. Elsevier, Amsterdam.
Wright, VP. and Alonso-Zarza, A,, 1990. Pedostratigraphic models for alluvial fan deposits: a tool for
interpreting ancient sequences. J. Geol. Soc. Lond., 147: 8-10.
Wright, VP, North, C.P., Hancock, PL., Curtis, M. and Robinson, D., 1988. Pedofacies variations across
an arid alluvial basin: a case study from the Upper Triassic of SW Britain. Int. Assoc. Sedimentol.,
Eur. Meet., Leuven, pp. 227-228.
Wright, VP. and Robinson, D., 1988. Early Carboniferousfloodplaindeposits from South Wales: a case
study of the controls on palaeosol development. J. Geol. Soc. London, 145: 847-857.
Yaalon, D.H., (Editor), 1971. Paleopedology. Int. Soc. Soil Sci., Israel University Press, Jerusalem, 350
pp.
501
Chapter 19
Underclays and related paleosols associated with

coals
R.E. HUGHES, PJ. DEMARIS and W.A. WHITE
Introduction
This chapter reviews the current state of knowledge of underclays and similar pa-
leosols associated with coals, primarily in the Illinois Basin in North America. Our
discussion focuses on paleosols in Pennsylvanian-age deposits because the largest
amount of information exists for these deposits. The term underclay is used for a
massive, argillaceous unit that underlies most coals of Carboniferous age through-
out the world. The absence of distinguishing features in underclays at the hand-
specimen and microscopic level is the basis of much of the controversy about their
origin. Proposed origins include sedimentary deposition, soil formation and prod-
ucts of diagenesis.
Underclays are laterally continuous over wide geographical areas (Logan, 1842).
In the Illinois Basin, they usually contain one of four distinctive mineralogical suites
— shale-type, gley-type, fireclay-type and sandstone-type (Hughes et al., 1987).
Hughes et al. (1987) showed that the distinctive features of the mineralogical suites
result from preservation of source material, alteration of source material in a gley-
ing environment, alteration of source material in a Si02-extracting soil environ-
ment, and alteration of feldspathic source material by post-depositional influx of
groundwater in sandy sediments.
In this review we emphasize approaches that may resolve differences in existing
points of view and summarize the distinctive mineralogical features that allow a
choice between various theories of underclay origin. For the purposes of this report,
the term soil or paleosol includes all materials altered by pedogenic processes,
including those beneath standing water.
Underclays are of special importance because they often contain the best ev-
idence of soil formation in the geologic record; they reveal patterns of mineral
alteration associated with primitive plants, such as K+ extraction, desilicification
and pH depression, contain a wide range of clay and nonclay minerals, reveal some
effects of water logging of peat-associated soils that are commonly overlooked, pre-
502 R.E. Hughes, RI Demons and WA. White
serve detailed evidence of geologic and pedologie history, may reveal information
about the paleolandscape, aid in the search for and exploitation of coal, oil and
gas, clay and other resources, and provide information important for the design of
safe mines and wells, and the protection of the environment. Because underclays
are widespread (Logan, 1842) and because it is reasonable to assume that peat-
forming vegetation spread rapidly on the underclay, the top of the underclay may
approximate a time Une.
Definition of underclay and related units

Background information on underclays and associated cyclothems is found in
Patterson and Hosterman (1960), Parham (1964), Odom and Parham (1968), Wil
liams et al. (1968), Hughes et al. (1987) and Gardner et al. (1988). For this re
port, underclays are defined as the massive, nonbedded, nondescript, argillaceous
units that typically underlie coals. Underclays usually contain cracks with polished
surfaces called slickensides and often contain fossil roots. Like Logan (1842) and
Gardner et al. (1988) we accept the definition of underclay to include clays in some
places where the overlying coal is missing, but only when the underclay is near the
stratigraphie position normally occupied by underclay and coal (Figure 19.1).
Underclays are also distinguished for what they are not — typical shales and
claystones. Conventional particle-size analyses of underclays have no meaning be
cause they reflect the degree of mechanical breakdown of aggregates during sample
preparation, rather than the particle size of the original sediments. The original sed
iment was made up of clayflocculesand individual grains in unknown proportions.
Underclays can be primarily unaltered deposits of sediment or extensively altered
paleosols. If another rock type is present below the coal, we designate this a seat
rock.
Underclays extend laterally into areas where the coal seam is missing. Logan
(1842) recognized the lateral continuity of underclays when he stated: "Immediately
below every regular seam of coal, in South Wales, (and nearly 100 are known to
exist) is constantly found a bed of clay, varying in thickness from six inches to
more than ten feet, and called the underclay, undercliff, understone, pouncer, or
bottom stone. It is so well known to the collier, that he considers it an essential
accompaniment of the coal; and only where it ceases, does he give up his expectation
offindingcoal. Seams which have thinned out in one portion of a work, have been
recovered in another by following this bed."
The top of an underclay is normally the contact with the overlying coal. The
bottom of an underclay is less well defined. In general, underclay extends downward
to a major lithologie change or until the massive, slickensided character is lost.
Claystones and shales up to several meters thick sometimes underlie underclays. In
these cases, the bottom is chosen to correspond to the level at which characteristics
of underclay are no longer observed and the strata clearly are normal claystone
Underclays and related paleosols associated with coals 503
or shale. Hughes et al. (1987) showed that underclays are developed on top of
one another in areas of slow deposition. The Cheltenham Member of the Spoon
Formation of Illinois and Missouri is thought to represent this type of multiple
underclay formation (Hughes et al., 1987).
Other useful definitions:
Claystone is used in a general sense for massive clay-rich units that lack the
laminations of shale and arefinerthan siltstone.
Claystone partings in coals are separable layers, often a few centimeters thick,
which contain noticeably less organic material than the associated coal.
Flint clays are deposits of intergrown crystallites of well-crystallized kaolinite that
are harder than claystone and whose crystallites will not disperse in water. Flint
clays often contain boehmite and diaspore in the most refractory zone of the deposit
and illite, mixed-layered illite/smectite (I/S) and quartz in less refractory zones
transitional with associated strata.
Tonsteins are thin partings, often in coals, that usually contain well-crystallized
and tightly intergrown kaolinite with a hardness equivalent to flint clays. The term
tonstein can be applied to distinctive layers in strata above and below coals when
the characteristic mineral content and hardness are present.
Hypotheses of origin
Hypotheses proposed to explain the origin of underclays suggest that underclays

could be:
(a) soils of the plants that formed coals (Logan, 1842);
(b) soils of plants that grew mostly before coal formation (Stout, 1923; Hughes et
al., 1987);
(c) residues of plant-derived inorganic material that builds up from the decay of
plant matter (Stout, 1923);
(d) primarily waterlaid deposits reflecting source area (Grim and Allen, 1938) or
differentially sorted into clay mineral "facies"(Parham, 1964);
(e) eolian deposits, possibly volcanic ash in some cases (Burger and Damberger,
1985);
(f) acid alteration zones due to downward migration of fluids from the coal
swamp or peat (Julien, 1880, Staub and Cohen, 1978; Rimmer and Eberl, 1982), or
from postdepositional influx of groundwater; or
(g) products of diagenesis, at least in part (Parham, 1964; Robinson and Wright,
1987).
Cyclothem association
Pennsylvanian-age underclays are unique in being part of a series of units called a

cyclothem (Figure 19.1) by Wanless and Weiler (1932). Although all the units of the
504 R.E. Hughes, RI Demaris and WA. White
Shale, gray, sandy at top; contains marine

fossils and ironstone concretions, especially
- I in lower part.
Limestone; contains marine fossils.
Shale, black, hard, fissile, "slaty"; contains

large black spheroidal concretions and
marine fossils.
Limestone; contains marine fossils.
Shale, gray; pyritic nodules and ironstone

concretions common at base; plant fossils
locally common at base; marine fossils rare.
Coal; locally contains clay or shale partings.

Underclay, mostly medium to light gray ex
cept dark gray at top; upper part noncalcare-
ous. lower part calcareous.
Limestone, argillaceous; occurs in nodules

or discontinuous beds; usually nonfossilifer-
ous.
Shale, gray, sandy.
Sandstone, fine-grained, micaceous, and

siltstone, argillaceous: variable from massive
to thin-bedded; usually with an uneven lower
surface.
Fig. 19.1. Schematic representation of an ideal cyclothem showing vertical variation from sandy, fresh
water sediments at base through more marine units above coal (after Willman and Payne, 1942).
ideal cyclothem are rarely present at any one site, the progression from fresh water
strata at the base of the cyclothem to marine strata above the coal is commonly ob
served in the Illinois Basin. Cyclothems of the Pennsylvanian system in North Amer
ica range from mostly continental freshwater deposits in Pennsylvania and West
Virginia on the east, to mixed continental/marine sequences in the Illinois Basin, to
marine-dominated sequences in Iowa and Kansas to the west. The cyclothems in the
Illinois Basin also range from sandier and more continental kaolinite-rich deposits
at the base of the section tofinerelastics and limestones indicating more marine
conditions at the top. These are only general trends, however, and kaolinite-rich
deposits and relatively coarse sands do occur in the upper part of the section. There
are more than 50 named cyclothems in the Illinois Basin.
In the eastern and central United States, these strata also contain flint clays
and tonsteins (Burger and Damberger, 1985). It has been suggested that an under-
clay and coal sequence may represent a pedological unit. In general, the standard
cyclothem might be studied from the perspective that the limestone below the un-
derclay, the underclay, and the first 10 or so centimeters of coal constitute a single
soil profile. However, distinct differences between palynofloras from underclays and
from coal immediately above have been noted previously (Peppers, 1964). Recent
work (Willard et al., in press) suggests that the underclay palynofloras are quite
distinct and that these pre-coal swamp floras may have been much more responsible
for the mineralogical alteration of the underclay than were the plants of the coal
swamp.
Methods of investigation
Part of our purpose in writing this chapter was to point out uncertainties and
gaps in our current knowledge and differences in interpretation of the clastic units
associated with coals and to suggest approaches that might resolve those issues.
In past studies, researchers have often made use of the best methods and instru-
ments available. However, many investigators have failed to consider the various
conditions and processes that might produce underclays and have provided data and
illustrations that are incomplete, are summarized to show support for or evidence
against a particular view, or lack raw data. For example, investigators who exclude
illustrations of X-ray diffraction spectra make it impossible for others to compare
results and test new ideas. For these reasons, we include a review of methods and
assess their relative value.
Field measurements, sampling, instrumental analyses and selection of multiple
working hypotheses (Chamberlin, 1890) should be carried out in as much de-
tail as possible. Detailed sampling should be done, and samples should include
unaltered material from above and below the underclay, because partings, clay-
stones and shales may preserve most of the original detrital mineral suite. Field
description, radiography, optical and electron microscopy [especially scanning elec-
tron microscopy/energy dispersive X-ray analysis/(SEM/EDX), Figure 19.2], X-ray
diffraction (XRD), paleontological description and chemical analysis can provide
critical data for an underclay study. Removal of organic material by low-temperature
oxidation, dating by isotopes, measurements of pH and Eh, studies of stable iso-
topes, and trace-element analyses can often add valuable information about certain
samples. It should be noted that samples with pyrite and marcasite oxidize rapidly
on exposure to air. Measurements of pH, Eh and mineralogical and chemical con-
tent can be significantly affected by the breakdown products of pyrite and marcasite.
::;
^%. fi
ΙΙβ I * i
Fig. 19.2. Scanning electron micrograph of the authigenic minerals in the centre of an ironstone con
cretion from the Francis Creek Formation (a roof shale) near Mazon Creek, Illinois. Large crystals in
centre are pyrite; well-crystallized kaolinite is present to the right and upper right, and siderite and
detrital clay minerals make up the background.
Essential parts of an underclay study are sampling of the interval from unaltered
claystone below the underclay, detailed sampling of the underclay, collection of clay
partings within overlying coal and selection of representative samples of unaltered
units above the coal. XRD, SEM/EDX, paleobotanical and often chemical analyses
are required for adequate characterization of the samples.
Current knowledge and interpretations
Pennsylvanian underclays in Illinois
Four types of underclays are recognized in Illinois by their mineralogical compo

sition.
(a) Shale-type underclay (Figure 19.3a) appears to be mostly a detrital deposit of
unaltered source material.
(b) Gley-type underclay (Figure 19.3b) represents pedogenic alterations under
relatively saturated conditions and with particular types of plants.
(c) Fireclay-type underclay (Figure 19.3c), previously referred to as "soil type"
(Hughes et al, 1987), contains mixed-layered kaolinite/smectite (K/S) and/or poorly
crystallized kaolinite. This type appears to be the result of intense conditions of
weathering by silica-extracting plants, perhaps in some cases with downward water
flow.
(d) Sandstone-type underclay (Figure 19.3d) is the result of pedogenic weather
ing or postburial infiltration of fresh water that has altered abundant feldspar in
sandy sediments to authigenic kaolinite (Hughes et al., 1987).
The mineral suites of each type are distinctive, and the mineral fractions of coals,
partings within coals, and of some anomalous claystones within the normally marine
part of the cyclothem may be described as comparable to one of the four basic types.
Choosing a boundary between types is arbitrary. However, the gley, fireclay or
sandstone designation is generally applied as soon as it is obvious which process of
mineral alteration is taking place: gley-type K+ extraction,fireclay-typealteration of
clays to K/S and/or poorly crystallized kaolinite, or formation of authigenic kaolinite
in sandy units. The XRD traces in Figure 19.3 illustrate the mineral suites of the
four underclay types. A trace for fireclay underclay rich in both K/S (Figure 19.3c,
upper) and poorly crystallized kaolinite (Figure 19.3c, lower) illustrates the range of
composition observed for fireclay-type underclays. The diagrams in Figure 19.4 are
representations of the vertical zonation of these four mineral compositions at four
locations within the Illinois Basin.
Our reasoning about the origin of each of the underclay types is as follows.
(a) The shale-type underclay, with its illite-chlorite-rich composition, represents
at least one type of detrital source material, because it consists mostly of the meta-
morphic polymorph of illite 2Μι· The depth of burial of Pennsylvanian strata was
insufficient to produce 2Mi illite diagenetically (Damberger, 1971); therefore 2Mi
illite must be derived from metamorphic source areas. Both pre-existing shales and
igneous and metamorphic materials in the source area should have produced an
illite-chlorite-rich suite of sediments of the type observed.
(b) The clay minerals in the gley-type underclay could be the products of smec-
tite-to-illite diagenesis, but, if the source material was deposited in place as volcanic
ash, kaolinite, rather than smectite, would be the expected product of weather-
508 R.E. Hughes, RJ. Demaris and WA. White
FIRECLAY TYPE
K
SANDSTONE TYPE
<2^m,
ethylene glycol
¹ ^
i 1— ~~i 1 r- "1 1 i r -1 1
30 25 20 15 10 5 2 20 15
Degrees.20 Degrees 2Θ
Underclays and related paleosols associated with coals
20 40 60 80 20 40 60 80 100
Clay mineral content (%) Clay mineral content (%)
I + I in l/S
1.4
20 40 60 80 20 40 60 80 100
Clay mineral content (%) Clay mineral content (%)
Fig. 19.4 (above). Clay mineralogical profiles for underclay types (after Hughes et al., 1987). Each
clay mineral component percentage is represented by the width of itsfieldat each depth, a. Shale-type
underclay (Sh) below weakly developed gley-type underclay (Gl) zone. b. Gley-type underclay (modified
from Rimmer and Eberl, 1982) with shale-type zone at base; Ca and C mark their highest occurrence.
c. Fireclay-type underclay (Fc) zone above shale-type zone. d. Two zones offireclay-typeunderclay
separated by a shale-type zone. (Ca = calcite; other symbols as in Figure 19.3.)
Fig. 19.3 (left). X-ray diffraction (XRD) patterns of four types of underclays (after Hughes et al., 1987).
a. Relatively unaltered source material of shale-type underclay. b. I/S and authigenic kaolinite-rich suite
in gley-type underclay. c. Two XRD traces of the well and poorly crystallized kaolinite and K/S in
fireclay-type underclay. d. Authigenic kaolinite in sandstone-type seat rock. (C - chlorite; I = illite; K
= kaolinite; K/S = mixed-layered kaolinite/smectite; 7/5" = mixed-layered illite/smectite; V = vermi-
culite; G = gypsum; Q = quartz; A = anatase; K-P = K-feldspar, P = plagioclase.)
510 R.E. Hughes, PJ. Demons and WA. White
ing (Burger and Damberger, 1985). Furthermore, no detrital source exists for a
smectite-rich sediment. The chlorite content of gley-type underclays is similar to
that of the shale type, supporting the idea that most of the gley-type weathering
involves illite alteration to I/S. The euhedral, well-crystallized kaolinite in these un
derclays is diagenetic and would not be expected to be present under conditions in
which smectite was altered to illite by diagenesis as in the Gulf Coast model of Perry
and Hower (1970). The removal of K+ from illite and orthoclase (by growing plants)
explains all the characteristics of the gley-type underclays.
(c) Thefireclay-typeunderclays could result from differential sorting or floccu-
lation and deposition in a near-shore environment (Parham, 1964). However, the
associated clay minerals are as fine as other underclay types, and, if depositional,
fireclay types should also contain chlorite and feldspar. Intense in situ weathering by
K+- and SiC>2-extracting plants completely explains the mineral suite.
(d) Sandy seat rocks contain abundant kaolinite crystals that are euhedral and
well-crystallized (Glass et al., 1956): evidence of in situ formation. The loss of
feldspar and perhaps micas from these deposits as well as preserved clay mineral
suites supports an origin based on postdepositional fluid influx.
Shale-type underclays have a composition similar to that of most Paleozoic shales
in the United States, especially in the Appalachian region, the principal source area
for Pennsylvanian sediments in the Illinois Basin. The clay minerals are dominated
by illite, and chlorite is present in significant amounts usually along with a small
amount of mixed-layered illite/smectite and kaolinite (Table 19.1). Most of the illite
has a 2Mi structure indicative of unaltered source material and little diagenetic
illite. Near the contact with coal and immediately above the coal in the roof shale,
chlorite is often missing and somewhat larger amounts of I/S may be present. The
shale-type composition is transitional with the gley,fireclayand sandstone types as
weathering intensity increases or the units become sandier. The common nonclay
minerals in a shale-type underclay include quartz, which is ubiquitous, plagioclase
and often orthoclase feldspars, a small amount of pyrite or marcasite, and calcite.
Calcite is usually absent near the coal.
Gley-type underclays contain abundant, well-crystallized I/S, common authigenic
kaolinite, and illite and chlorite (Table 19.1). These underclays are unique in that
they generally contain only plagioclase feldspar, and they appear on the whole to
represent source material (shale type) that has been altered by K+ extraction from
illite and orthoclase to produce I/S and kaolinite (Figures 19.3b and 19.4b). The
persistence of chlorite (normally thefirstmineral to be lost as pH is lowered during
weathering) suggests that little excess H+ was produced by the growth or decay of
plants. Water-saturated conditions are suggested by more abundant calcite, ferroan
dolomite, siderite, pyrite, marcasite and organic matter. In places where alteration
lasted longer or perhaps a different group of plants caused the alteration, a smectite
like I/S andflintclay resulted (Hughes and White, 1969). Some silty underclays have
this composition, which suggests that soil-forming environment and plant type con-
TABLE 19.1
Typical compositions of end member clay mineral suites in Pennsylvanian strata of Illinois
Suite name Minerals3

I Kb C I/S,V Q KF P Cac Py, Ma
(%) (%) (%) (%)
Shale-type 70 10 10 10 cd m-c m-c m m
Gley-type 15 15 5 65 c - m-c m-c m-c
Fireclay-type 10 80 0 10 c - - r r-m
Sandstone-type 10 75 5 10 a r r r? r-m
a
I = illite; K = kaolinite; C = chlorite; I/S, V = illite/smectite and vermiculite; Q = quartz; KF =
potassium feldspar, P = plagioclase feldspar; Ca = calcite, ferroan dolomite and siderite; Py, Ma =
pyrite and marcasite.
b
Kaolinite in fireclay-type suite is partly K/S or poorly crystallized; kaolinite in shale-, gley- and
sandstone-type suites is well crystallized and euhedral.
c
Occurrence in clastic units (excluding limestone and dolomite).
d
c = common; m = minor; r = rare; a = abundant; - = absent.
Source: Hughes et al. (1987).
trolled the mineral content. Gley-type underclay profiles (Figure 19.4b) tend to
parallel the clay mineral compositional patterns observed for smectite-to-illite di-
agenesis (Reynolds, 1980). This parallelism is present as a progressive downward
increase in content of discrete illite and in the illite content of I/S, and in increased
structural order of the I/S (RO, RI, R3). It appears that the sequence of I/S min
erals produced by weathering is nearly the exact reverse of the sequence produced
by diagenesis. This similarity between diagenetic and weathering products suggests
underlying mechanisms in common and it suggests that insights in one area may
improve understanding in the other.
Fireclay-type underclays represent an alteration pathway different from that which
produced gley types. Poorly crystallized kaolinite and K/S are usually associated with
poorly crystallized I/S, well-crystallized kaolinite, and variable amounts of unaltered
illite. Chlorite is absent (Table 19.1). The absence of feldspars and carbonates and
the nature of the clay mineral suite suggest that intense alteration over a fairly long
period may have been required to form these paleosols (Figure 19.4c). The amount
of well-crystallized kaolinite and the close correspondence between loss of illite and
chlorite and gain of K/S and poorly crystallized kaolinite suggest that poorly crys
tallized kaolinite and K/S form from illite and chlorite, whereas well-crystallized
kaolinite is derived from feldspars (Figure 19.3c). The elevated background in fire
clay underclays between 7 and 10 Â on the XRD trace of a sample (after heating
to 325°C) is due to K/S derived from several varieties of 2:1 clays rather than only
from smectite. The close vertical proximity of chlorite and K/S — minerals with
distinctly different fields of geochemical stability (Figure 19.4c and d) — indicates
that silica-extracting plants were the dominant alteration factor in the formation of
512 R.E. Hughes, El Demaris and WA. White
K/S; S1O2 removal by downward water flow would have obliterated the illite- and
chlorite-rich zones. Close proximity of K/S and chlorite also implies that relatively
small changes in rate of deposition, plant growth, or both had major effects on the
mineral suite. If thick enough, the best examples of this type of underclay have
commercial value asfireclays.Although these clays are not normally associated with
flint clays or authigenic kaolinite in the Illinois Basin, Gardner et al. (1988) indicate
that flint clays in the eastern United States are associated with this type of under
clay. Possible explanations for the origin of these flint clays will be discussed in the
section aboutflintclays and tonsteins.
Sandstone-type suites seem to be normal sandy sediments that have been affected
by movement of fresh water during pedogenesis or after burial (Figure 19.3d).
Because feldspars alter relatively rapidly during infiltration of fresh water under
pedogenic or postburial conditions, the clay mineral fraction, minus newly formed,
authigenic kaolinite, may have a shale-type, gley-type orfireclay-typesuite similar
to one of the other underclay types. It is also possible that any of the four underclay
types can be altered by relatively recent (Tertiary to present) postdepositional fluid
intrusion, particularly fluids rich in sulfuric acid derived from the oxidation of as
sociated, iron sulfide-rich strata. This recent alteration has been observed to result
only in the formation of vermiculite, gypsum and jarosite, and therefore is unlikely
to be confused with a true underclay alteration process.
Seat rocks
Other types of strata can underlie a coal. When a shale, sandstone or limestone
underlies a coal, the unit is designated a seat rock. If the floor is shale or limestone,
the clay mineral suite should be shale-type. Relatively unaltered claystones might be
classified as seat rocks, but our normal practice is to designate them as shale-type
underclays, particularly those with slickensides and minor alteration zones near the
coal.
Underclay profiles
We have studied underclay profiles only for the last 10 to 15 years. In most of the
earlier studies at the Illinois State Geological Survey (as in clay studies elsewhere)
underclay was considered as a single unit or separated into only a few benches.
However, the results of Rimmer and Eberl (1982), Gardner et al. (1988) and our
own investigations indicated that many underclays may not show uniform and pro
gressive mineralogical zonation in profile. Patterson and Hosterman (1960) thought
that underclays lack profiles because they were water saturated much of the time.
The lack of consistent mineralogical profile development in many underclays is most
likely the result of a continuous series of depositional and soil-forming events lead
ing to a single underclay. In this sense, an individual underclay represents multiple
soils rather than a single pedogenic event. Wnuk and Pfefferkorn (1987) noted that
underclays represent a series of alternating sediment floods and paleosol forma
tions during which organic matter from previous soil and sedimentary structures
that had formed during flooding were destroyed by the rooting action of plants.
The tendency of composite samples (single units or benches) of earlier investiga
tors to contain mineral suites (Parham, 1964; Odom and Parham, 1968) similar to
individual samples from profiles (Rimmer and Eberl, 1982; Hughes et al., 1987)
further supports the hypothesis that most underclays lack mineralogical profiles and
represent multiple units of soil formation with an average degree of mineral alter
ation characteristic of local paleoenvironmental conditions. The profiles in Figure
19.4b-d, which indicate the presence of zones of maximum alteration other than the
zone immediately below the coal, also support a concept of underclay development
involving multiple depositional and pedogenic events.
Underclays of other places and ages
Parham (1964) mapped underclay types for the Colchester Coal from western In
diana and Kentucky, through the Illinois Basin, through Missouri, to parts of eastern
Kansas and Oklahoma. Parham (1964) also mapped type distributions for the under
clay of the Middle Kittanning Coal of Ohio and Pennsylvania. Parham's underclay
types are directly convertible to the four types we described earlier (Parham's types
A-G = fireclay, types H-L = shale and types M-Q = gley), and these classifica
tions have been applied to underclays in most of North America. Unfortunately, the
underclays of both coals studied by Parham outside the Illinois Basin include only
samples in transition betweenfireclay-typeand shale-type underclays. Furthermore,
compositional information from associated coals and marine/fresh water shales and
limestones is lacking, and it is not known whether examples east and west of the Illi
nois Basin had similar source material compositions (Parham, 1964). However, the
occurrence of shale-type underclays throughout North America suggests equivalent
source material for most localities.
Gardner et al. (1988) provided data that suggest that the underclays they investi
gated in eastern Kentucky and western Pennsylvania are composed of fireclay-type
to shale-type underclays. A direct comparison of our results with those of Gardner
et al. (1988) is difficult because the former used kaolinite : illite, kaolinite : mica or
kaolinite : mica + vermiculite XRD peak ratios, the apparent crystallite thickness
of mica (illite), and absolute amounts of chlorite and quartz to characterize their
profiles. Figure 19.9 of Gardner et al. (1988) shows the distribution of the kaolin-
iterillite XRD peak ratio and the closely correlated measure of kaolinite crystallinity
for an underclay that occurs within a shale approximately two meters below a coal.
The lower kaolinite : illite ratio and lower kaolinite crystallinity typical of fireclay-
type underclays is clearly associated with the zone of maximum weathering by plants
(Gardner et al., 1988). The apparent drop in the kaolinite : illite ratio demonstrates
514 R.E. Hughes, El Demons and WA. White
the difficulty of using this type of analytical method to investigate underclay pro
files. It is reasonable to assume that the kaolinite : illite ratio drops in the paleosol
because chlorite is lost and, as Gardner et al. (1988) point out, because the newly
formed kaolinite has a broad X-ray diffraction peak which probably produced a
lower kaolinite : illite peak height ratio. The degree of alteration in this^ paleosol
could have been more clearly assessed by quantifying the distribution of % kaolin
ite, % chlorite, and kaolinite crystallinity.
Wilson (1965) describes Carboniferous underclays from South Wales that appear
to be comparable to shale- andfireclay-typeunderclays of Parham (1964), Gardner
et al. (1988) and to Illinois underclays we studied. Unfortunately, it is unclear exactly
how Wilson's (1965) methods and results compare to those in North America.
Underclays younger and older than the Pennsylvanian-age underclays discussed
in this chapter offer the opportunity to investigate the effects on mineral suites of
different climates, plant species, source materials and environments of deposition.
Plant evolution is especially important to the differences observed for underclays
younger than Paleozoic. The Tfertiary coal-bearing strata of the western United
States received large inputs of volcanic ash and other coals may have had source
materials that were particularly rich in other rock types. These contrasts with the
source materials of Pennsylvanian-age deposits will be a particularly interesting
topic for future studies.
The paleobotany of younger coals is better understood, and therefore, detailed
studies could be done on the types of alteration that result from growth of particular
plants on a mineral substrate. The rise of angiosperms in the Mesozoic saw the grad
ual loss or substantial reduction in numbers of species from plant groups common in
Paleozoic swamps, such as lycopods, sphenopsids and early gymnosperms (Stewart,
1983). These changes probably increased both the net efficiency of nutrient extrac
tion from the substrate and the range of substrates in which plants could survive,
and should further modify the duration and type of alteration of minerals.
The underclay types described in this chapter should be applicable to under
clays of different localities and ages because the types result from processes that
include preservation of relatively unaltered source material (shale-type), soil for
mation in environments typical of intense weathering and/or downward drainage
(fireclay-type) and soil formation in a gley environment. Processes that might affect
the chemical and mineralogical content of underclays of other ages and locations,
after pedogenesis, include diagenesis of kaolinite (sandstone-type and associated
with the formation of flint clay) and burial diagenesis and metamorphism, such as
smectite and I/S —▲ illite diagenesis and diagenesis of kaolinite —▲ illite, chlorite
and feldspars. Although most underclays do not show evidence of postburial effects,
these processes change the depositional or pedologically produced composition.
Therefore, the effects of diagenesis and metamorphism must be identified and cor
rected for before an evaluation can be made of the kind and degree of paleosol
formation.
Because coal formation requires a relatively narrow set of climatic conditions, cli
mate is unlikely to have an overwhelming effect on the type of underclay formation.
However, it may be possible to distinguish ever-wet from monsoonal climatic local
ities. A high water table is assumed essential for development of most underclays.
Furthermore, the weathering products of cool, temperate and tropical environments
of underclay formation are distinguished by the degree of alteration, not by basic
processes. As a result, temperature will not affect the underclay type.
The nature of the source material for underclays of other ages and localities
only changes the absolute mineral and chemical content. The main characteristics of
the pedologie environment will be constant: rapid burial for shale-type, weathering
by S1O2 and alkali and alkaline earth removal forfireclay-type,or weathering by
K+ removal in a waterlogged environment for gley-type. Certain source materials
will mask the environment of soil formation. For example, a volcanic ash source
material will likely be altered to flint clay in any environment associated with coals,
while a smectitic source material will be distinctly modified to K/S and poorly
crystallized kaolinite in afireclay-typeenvironment and little changed in a gley-type
environment.
Because the process of underclay formation should be similar for underclays of
different geologic ages and localities, it should be possible to make a reasonable
comparison between underclay formation in different localities and ages with that in
the Pennsylvanian of Illinois.
Mineral matter associated with coals
Minerals occur in coals in two basic ways (Rao and Gluskoter, 1973, Stach et al,
1982). First, they occur in relatively discrete partings that are commonly claystone,
with or without diagenetic pyrite and/or marcasite layers or nodules (DeMaris et
al., 1983). The mineral suites of partings are similar to those offireclay-typeunder
clays and may therefore reflect alteration by plants and conditions similar to those
that produced thefireclay-typeunderclays. The lack of oxidation of coal below most
partings suggests that plant type was more important in determining the type of min
eral alteration than was downward-moving water. Alteration by downward-moving
water would have produced a profile with extensive oxidation of the underlying peat.
Second, minerals occur in coal as disseminated mineral matter. Disseminated
mineral matter includes biogenic silicates (produced in the original plants) and fine
grained pyrite and/or marcasite, and variably altered detrital mineral matter, both
eolian and water deposited. Disseminated mineral matter in coals often includes
fine clays within macérais, a biogenic diagenetic component such as quartz (Ruppert
et al., 1985) or kaolinite, kaolinite and pyrite cleat fillings and calcite fracture fill
ings. Taken as a whole, the suite offinelydisseminated clay minerals is similar to that
of the gley-type underclays, but, in general, the alteration has not proceeded as far,
presumably because the rate of deposition during peat formation was more rapid
than it was during formation of the gley-type underclay (Hughes, 1971; Hughes et
al, 1987).
Diagenetic and epigenetic minerals (such as calcite, kaolinite, pyrite and mar
casse) are often found in coal balls, cleats and other fractures. Kaolinite fillings
of cleat fractures were interpreted by Hughes (1971) to be derived from silica and
alumina in peat that mobilized during epigenesis. If correct, this interpretation
raises the question of whether well-crystallized kaolinite in other strata could be
similarly derived from plant matter. Some gley-type underclays and related breccia
flint clays (Hughes and White, 1969) contain more kaolinite than can be easily
accounted for by the alteration of detrital feldspar to kaolinite. Therefore, plant-
derived kaolinite may be an important source of kaolinite in these strata.
Further clarification of the interaction between mud partings and ancient swamp
plants is needed. Studies of the paleobotany of the Herrin Coal (Phillips and
DiMichele, 1981) indicated that pteridosperms and other non-lycopod plants in
crease in abundance in the coal immediately above claystone partings (Figure 19.5).
Data from paleobotanical investigations of the Herrin Coal in western Kentucky
showed a significant relationship between abundances of pteridosperms and clay-
roof shale
220
underclay
10 20 30 40 50 60 70%
Fig. 19.5. Stratigraphie distribution of claystone partings and percentage non-lycopod plant biomass in
Herrin Coal from site in Saline County, Illinois (modified from Phillips and DiMichele, 1981).
stone partings or clay-rich coal (Winston, 1988).

The influence of the partings on the composition of the plant community is
probably related to increased nutrient levels, a change to a wetter or drier habitat,
or both. These factors need to be evaluated by considering the type and extent of
alteration of associated mineral matter, both in partings and disseminated in the
coal. The pattern of mineral alteration suggests that fresh water sediment floods
were followed by the spread of silica-extracting and perhaps pH-lowering plants on
a waterlogged sediment or a sediment that was slightly above the water table.
The alteration of the peat soon after deposition by permineralization of the plant
tissues to form coal balls also deserves mention. This permineralization is usually an
early diagenetic process, is local in extent and requires a change in the composition
of interstitial waters, such as an increase in dissolved silica to form silicate coal
balls and enrichment of Ca2+ ions to bond with carbonate ions generated in the
peat to form carbonate coal balls (DeMaris et al., 1983). The detailed geochemical
properties of local swamp water and their role in coal-ball formation are poorly
understood and deserve further study.
Roof and overlying strata

Rocks overlying coals are only slightly altered by growing plants, and the degree
of alteration decreases upward as the environment of deposition becomes more
marine. Some I/S probably forms from illite, and a small amount of well-rystallized
kaolinite forms, particularly in concretions with plants (Figure 19.2). The dissem
inated kaolinite in these strata may originate from reworking of paleosols in the
sourceward direction, kaolinite in the detrital source material, plant growth on the
sediments, or (diagenetically) from fluids derived from the underlying peat during
coalification. A thorough study of the variety of kaolinite and its pétrographie re
lationships to other minerals is required to determine the origin of the kaolinite
in roof strata. Plagioclase is selectively preserved in gley-type underclays. We have
assumed that albite is preserved because Na+ is not required for plant growth. How
ever, Moore (1982) has suggested that albite in the Purington Shale (roof shale)
formed diagenetically. Similarly, further study is required to determine whether the
increased IM illite content in roof strata, compared with that in underclays, is the
result of diagenesis or better preservation of source materials. The frequency of
volcanic tonsteins in European coal strata cited by Burger and Damberger (1985)
suggests that IM illite in marine parts of the cyclothem may have formed by diagen
esis of smectite derived from volcanic ash.
Flint clays and tonsteins

In areas in which breccia flint clays were found in roof strata, Hughes and White
(1969) observed more alteration of roof strata than is usually found where these
clays are not present. It is likely that these areas were unique environments. The flint
clays themselves seem to occur in abandoned oxbows, and alteration in the region
was thought to result from continued plant growth during that part of cyclothem
formation when marine shales and limestones were being deposited elsewhere in
the basin (Hughes and White, 1969).
Hint clays and related tonsteins are commonly associated with Pennsylvanian
strata (Patterson and Hosterman, 1960; Smith and O'Brien, 1965; Williams et
al., 1968; Gardner et al., 1988). Large deposits occur in Missouri; circular pat
terns of mineral zonation in these deposits were thought by Keller (1968) to
represent long-continued alteration of illite-rich source material by fresh water
dialysis in sinkhole ponds. Tburtelot et al. (1986) have recently suggested that
these deposits represent altered Pennsylvanian sediments in sinkholes where al
teration was by downward moving fresh water of unknown age(s). Flint clays are
unusual in containing well-crystallized kaolinite and often being associated with
fireclay-type materials containing poorly crystallized kaolinite (Keller, 1968, Gard
ner et al, 1988). The close association between flint clays and underclays ob
served by Patterson and Hosterman (1960), Williams et al. (1968) and Gardner
et al. (1988) indicates a similarity of process. Our recent discovery of nondis-
persibleflintlikegrains in well-developedfireclay-typeunderclays of Illinois shows
that intense weathering in a fireclay-type environment may eventually produce
flint clay from shale-type source material. Burger and Damberger (1985) propose
that most flint clays and "flint clay underclays" in eastern North America result
from alteration of volcanic ash. If this is true, a comprehensive model of alter
ation of various types of source materials to kaolinite in soil environments will be
possible.
Tbnsteins are very common in Pennsylvanian coals of Europe and younger coals
in the western United States. They are interpreted by Burger and Damberger (1985)
to result from the deposition of volcanic ash in the coal swamp, and they show the
degree to which mineral matter is both altered and retained in these environments.
Tbnsteins and the authigenic kaolinites associated with underclays and coals suggest
that the alteration of volcanic glass and feldspars to well-crystallized kaolinite is
much easier than from 2:1 clay mineral source materials such as illite, chlorite,
I/S and smectite. The latter 2:1 clay minerals mainly produce K/S and poorly
crystallized kaolinite. Therefore, the common assumption that volcanic ash alters
first to smectite and then to well-crystallized kaolinite in swamps is not supported by
the evidence.
The role ofplants in the formation ofpaleosols

The role of plants in the formation of ancient soils was undoubtedly great, but has
not been well studied. The depth of root penetration and nutrient extraction mech
anism^) of ancient plants are especially important to investigations of underclays,
because the extent of possible alteration by plants is constrained by their growth

habits. Recent advances in the understanding of the morphology of Pennsylvanian-
age swamp plants have resulted from studies of concretions of mineralized plants in
coal balls (Phillips and Peppers, 1984; Phillips et al. 1985; DiMichele and Phillips,
1988). The nature and extent of the root systems of ancient plants in coal swamps
are subjects that have often been neglected or ignored. In general, however, it can
be said that the root systems of Pennsylvanian-age plants were more primitive than
modern plants, and these plants were generally not adapted to deep rooting. Lyco-
pod trees, which did well in both peat-swamp and underclay-swamp environments,
had extensive lateral root systems, with rootlets typically extending about 40 cm
from the main roots. Calamités, the only major group of plants which reproduced
vegetatively (rhizomes), were a minor but ubiquitous component of the peat swamp
and are well represented in non-swampflorasas well. Psaronius produced an aerial
root mantle which makes evaluation of the amount and function of its true roots
difficult. Other plants such as the tree ferns, represented by Medulosa, appeared to
lose their roots quickly to decay and are therefore difficult to study. In the living
swamp, root biomass was significant, giving the dominant lycopods, for example,
very large exchange surfaces for respiration (aeration) and nutrient intake from the
water column and from the decaying plant substrate.
Modern examples of the contribution of plants to soil formation (Lovering,
1959), the authigenic kaolinite associated with the organic fraction of coals (Hughes,
1971) and selective removal of K+ in gley-type underclays suggest the importance
of growing plants in the alteration of mineral matter. It seems likely that growing
plants transported significant amounts of inorganic material (such as K+, Ca2+,
Mg2+, Si0 2 , AI2O3) in an upward direction. Because underclays were near the wa
ter table at the time of their formation, they lack the features of a well-drained soil,
which include illuviation offinesand downward transport of nutrients. The evidence
suggests that inorganic material was transported to the surface litter zone during
each increment of plant growth. As new sediments were deposited, the process con
tinued. The apparent drop in intensity of alteration near the top of the profiles
shown in Figure 19.4b and c (as illustrated by the decrease in smectite and K/S)
probably reflects the increased organic deposition rate and corresponding decrease
in amount of alteration and upward transport of inorganic nutrients by plants of the
coal swamp.
Growing plants may represent the only process that produces K/S. Lovering
(1959) pointed out that plants may be passive or aggressive in their uptake of
inorganic materials. He suggested that some modern plants can accumulate S1O2
aggressively enough to maintain very low levels of S1O2 in solution. Furthermore,
the rates of alteration suggested by Lovering (1959) are high enough at normal pH
conditions that alteration of 2:1 clays to K/S by inorganic processes (or by organic
acids from decay of plant matter) other than Si02-extracting plants would result in
kaolinite or mineral alteration products other than K/S. Therefore, K/S formed in
underclays, related paleosols, and soils and paleosols in general indicates alteration
primarily by growing plants that extract S1O2.
Future research needs
Research on underclays and related paleosols will require a broader perspective

and a more interdisciplinary scope in the future than have been applied in the past.
More effort needs to be made to correlate mineral alteration patterns with plant
populations, and with other changes in depositional environments, and to correlate
mineral alteration patterns among equivalent basins of coal formation. Geochem-
ical information, including data from stable isotope determinations, is needed to
clarify mineral alteration pathways and to improve quantitative estimates of mineral
content. In the future, geologists should increase vertical sampling frequency to in-
crease chances of finding and detecting mineralogical profiles. Further comparison
of source materials and alteration products should improve the accuracy of mass
balances and thus improve estimates of intensity and duration of weathering as well
as correlation of those factors on a regional and global scale.
An improved understanding of how modern plants extract nutrients from mineral
matter and which plants extract particular constituents might make possible a better
model of the origin of individual types of alteration. More studies of illite (and
chlorite) structural polytypes are needed, especially studies that incorporate K/Ar or
similar dating techniques, because polytypes distinguish clearly between detrital and
diagenetic origins. Further studies are also needed on the distribution of biogenic
quartz (by cathodoluminescence) and on the geochemical fingerprint of volcanic ash
versus that of typicalfluvialsource materials.
Summary
Four types of underclays are distinguished: (a) shale-type underclays, primarily

unweathered source material — illite, chlorite, quartz, orthoclase and plagioclase;
(b) gley-type underclays in which source materials were altered by K+ extraction
in a water-saturated environment producing well-crystallized I/S and kaolinite, and
preservation of plagioclase; (c) fireclay-type underclays representing comparatively
long periods of weathering and the effects of silica-extracting plants in an indeter-
minate but possibly less-waterlogged environment, which led to poorly crystallized
I/S, vermiculite, and kaolinite, K/S and the total removal of chlorite and feldspars;
and (d) sandstone-type underclays, which developed where an influx of fluids over-
printed the clay mineral suite with abundant authigenic kaolinite, primarily devel-
oped from feldspars in the source material.
Most underclays are parts of paleosol profiles, which are similar to the substrates
of modern organic soils. However, the plants that grew on and altered the underclay,
at least Pennsylvanian plants described, left little organic material in the underclay
itself other than fossil roots. Where coal formed, organic material began to accu-
mulate on the underclay, and where coal is thicker than about 10 centimeters, the
organic material eventually became the rooting medium for growing plants. Partings
within Pennsylvanian coals have weathering patterns similar to those of fireclay-type
underclays (because of silica-extracting plants), while the growth of coal-forming
plants on peat caused alteration of the minerals similar to that observed in gley-type
underclays. Roof materials generally were weathered so little that they are compa-
rable to shale-type underclays. Tbnsteins are primarily the result of plant-induced
alteration of volcanic ash. Flint clays may result from alteration of igneous source
materials, weathering of feldspar-rich sediments, or from unusually long periods of
alteration of normal sediments by plants in a gley-type or fireclay-type environment.
The effects of plant growth seem to be the dominant agent in formation of all of
the paleosols. In gley-type underclays, most of the authigenic kaolinite probably re-
sults from K+ and Ca2+ removal from feldspars, although some kaolinite may result
from recrystallization of plant-derived S1O2 and AI2O3. In fireclay-type underclays,
S1O2 removal by plants and possibly lowered pH leads to a greater degree of alter-
ation and to poorly crystallized clay minerals (especially K/S). Significant amounts
of inorganic nutrients such as K+ and Ca2+ may have been transported upward
by plant growth as the underclay accumulated. The pH of pyrite- and marcasite-
containing underclays mostly reflects oxidation and hydrolysis of iron sulfides, unless
pH is measured on samples never exposed to air.
Weathering in gley-type environments produces a suite of I/S clay minerals that
parallels those formed by smectite-to-illite diagenesis. This parallelism may clarify
mechanisms in weathering and diagenesis. The rapid alteration of volcanic ash to
kaolinite in tonsteins suggests that ash may be an important source material in
some underclays and that ash weathers directly to kaolinite (rather than by way of
smectite) in environments associated with coals. K/S in fireclays has unique XRD
properties because it is derived from illite, chlorite, I/S and vermiculite rather than
only from smectite. S1O2 removal from 2:1 clay minerals by plants may be the only
process capable of producing K/S and therefore, K/S occurrence is unique to certain
types of pedogenesis.
References
Burger, K. and Damberger, H.H., 1985. Tonsteins in the coal fields of western Europe and North
America. In: A.T. Cross (Editor), Economic Geology, Coal, Oil and Gas. C. R. IX Int. Congr.
Carboniferous Stratigraphy and Geology, Southern Illinois University Press, Carbondale, 111., pp.
433-448.
Chamberlin, T.C., 1890. The method of multiple working hypotheses. Science, 15: 92-96.
Damberger, H.H., 1971. Qualification pattern of the Illinois Basin. Econ. Geol. 66: 488-494.
DeMaris, P.J., Bauer, R.A., Canili, R.A. and Damberger, H.H., 1983. Geologic investigation of roof and
522 R.E. Hughes, RJ. Demaris and WA. White
floor strata: Longwall demonstration, Old Ben Mine No. 24. Prediction of coal balls in the Herrin
Coal. Illinois State Geol. Surv. Contract/Grant Rep. 1983-2, USDOE, FG01-78ET12177, 69 pp.
DiMichele, WA. and Phillips, TL., 1988. Paleoecology of the Middle Pennsylvanian-age Herrin Coal
swamp (Illinois) near a contemporaneous river system, the Walshville paleochannel. Rev. Paleobot.
Palynol., 56: 151-176.
Gardner, T.W., Williams, E.G. and Holbrook, P.W, 1988. Pedogenesis of some Pennsylvanian under-
clays; ground-water, topographic, and tectonic controls. In: J. Reinhardt and WR. Sigleo (Editors),
Paleosols and Weathering through Geologic Time: Principles and Applications. Geol. Soc. Am.,
Spec. Pap., 216: 81-101.
Glass, H.D., Potter, P.E. and Siever, R., 1956. Clay mineralogy of some basal Pennsylvanian sandstones,
clays, and shales. Bull. Am. Assoc. Pet. Geol., 40: 751-754.
Grim, R.E. and Allen, V.T., 1938. Petrology of Pennsylvanian underclays of Illinois. Geol. Soc. Am.
Bull., 49:1485-1514.
Hughes, R.E., 1971. Mineral matter associated with Illinois coals. Ph.D. Thesis, University of Illinois,
145 pp. (unpublished).
Hughes, R.E. and White, W.A., 1969. A flint clay in Sangamon County, Illinois. In: L. Heller (Editor),
Proc. Int. Clay Conference, Tokyo, I, pp. 291-303.
Hughes, R.E., DeMaris, P.J., White, WA. and Cowin, D.K., 1987. Origin of clay minerals in Pennsylva
nian strata of the Illinois Basin. In: L.G. Schultz, H. van Olphen and EA. Mumpton (Editors), Proc.
Int. Clay Conference, Denver, Colo., pp. 97-104.
Julien, A.A., 1880. On the geological action of the humus acids. In: Proc. Am. Assoc. for the Advance
ment of Science, 28th Annu. Meet. 1879, Saratoga Springs, N.Y., pp. 311-410.
Keller, W.D., 1968. Flint clay and a flint-clay facies. Clays Clay Miner., 16: 113-128.
Logan, W.E., 1842. On the character of the beds of clay lying immediately below the coal seams of South
Wales. Proc. Geol. Soc. London, 3: 275-277.
Lovering, T.S., 1959. Significance of accumulator plants in rock weathering. Geol. Soc. Am. Bull., 70:
781-800.
Moore, D.M., 1982. Shallow burial diagenesis of the Purington Shale. Program with Abstracts, 19th
Annu. Meet., Clay Miner. Soc., Hilo, Hawaii, p. 73.
Odom, I.E. and Parham, WE., 1968. Petrography of Pennsylvanian underclays in Illinois and their
application to some mineral industries. 111. Geol. Surv. Circ, 429, 36 pp.
Parham, W.E., 1964. Lateral clay mineral variations in certain Pennsylvanian underclays. In: WF. Bradley
(Editor), Proc. 12th Nat. Conf. Clay Miner. Soc. Am., Atlanta, Georgia, Pergamon Press, New York,
N.Y., pp. 581-602.
Patterson, S.H. and Hosterman, J.W, 1960. Geology of the clay deposits in the Olive Hill District,
Kentucky. Proc. Seventh Nat. Conf. Clays Clay Miner., 5: 178-194.
Peppers, R.A, 1964. Spores in strata of Late Pennsylvanian cyclothems in the Illinois Basin. 111. State
Geol. Surv. Bull., 90, 89 pp.
Perry, E. and Hower, J., 1970. Burial diagenesis in Gulf Coast pelitic sediments. Clays Clay Miner., 18:
165-177.
Phillips, T.L. and DiMichele, W.A., 1981. Paleoecology of Middle Pennsylvanian age coal swamps in
southern Illinois — Herrin Coal Member at Sahara Mine No. 6. In: K.J. Miklas (Editor), Paleob-
otany, Paleoecology and Evolution, Praeger, New York, N.Y., pp. 231-284.
Phillips, T.L. and Peppers, R.A., 1984. Changing patterns of Pennsylvanian coal-swamp vegetation and
implications of climatic control on coal occurrence. Int. J. Coal Geol., 3: 205-255.
Phillips, TL., Peppers, R.A. and DiMichele, W A , 1985. Stratigraphie and interregional changes in
Pennsylvanian-age coal swamp vegetation: Environmental inferences. In: T.L. Phillips and C.B. Cecil
(Editors), Paleoclimatic Controls on Coal Resources of the Pennsylvanian System of North America.
Int. J. Coal Geol., 5: 43-109.
Underclays and related paleosoh associated with coals 523
Rao, C.P. and Gluskoter, H.J., 1973. Occurrence and distribution of minerals in Illinois coals. 111. Geol.
Surv. Circ, 476, 56 pp.
Reynolds, R.C., Jr., 1980. Interstratified clay minerals. In: G.W Brindley and G. Brown (Editors),
Crystal Structures of Clay Minerals, Mineral. Soc. London, pp. 249-303.
Rimmer, S.M. and Eberl, D.D., 1982. Origin of an underclay as revealed by vertical variations in
mineralogy and chemistry. Clays Clay Miner., 30: 422-430.
Robinson, D. and Wright, V.P., 1987. Ordered illite-smectite and kaolinite-smectite. Pedogenic minerals
in a Lower Carboniferous paleosol sequence, South Wales. Clay Miner., 22: 109-118.
Ruppert, L.E, Cecil, C.B., Stanton, R.W. and Christian, R.P., 1985. Authigenic quartz in the Upper
Freeport Coal bed, west-central Pennsylvania. J. Sediment. Petrol., 55: 334-339.
Smith, W.H. and O'Brien, N.R., 1965. Middle and Late Pennsylvanian flint clays. J. Sediment. Petrol.,
35: 610-618.
Stach, E., Mackowsky, M.-TH., Teichmuller, M., Taylor, G.H., Chandra, D. and Teichmuller, R., 1982.
Stach's Textbook of Coal Petrology, 3rd edition. Borntraeger, Berlin and Stuttgart, 533 pp.
Staub, J.R. and Cohen, A.D., 1978. Kaolinite-enrichment beneath coals: a modern analog, Snuggedy
Swamp, South Carolina. J. Sediment. Petrol., 48: 203-210.
Stewart, W.N., 1983. Paleobotany and the evolution of plants. Cambridge University Press, New York,
N.Y., 405 pp.
Stout, W, 1923. Origin of coal formation clays. In: W. Stout, R.T. Stull, W. J. McCaughey and D.J.
Demorest. Coal Formation Clays of Ohio, Geol. Surv. Ohio, Bull., 26: 533-568.
Tourtelot, H.A., Goldhaber, M.B. and Fitzpatrick, J.J., 1986. Filled-sink deposits of iron-rich clay, South
central Missouri: New data from drilling. Program with Abstracts, 23rd Annu. Meet., Clay Miner.
Soc., Jackson, Mississippi, p. 61.
Wanless, H.R. and Weller, J.M., 1932. Correlation and extent of Pennsylvanian cyclothems. Geol. Soc.
Am. Bull., 43: 1003-1016.
Willard, D.A, DeMaris, PJ. and Nelson, WJ., in press. Paleoecology of early Herrin swamp develop
ment within a small paleochannel, Christian County, Illinois.
Williams, E.G., Bergenback, R.E., Falla, W.S. and Vdagawa, S., 1968. Origin of some Pennsylvanian
underclays in western Pennsylvania. J. Sediment. Petrol., 38: 1179-1193.
Willman, H.B. and Payne, J.N., 1942 Geology and mineral resources of the Marseilles, Ottawa, and
Streator quadrangles. 111. State Geol. Surv. Bull., 66, 388 pp.
Wilson, M.J., 1965. The origin and geological significance of the South Wales underclays. J. Sediment.
Petrol., 35: 91-99. Winston, R.B., 1988. Paleoecology of Middle Pennsylvanian-Age Peat Swamp
Plants in Herrin Coal, Kentucky, U.S.A. Int. J. Coal Geol., 10: 203-238.
Wnuk, C. and Pfefferkorn, H.W., 1987. A Pennsylvanian-age terrestrial storm deposit: using plant fossils
to characterize the history and process of sediment accumulation. J. Sediment. Petrol., 57: 212-221.
525
Chapter 20
Mesozoic and Tertiary paleosols

MARY J.KRAUS
Introduction
Early investigations of Mesozoic and Tertiary paleosols generally focused on soli-

tary, well developed profiles at unconformities (Wanless, 1923; Pettyjohn, 1966).
The fossil soils were deemed important as stratigraphie markers of local and even
regional extent, and also provided information on past climates. Although these
studies emphasized soils that developed during lengthy periods of landscape stabil-
ity, interest has now expanded to include vertically stacked paleosols that formed
in aggrading sedimentary systems. Not surprisingly, this advance in paleosol re-
search occurred at approximately the same time that more rigorous sedimentologie
techniques, especially facies analysis, were coming to the fore in reconstructing
ancient sedimentary environments. Many of the early descriptions of vertically suc-
cessive paleosols originated from broader paleoenvironmental studies, especially
of ancient fluvial deposits (Allen, 1974; McBride, 1974; Leeder, 1975). Some of
these investigations demonstrated the great promise of paleosols for understanding
the different intrabasinal and extrabasinal factors that controlled the development
of stratigraphie packages (Allen, 1974; Leeder, 1975). Since these earlier investi-
gations, studies of paleosols have been applied to diverse geologic problems in-
cluding changes in paleoclimate elucidation of the geologic history of individual
stratigraphie sequences and analysis of taphonomy, evolution and major extinctions
(Bown and Kraus, 1981a, b; Thompson et al., 1982; Retallack, 1983a, 1985; Atkin-
son, 1986; Bown and Rose, 1987; Kraus, 1987; Retallack et al., 1987; Wright and
Robinson, 1988). In fact, vertically successive paleosols are now a major focus of
research on Mesozoic and Tertiary paleosols (Bown and Kraus, 1987; Fastovsky and
McSweeney, 1987; Kraus, 1987; Retallack et al., 1987; Wright and Robinson, 1988).
This chapter emphasizes the sedimentologie aspects of paleosols because the
Mesozoic and Tertiary record provides important insights concerning how soil de-
velopment and sedimentary processes are integrated over long time periods. First
considered is the broad spectrum of paleosols that develop in sedimentary systems,
526 Mary J. Kraus
depending on the balance between deposition and nondeposition. The spectrum

ranges from paleosols representing short time spans because they developed in
rapidly aggrading environments to very thick and well developed paleosols at un-
conformities. The different processes that regulate the occurrence and duration of
nondepositional episodes are also examined. Examples from the growing literature
of Mesozoic and Tertiary paleosols demonstrate these concepts. The second part
of the chapter focuses more closely on vertically stacked paleosols, in particular,
alluvial paleosols. The Eocene Willwood Formation of Wyoming is used to illustrate
lateral variability within individual paleosols and among thicker paleosol packages,
and to discuss causal factors. Alluvial paleosols also exhibit complex patterns of
vertical variability at several different scales. These are considered in light of the di-
verse and interrelated processes that govern fluvial systems. A final section explores
some of the new and exciting directions in paleosol research. These include the
application of new techniques and perspectives, such as remote sensing, to paleosol
studies, and new applications of paleosol analysis to determine sediment accumula-
tion rates and to predict the location of hydrocarbon reservoirs.
Paleosols and the Mesozoic-Tertiary stratigraphie record
Soil formation versus sedimentation and erosion
Stratigraphie sequences reflect a complex interplay among sedimentation, ero-

sion and non-deposition. Although the very presence of a stratigraphie column
emphasizes aggradation, of most interest to paleopedologists are times of non-
deposition during which pedogenic processes modified rock or sediment exposed
at the surface (Kraus and Bown, 1986). Intervals of non-deposition can be very
short, lasting approximately a year if related to some seasonal process. In fluvial
environments, for example, both floodplain sedimentation and erosion result from
overbank floods which are generally of short duration and recur every year or two.
Between flood events, pedogenesis modifies the alluvium. Superimposed on these
short term patterns are longer term trends of nondeposition and erosion. Lengthy
episodes of nondeposition, generally in conjunction with erosion, produce major
gaps in the stratigraphie record that are commonly regional in extent. Significant
unconformities result from processes that are extrinsic rather than inherent to a
sedimentary environment. Tfectonic activity and sea level fluctuations are examples
of such major extrinsic controls. Sedimentary sequences also contain gaps that are
intermediate between seasonal hiatuses and major unconformities. These episodes
of non-deposition and/or erosion also may be produced by extrinsic factors, for
instance, 20,000 or 40,000 year climatic cycles.
Relatively continuous aggradation, such as yearly overbank flooding by streams,
yields sequences of vertically stacked paleosol profiles that are termed cumulative
Mesozoic and Ternary paleosols 527
because there are regular additions of new parent material during soil develop
ment. Cumulative soils can be composite if they show the effects of two or more
episodes of pedogenesis, because sedimentation was slow relative to the pedogenic
rate (Morrison, 1967). If sedimentation is rapid relative to the rate of pedogenesis,
multistory soils, showing little or no overlap between successive profiles, develop
(Morrison, 1967). Moderate interruptions in deposition can produce thicker and
more mature paleosols. In afluvialsystem for example, a short-lived climatic change
could result in channel incision and terrace development. This event would be rep
resented in the stratigraphie record as a more mature paleosol than either the over
lying or underlying cumulative floodplain paleosols (Kraus and Middleton, 1987).
At major unconformities, paleosols generally consist of a single strongly developed
profile, which is commonly very thick. Obviously, these different paleosols are not
discrete types but rather form a continuous spectrum in response to the chang
ing proportions of aggradation, degradation and nondeposition. The Mesozoic and
Tfertiary paleosol record clearly illustrates this continuum.
Ancient aggradational systems

Cumulative paleosols are probably best developed in alluvial settings, including
alluvial fans, delta plains and, most notably, streamfloodplains.Therefore, it is not
surprising that alluvial paleosols represent the majority of Mesozoic and Tfertiary
examples described in the literature. Channel deposits can undergo pedogenic mod
ification; however, paleosols are better developed on overbank deposits because the
floodplain is emergent and subject to pedogenic modification for most of each year
(McBride, 1974; Hubert, 1977; Bown and Kraus, 1981a, 1987; Goldbery, 1982; Re-
tallack, 1983a, b). Cumulative paleosols have also formed on delta plains and these
may show a marine influence (McBride, 1974; Fastovsky and McSweeney, 1987).
On alluvial fans, deposition is typically restricted to an actively aggradational area
or fan lobe and the remainder of the fan is inactive until the locus of sedimenta
tion changes. Because periods of deposition and nondeposition alternate, vertically
successive paleosols can develop; and pedogenic features are typical of modern
and ancient alluvial fan deposits (Gilè and Hawley, 1966; Nilsen, 1982). Although
fans develop under a variety of climatic conditions, many occur in semiarid or arid
settings in which calere tes and reddened horizons are produced (Abbott, 1981; Hay-
ward, 1983).
Soil development is common in modern coastal eolian sediments because there
is some moisture (Amiel, 1975). Likewise, paleosols have been described from an
cient coastal eolian deposits (Bown et al., 1988). Paleosols are less frequently re
ported from ancient erg or sand sea deposits, which are common in the Mesozoic
record. Calcretes and gypcretes are typical of moist depositional interdunes and
sand sheets associated with modern ergs (Ahlbrandt and Fryberger, 1981; Fryberger
et al, 1983). In the ancient record, the Upper Tïiassic Dolores Formation in Col-
528 Mary J. Kraus
orado contains paleosols that developed on a sand sheet deposit at the margin of a
large erg, probably in a seasonal humid-dry climate (Blodgett, 1988). Many of the
ancient erg sequences, however, are dominated by dune deposits, which unless sta
bilized by a change in climate or sediment supply, appear less conducive to paleosol
development than interdune deposits. For example, Quaternary paleosols in Saudi
Arabian dune deposits, described by Anton (1983), represent episodes of increased
humidity; and paleosols in Late Paleozoic eolianites are associated with extensive
deflation surfaces and indicate that periods of dune migration were interspersed
with significant periods of landscape stability (Loope, 1984, 1985). In addition, be
cause dunes truncate a considerable part of the preceding dune as they migrate, the
preservation potential of any paleosols that do develop on dunes should be low.
Paleosols have developed on both marginal marine and lacustrine deposits in
response to regression and subaerial exposure of the sediment. Regressive events
range in time span and recurrence frequency; consequently, regression produced
paleosols vary in number and vertical spacing. A case in point is a sequence in
France of Late Cretaceous and early Tbrtiary paleosols, which developed on pe
ripheral lacustrine deposits that were frequently emergent, probably in response
to seasonal lake levelfluctuations(Freytet, 1973; Freytet and Plaziat, 1982). Pedo
logie features include burrows and root casts, calcareous and ferruginous nodules,
and mottling. On a longer time scale, lowermost Cretaceous carbonates in Switzer
land exhibit repetitive small-scale facies sequences (less than 2 m thick) and these
are grouped into medium-scale sequences (Strasser, 1988). One type of small-scale
sequence consists of subtidal to intertidal carbonates capped by pedogenic lami
nar calcretes with rhizoliths. Strasser (1988) attributed the small-scale sequences to
episodic exposure of the platform in response to 20,000 year climatic cycles and the
larger sequences, which have laterally extensive emersion surfaces, to 100,000 and
400,000 year cycles.
Ancient erosional landscapes

Not all unconformities in continental settings are marked by paleosols. Their ab
sence suggests that there was little lag time in the transition from degradational to
aggradational conditions. The particular attributes of a paleosol at an unconformity
provide important clues both as to the length of time between erosion and renewed
aggradation and for the cause of the erosional episode. Development of unconfor
mities and their associated paleosols have been variously attributed to tectonism,
sea levelfluctuationsand climatic change.
Probably the best known paleosol associated with an unconformity is the "Inte
rior Paleosol", which has been observed on rocks ranging in age from Early Creta
ceous to Eocene across a broad area of North Dakota, South Dakota, Nebraska and
Wyoming (Pettyjohn, 1966). In Badlands National Park, Retallack (1983a) observed
that this weathering zone is relatively complicated with both a late Eocene yellowish
Mesozoic and Tertiary paleosoh 529
zone and an earliest Oligocene reddish zone which he assigned to the Ultisol and
Alfisol orders, respectively. Paleosols associated with regional unconformities can
be extremely well developed and thick, indicating lengthy periods of development.
The "Interior Paleosor is up to 7 m thick (Pettyjohn, 1966). Even more impressive
is a Paleogene Oxisol, developed in southern California and Baja California. Its
thickness of 30 m is a minimum because its upper contact is an erosion surface (Ab
bott, 1981). Some of these thick paleosols can be correlated over great distances,
indicating that the conditions responsible for their development were widespread.
Of probably similar age to the "Interior Paleosol" are an Oxisol in California and a
deeply "weathered zone" associated with an erosion surface that can be correlated
from Colorado and New Mexico to Wyoming, the Pacific Northwest, and possibly as
far south as Mexico (Epis and Chapin, 1975; Scott, 1975; Ely et al., 1977; Singer and
NkediKizza, 1980; Gresens, 1981).
Tfectonic activity is a major cause of unconformities, especially on a regional scale.
Gresens (1981) attributed development of the above described late Eocene uncon
formity and associated paleosol or weathering zone to the slower movement and
shifted orientation of the North American plate that is believed to have occurred at
approximately 40 Ma. In Brazil, thick red paleosols of tropical affinity are developed
on a regional unconformity spanning Middle THassic through Late Jurassic time.
The stratigraphie gap formed in response to the separation of South America from
Africa. Mabesoone et al. (1977) ascribed these paleosols to a subsequent period of
tectonic quiescence and landscape stability.
Various workers have invoked climatic change to explain unconformities and
paleosol formation. Paleosols interpreted as latérites mark two unconformities of
Eocene and Miocene age in alluvial fan and braided stream deposits in Montana
and Idaho (Thompson et al., 1982). Erosion and subsequent development of kaoli-
nite rich soils at the unconformities were attributed to high rainfall, and deposition
of smectite-rich sediments to relatively arid conditions. Areally extensive "regional
bounding surfaces" (Tàlbot, 1985) or "super bounding surfaces" (Kocurek, 1988)
can develop in eolian deposits as a result of increased humidity which promotes erg
stabilization or diminishes sediment supply. Because these surfaces are the result
of erosion and a subsequent period of non-deposition and because they develop
under more humid conditions, Kocurek (1988) suggested that they should show ev
idence of soil modification and be regional to subcontinental in scale. In fact, soils
or paleosols associated with climatically-produced unconformities have been docu
mented in Saudi Arabia and the Sahel of the southern Sahara (Anton, 1983; Tàlbot,
1985).
Paleosols associated with extensive unconformities also form in response to sea
level changes, in many cases because newly emergent sediments are available for
pedogenic modification. A case in point is widespead paleosol development in Pale
ocene rocks of North Dakota and South Dakota, which Cherven and Jacob (1985)
attributed to a marine regression that corresponded to a global sea level fall approx-
530 Mary J. Kraus
imately 60 million years ago. Falling sea level can also result in stream incisement
of recently exposed marine sediments; when base level rises during a subsequent
transgression, the streams aggrade and fill in the valleys. Weimer (1983) suggested
that if the resultant paleovalleys have paleosols or root zones developed on them,
they are excellent evidence for marine regression. Because sea level fluctuations
affect sediment supply and sandflow paths, ergs can also be truncated and stabilized
in response to regression (Kocurek, 1988). Many of the eolian deflation surfaces
observed by Loope (1984, 1985) are capped by marine carbonates. The associated
paleosols indicate that erosion was a subaerial process that occurred during regres-
sion rather than marine incursion.
Paleosols in alluvial rocks
The developmental history of cumulative soils is complex because C-horizon ma-

terials are continuously added at the surface as new alluvium and B horizons form
in what were once A horizons. New increments of sediment can vary in texture and
mineralogy producing vertically heterogeneous parent material; and some additions
are thick enough to hinder or completely arrest soil development. Just as soils in a
modern landscape are diverse, so too are cumulative paleosols in a particular allu-
vial sequence. In the Tertiary White River and Arikaree Groups of South Dakota,
for example, Retallack (1983a) recognized ten different paleosol types, each rep-
resenting different sets of ancient soil forming conditions. Important new trends
in paleosol research include documenting the diverse paleosol types in a particular
alluvial sequence and analyzing their lateral and vertical relationships. Such stud-
ies are proving valuable for reconstructing ancient landscapes and examining the
intrabasinal and extrabasinal processes that controlled fluvial systems over time.
Lateral variability in paleosols

In the Eocene Willwood Formation of the Bighorn Basin, Wyoming, Bown (1985)
and Bown and Kraus (1987) recognized five basic paleosol types developed on
overbank deposits (Figure 20.1). These were interpreted as representing consecutive
stages of pedogenic development or maturity. Field observations demonstrate a
progressive lateral transition from the least mature stage-1 paleosols to the most
mature stage-5 paleosols with increased distance from a coeval channel sandstone.
The term "pedofacies" was introduced to designate these lateral changes in paleosol
type relative to a channel sand-body (Bown and Kraus, 1987).
Different pedofacies developed partly in response to variable environmental con-
ditions (such as soil moisture, vegetation) across the ancient floodplain. Pedofacies
also reflect differential rates of sediment accumulation in local drainage basins.
Floodplain accumulation rates decrease away from a channel because new additions
EXPLANATION
| I Gray
I I Orange
\ Light red
% Dark red
| Purple
lllljyljlji Clay-rich zones
* ·" Fe/AI glaebules
0
e, Mn glaebules
°.f CaC03 glaebules
V ^ Rhizoliths
<!: ::?
^ Burrows
STAGE 1 STAGE 2 STAGE 3
PALEOSOL PROFILE
1 ri i r I I Ã ' r~i r
Fig. 20.1. Five basic Willwood paleosol types, a. Field attributes, including color, trace fossils and
glaebules, characteristic of stages 1-4. b. Field attributes and corresponding distributions of interstitial
iron, aluminum and manganese, organic carbon and clay in a stage-5 paleosol. Type-i paleosols show
only incipient pedogenic modification and are formed on levee deposits next to a channel; type 5, the
most mature, developed farthest from a coeval channel on distalfloodplainsediments. A = A horizon;
(Ao) = epipedon in A horizon; B = B horizon; Bea = calerete in B horizon; Bg = gleyed B horizon; Bt
= clay enrichment in B horizon; Bfe = iron enrichment in B horizon (maximum); C = C horizon; Cca
= calerete in C horizon; Cg = gleyed C horizon; Eb = eluviated (albic) zone in A horizon. (From Bown
and Kraus, 1987.)
532 Mary J. Kraus
of alluvium from a single flood typically thin away from a channel and areas farther
from the channel are flooded less frequently than are areas close to the channel
(Bridge and Leeder, 1979). The maturity of accretionary soils depends in part on
the rate of sediment accumulation. This is borne out by a study in the Tïent Valley of
England, where the degree of soil development increases progressively from active
channel sediments to levee and then tofloodbasindeposits (Bridges, 1973).
As originally depicted, pedofacies changes are relatively systematic and simple
because a solitary stage-1 profile commonly grades laterally into a solitary stage-5
profile. Furthermore, the alluvial deposits were represented in only two dimen
sions, in a cross-section perpendicular toflowin the paleochannel. Additional study
demonstrates that pedofacies are more complex because of the pronounced spatial
and temporal variability infloodplainprocesses and deposits, including:
Lateral thickness variability. The thickest and coarsest grained alluvium (fine sands
and silts) accumulates directly adjacent to a channel as levees. Because levees ag
grade more rapidly than the surrounding flood plain, they contain several vertically
successive and poorly developed paleosols, these passing laterally into coeval flood-
plain deposits with fewer but more strongly developed paleosols (Figure 20.2). Thus
pedofacies form a prism of essentially contemporaneous paleosols in which the
type and number of vertically stacked profiles varies laterally, generally increasing in
maturity but decreasing in number away from the channel. Individual Willwood ped
ofacies thin from approximately 10 m to only several meters over lateral distances of
as much as 15 km (which represents half the width of the floodplain).
Crevasse-splay deposition. Crevasse-splay deposits originate when flood waters
break through the channel levee and spread onto the floodplain as a combination
of channelized and sheetflow.Over the course of approximately 100 years, a typical
Fig. 20.2. Schematic diagram showing lateral relations of Willwood paleosols to different parts of the
floodplain. Numbers refer to paleosol types recognized by Bown and Kraus (1987). Type 1 paleosols
show only incipient pedogenic modification and are formed on levee deposits; type 5, the most mature,
developed on distalfloodplainsediments. Plus or minus signs (3+) indicate a lesser or greater degree
of development than is typical for a particular paleosol type. A major crevasse splay sandstone forms an
essentially isochronous base to the pedofacies package. Thinner, less extensive splays separate individual
profiles.
splay enlarges into a network of distributary channels and then is gradually aban
doned and vegetated (Smith, 1983). The deposits are lobate to elliptical in plan view
and may extend onto the floodplain several km to nearly 10 km and cover areas
ranging from only a few square km to several tens of square km (Smith, 1983; Far
rell, 1987). Individual crevasse-splay complexes thin and become finer grained away
from the main channel (Smith, 1983; Farrell, 1987). Consequently, the influence of
individual crevasse-splay deposits on soil development varies laterally (Figure 20.3).
Successive Willwood paleosols are commonly separated by crevasse-splay de
posits ranging in thickness from 0.5 to 3 m and consisting offinegrainedsandstone
and siltstone (Kraus, 1987). These splay deposits were thick enough to insulate the
underlying soil from further pedogenesis and also formed the initial sediment pack
age for the overlying paleosol. Farther from the channel, splay deposits are thinner
(usually less than 0.5 m) and finer grained (siltstones or sandy silstones). These
occur within, rather than capping a paleosol profile, indicating that they were thin
enough to be incorporated into the developing soil. In some cases, the original splay
deposit became so well mixed with surrounding sediment, probably as a result of
bioturbation, that it is now barely perceptible. As a result, splays enhance the effect
of sediment thinning on pedofacies variability in two dimensions; paleosols sepa
rated by splay deposits form adjacent to the channel and pass laterally into fewer
paleosols, many of which incorporate thin splay deposits. In three dimensions, the
limited areal extent of individual splay deposits introduces additional heterogene
ity (Figure 20.3). Even in the well-exposed Willwood Formation, this complexity is
exceptionally difficult to document.
Crevasse-splay deposits of unusual thickness and lateral extent act as boundaries
between successive pedofacies packages (Figure 20.2). Such thick and widespread
Fig. 20.3. Schematic diagram showing three dimensional heterogeneity in paleosol development caused
by crevasse-splay deposits. Splay deposit 1 halted development of paleosols in columns A and B. In
column C, the splay was too thin to insulate the underlying soil and was incorporated into a developing
profile. Splay deposit 2 buried soils in columns B and C but do not extend to column A. As a result,
column/I has one buried paleosol and a still developing soil; column B has two paleosols; column C has
a single thick paleosol.
534 Mary J. Kraus
splay deposits may represent unusual floods of high magnitude and low recurrence
frequency. Alternatively, they may have been deposited as the channel avulsed, or
relocated to a different, lower part of the floodplain and generated widespread
discharge of water and sediment onto the floodplain.
Levee-building. As levees grow upward, they also prograde away from the channel
(Farrell, 1987). This results in more proximal paleosols eventually forming above
more distal paleosols. In addition, youthful levee systems flank both sides of a
meandering channel, whereas mature levees are developed preferentially on the
cutbank side of the channel. With lateral migration of the channel, levee deposits
move in a downstream direction and eventually form an interfingering network
with otherfloodplaindeposits (Farrell, 1987). This levee asymmetry and migration
should produce complex lateral and vertical variablity in paleosol types, a hypothesis
that is only now being tested.
Levee building is also self-limiting; with continued growth of levees above the
level of the channel, yearly floods are less likely to overtop the banks. As the
interval between depositional events on the floodplain increases, the amount of
time for pedogenic modification also increases (Taylor and Woodyer, 1978). Thus
the vertically stacked paleosols that develop on Willwood levee deposits commonly
show an upward increase in their degree of development (Figure 20.2). Likewise,
sediment influxes farther out on the floodplain should decline in frequency over
time, affording more time for pedogenesis later in the life of a soil, prior to burial by
a splay. This may explain how distinct profiles are preserved in alluvial paleosols.
Vertical sequences of paleosols. Vertical sequences of paleosols also exhibit vari
ability and the changes are commonly systematic. Kraus (1987) described three
orders of nested pedofacies sequences.
Simple sequences are less than 10 m thick and consist of one or more paleosols
bounded below and above by major crevasse-splay deposits. In many instances,
these are the unusually thick and extensive crevasse-splay deposits discussed above.
Simple sequences represent an alternation between seasonal overbank alluviation
accompanied by soil modification and the influx of crevasse-splay deposits thick
enough to arrest soil development. Several simple sequences are grouped to form a
compound pedofacies sequence, which is typically tens of meters thick and bounded
by major channel sand-bodies. Compound sequences show an upward change from
paleosols characteristic of levee deposits to progressively more distalfloodplainpa
leosols followed by a return to channel-proximal paleosols. Kraus (1987) attributed
compound sequences to a series of avulsions, which caused the channel tofirstshift
away from and then back towards the developing soil column (Figure 20.4). Conse
quently, upward changes through a compound sequence should reflect the pattern
or style of avulsion, for example, whether the channel moved in fairly regular,
stepwise jumps or in a series of random motions. Compound sequences have also
been documented by Wright and Robinson (1988) in Early Carboniferous flood-
plain deposits. Simple and compound pedofacies sequences are caused by processes
Fig. 20.4. Schematic diagram showing development of compound pedofacies sequence. Arabic numerals
indicate the paleosol stage (Figure 20.1); Roman numerals denote successively younger channel sand
stones. Stage 1 paleosols at the base of the column are related to sandstone /. The change from stage
1 to stage 2 indicates channel avulsion to location //. The most mature soil (stage 3) formed when the
channel occupied position IV. Development of this soil ended when the channel avulsed to position V.
The upper part of the sequence shows a decrease in pedogenic maturity related to the return of the
channel towards its original position. A stage-2 soil formed as channel sandstone V was deposited and
stage-i soils directly below sandstone VI are probably related to gradual avulsion of the channel from
V to VI. (From Kraus, 1987.)
inherent to the channel and thus record changes in the local pattern of overbank
deposition. Superimposed on these are longer term aggradational trends, which are
basin-wide or regional in extent, and generated pedofacies megasequences. Megase-
quences are hundreds of meters thick and encompass at least several compound
sequences. They are characterized by a distinct upward change in the overall matu
rity of their constituent compound sequences. Megasequences provide evidence for
major changes in the rate of sediment accumulation or soil forming conditions as a
result of allocyclic processes including tectonic activity and climatic change (Kraus,
1987). For example, Willwood deposits in the northern Bighorn Basin exhibit a
significant upward increase in pedogenic maturity due to declining rates of basin
subsidence. The degree of pedogenic development was greater during periods of
slow basin subsidence when sediment was accumulating slowly than during periods
of rapid subsidence during which sediment accumulation was more rapid.
Intrabasinal variability in paleosols

Vertical sequences of paleosols can also be used to compare and contrast sedi
ment accumulation in different parts of a sedimentary basin. Within a tectonically
evolving basin, the locus of maximum subsidence varies areally and temporally.
Differential subsidence influencesfluvialsystems and the resulting lithofacies pack
ages (Allen, 1978; Bridge and Leeder, 1979). For example, channel belts tend to
localize in areas of greater subsidence, along a structural axis or near active fault
zones (Bridge and Leeder, 1979). As the following examples illustrate, differential
subsidence also produces spatial variability influvialpaleosols. The middle Eocene
536 Mary J. Kraus
Capella Formation in Spain shows the effects of floodplain tilting and differential
subsidence in response to synsedimentary thrust faulting (Atkinson, 1986). Thrust-
ing caused westward tilting of the floodplain and more rapid subsidence of the
western part of the basin. Atkinson concluded that a thick sequence of poorly
drained, immature paleosols (Aquents and Fluvents) in downstream areas to the
west developed in response to low topography and relatively rapid rates of sediment
accumulation. In contrast, a thinner sequence of well drained, moderately mature
Vertic Fluvent paleosols in eastern upstream areas reflected better drainage and
slower rates of aggradation. In a second study, differences including the color, de-
gree of mottling and abundance of carbonate nodules in paleosols and the presence
or absence of carbonaceous shales were documented among penecontemporaneous
Willwood rocks in four different areas of the Bighorn Basin (Wing and Bown, 1985).
The lithologie variability of these sediment packages was attributed to structurally
controlled differences in topography. Wing and Bown (1985) concluded that a thick
section of drab, intensely mottled paleosols in the east central part of the basin
resulted from poor drainage and rapid sediment accumulation, possibly in response
to downwarping along a major basement lineament.
Future trends
A significant aspect of paleopedology is that it intrigues geoscientists with diverse

areas of expertise. These workers are supplying fresh and exciting methods for
analyzing paleosols and stimulating new questions and additional problems to be
examined. This final section explores several new directions in paleosol research.
Remote sensing
With few exceptions, studies of Mesozoic and Cenozoic paleosols have been
local in scope. We face a growing need for basin wide investigations to evaluate
the complex factors that influence larger scale spatial and temporal variability in
paleosols. Regional studies are difficult, however, because they require extensive,
good exposures and are labor intensive. One approach of enormous potential is
remote sensing, used in conjunction with established methods of paleosol analysis.
Remote sensing is feasible because materials at the surface differ in the way they
scatter and absorb incoming solar energy. These differences can be recorded by
airborne or satellite instruments as functions of wavelength and intensity. Different
sensors record these spectral data in different wavelength regions and with different
surface resolution (Goetz et al., 1985). The spectral data gathered by remote sensors
can be used to discriminate different lithologies at the surface and determine their
distribution, with the distinct advantage that compositional variations are recorded
over very large areas (Lang et al., 1987). Consequently, the geologic applications of
remote sensing include mapping and structural interpretation, mineral exploration

and stratigraphie analysis. Well exposed paleosol sequences that exhibit large-scale
spatial and/or temporal variability and that contain paleosols with distinct spectral
signatures are potentially well suited to remote sensing. Willwood paleosols, for
example, vary in iron minerals such as hematite and goethite, minerals that can
be distinguised spectrally (Hunt et al., 1971). Work now in progress indicates that
different paleosol packages in the Willwood Formation, such as those observed
in the field by Wing and Bown (1985), can be recognized using remote sensing
techniques. The advantages to this approach are manifold. Field collection of data is
subject to time limitations and to the quality and accessibility of exposures. Certain
types of lithologie information can be more efficiently gathered or more readily
grasped using remote sensing than conventional field methods. Detailed field study,
including paleosol analysis, concentrates on small areas and remote sensing can
expand a study to cover larger areas. This is an especially important consideration
for workers interested in applying paleosols to basin analysis. Furthermore, the
broad overview shown in airborne or satellite images has the potential to reveal
large scale features not visible from the ground, and thus to stimulate new questions
and insights in paleosol investigations.
Sediment accumulation rates
Estimation of sediment accumulation rates can add significantly to understanding

the development of sedimentary basins. Changes in accumulation rates through
time and across space record changes in the different extrinsic and intrinsic factors
that control depositional systems. Rates are generally calculated by dividing the
thickness of a stratigraphie section by its known or estimated time span, a method
that produces time-averaged rates. A major problem for most nonmarine sequences
is determining how much time is represented. With few exceptions, paleomagnetic
dating is still relatively coarse and radiometrie dates are generally rare or widely
spaced stratigraphically. As a consequence, even if the time interval represented by
a stratigraphie section can be estimated, the interval is quite lengthy (on the order
of 1,000,000 years or more) and incorporates not only periods of deposition but also
significant hiatuses. In addition, important short term changes in accumulation rate
that occur during a 1,000,000 year period are not apparent.
Based on the assumption that paleosol maturity reflects a balance between the
rates of sediment accumulation and pedogenesis, cumulative paleosols provide a
new and important technique for evaluating short term sediment accumulation rates
in nonmarine deposits. Leeder (1975) and Retallack (1983b, 1984) have taken a
quantitative approach to accumulation rates by estimating development times for
different paleosol types through analogy to modern soils. Allen (1974), Atkinson
(1986) and Kraus (1987), on the other hand, used palesols to evaluate relative
accumulation rates and to compare rates for different areas and parts of the section.
538 Mary J. Kraus
Using paleosols for time resolution suffers from several assumptions that underlie
this approach, most importantly:
(a) The paleosols are not composite and have not been eroded. In aggradational
fluvial sequences, the latter assumption is probably not unfounded; however, if
paleosols are composite, accumulation rates can be grossly overestimated.
(b) Modern analogues for the paleosols can be identified and their matura
tion times are established. It is generally difficult to classify paleosols according to
modern soil schemes because important data, such as pH and base saturation, are
lacking and because most fossil soils have undergone some degree of pedogenic
overprinting and/or diagenetic modification. Furthermore, even if analogues can
be confidently determined, it is probable that accurate estimates of their matura
tion time are currently lacking, although relative, order of magnitude estimates are
probably not far off.
Despite these drawbacks, the quantitative approach to accumulation rates merits
further study together with improved knowledge of modern soil maturation rates.
Pre-Quaternary workers should follow Bilzi and Ciolkosz (1977) by developing pale-
osol indices, which will enable the degree of pedogenic development to be easily and
consistently assessed. Although soil indices are implicit in the studies of Retallack
(1983a, b) and Bown and Kraus (1987) those workers did not quantify them.
Reservoir prediction
Fluvial facies tend to change over very short distances; thus predicting the po
sition of fluvial reservoirs is difficult unless well control is exceptional. Well log
studies or coring programs typically focus on correlating and describing channel
sand-bodies rather than non-producing overbank deposits; however, the lateral and
vertical heterogeneity in sequences offluvialpaleosols is potentially valuable for as
sessing the location of channel sand-bodies (Kraus and Bown, 1988). The different
paleosol types in the Willwood Formation differ in features recognizable in core,
including lithology, color and color mottling, geochemical properties, type of soil
nodules and trace fossils (Bown and Kraus, 1987). Tb use paleosols in exploration,
different maturation stages, such as observed in the Willwood Formation, must be
identified. Because lateral ordering of paleosols is similar to that encountered verti
cally, this can be accomplished from vertical profiles (exposures or cores) alone.
The type of paleosol at a specific location (exposure or core) provides an estimate
of the distance to a channel sand-body at the same stratigraphie level. A mature pa
leosol indicates that a sand-body is relatively far from the core or outcrop and an
immature paleosol indicates that a laterally equivalent sand-body is close. Immature
paleosols are concentrated along the margins of sand bodies and thus may be help
ful in predicting the trend of ribbon sand-bodies, which are especially difficult to
follow in the subsurface. Vertical sequences of paleosols are valuable because they
provide a record of channel movement over time. A change in maturity between two
successive paleosols indicates that a channel sand-body moved relative to the core
position. Information from additional cores are needed to show the actual direction
of movement. The particular stage of paleosol development can then be used to
estimate distance to the sand-body. Larger packages of paleosols can help predict
sand-body location on a basin wide scale. As the Willwood Formation demonstrates,
irregular basin subsidence is a major cause of largescale lateral and vertical vari-
ability in paleosols. The subsidence pattern also influences channel location. Over
time, channel sandstones tend to concentrate in areas of more rapid subsidence,
close to a basin axis or near active fault zones (Bridge and Leeder, 1979). The re-
sult is that paleosol packages dominated by immature types highlight areas where
channel sand-bodies are most abundant. In contrast, large packages dominated by
strongly developed paleosols characterize the more slowly subsiding areas of the
basin, where sand-bodies are apt to be rare.
Summary
Mesozoic and Tfertiary rocks are replete with diverse fossil soils, most of which
formed on recently deposited sediment or on sedimentary rocks. Such paleosols
constitute a continuum representing non-depositional periods of various lengths.
Vertically successive cumulative profiles characterize aggrading systems and record
relatively short and frequent intervals of sedimentologie inactivity. With increas-
ingly lengthy episodes of non-deposition, these grade into very mature profiles that
formed at unconformities. The presence of a paleosol at an unconformity and its
particular attributes provide important clues concerning the process responsible for
the stratigraphie gap.
The profusion of well exposed alluvial paleosols of Mesozoic-Tertiary age af-
fords a remarkable opportunity to explore aspects of paleopedology beyond mere
identification and classification. As discussed in this chapter, analyzing the remark-
able lateral and vertical variability of paleosols is proving helpful in evaluating the
depositional and structural history of alluvial basins and the paleoclimatic setting.
Paleosols exhibit complex lateral variability caused in part by the normal processes
of floodplain building such as seasonal floods, levee construction, and crevasse-
splay deposition. Differential subsidence or local climatic differences may yield re-
gional variability in paleosol types. Vertically successive paleosols also show upward
changes effected by episodic avulsion, and, on a grander scale, by tectonic or cli-
matic fluctuations. Pre-Quaternary paleosols are increasingly used to attack sundry
problems including interpreting paleoclimates, reconstructing ancient landscapes
and depositional environments, and understanding the advent or demise of various
plants and animals in addition to their taphonomic and paleoecologic setting. With
future advances in our knowledge of paleosols and the methodology to examine
them, especially at a regional scale, they will facilitate research in areas such as
nonmarine basin analysis and reservoir prediction.
540 Mary J. Kraus
References
Abbott, P.L., 1981. Cenozoic paleosols San Diego area, California. Catena, 8: 223-237.
Ahlbrandt, T.S. and Fryberger, S.G., 1981. Sedimentary features and significance of interdune deposits.
In: F.G. Ethridge and RM. Flores (Editors), Recent and Ancient Nonmarine Depositional Environ
ments: Models for Exploration. Soc. Econ. Paleontol. Mineral., Spec. Pubi., 31: 293-314.
Allen, J.R.L., 1974. Studies in fluviatile sedimentation: implications of pedogenic carbonate units, Lower
Old Red Sandstone, Anglo-Welsh outcrop. Geol. J., 9: 181-208.
Allen, J.R.L., 1978. Studies in fluviatile sedimentation: an exploratory quantitative model for the archi
tecture of avulsion controlled alluvial suites. Sedimentol. Geol., 21: 129-147.
Amiel, AJ., 1975, Progressive pedogenesis of eolianite sandstone. J. Sediment. Petrol., 45: 513-519.
Anton, D., 1983. Modern eolian deposits of the eastern province of Saudi Arabia. In: M.E. Brookfield
and T.S. Ahlbrandt (Editors), Eolian Sediments and Processes. Developments in Sedimentology, 38.
Elsevier, Amsterdam, pp. 365-378.
Atkinson, CD., 1986. Tectonic control on alluvial sedimentation as revealed by an ancient catena
in the Capella Formation (Eocene) of northern Spain. In: VP. Wright (Editor), Paleosols: Their
Bilzi, A.F. and Ciolkosz, E.J., 1977. A field morphology rating scale for evaluating pedological develop
ment. Soil Sci., 124: 45-48.
Blodgett, R.H., 1988. Calcareous paleosols in the Triassic Dolores Formation, southwestern Colorado.
In: J. Reinhardt and WR. Sigleo (Editors), Paleosols and Weathering Through Geologic Time:
Principles and Applications. Geol. Soc. Am., Spec. Pap., 216: 103-121.
Bown, TM., 1985. Maturation sequences in lower Eocene alluvial paleosols, Willwood Formation. In:
R.M. Flores and M. Harvey (Editors), Field Guide to Modern and Ancient Fluvial Systems in the
United States. 3rd Int. Fluvial Sedimentol. Conf., Guidebook, pp. 20-26.
Bown, TM. and Kraus, M.J., 1981a. Lower Eocene alluvial paleosols (Willwood Formation, northwest
Wyoming, U.S.A.) and their significance for paleoecology, paleoclimatology, and basin analysis.
Palaeogeogr., Palaeoclimatol., Palaeoecol., 34: 1-30.
Bown, TM. and Kraus, M.J., 1981b. Vertebrate fossil-bearing paleosol units (Willwood Formation, lower
Eocene, northwest Wyoming, U.S.A.): implications for taphonomy, biostratigraphy, and assemblage
analysis. Palaeogeogr., Palaeoclimatol., Palaeoecol., 34: 31-56.
Bown, TM. and Kraus, M.J., 1987. Integration of channel and floodplain suites, I. Developmental
sequence and lateral relations of alluvial paleosols. J. Sediment. Petrol., 57: 587-601.
Bown, TM. and Rose, K.D., 1987. Patterns of dental evolution in early Eocene anaptomorphine pri
mates (Omomyidae) from the Bighorn Basin, Wyoming. Paleontol. Soc, Mem., 23, 161 pp.
Bown, TM., Larriestra, C.N., and Powers, D.W., 1988. Analysis paleoambiental de la Formation Pinturas
(Mioceno Inferior), Provincia de Santa Cruz. Segunda Reunion. Argent. Sediment., Actas, 1: 31-35.
Bridge, J.S. and Leeder, M.R., 1979. A simulation model of alluvial stratigraphy. Sedimentology, 26:
617-644.
Bridges, E.M., 1973. Some characteristics of alluvial soils in the Trent Valley, England. Int. Soil Sci. Soc.
Trans., 5th and 6th Comm., pp. 247-253.
Cherven, VB. and Jacob, A F , 1985. Evolution of Paleogene depositional systems, Williston Basin, in
response to global sea level changes. In: R.M. Flores and S.S. Kaplan (Editors), Cenozoic Paleogeog-
raphy of the West Central United States. Rocky Mt. Sect. Soc. Econ. Paleontol. Mineral., Denver,
Colo., pp. 127-170.
Ely, R.W., Grant, TA. and McCleary, J.R., 1977. Exhumed early Tertiary paleotopographic surfaces on
the lone Formation, northern San Joaquin Valley, California. Geol. Soc. Am., Abstr. with Progr., 9:
417.
Epis, R.C. and Chapin, CE., 1975. Geomorphic and tectonic implications of the post Laramide, Late
Eocene erosion surface in the Southern Rocky Mountains. Geol. Soc. Am. Mem., 144: 45-74.
Mesozoic and Tertiary paleosols 541
Farrell, K.M., 1987. Sedimentology and facies architecture of overbank deposits of the Mississippi River,
False River Region, Louisiana. In: EG. Ethridge, R.M. Flores and M.H. Harvey (Editors), Recent
Developments in Fluvial Sedimentology. Soc. Econ. Paleontol. Mineral., Spec. Pubi., 39: 112-120.
Fastovsky, D.E. and McSweeney, K., 1987. Paleosols spanning the Cretaceous Paleogene transition,
eastern Montana and western North Dakota. Geol. Soc. Am. Bull., 99: 66-77.
Freytet, P., 1973. Petrography and paleoenvironment of continental carbonate deposits with particular
reference to the Upper Cretaceous and Lower Eocene of Languedoc (southern France). Sedimentol.
Geol., 10: 25-60.
Freytet, P. and Plaziat, J.C., 1982. Continental Carbonate Sedimentation and Pedogenesis Late Creta
ceous and Early Tertiary of Southern France. Contrib. Sedimentol., 12, 212 pp.
Fryberger, S.G., AlSari, A.M. and Clisham, T.J., 1983. Eolian dune, interdune, sand sheet, and silici-
clastic sabkha sediments of an offshore prograding sand sea, Dharan area, Saudi Arabia. Am. Assoc.
Pet. Geol. Bull., 67: 280-312.
Gilè, L.H. and Hawley, J.W., 1966. Periodic sedimentation and soil formation on an alluvial-fan pied
mont in southern New Mexico. Soil Sci. Soc. Am. Proa, 30: 261-268.
Goetz, A.F.H., Vane, G., Solomon, J.E. and Rock, B.N., 1985. Imaging spectrometry for earth remote
sensing. Science, 228: 1147-1153.
Goldbery, R., 1982. Palaeosols of the Lower Jurassic Mishhor and Ardon Formations ("Laterite Deriva
tive Fades"). Sedimentology, 29: 669-704.
Gresens, R.L., 1981. Extension of the Telluride erosion surface to Washington state, and its regional
and tectonic significance. Tectonophysics, 79:145-164.
Harden, J.W., 1982. A quantitative index of soil development from field descriptions: examples from a
chronosequence in central California. Geoderma, 28: 1-28.
Hayward, A.B., 1983. Coastal alluvial fans and associated marine facies in the Miocene of S.W. Turkey.
In: J.D. Collinson and J. Lewin (Editors), Modern and Ancient Fluvial Systems. Int. Assoc. Sedi
mentol. Spec. Pubi. 6, Blackwell, Oxford, pp. 323-336.
Hubert, J.F., 1977. Paleosol caliche in the New Haven Arkose, Connecticut: record of semiaridity in
Late Triassic-Early Jurassic time. Geology, 5: 302-304.
Hunt, G.R., Salisbury, J.W. and Lenhoff, C.J., 1971. Visible and near infrared spectra of minerals and
rocks: III. Oxides and hydroxides, Modern Geol., 2: 195-205.
Kocurek, G., 1988. First order and super bounding surfaces in eolian sequences — Bounding surfaces
revisited. Sedimentol. Geol., 56:193-206.
Kraus, MJ., 1987. Integration of channel and floodplain suites, II. Vertical relations of alluvial paleosols.
J. Sediment. Petrology, 57: 602-612.
Kraus, M.J. and Bown, T.M., 1986. Paleosols and time resolution in alluvial stratigraphy. In: VP. Wright
(Editor), Paleosols: Their Recognition and Interpretation. Blackwell, Oxford, pp. 180-207.
Kraus, M J. and Bown, T.M., 1988. Using paleosols to predict fluvial reservoir location and continuity.
Geol. Soc. Am., Abstr. with Progr., 20: 260.
Kraus, M.J. and Middleton, L.T., 1987. Dissected paleotopography and base-level changes in a THassic
fluvial sequence. Geology, 15:18-21.
Lang, H.R., Adams, S.L., Conel. J.E., McGuffie, B.A., Paylor, E.D. and Walker, R.E., 1987. Multispec-
tral remote sensing as stratigraphie and structural tool, Wind River Basin and Big Horn Basin areas,
C o m i n g . Am. Assoc. Pet. Geol. Bull., 71: 389-402.
Leeder, M.R., 1975. Pedogenic carbonates and flood sediment accretion rates: a quantitative model for
alluvial arid-zone lithofacies. Geol. Mag., 112: 257-270.
Loope, D.B., 1984. Origin of extensive bedding planes in aeolian sandstones: a defence of Stokes'
hypothesis. Sedimentology, 31: 123-125.
Loope, D.B., 1985. Episodic deposition and preservation of eolian sands: a Late Paleozoic example
from southwestern Utah. Geology, 13: 73-76.
542 Mary J. Kraus
Mabesoone, J.M., Rolim, J.L. and de Castro, C , 1977. Late Cretaceous and Cenozoic history of north
eastern Brazil. Geol. Mijnbouw, 56:129-139.
McBride, E.E, 1974. Significance of color in red, green, purple, olive, brown, and gray beds of the
Difunta Group. J. Sediment. Petrol., 44: 760-773.
Morrison, R.B., 1967, Principles of Quaternary soil stratigraphy. In: R.B. Morrison and H.E. Wright
(Editors), Means of Correlation of Quaternary Successions. Int. Assoc. Quat. Res. (INQUA), Proc.
VII Congr., 9:1-69.
Nilsen, T., 1982. Alluvial fan deposits. In: P.A. Scholle and D.R. Spearing (Editors), Sandstone Deposi-
tional Environments. Am. Assoc. Pet. Geol. Mem., 31: 49-86.
Pettyjohn, W.A., 1966. Eocene paleosol in the northern Great Plains. U.S. Geol. Surv., Prof. Pap., 550C:
61-65.
Retallack, G.J., 1983a. Late Eocene and Oligocene paleosols from Badlands National Park, South
Dakota. Geol. Soc. Am., Spec. Pap., 193, 82 pp.
Retallack, G.J., 1983b. A paleopedological approach to the interpretation of terrestrial sedimentary
rocks: the mid Tertiary fossil soils of Badlands National Park, South Dakota. Geol. Soc. Am. Bull.,
94: 823-840.
Retallack, G.J., 1984. Completeness of the rock and fossil record: some estimates usng fossil soils.
Paleobiology, 10: 59-78.
Retallack, G.J., 1985. Fossil soils as grounds for interpreting the advent of large plants and animals on
land. Philos. Trans. R. Soc. London, Ser. B, 309: 105-142.
Retallack, G.J., Leahy, G.D. and Spoon, M.D., 1987. Evidence from paleosols for ecosystem changes
across the Cretaceous/Tertiary boundary in eastern Montana. Geology, 15: 1090-1093.
Scott, G.R., 1975. Cenozoic surfaces and deposits in the Southern Rocky Mountains. Geol. Soc. Am.
Mem. 144, 227-248.
Singer, M.J. and NkediKizza, P., 1980. Properties and history of an exhumed Tertiary Oxisol in Califor
nia. Soil Sci. Soc. Am. J., 44: 587-590.
Smith, D.G., 1983, Anastomosed fluvial deposits: modern examples from Western Canada. In: J.D.
Collinson and J. Lewin (Editors), Modern and Ancient Fluvial Systems. Int. Assoc. Sedimentol.,
Spec. Pubi., 6:155-168.
Strasser, A., 1988. Shallowing-upward sequence in Purbeckian peritidal carbonates (lowermost Creta
ceous, Swiss and French Jura Mountains). S ed im en toi ogy, 35: 369-383.
Talbot, M.R., 1985. Major bounding surfaces in aeolian sandstones a climatic model. Sedimentology, 32:
257-265.
Taylor, G. and Woodyer, K.D., 1978. Bank deposition in suspended load streams. In: A.D. Miall (Edi
tor), Fluvial Sedimentology. Can. Soc. Pet. Geol. Mem., 5: 257-275.
Thompson, G.R., Fields, R.W. and Alt, D., 1982. Land-based evidence for Tertiary climatic variations:
northern Rockies. Geology, 10: 413-417.
Winless, H.R., 1923. The stratigraphy of the White River Beds of South Dakota. Am. Philos. Soc. Proc,
62: 663-669.
Weimer, RJ., 1983. Relation of unconformities, tectonics, and sea level changes, Cretaceous of the Den
ver Basin and adjacent areas. In: M.W Reynolds and E.D. Dolly (Editors), Mesozoic Paleogeography
of the West Central United States. Rocky Mt. Sect. Soc. Econ. Paleontol. Mineral., Denver, Colo.,
pp. 359-376.
Wing, S.L. and Bown, T.M., 1985. Fine scale reconstruction of Late Paleocene Early Eocene paleo
geography in the Bighorn Basin of northern Vtyoming. In: R.M. Flores and S.S. Kaplan (Editors),
Cenozoic Paleogeography of the West Central United States. Rocky Mt. Sect. Soc. Econ. Paleontol.
Mineral., Denver, Colo., pp. 93-106.
Wright, V.P. and Robinson, D., 1988. Early Carboniferous floodplain deposits from South Wales: a case
study of the controls on palaeosol development. J. Geol. Soc. London, 145: 847-857.
543
Chapter 21
Paleozoic paleosols
GJ. RETALLACK
Introduction
Soils and life are so intimately interconnected that it is difficult to conceive of

many kinds of life without soils or of many kinds of soils without life. Nevertheless,
there is life in both sea and air that is only indirectly dependent on soil. There also
are soils on mountain tops and in Antarctic deserts that are only minimally affected
by life. The surficially altered surfaces of the Moon, Mars and Venus also have
been called soils and, so far, all appear devoid of life. Comparison between these
various modern soils and organisms is one way of assessing the interrelationship
between soil and life. But there is another way. The fossil record of soils, particularly
those of Paleozoic age, some 245 to 570 million years old, can be viewed as a long
term natural experiment in soil formation under increasingly complex terrestrial
ecosystems. During this period of geological time, multicellular land plants and
animals appeared and diversified to create most of the main kinds of ecosystems
now gracing the land. Along with them, diversified different kinds of soils. Paleozoic
paleosols can be regarded as trace fossils of ancient ecosystems, providing a long
term record of increasing biotic influence on soil formation.
The Paleozoic diversification of soils could be assessed by a variety of methods.
Most of these are quite premature, given the level of available information on Pale-
ozoic paleosols (Retallack, 1986a) and the progress yet to be made in understanding
how much of soil-forming processes are preserved in such ancient and altered pa-
leosols (Retallack and Wright, 1990). For these reasons, it is not yet possible to
make compilations of named paleosols, comparable with plots of changing diversity
through geological time available for fossil land plants (Niklas et al., 1985), insects
(Sepkoski and Hulver, 1985) or vertebrates (Benton, 1985). It is, however, feasible
to analyze the first appearance of major kinds of paleosols. These are sufficiently
distinctive that they can be recognized among a large number of paleosols exam-
ined in the field or assessed from scientific literature. This preliminary effort to
document first appearances of paleosol types serves as a reconnaissance outline of
544 GJ. Retaüack
soil diversification during Paleozoic time. Later studies may refine these conclusions
and perhaps prove that some peculiar Paleozoic paleosols represent extinct types of
soil. Until such a surplus of information accrues, we must settle only for tracing the
origins of modern soil types, such as the orders of the U.S. Soil Conservation Ser-
vice or "Soil Taxonomy". There are only ten of these orders, and only nine need to
be considered here because Mollisols are unknown in deposits older than Eocene,
some 53 million years ago (Retallack, 1982).
In considering the diversification of life on land during Paleozoic time, the tab-
ulation of numbers of genera or families of terrestrial fossils already mentioned
are not readily comparable with soil types. The environmental and historical con-
straints of terrestrial ecosystems are such that their complement of species, genera
and families appears almost random (Bernado and Webb, 1977). Instead, emphasis
will be placed here on the overall structure of plant communities; that is to say, on
distinctive kinds of plant formations. These are not the same as rock formations
in geology, but rather refer to the general appearance and structure of vegetation.
Wooded grassland (or savanna in a very restricted sense) can appear quite similar
in South America, Africa and Australia, despite the very different plant and an-
imal species involved on each continent (Cole, 1985). It is plant formations that
are reflected best in the major features of soils (Foth and Schäfer, 1980), rather
than plant species or communities in a taxonomic sense. My nomenclature of plant
formations follows that developed in recent vegetation mapping of Africa (White,
1983), with a few additions (Retallack, 1990). The Paleozoic was a time of transition
from simple, visually unimpressive, microbial earths and rocklands, to many of the
plant formations found today, with a concomitant diversification of soils.
Entisols
These kinds of soils showing minimal profile development probably formed well
back in geological time. They must have existed as early stages in the develop-
ment of better developed paleosols now known as old as 3400 million years (Lowe,
1983). Fossilized Entisols are difficult to recognize in Precambrian rocks because
their degree of weathering is slight. Fossil root traces and burrows are needed to
distinguish ancient Entisols from associated sedimentary, metamorphic or igneous
rocks. In a few cases, some as old as 2300 million years, periglacial soil features such
as sand wedges can be used to recognize Precambrian Entisols (Cryorthents and
Cryopsamments: Williams, 1986).
Among the most ancient Entisols known to have supported multicellular land life
are burrowed red sandstones and siltstones, such as the Faust Flat silty clay paleosol
(a Fluvent) of Late Ordovician (Ashgillian) age in the Juniata Formation of central
Pennsylvania (Figure 21.1). These paleosols show no evidence of gleying, and pre-
sumably formed in well-drained parts of an extensive coastal floodplain (Feakes and
Paleozoic paleosols 545
Fig. 21.1. Outcrop of type Potters Mills clay (below upper arrow) and type Faust Flat silty clay (be
low lower arrow) paleosols in the Late Ordovician (Ashgillian) Juniata Formation near Potters Mills,
Pennsylvania (at 40°45'32"N and 77°37 / 04 // W on U.S. Geological Survey 7.5 minute Spring Mills Quad
rangle). Hammer handle is 25 cm.
Retallack, 1988). The conspicuous (1-21 mm diameter) and abundant burrows are
superficially similar to those of millipedes (Retallack and Feakes, 1987). There is
no evidence of fossil roots or rhizomes, but there are surficial reduction spots that
may have formed around buried organic matter. Any plants present were presum
ably non-vascular and perhaps liverwort-like (Retallack, 1985). Similar paleosols
are found in rocks of most succeeding geological periods (Retallack, 1986a). Their
plants probably were early successional stages in the colonization of bare ground.
Preservation of fossil plants in Entisols is much better where soil development
and plant decay have been inhibited by waterlogging, by extreme cold or by fire.
A very early example of a salt-marsh paleosol has been found in the Early Silurian
(Llandoverian) Frenchville Formation near Stockholm, Maine. It contains erect axes
546 GJ. Retallack
of the early land plant Eohostimella heathana (Schopf et al., 1966; Niklas, 1982).
The plant-bearing horizon is associated with marine fossils.
Sandy Entisols (Psamments) have been found among deposits of coastal dunes in
the Mid-Pennsylvanian (Desmoinesian, equivalent to Moscovian) Hermosa Forma-
tion at several localities in Utah, as well as in the Early Permian (Wolfcampian, or
Sakmarian) Cedar Mesa Formation of Utah and Fountain Formation of Wyoming
(Loope, 1988). The size distribution of calcareous root traces in these paleosols are
indications of desert shrublands and scrub.
Silty Entisols with shallow carbonaceous root traces (Cryofluvents) have been
found at several levels in earliest Permian (Asselian and Sakmarian) fluvioglacial
deposits in eastern New South Wales, Australia, which was at that time close to the
south pole. Some of these weakly developed paleosols, such as those near Lochin-
var, contain low-diversity assemblages of leaves of Botrychiopsis, which may have
formed a kind of polar tundra (Retallack, 1980). Other paleosols at slightly higher
stratigraphie levels in this postglacial sequence have a more diverse assemblage
of leaves dominated by Gangamopteris, in an ancient vegetation that was an early
version of taiga. The similar alpine plant formations of fellfield and krummholz
also may date back to late Paleozoic time, but no evidence of them has yet been
discovered.
Some shallow stony soils (Orthents) support fireprone shrublands, variously
known as chaparral (U.S.A.), maquis (Europe) and matorral (Chile). Such soils
and vegetation are indicated by dolines in paleokarst of latest Tïïassic or early Juras-
sic (Rhaeto-Liassic) agefilledwith fossil charcoal (Harris, 1957). Fossil charcoal has
been found as far back in the rock record as earliest Mississippian (Tburnaisian:
Cope and Chaloner, 1985), but the association of charcoal with bedrock paleosols
has not yet been established so far back in time.
Inceptisols
These soils of somewhat stronger development than Entisols also can be difficult
to recognize in Precambrian rocks. Surprisingly, many Precambrian paleosols that
are exceptionally thick and well developed often turn out to be best classified as
Inceptisols, mainly because they lack the diagnostic properties of other modern soil
orders (Retallack, 1986a). As more is learned of Precambrian paleosols, some of
these profiles probably will become recognized as extinct orders.
The most ancient Inceptisols with remains of multicellular life yet recorded are
of Late Ordovician age in Pennsylvania and Nova Scotia. The Potters Mills clay pa-
leosol in central Pennsylvania is a common kind of profile in fluvial deposits of the
Late Ordovician (Ashgillian) Juniata Formation (Figure 21.1). These profiles are
distinguished by a surficial network of conspicuous (1-15 mm diameter) burrows
and a subsurface horizon of comparably-sized calerete nodules (Feakes and Retal-
lack, 1988). The nodules truncate and are truncated by the burrow walls indicating
that the burrows are part of the original soil. From their geological setting these
paleosols appear to have formed on well drained terraces of an extensive piedmont,
flanking a major mountain range of folded sedimentary and metamorphic rocks.
The nature of these paleosols, together with information from marine faunas of
equivalent age, are evidence of a subhumid, perhaps seasonally dry, tropical cli
mate. As in comparable modern soils, the paleosols show evidence of calcification,
ferruginization and clay formation. The fossil burrows in these Ordovician paleosols
provide some evidence of large land animals, probably millipede-like arthropods
(Retallack and Feakes, 1987). There is no clear evidence of the likely vegetation of
these soils, such as traces of roots, rhizomes or rhizoids. If this negative evidence can
be used to exclude the possibility of vascular land plants, the paleosols could have
supported cyanobacteria and other microbes, or even forms of a comparable grade
of evolution to lichens and liverworts. For the following reasons, my own opinion is
that they supported thallose multicellular plants. The fecal pellets in burrows and
the abundance of the burrows indicate a productive ecosystem that included a con
siderable biomass of solid food. Furthermore, trace elements normally associated
with organic matter in soils (such as Li, Nb, Ni, P, Sr and Y) show a pattern of
surficial depletion that is more like that found in soils with a surface cover of plants
than those with an interstitial fauna or a richly humified surface layer.
Looking for evidence of this ancient vegetation further afield, land plant spores
of a kind comparable to those of living liverworts are widespread in Late Ordovi
cian rocks (Gray, 1985), well before spores of vascular land plants and megafossil
remains of land plants appear in geologically younger (Silurian) rocks. There also
is another paleosol of Late Ordovician (Caradocian) age developed on an andésite
lava flow in Nova Scotia, which has reduction spots in mounds of paleosol between
erosion scours filled with redeposited material (Figure 21.2). These features have
been interpreted (Dewey in Boucot et al., 1974) as evidence for clumps of early land
plants which locally stabilized the soil against erosion, and whose remnant organic
matter decayed anaerobically to create the green reduction spots. Lichens and liver
worts are common in very dry sites today: on tree trunks, on cliffs and among alpine
boulder fields. Extensive cover of such thallose vegetation in well-drained parts of
lowland fluvial settings, as envisaged here for Late Ordovician times, is not found
today. This distinctive archaic kind of plant formation can be called a polsterland
(Retallack, 1990).
The advent of vascular land plants in well-drained soils also may be recorded
in Paleozoic paleosols. Late Silurian (Ludlovian) paleosols of the Bloomsburg For
mation near Palmerton, Pennsylvania, are in many ways similar to those of Late
Ordovician age already described from Pennsylvania. These Silurian paleosols also
formed on quartzo-feldspathic floodplains flanking a major mountain range in a
subhumid tropical climate. Unlike the Ordovician paleosols, the Silurian ones show
irregular surficial bioturbation that is similar to what one would expect for rhizomes
of land plants (Retallack, 1985). These traces are indistinct and have not been stud-
548 G.J. Retaüack
J scale in meters
^^Ë red claystone red shale gray claystone
calcareous nodules andésite corestone
Fig. 21.2. Field sketch of surficial reduction spots and scour-and-fill structure, and of subsurface calerete
nodules and corestones in a Late Ordovician paleosol developed on an andésite flow near Arisaig, Nova
Scotia (from Boucot et al., 1974).
ied in such detail or numbers to be certain of this interpretation. Nevertheless, they

are a promising new line of evidence in the quest for understanding early vegetation
on dry land. Colonization of the land by vascular plants in Silurian time also is in
evidence from globally widespread and diverse fossil spores, sporangia and stems
of vascular land plants of this geological age (Gray, 1985; Gensel and Andrews,
1984). Such vegetation can be imagined as comparable to fields of bracken fern
(Pteridium aquilinum) that spring up after forest fires, or to patches of clubmosses
(Lycopodium) in alpine meadows. This distinctive early kind of vegetation can be
called brakeland (Retallack, 1990).
Polsterlands and brakelands of the early Paleozoic were supplanted by more
modern kinds of vegetation and soils during mid-Paleozoic time. The large woody
roots of trees leave clear and distinctive traces in soils, and a good record of the ad
vent of woodlands and forests can be expected from paleosols. There is paleobotan-
ical evidence for a variety of trees by Middle Devonian (Givetian) time (Banks et
al., 1985) and paleosols of this geological age confirm that these trees formed wood
lands and forests. The Peas Eddy clay paleosol in the Late Devonian (Frasnian)
Walton Formation near Hancock, New York, has been studied in some detail. The
large root traces penetrating this purple-brown siltstone have not entirely disrupted
relict bedding (Retallack, 1985). Although it was not a well developed paleosol,
there is evidence of slight clay illuviation and of incipient translocation of iron: both
indicators that it was an Inceptisol. This profile also formed in quartzo-feldspathic
alluvium of a major mountain range, but it is non-calcareous, unlike associated Late

Devonian paleosols of former floodplains away from streams. Early forest of this
weakly podzolized paleosol appears to have formed galleries along streams.
By Late Devonian time there are indications of considerable local variation in
vegetation. In the same Late Devonian (Frasnian) outcrops as the Peas Eddy clay
there are other fossilized Inceptisols. These have carbonized woody root traces in
a gray massive clay and a horizon of pyrite nodules. Such soils (Sulfaquepts) are
typical of intertidal woodland, or mangal. Exactly what kind of trees formed this
early mangal has not yet been investigated, but such reducing environments are very
favorable for preserving plant fossils. Slightly older (Middle Devonian or Givetian)
siltstones in New York containing marine fossils have within them large stumps
and rootlets of trees (Johnson, 1972). These stumps of Eospermatopteris erianium
have been considered progymnospermous, but may in fact be lycopods (Scheckler,
pers. commun., 1988). A variety of other kinds of plants formed mangal vegetation
long before the golden fern (Acrostichum aureum), palm (Nipa) and angiosperms
(Avicennia, and Rhizophora among others) that now are recognized as mangroves
(DiMichele et al., 1987).
Histosols
Peaty soils of permanently waterlogged ground have an excellent fossil record in

the form of coal seams and petrified peat, within which often are preserved fossils
of the plants that grew in them. There may be evidence of freshwater marshes as
early as Late Silurian (Ludlovian) if local accumulations of early land plants in the
Bloomsburg Formation of Pennsylvania (Willard, 1938; Strother and Traverse, 1979)
prove to be sufficiently thick and in place of growth. Freshwater marsh certainly
existed by Early Devonian (Seigenian) time. The Rhynie Chert of this age in Scot-
land is a well known permineralized peat preserving in exquisite detail a variety
of herbaceous early vascular {Rhynia, Horneophyton, Asteroxylon) and non-vascular
(Aglaophyton, Lyonophyton) plants (Kidston and Lang, 1921; Edwards and Edwards,
1986). Of about the same geological age near Barzass, in central Siberian U.S.S.R.,
there are thick (4 m) cuticle coals of the early land plant Orestovia devonica (Stach
et al., 1975; Krassilov, 1981).
The appearance of trees in waterlogged peaty substrates may have occurred
somewhat later in geological time. The oldest woody coal of a thickness that would
qualify as a Histosol after considering its compaction due to burial is of latest De-
vonian (Famennian) age in West Virginia (Gillespie et al, 1981). It is supported
salt-marsh of the zygopterid fern Rhacophyton ceratangium. Other pyritic nodules
with remains of tree cordaites in coals of Late Carboniferous age appear to rep-
resent ancient mangal vegetation (Raymond and Phillips, 1983). Woody coals are
abundant and widespread in Carboniferous, Permian, Cretaceous and early Tertiary
rocks (DiMichele et al., 1987).
550 GJ. RetaUack
Most coals appear to have formed in acidic swamps and marshes. However, there
are in the United States and western Europe, many Carboniferous coals containing
calcite or dolomite nodules that may have formed during accumulation of the peat.
Various lines of evidence can be marshalled to support the idea that these were
soil nodules rather than clasts eroded from other rocks or phenomena of alteration
of the peat after its burial (Scott and Rex, 1985). The most compelling evidence
is the exquisite cellular preservation of uncrushed fleshy fossil plant tissues, even
at the base of coal seams where plant debris in modern comparably thick peats
is compacted. These ancient wooded wetlands were thus chemically neutral to al-
kaline. Thus, vegetation of coal balls with tree ferns and tree lycopods should be
called carr, rather than swamp. Other coal balls containing only small ferns (Rex
and Scott, 1987) represent a vegetation most like modern fens. Alkaline wetlands
with fen and carr vegetation are widespread today, but only a few show evidence of
carbonate permineralization of plant tissues, and in no known case is preservation
as exquisite as in Carboniferous coal balls (Retallack, 1986b). Subhumid climate
and nearby limestone bedrock promote fens in wetlands, but it also could be that
phenolic compounds and other acidifying agents of Paleozoic plants were less effec-
tive in controlling soil pH than are those of many modern plants. Certainly the tree
lycopods (Lepidodendron) and horsetails (Calamités), so well represented in Pennys-
lvanian coal balls, were very different plants from the woody seed ferns and conifers
of Permian and geologically younger coals.
Aridisols
Shallow, stony, salty and calcareous soils, like those now forming in deserts,
have a long fossil record. The earliest calcrete-bearing paleosols are some 1800
million years old in the Canadian Northwest Territories, and other examples provide
a fossil record of these soils reaching to the present (Retallack, 1986a). When
these soils began to support different kinds of vegetation is a project amenable
to study, considering the common preservation of root traces in modern calcretes
(Klappa, 1980). A comprehensive study of this kind has not yet been done. Many
of the great array of calcareous paleosols from the latest Silurian (Pridolian) to
earliest Carboniferous (Tournaisian) Old Red Sandstone of Britain, probably were
Aridisols. Some contain fossil root traces and burrows (Allen, 1986). More detailed
studies have been reported of Aridisols from South Wales of Early Carboniferous
(Visean) age (Wright, 1987). By this geological time such soils certainly had been
invaded by plants of moderate stature. Root traces in the Heatherslade Geosol at
Three Cliffs Bay and Miskin reach a diameter of 3 mm (Wright, 1986). Associated
with these root traces are needles of calcite of a kind identical to those now formed
by fungi associated with roots. Aridisols containing large woody root traces are
known in Early Carboniferous (Visean) limestones of Kentucky (Ettensohn et al.,
1988), and Tttassic sandstones of Tfexas (Blodgett, 1988), Connecticut (Hubert, 1977)
and West Germany (Martins and Pfefferkorn, 1988). Fossil Aridisols in the latest
Jurassic (Tithonian) limestones of Dorset in England, are notable for fossil stumps
and root traces as well as plant fossils of a dry conifer-bennettite woodland (Francis,
1986).
Vertisols
Swelling clay soils also are of great antiquity. The oldest example currently known
is a paleosol some 2200 million years old in a deep roadcut near the town of
Waterval Onder, South Africa (Retallack, 1986c). In this large exposure can be seen
the swalelike thickening and thinning of its surface horizon and deep cracking of
the ridges (mukkara structure) that is a characteristic subsurface expression of gilgai
microrelief. Large exposures are needed to confidently identify fossil Vertisols, so
that paleosols more ancient than the Waterval Onder profile but known only from
drill cores (Grandstaff et al., 1986) may yet turn out to be fossil Vertisols.
When Vertisols began to support plants remains an open question. Probably this
occurred at least by Devonian time. Fossil Vertisols with tubular features are known
among Devonian paleosols of Britain (Allen, 1986) and Antarctica (McPherson,
1979). Detailed studies of their roots or burrows remain to be reported. It is con-
ceivable that they supported shrublands, wooded shrublands or dry woodland. Open
grasslands and wooded grasslands, so widespread on modern Vertisols, probably
were a much later addition to the array of vegetation on Vertisols. Grasses have
been found fossilized no older than Cretaceous, and the main kind of grassland soils
(Mollisol) are known no older than Eocene (Retallack, 1982).
Alfisols
Base-rich soils with a clayey subsurface (Bt) horizon today are largely under
woodland and forest. These soils also seem to date back to about the time of the
earliest forests. The earliest example known to me is in the uppermost Old Red
Sandstone on the rock platform north of Pease Bay, southeastern Scotland (Figure
21.3). This alluvial sequence is of Late Devonian (Famennian) age, judging from
associated fossil fish of Holoptychius nobilissimus (Craig, 1975). The paleosol profile
has a relict shale bed, much disrupted by burrows and root traces. This clayey part
of the profile is not an argillic horizon, but the down-profile increase in clay and clay
skins in the sandy overlying layer does appear to be pedogenic. The paleosol also has
carbonate nodules, in this case dolomitic and mainly in that part of the profile below
the clayey layer. In pétrographie thin sections the nodules consist of rhombohedral
dolomite (Figure 21.4), a crystal habit often taken as evidence of dolomitization of
552 GJ. Retaüack
Fig. 21.3. The type Pease Bay loamy sand paleosol beneath cross-bedded sandstone in the Late Devonian
(Fammennian) Old Red Sandstone on the rock platform north of Pease Bay, Scotland (British National
Grid NT802713). Sledge hammer handle on Bt horizon is 23 cm.
Fig. 21.4. Photomicrograph, under crossed niçois, of recrystallized dolomite (rhombs) in small calerete
nodule of the type Pease Bay loam sand paleosol in the Late Devonian (Famennian) Old Red Sandstone
near Pease Bay, Scotland. Scale bar is 1 mm.
other carbonates after burial (Zenger and Mazzullo, 1982). Dolomite is found in
very few modern soils (Cline, 1955), but is common in paleosols of Devonian age
and older (Retallack, 1986a). The nodules themselves are original features, judging
from their avoidance by some of the burrows and root traces. Whether they were
originally dolomite or calcite, the nodules indicate that this was a soil of dry climate
and high base status, a conclusion compatible also with the abundance of plagioclase
and other easily weatherable minerals in the profile. The depth to the carbonate
nodules (Bk horizon) is more like that of Alfisols than Aridisols, and this depth
would have been even greater before compaction during burial. The upper part of
the paleosol contains irregular ferruginized root traces up to 6 cm in diameter and
burrows up to 2 cm. Its vegetation probably was dry woodland. Such vegetation may
have been widespread in quartzo-feldspathic alluvium of the "Old Red Continent",
an ancient land mass that included what is now Britain, western Europe, Greenland
and North America.
Fossil Alfisols are common in non-marine rocks of younger geological periods
(Retallack, 1986a). In many cases they also preserve fossil evidence of their ecosys-
tems. Gleyed Alfisols may contain remains of stumps and leaf litter. Well drained
fossil Alfisols, on the other hand, commonly preserve fossil bones, teeth, snails and
phytoliths of plants (Retallack, 1984). Fossiliferous paleosols of these kinds show
special promise for unravelling the coordinated evolution of life and landscapes
(Retallack, 1988).
Spodosols
Quartzose, sandy, acidic soils may have a very distinctive subsurface spodic (Bs)
horizon in which the sand grains are coated with iron and humus compounds,
usually with deep radial cracking (de Coninck et al., 1974). Such characteristic
horizons of Spodosols are thought to form preferentially on initially siliceous parent
materials by the leaching effects of chemicals such as phenols washed from leaves
by rain (Fisher and Yam, 1984). The oldest currently known Spodosols are Eocene
in age (Pomerol, 1964), but the process of their formation (podzolization) was
clearly initiated during Paleozoic time. Incipient podzolization can be seen in the
Peas Eddy clay paleosol of Late Devonian age (Retallack, 1985). More advanced
podzolization is found in the Rogerly sandy clay loam paleosol (Figures 21.5 and
21.6) of mid-Carboniferous (Namurian) age in a quarry high on the hill southwest of
Stanhope, England (Percival, 1986). This profile is an Inceptisol (Dystrochrept more
precisely), but is so close to qualifying as a Spodosol that discovery of Spodosols can
be expected in Carboniferous rocks eventually.
The Rogerly sandy clay loam is a thick (at least 2 m) profile. Its surficial coal (O
horizon) may reflect more waterlogged conditions on burial of the paleosol. There
is no evidence of waterlogging during its formation. The bleached (E) horizon is
554 GJ. RetaUack
other carbonates after burial (Zenger and Mazzullo, 1982). Dolomite is found in
very few modern soils (Cline, 1955), but is common in paleosols of Devonian age
and older (RetaUack, 1986a). The nodules themselves are original features, judging
from their avoidance by some of the burrows and root traces. Whether they were
originally dolomite or calcite, the nodules indicate that this was a soil of dry climate
and high base status, a conclusion compatible also with the abundance of plagioclase
and other easily weatherable minerals in the profile. The depth to the carbonate
nodules (Bk horizon) is more like that of Alfisols than Aridisols, and this depth
would have been even greater before compaction during burial. The upper part of
the paleosol contains irregular ferruginized root traces up to 6 cm in diameter and
burrows up to 2 cm. Its vegetation probably was dry woodland. Such vegetation may
have been widespread in quartzo-feldspathic alluvium of the "Old Red Continent",
an ancient land mass that included what is now Britain, western Europe, Greenland
and North America.
Fossil Alfisols are common in non-marine rocks of younger geological periods
(RetaUack, 1986a). In many cases they also preserve fossil evidence of their ecosys-
tems. Gleyed Alfisols may contain remains of stumps and leaf litter. Well drained
fossil Alfisols, on the other hand, commonly preserve fossil bones, teeth, snails and
phytoliths of plants (RetaUack, 1984). Fossiliferous paleosols of these kinds show
special promise for unravelling the coordinated evolution of life and landscapes
(RetaUack, 1988).
Spodosols
Quartzose, sandy, acidic soils may have a very distinctive subsurface spodic (Bs)
horizon in which the sand grains are coated with iron and humus compounds,
usually with deep radial cracking (de Coninck et al., 1974). Such characteristic
horizons of Spodosols are thought to form preferentially on initially siliceous parent
materials by the leaching effects of chemicals such as phenols washed from leaves
by rain (Fisher and Yam, 1984). The oldest currently known Spodosols are Eocene
in age (Pomerol, 1964), but the process of their formation (podzolization) was
clearly initiated during Paleozoic time. Incipient podzolization can be seen in the
Peas Eddy clay paleosol of Late Devonian age (RetaUack, 1985). More advanced
podzolization is found in the Rogerly sandy clay loam paleosol (Figures 21.5 and
21.6) of mid-Carboniferous (Namurian) age in a quarry high on the hill southwest of
Stanhope, England (Percival, 1986). This profile is an Inceptisol (Dystrochrept more
precisely), but is so close to qualifying as a Spodosol that discovery of Spodosols can
be expected in Carboniferous rocks eventually.
The Rogerly sandy clay loam is a thick (at least 2 m) profile. Its surficial coal (O
horizon) may reflect more waterlogged conditions on burial of the paleosol. There
is no evidence of waterlogging during its formation. The bleached (E) horizon is
Paleozoic paleosols
Fig. 21.5. The type Rogerly sandy clay loam paleosol of the mid-Carboniferous (Namurian) Firestone Sill
Sandstone exposed in the upper quarry east of Stanhope, England (British National Grid NZ0122383).
Hammer handle in A horizon is 25 cm.
thick over an incipient spodic (Bs) horizon (Figure 21.5) and stigmarian rootlets
penetrate very deeply into the subsurface. The paleosol appears to have formed on
sands of the levee of a large deltaic distributary that later became a swamp and ul-
timately a marine bay, as do modern delta lobes of the Mississippi Delta (Coleman,
1988). Unlike the conifer forest of modern boreal Spodosols or the angiospermous
dwarf forest of modern "giant podzols" in tropical regions (Richards, 1952), this
paleosol supported a forest of tree lycopods whose roots (Stigmariaficoides)were
quite stout (up to 5 cm in diameter). Could this difference in vegetation explain
the weak mobilization of iron and humus in this mid-Carboniferous Inceptisol com-
pared to modern Spodosols? Weak development also could be due to a short time
of formation, but this does not tally with the depth of leaching and abundance of
root traces in this paleosol which would be unusual for a weakly developed profile.
556 GJ. Retaüack
Fig. 21.6. Pétrographie thin section (crossed niçois) of thin iron and humus coatings on angular grains,
since overgrown with silica during burial diagenesis. From the subsurface (Bs) horizon of the type
Rogerly sandy clay loam paleosol, of the mid-Carboniferous (Namurian) Firestone Sill Sandstone, near
Stanhope, England. Scale bar is 0.1 mm.
A relationship between degree of expression of spodic horizons and plant evolution

cannot yet be proven, but has some theoretical appeal. Phenolic substances, often
implicated in podzolization, are widespread in plants as biochemical byproducts of
wood production that also are effective as mild insecticides (Swain and Cooper-
Driver, 1981). It is conceivable that phenolic concentrations and effectiveness have
increased in conifers and angiosperms over hundreds of millions of years of evo-
lution in the face of the ever present threat of insect attack. The fossil record of
Spodosols, and related soils, may offer clues to this loosely coordinated evolution of
insects, plants and soils.
Ultisols
Base-poor soils with clayey subsurface (Bt) horizons present problems of low
nutrients for plants, similar to Spodosols. Ultisols also may date back to Carbonifer-
ous time (Retallack, 1986a), although none have been reported older than Tïiassic
(Lucas, 1976) and well documented examples are no older than Eocene (Retallack,
1983). The exact timing of the afforestation of such soils is of some interest. It could
be argued, as has Weaver (1967) for example, that the stabilizing influence of forests
was necessary in order to promote such deeply weathered and differentiated pale-
osols. On the other hand, forests could have been established geologically early on
nutrient-rich lowland soils, but only later developed microbially mediated nutrient
procurement systems, such as mycorrhizae, that allowed them to grow in nutrient
poor upland soils. Current information remains insufficient to decide this and re-
lated questions, but study of the trace fossils of fossil Ultisols and related soils can
now be seen as a way to address such formerly intractable problems of historical
geology.
Oxisols
These uniform deeply weathered soils of ancient stable land surfaces are surficial
equivalents of many paleosols along major geological unconformities. Such deeply
weathered soils appear to be geologically ancient. High grade metamorphic rocks
at least as old as 2600 million years have the highly aluminous and ferruginous
composition of Oxisols (Dash et al., 1987). The oldest little metamorphosed profiles
showing such oxidation and deep weathering are some 1660 million years old in the
Thelon and Athabaska Basins of north central Canada (Ramaekers, 1981; Chiaren-
zelli et al., 1983; Ttemblay, 1983; Holland, 1984). Oxisols commonly are associated
with laterite and bauxite. These also are geologically ancient. Banded iron forma-
tions have been found to have been enriched to laterite as long ago as 2070 million
years in South Africa and 1800 million years in Western Australia (Morris, 1985).
Highly aluminous metamorphic rocks as old as 3500 million years in the Aldan
Shield region of Siberian U.S.S.R. (Salop, 1983) have been interpreted as former
bauxites (Serdyuchenko, 1968). Similar explanations were advanced for corundum
ores some 2700 million years old in Zimbabwe and in Western Australia, but both
have been interpreted later as hydrothermally altered rocks around volcanic hot
springs (Schreyer et al., 1981; Martyn and Johnston, 1986). More promising candi-
dates for ancient bauxites are rocks rich in pyrophyllite and kyanite within the 3000
million year old Nsuze Group of the Barberton Mountain Land of South Africa
(Reimer, 1986). Bauxites have been found in rocks of most geological periods up to
the present (Bardossy, 1982).
When these ancient kinds of soils became colonized with plants and animals,
and what effects they had on them, remain open questions. Latentes of Sri Lanka,
probably dating to Late Cretaceous or Early Tertiary time have extensive channeling
attributed to roots and burrowing animals (Vermaat and Bentley, 1955). Such traces
need to be sought in older bauxites, latérites and Oxisols.
Both life and soils on these unusually stable cratonic landscapes have changed
considerably over geological time. Although many Precambrian paleosols at un-
conformities are impressive in their thickness and degree of weathering, they pale
against the extreme weathering and horizon differentiation extending to many tens
of meters in lateritic and bauxitic profiles of Tertiary and Mesozoic age (Senior and
558 G.l Retaüack
Mabbutt, 1979). In humid regions some Oxisols now support tropical rainforest in
which nutrient difficulties of the soil are circumvented by efficient recycling in a shal-
low root mat (Sanford, 1987). Many deeply weathered Oxisols and latérites formed
under forest now are exposed in regions where they support open grassy woodlands,
such as the miombo of East Africa (Cole, 1986), or desert vegetation, such as mulga
of Australia (Beadle, 1981). These are soils that form over long periods of time.
Many of them have had complex histories.
Coordinated evolution of soil and life
Although studies pertinent to the evolution of soil types during Paleozoic time
are few, it seems clear that this was a time of diversification of soils hand in hand
with the evolution of new kinds of vegetation (Figure 21.7). Other factors could
have been involved in this diversification of soil types, but probably did not have as
direct an influence. Continental geography and parent materials were by and large
as varied during Paleozoic time as they are today. So also was the tectonic stability
of landscapes, ranging from stable ancient cratonic land surfaces to young alluvial
plains and volcanic mountains (Zeigler et al, 1979). In addition, the oxygénation of
the atmosphere, that left a record in the changing nature of Precambrian paleosols,
appears to have been largely completed by early Paleozoic time (Feakes et al., 1989).
green clay
bauxite
karst
laterite
calerete
silcrete
ENTISOL
INCEPTISOL
VERTISOL
ARIDISOL
OX I SOL
HISTOSOL
ALFISOL
ULTISOL
SPODOSOL
MOLLISOL
PRECAMBRIAN Mollici UH
3 2 I
thousands of millions of years before present
Fig. 21.7. Geological time ranges of U.S.D.A. soil orders and other weathering features. Green clay is
an informal name for certain Precambrian paleosols. (Retallack, 1986a.)
Paleozoic paleosob 559
These other variables in soil formation do not account for the observed diversifica
tion of Paleozoic soil types as well as the diversification of plant formations.
Paleozoic paleosols thus offer a unique view of major events in the progres
sive cover of the landscape by different kinds of vegetation (Retallack, 1985). Not
enough is yet known to decide whether plants and animals invaded the land in suc
cessive waves from lakes and oceans (Knoll et al, 1984) or whether they evolved
on land from pre-existing soil microbes (Stebbins and Hill, 1980). There are, how
ever, hints that terrestrial ecosystems were built up by degrees: from microbial
earths to polsterlands, brakelands, shrublands and woodlands. There also are indi
cations that nutrient-rich soils, with carbonates and with easily weathered minerals
such as plagioclase, were colonized earlier than siliceous sands or deeply weath
ered clay. Well-drained soils were occupied by polsterlands surprisingly early (late
Ordovician). Marine-influenced marsh also is quite ancient (Early Silurian). Woody
vegetation appeared almost simultaneously (geologically speaking, during the mid-
Devonian) in marine influenced, waterlogged and well-drained soils. Afforestation
of nutrient-poor peats, silica sands, kaolin clays, latérites and bauxites are less well
constrained, but may have been somewhat later (latest Devonian to TMassic). Such a
stepwise development of different vegetation types is understandable from ecologi
cal structure and succession of ecosystems. Microbial nutrient procurement systems,
especially for phosphorus, potassium and nitrogen, probably were necessary before
larger plants could thrive. A landscape stabilized by herbaceous vegetation may have
been necessary before trees could grow to any stature.
The advents of these various plant formations were major biological events and
were also crucial to weathering regimes. Forests, for example, create their own mi
croclimate in which diurnal and seasonal variation in humidity and temperature are
dampened to more equable levels. Forests also persist in the landscape by resisting
the perturbing effects of floods, and forests also maintain global atmospheric oxida
tion levels. They are the most significant producers of oxygen, both as a by-product
of photosynthesis and by helping to maintain and bury in swamps organic carbon
that would otherwise burn or decay to carbon dioxide. When oxygen levels increase,
forest fires rage. When carbon dioxide levels increase, theflammabilityof trees and
forest fires is suppressed (Cope and Chaloner, 1985). It was during the Paleozoic
that these environmental feedback systems became established. It was thus a period
of geological history when there was increasing biotic control of soil formation.
TWo equally extreme views on the role of life in surficial processes can be imag
ined, and both probably are equally incorrect. According to the Gaia hypothesis, life
has always, and will continue, to exert a homeostatic influence on surficial processes
that serves to promote its own persistence (Lovelock, 1979). This is an optimistic
and comforting view, though somewhat unscientifically, almost mystically, holist. An
equally extreme alternative suggested mainly by earth scientists (Holland, 1984) is
that life is a trivial epiphenomenon that adapts as best it can through evolution to
wider environmental forces, such as the differentiation of the Earth's interior and its
560 G.l Retallack
surface expression as volcanoes and earthquakes. This pessimistic view of life's role
on Earth can be characterized as Ereban, after the primeval god of darkness (Figure
21.8; Retallack, 1990). These extreme views are not much more than a deification of
an old debate in ecology (O'Neill et al., 1986). Are communities in any sense like
superorganisms or are they simply random associations of species that happen to
have similar environmental requirements? The answer is probably somewhere be
tween these extremes. This can be seen from the example of forests, whose specific
composition varies widely according to soil type and regional climatic history. On
the other hand, forests are environmentally regulated habitats for other creatures
and play a role as regulators of global atmospheric composition.
The question now before us is not whether life has absolute or no control over
surficial processes, but rather how much? A good answer to this question is still
a long way off, but some approaches now can be seen. One way is to evaluate
the fluxes of volatile elements in modern ecosystems. Such work is showing that
life does not have a monopoly on surficial processes over physical forces such as
volcanoes, but life's influence is still far from negligible (Mooney et al., 1987).
Another way to assess fife's role is by historical evaluation of biotic influences in
soil formation through geological time. As outlined here, there does appear to have
been a general trend of increasing biotic influence in Paleozoic paleosols. Life does
not seem to have rapidly assumed a hegemony of Earth resources soon after it
evolved (an extreme Gaian view; Figure 21.8). Nor does it seem likely that life
has always been at the whim of every volcanic eruption, change in sea level or
major meteorite impact (an extreme Ereban view). Additional studies of Paleozoic
paleosols, controlled for soil forming factors other than biotic ones, show some
promise for probing the intricate interrelationship between life and soil.
Thousands of millions of years before present
Fig. 21.8. A conceptual model for the historical degree of control by extreme Gaian and Ereban views,
together with a range of more reasonable intermediate views.
References
Allen, J.R.L., 1986. Pedogenic carbonates in the Old Red Sandstone fades (Late Silurian-Early Car
boniferous) of the Anglo-Welsh area, southern Britain. In: P.V. Wright (Editor), Paleosols; Their
Banks, H.P., Grierson, J.D. and Bonamo, P.M., 1985. The flora of the Catskill clastic wedge. In: D.L.
Woodrow and WD. Sevon (Editors), The Catskill Delta. Geol. Soc. Am., Spec. Pap., 201:125-141.
Bardossy, G., 1982. Karst Bauxites. Developments in Economic Geology 14. Elsevier, Amsterdam, 441
pp.
Beadle, N.C.W., 1981. The Vegetation of Australia. Cambridge University Press, London, 690 pp.
Benton, M.J., 1985. Mass extinction among non-marine tetrapode. Nature, 316: 811-814.
Bernado, J.C. and Webb, T, 1977. Changing patterns in the Holocene pollen record from northeastern
North America: a mapped summary. Quat. Res., 8: 64-96.
Blodgett, R.H., 1988. Calcareous paleosols in the Triassic Dolores Formation, southwestern Colorado.
In: J. Reinhardt and W.R. Sigleo (Editors), Paleosols and Weathering through Geological Time:
Principles and Applications. Geol. Soc. Am., Spec. Pap., 216: 103-121.
Boucot, A.J., Dewey, J.E, Dineley, D.L., Fletcher, R., Fyson, W.K., Griffin, J.G., Hickox, D.F., McK-
errow, W.S. and Zeigler, A.M., 1974. The Geology of the Arisaig Area, Antigonish County, Nova
Scotia. Geol. Soc. Am., Spec. Pap., 139,191 pp.
Chiarenzelli, J.R., Donaldson, J.A. and Best, M., 1983. Sedimentology and stratigraphy of the Thelon
Formation and the sub-Thelon regolith. In: Current Research, Part A. Geol. Surv. Can., Pap., 83-1A:
443_445.
Cline, M.G., 1955. Soil survey of the Territory of Hawaii. U.S. Dept. Agric. Soil Surv. Ser. 1939, No. 25,
644 pp.
Cole, M.M., 1986. The Savannas. Academic Press, London, 438 pp.
Coleman, J.M., 1988. Dynamic changes and processes in the Mississippi River delta. Geol. Soc. Am.
Bull., 100: 999-1015.
Cope, M.J. and Chaloner, WG., 1985. Wildfire: an interaction of biological and physical processes: In:
B.H. Tiffney (Editor), Geological Factors and the Evolution of Plants. Yale University Press, New
Haven, Conn., pp. 257-277.
Craig, G.Y., 1975. Siccar Point, Cove, Catcraig. In: G.Y. Craig and P. McL. D. Duff (Editors), The
Geology of the Lothians and Southeast Scotland. Scottish Academic Press, Edinburgh, pp. 107-117.
Dash, B., Sahu, K.N. and Bowes, D.R., 1987. Geochemistry and original nature of Precambrian khon-
dalites in the eastern Ghats, Orissa, India. Trans. R. Soc. Edinburgh, Earth Sci., 78: 115-127.
de Coninck, E, Righi, D., Maucorps, J. and Robin, A.M., 1974. Origin and micromorphological nomen
clature of organic matter in sandy spodosols. In: G.K. Rutherford (Editor), Soil Microscopy. Lime
stone Press, Kingston, Ont., pp. 263-280.
DiMichele, W.A., Phillips, T.L. and Olmsted, R.G., 1987. Opportunistic evolution: abiotic environmental
stress and the fossil record of plants. Rev. Palaeobot. Palynol., 50: 151-178.
Edwards, D. and Edwards, D.S., 1986. A reconsideration of the Rhyniophytina Banks. In: R.A. Spicer
and B.A. Thomas (Editors), Systematic and Taxonomic Approaches in Palaeobotany. Systematics
Assoc. Spec. Vol. 31, Oxford University Press, Oxford, pp. 201-222.
Ettensohn, F.R., Dever, G.R. and Grow, J.S., 1988. A paleosol interpretation for profiles exhibiting sub-
aerial exposure "crusts" from the Mississippian of the Appalachian Basin. In: J. Reinhardt and W.R.
Sigleo (Editors), Paleosols and Weathering Through Geologic Time: Principles and Applications.
Geol. Soc. Am., Spec. Pap., 216: 49-79.
Feakes, C.R. and Retallack, G.J., 1988. Recognition and chemical characterization of fossil soils devel
oped on alluvium: a late Ordovician example. In: J. Reinhardt and W.R. Sigleo (Editors), Paleosols
562 GJ. Retaüack
and Weathering Through Geologic Time: Principles and Applications. Geol. Soc. Am., Spec. Pap.,
216: 35-48.
Feakes, C.R., Holland, H.D. and Zbinden, E.A., 1989. Ordovician paleosols at Arisaig, Nova Scotia and
the evolution of the atmosphere. In: A. Bronger and J.A. Catt (Editors), Paleopedology. Catena
Supp., 16: 207-232.
Fisher, G.C. and Yam, O.-L, 1984. Iron mobilization by heathland plant extracts. Geoderma, 32: 339-
345.
Foth, H.D. and Schäfer, J.W., 1980. Soil Geography and Land Use. Wiley, New York, N.Y., 484 pp.
Francis, J.E., 1986. The calcareous paleosols of the basal Purbeck Formation (Upper Jurassic), south
ern England. In: P.V. Wright (Editor), Paleosols: Their Recognition and Interpretation. Blackwell,
Oxford, pp. 112-138.
Gensel, P.G. and Andrews, H.N., 1984. Devonian Paleobotany. Praeger, New York, N.Y., 381 pp.
Gillespie, WH., Rothwell, G.W. and Scheckler, S.E., 1981. The earliest seeds. Nature (London), 293:
462-464.
Grandstaif, D.E., Edelman, M.J., Foster, R.W., Zbinden, E.A. and Kimberley, M., 1986. Chemistry and
mineralogy of Precambrian paleosols at the base of the Dominion and Pongola Groups. Precambrian
Res., 32: 97-131.
Gray, J., 1985. The microfossil record of early land plants: advances in understanding of early terrestri-
alization, 1970-1984. Philos. Trans. R. Soc. London, Ser. B, 309: 167-195.
Harris, T.M., 1957. A Rhaeto-Liassic flora in South Wales. Proc. R. Soc. London, Ser. B, 147: 289-308.
Holland, H.D., 1984. The Chemical Evolution of the Atmosphere and Oceans. Princeton University
Hubert, J.F., 1977. Paleosol caliche in the New Haven Arkose, Newark Group, Connecticut. Palaeo-
geogr., Palaeoclimatol., Palaeoecol., 24: 151-168.
Johnson, K.G., 1972. Evidence for tidal origin of Late Devonian elastics in eastern New York State.
In: J.E. Gill (Editor), Proc. 24th Int. Geol. Congr., Montreal, Sect. 6, Harpells Press, Gardenvale,
Quebec, Que., pp. 285-293.
Kidston, R. and Lang, W.H., 1921. Old Red Sandstone plants showing structure from the Rhynie Chert.
Part V. The Thallophyta occurring in the peat bed: the succession of plants through a vertical
section of the bed and the condition of accumulation and preservation of the deposit. Trans. R. Soc.
Edinburgh, 52: 855-902.
Klappa, CF., 1980. Rhizoliths in terrestrial carbonates: classification, recognition, genesis and signifi
cance. Sedimentology, 27: 613-629.
Knoll, A.H., Niklas, K.T., Gensel, P.G. and Tiffney, B.H., 1984. Character diversification and patterns of
evolution in early vascular plants. Paleobiology, 10: 34-47.
Krassilov, V.A., 1981. Orestovia and the origin of vascular plants. Lethaia, 14: 235-250.
Loope, D.B., 1988. Rhizoliths in ancient eolianites. In: G. Kocurek (Editor), Late Paleozoic and early
Mesozoic Eolian Systems of the Western United States. Sediment. Geol., 56: 301-314.
Lovelock, J.E., 1979. Gaia: A New Look at Life on Earth. Oxford University Press, Oxford, 157 pp.
Lowe, D.R., 1983. Restricted shallow-water sedimentation of Early Archaean stromatolitic and evapor-
itic strata of the Strelley Pool chert, Pilbara Block, Western Australia. Precambrian Res., 19: 239-
283.
Lucas, C , 1976. Vestiges de paleosols dans le Permien et le Trias inférieur des Pyrenees et de l'Aquitaine.
C.R. Acad. Sci. Paris, 282: 1419-1422.
Martins, VP and Pfeiferkorn, H.W, 1988. Genetic interpretation of a lower Triassic paleosol complex
based on soil micromorphology. Palaeogeogr. Palaeoclimatol. Palaeoecol., 64: 1-14.
Martyn, J.E. and Johnson, G.I., 1986. Geological setting and origin of fuchsite-bearing rocks near
Menzies, Western Australia. Aust. J. Earth Sci., 33: 1-18.
McPherson, J.G., 1979. Calerete (caliche) palaeosols in fluvial redbeds of the Aztec Siltstone (Upper
Devonian), southern Victoria Land, Antarctica. Sediment. Geol., 22: 319-320.

Mooney, H.A., Vitousek, P.M. and Matson, P.A., 1987. Exchange of materials between terrestrial ecosys
tems and the atmosphere. Science, 238: 926-932.
Morris, R.C., 1985. Genesis of iron ore in banded-iron-formation by supergene and supergene-meta-
morphic processes. In: K.H. Wolff (Editor), Handbook of Stratabound and Stratiform Ore Deposits.
Vol. 13. Regional Studies and Specific Deposits. Elsevier, Amsterdam, pp. 73-235.
Niklas, KJ., 1982. Chemical diversification and evolution of plants as inferred from paleobiochemi-
cal studies. In: M.H. Nitecki (Editor), Biochemical Aspects of Evolutionary Biology. University of
Chicago Press, Chicago, 111., pp. 29-91.
Niklas, K.J., Tiffney, B.H. and Knoll, A.H., 1985. Patterns in vascular land plant diversification: a factor
analysis at the species level. In: J.W. Valentine (Editor), Phanerozoic Diversity Patterns: Profiles in
Macroevolution. Princeton University Press, Princeton, N.J., pp. 97-128.
O'Neill, R.V, de Angelis, D.L., Waide, J.B. and Allen, T.F.H., 1986. A Hierarchical Concept of Ecosys
tems. Princeton University Press, Princeton, N.J., 253 pp.
Percival, CJ., 1986. Paleosols containing an albic horizon: examples from the Upper Carboniferous
of northern England. In: V.P. Wright (Editor), Paleosols: Their Recognition and Interpretation.
Blackwell, Oxford, pp. 87-111.
Pomerol, C , 1964. Découverte de paleosols de type podzol au sommet de l'Auversien (Bartonien
inférieur) de Moisselles (Seine-et-Oise). C. R. Acad. Sci. Paris, 258: 974-976.
Ramaekers, P., 1981. Hudsonian and Helikian Basins of the Athabaska region, northern Saskatchewan.
In: EH.A. Campbell (Editor), Proterozoic Basins of Canada. Geol. Surv. Can., Pap., 81-10: 219-233.
Raymond, A. and Phillips, T.L., 1983. Evidence for an Upper Carboniferous mangrove community. In:
H.J. Teas (Editor), Tasks for Vegetation Science, Vol. 8. Junk, The Hague, pp. 19-30.
Reimer, TO., 1986. Alumina-rich rocks from the Early Precambrian of the Kaapvaal Craton as indica
tors of paleosols and as products of decompositional reactions. Precambrian Res., 32:155-179.
Retallack, GJ., 1980. Late Carboniferous to Middle Triassic megafossil floras from the Sydney Basin.
In: C. Herbert and R.J. Helby (Editors), A Guide to the Sydney Basin. Bull. Geol. Surv. N.S.W., 26:
383-430.
Retallack, G J., 1982. Paleopedological perspectives on the development of grasslands during the Ter
tiary. In: B. Mamet and M.J. Copland (Editor), Proc. 3rd North Am. Paleontol. Conv., Business and
Economic Services, Toronto, Ont., 2: 417-421.
Retallack, G.J., 1983. Late Eocene and Oligocene Paleosols from Badlands National Park, South
Dakota. Geol. Soc. Am., Spec. Pap., 193, 82 pp.
Retallack, G.J., 1984. Completeness of the rock and fossil record: some estimates using fossil soils.
Paleobiology, 10: 59-78.
Retallack, G J., 1985. Fossil soils as grounds for interpreting the advent of large plants and animals on
land. Philos. Trans. R. Soc. London, Ser. B, 309: 105-142.
Retallack, G.J., 1986a. The fossil record of soils. In: P.V. Wright (Editor), Paleosols: Their Recognition
and Interpretation. Blackwell, Oxford, pp. 1-57.
Retallack, G.J., 1986b. The Holocene Hitchcox limey peat soil as an analog for Pennsylvanian coals
bearing coal balls. 98th Annu. Meet. Geol. Soc. Am., Abstr. Progr., 16: 728.
Retallack, G.J., 1986c. Reappraisal of a 2200-Ma-old paleosol from near Waterval Onder, South Africa.
Precambrian Res., 32: 195-232.
Retallack, G J., 1988. Down to Earth approaches to vertebrate paleontology. Palaios, 3: 235-334.
Retallack, G.J., 1990. Soils of the Past. Unwin-Hyman, London, 520 pp.
Retallack, G.J. and Feakes, C.R., 1987. Trace fossil evidence for Late Ordovician animals on land.
Science, 235: 61-63.
Retallack, GJ. and Wright, VP, 1990. Micromorphology of lithified paleosols. In: L.A Douglas (Editor),
Soil Micromorphology, A. Basic and Applied Science. Elsevier, Amsterdam, pp. 641-652.
564 GJ. Retaüack
Rex, G.M. and Scott, A.C., 1987. The sedimentology, paleoecology and preservation of the Lower
Carboniferous plant deposit at Pettycur, Fife, Scotland. Geol. Mag., 124: 43-66.
Richards, P.W., 1952. The Tropical Rainforest. Cambridge University Press, Cambridge, 450 pp.
Salop, L.J., 1983. Geological Evolution of the Earth During the Precambrian. Springer, Berlin, 459 pp.
Sanford, R.L., 1987. Apogeotropic roots in an Amazon rain forest. Science, 235: 1061-1064.
Schopf, J.M., Mencher, E., Boucot, AJ. and Andrews, H.N., 1966. Erect plants in the early Silurian of
Maine. U.S. Geol. Surv., Prof. Pap., 550D: D69-D75.
Schreyer, W, Werding, G. and Abraham, K., 1981. Corundum-fuchsite rocks in greenstone belts of
southern Africa: petrology, geochemistry and possible origin. J. Petrol., 22: 191-231.
Scott, A.C. and Rex, G.M., 1985. The formation and significance of Carboniferous coal balls. Trans. R.
Soc. London, Ser. B, 311:123-1137.
Senior, B.R. and Mabbutt, J.A., 1979. A proposed method of defining deeply weathered rock units
based on regional geological mapping in southwest Queensland. J. Geol. Soc. Aust., 26: 237-254.
Sepkoski, J.J. and Hulver, M.L., 1985. An atlas of Phanerozoic clade diversity diagrams. In: J.W. Valen
tine (Editor), Phanerozoic Diversity Patterns: Profiles in Macroevolution. Princeton University Press,
Princeton, N.J., pp. 11-39.
Serdyuchenko, D.P., 1968. Metamorphosed weathering crust of the Precambrian, their metallogenic
and pétrographie fabric. In: M. Malkovsky and B. Hejtman (Editors), Proc. 13th Int. Geol. Congr.,
Prague, Sect. 4: 37-42.
Stach, E., Mackowsky, M.-Th. Teichmuller, M. and Teichmuller, R., 1975. Stach's Textbook of Coal
Petrology (translated by D.G. Murchison, G.H. Taylor and E Zierkie). Borntraeger, Berlin, 423 pp.
Stebbins, G.L. and Hill, G.J.C., 1980. Did multicellular plants invade the land? Am. Nat., 115: 342-353.
Strother, P.K. and Traverse, A., 1979. Plant microfossils from Llandoverian and Wenlockian rocks of
Pennsylvania. Palynology, 3: 1-21.
Swain, T and Cooper-Driver, G., 1981. Biochemical evolution in early land plants. In: K.J. Niklas
(Editor), Paleobotany, Paleoecology and Evolution. Praeger, New York, N.Y., pp. 103-134.
Tremblay, L.P., 1983. Some chemical aspects of the regolithic and hydrothermal alterations associated
with uranium mineralization in the Athabaska Basin, Saskatchewan. In: Current Research, Part A.
Geol. Surv. Can., Pap., 83-1A: 1-14.
Vermaat, J.G. and Bentley, CF., 1955. The age and channeling of Ceylon laterite. Soil Sci., 79: 239-247.
Weaver, CE., 1967. Potassium, illite and the oceans. Geochim. Cosmochim. Acta, 31: 281-295.
White, F, 1983. The Vegetation of Africa. UNESCO, Paris, 356 pp.
Willard, B., 1938. Evidence of Silurian land plants in Pennsylvania. Proc. Pa. Acad. Sci., 12: 121-124.
Williams, G.E., 1986. Precambrian permafrost horizons as indicators of palaeo-climate. Precambrian
Res., 32: 233-242.
Wright, V.P., 1986. The role of fungal biomineralization in the formation of Early Carboniferous soil
fabrics. Sedimentology, 33: 831-838.
Wright, VP, 1987. The ecology of two early Carboniferous paleosols. In: J. Miller, A.E. Adams and VP.
Wright (Editors), European Dinantian Environments. Geol. J., Spec. Issue, 12: 345-358.
Zenger, D.H. and Mazzullo, S.J. (Editors), 1982. Dolomitization. Hutchinson, Ross, Stroudsburg, Pa.,
426 pp.
Ziegler, A.M., Scotese, C.R., McKerrow, WS., Johnson, M.E. and Bambach, R.K., 1979. Paleozoic
paleogeography. Annu. Rev. Earth Planet. Sci., 7: 473-502.
567
Chapter 22
Classification of earth materials: a brief

examination of examples
W. CHESWORTH, G.A. SPIERS, LJ. EVANS and LP. MARTINI
Introduction
A classification is an abstract structure imposed on a collection of objects by the

human mind. If the classification is thought to reflect some intrinsic structure in the
collection of objects themselves, it is said to be a natural classification. If the order
imposed by the mind is not intrinsic to the objects, then the collection is said to be
artificial.
At its most general, a classification has the paramount purpose of describing
"the structure and relationships of the constituent objects, and to simplify these
relationships in such a way that general statements can be made about classes of
objects" (Sokal, 1974). In the case of a natural classification, this particular purpose
is pursued in order to discover order in nature. It is, in fact, the principal scientific
tool for doing so, and natural classification is taken here to be synonymous with
scientific classification.
By contrast, artificial classifications are ad hoc and serve such practical purposes
as to provide a convenient summary of observed facts, to provide systems for in-
formation storage and retrieval, as well as to provide in special cases, dedicated
systems to suit particular utilitarian needs. Some of these ends can, of course, be
met by natural classifications, but unlike the latter, an artificial classification con-
tains no heuristic content.
The procedure of classification involves the following steps:
(a) a population of objects (domain, universe of discourse) is defined;
(b) the classifier's purpose is stated;
(c) criteria (characteristics, properties) are selected to achieve that purpose;
(d) elements with like characteristics are grouped into classes;
(e) a degree of "likeness" is operationally defined; and
(f) names (labels, identities) are given to the classes.
A somewhat different sequence is given by Griffiths (1967).
568 W. Chesworth, G.A. Spiers, LJ. Evans and LE Martini
A classification should be comprehensive (capable of embracing all objects of the

domain classified) and the classes should be mutually exclusive (with a given object
assignable to one class only). For rocks and soils the second of these requirements is
never completely attained. In other words, the species concept, central to biological
classifications, is not applicable to rocks and soils. By their nature species are exclu-
sive groups whereas a population of rocks or soils is a smear or spectrum with no
precise boundaries dividing one rock or soil from another. Consequently, when the
spectrum is divided into segments, the boundaries between segments are more or
less arbitrary, so that demarcation disputes arise for those rocks or soils falling close
to a boundary.
Since the role of a natural, or scientific classification is to discover some underly-
ing order in nature, it is clear that the construction of a classification will lead to the
making and testing of hypotheses to explain whatever natural order is discovered.
Indeed, it must be a necessary first step. What types of hypotheses a classification
may lead to, will depend on the nature of the population. Biological classifications
for example, at least since "The Origin of Species" have played the role of pro-
voking hypotheses concerning the genetic relationships among organisms. Hennig
(1979) for example, emphasises this point when he says that systematics has to be
based on some all-inclusive explanatory principle, and that the only candidate in
biology is phylogeny. A similar generalization could be made concerning the classi-
fication of rocks, that is, the best all-inclusive explanatory principle for a population
of rocks is genetic.
Classification of earth materials
The set of earth materials of interest here is the mineral kingdom in its broadest
sense: minerals, rocks and soils. The domain is normally considered in terms of four
smaller domains, that is: minerals, igneous, sedimentary and metamorphic rocks,
and soils. It is possible to group some of these together, for example soils can be
classified with sedimentary rocks (Grout, 1932), but the most detailed classifications
so far, have arisen separately in the four domains.
The criteria chosen in the classification of earth materials are generally:
(a) intrinsic ones of an external type (for example, crystal-form in mineralogical
classifications, grain size in petrological ones, and horizonation in pedology);
(b) intrinsic ones of an internal type (for example, mineral structure, physico-
chemical constitution); or
(c) extrinsic (environmental) ones.
Classifications that look beyond surface appearances have depended and leaned
heavily on advances in more fundamental sciences. Chemical periodicity (itself a
triumph of the classifying instinct), X-ray crystallography, analytical chemistry, hy-
drodynamics and other branches of knowledge are all necessary components of
Classification of earth matenah: a brief examination of examples 569
the paradigm within which modern mineralogical, petrological or pedological clas

sifications are constructed. They provide us with criteria not available to earlier
classifiers.
For example, the simple classification of minerals given in Tkble 22.1 depends
directly on modern chemistry, with the further classification of the silicates (Table
22.2) depending also on early 20th century advances in X-ray diffraction. From
this extreme (that is, classifications based on criteria at the atomic or molecular,
constitutional level) we can pass to the grand scale of classification based on envi
ronmental characteristics, for example on criteria derived from the theory of plate
tectonics, again a source of inspiration not available to our forebears. At this com
prehensive scale of classification, great syntheses will become possible in the future.
For now we are still too close to the paradigm shift to see clearly where it is lead
ing. Probably, when the dust settles, plate tectonics will produce the same kind of
re-ordering of the material world of the earth sciences that Darwinian evolution did
for the organic world in the last century.
How elaborate a classification becomes, depends on how many criteria it is based
on. The simplest classification in igneous petrology for example is based on a single
criterion. Thus the division of igneous rocks into acid, intermediate, basic and ultra-
basic is based on Si0 2 content. More informative classifications are based on two
criteria: nature of feldspar, and saturation status with respect to silica. Such classi
fications may have a surprising heuristic content, because the criteria on which they
are based are chemical and mineralogical criteria that are sensitive to the physico-
chemical processes that give rise to igneous rocks. Furthermore, classifications of
this type are easy to use because they are easily represented on a two dimensional
page. However, there is no limit to the number of criteria used to delimit a class,
though if we proceed beyond three, we are entering the taxonomic equivalent of
TABLE 22.1
Chemical classification of mineral groups
1. Native elements
2. Sulfides
3. Sulfosalts
4. Oxides
(a) Simple and multiple
(b) Hydroxides
5. Halides
6. Carbonates
7. Nitrates
8. Borates
9. Phosphates
10. Sulfates
11. Tbngstates
12. Silicates
570 W. Chesworth, GA. Spiers, LJ. Evans and I.P. Martini
TABLE 22.2
Structural classification of silicate minerals

Geometry of linkage of SiO ratio Example Formula
S1O4 tetrahedra
Isolated tetrahedra: linked by cations 1:4 Olivine (Mg,Fe)2Si04
in octahedral coordination
Rings of tetrahedra: three-, four- 1:3 Beryl BeAl 2 (SÌ60i8)
or six-membered rings
Single chains: each tetrahedron 1:3 Pyroxene (Mg,Fe)SiC>3
linked to two others by shared
oxygens; chains bonded by cations
Double chains: two chains joined 4:11 Amphibole (Ca 2 Mg5)Si 8 0 2 2(OH)2
by shared oxygens as well as cations
Sheets: each tetrahedron linked to 2:5 Kaolinite Al 2 SÌ20 5 (OH)4
three others by shared oxygens;
sheets bonded by cations or
alumina sheets
Frameworks: each tetrahedron 3:8 Feldspar NaAlSi30g
shares all its oxygens with other (albite)
S1O4 tetrahedra (in quartz)
or AIO4 tetrahedra in feld- 1:2 Quartz Si02
spars and feldpathoids
hyperspace. In terms of matrix algebra and machine computation, there is no prob

lem in constructing classifications of this type (Perry, 1967), but the satisfaction of
having it represented pictorially is lost.
When it comes to setting the limits of a class, the ideal way would be to examine
the population statistically and to look for probability lows (in terms of discriminant
functions) in the frequency distribution of the objects in the domain. This is seldom
done so that setting class limits, which determine in Griffiths' (1967) word, the
amount of "likeness" tolerated within a class, is still largely an arbitrary procedure.
Arbitrariness on a grand scale was a feature of the petrological classifications of the
first half of the twentieth century and it can still be found, though in a subdued form
in modern classifications (compare, for instance, the definition of spodic horizon
used by the Americans with the equivalent used by the Canadians).
Naming classes is the last stage of classification, and although rock and mineral
terms display the vagaries of history, of geography and of the discoverer's identity,
a more systematic approach has been taken by modern pedologists [for example,
Soil Taxonomy and the FAO system, see FitzPatrick (1980) for a comprehensive
review].
In the next section, outline classifications are presented for each of the major
types of earth materials of interest to the readers of this book. In the space available
the choice cannot be comprehensive. It is offered simply to give cross-disciplinary
Classification of earth materials: a brief examination of examples 571
guidance to geologists and soil scientists who may not be conversant with each
others' terminology.
Classification of minerals
Criteria: 1. Major mineral groups are recognized on the basis of the presence (or
absence) of an anion or anionic group.
2. A given group can be subdivided on the basis of structure, for example
silicates are classed in terms of the presence of Si(Al)04 tetrahedra
and how they interconnect.
Table 22.1 shows the major mineral groups classed according to chemistry while
Table 22.2 shows the structural classification of silicates.
Classification of rocks
Criteria: 1. Tectonics.
2. Field relations or mode of occurrence.
3. Age.
4. Texture, structure and composition.
At the most fundamental level the last named are the most generally useful to the
earth scientist in the field or laboratory, and classification schemes based on them
are presented below.
Nomenclature of igneous rocks

Tectonic criteria: timing with respect to orogeny (prekinematic, synkinematic,
postkinematic); subduction zone versus oceanic ridge association.
Field relations or mode of occurrence: plutonic versus volcanic association; plu-
tonic types — katazonal, mesozonal, epizonal, volcanic types — shield volcanoes,
stratovolcanoes.
Age criteria: pétrographie and petrological provinces restricted in time and space.
Textural criteria: fine grained versus coarse grained — former considered vol-
canic, latter plutonic. Intermediate grain sizes originally considered hypabyssal, now
grouped in the volcanic association.
Compositional criteria: chemical composition, minera logical composition.
Figure 22.1 is the generalized (IUGS) classification of plutonic igneous rocks and
is based on the mode, that is, the actual, as opposed to the normative, mineralogy
of the rock. Since the mode may not always be determinable for a volcanic rock, the
alkali-silica diagram provides an alternative (Figure 22.2) for fine grained and glassy
rocks.
572 W. Chesworth, GA. Spiers, L.L Evans and IP. Martini
<60
2oy v20
A< VIII
10\
50
Px + Hbl 40
^60
Fig. 22.1. IUGS classification of plutonic igneous rocks (Murray, 1981). Q = quartz, A = alkali feldspar,
F = feldpathoid, Ol = olivine, Px = pyroxene, Hbl = hornblende. The fields: / = granitoids, / / =
syentoids, / / / = dioritoids, IV = gabbroids, V = foid syenitoids, VI = foid dioritoids and gabbroids,
VII = foidolites, VIII = anorthosites, IX = peridotites, X = pyroxenites, XI = hornblendites.
Na20 + K2O %
5 10
5 10
Fig. 22.2. Classification of volcanic igneous rocks: the alkali-silica diagram (Middlemost, 1985).
Classification of earth materìals: a brief examination of examples 573
The specific modal criteria used in the IUGS double triangle are: (a) presence
or absence of quartz, (b) presence or absence of a feldspathoid, and (c) the alkali
feldspar/plagioclase feldspar ratio. These have formed the basis of many simple
igneous rock classifications, and Figure 22.3 is a good example of these. This also
makes explicit the complementary character of the nature of any ferromagnesian
component with the nature of the feldspar.
Nomenclature of sediments and sedimentary rocks

Compositional criteria: relative importance of the three principal components:
terrigenous component (siliciclastic material), allochemical component (biochemi
cally produced materials) and chemical component (materials derived from chem
ical precipitation from an aqueous solution) (Table 22.3a). Examples of rocks pro
duced are, respectively: clastic rocks (conglomerate, sandstone, siltstone, shale; Ta
ble 22.3b; Figure 22.4a) and carbonate rocks [limestones (Table 22.3c; Figure 22.4b),
dolostones, coal (Figure 22.4c) and evaporites (gypsum and various salts and oth
ers; Table 22.4)]. These rocks are further subdivided on the basis of the mineralogy
100
FekJspar
50
Feldspathoids
Quartz
T T Ô¾"
Alkali granite Alkali syenite Feldspathoidal
Quartz syenites
Keratophyre
keratophyre
Phono lite
Granite Syenite E
Rhyolite co Trachyte
—e
Quartz monzonite co Monzonite
(0
cö
_i Latite
Quartz latite tt!
Granodiorite
Dacite Trachyandesite Urtite
Tonalité Diorite
Dacite Andésite Essexite
Tephrite
SS Gabbro Basanite
8 Basalt Theralite Melteigite
Peridotites
Pyroxenites
J_L
100 50 0 100 50 0
% Plagioclase o/o M a f j c
of total Feldspar
Fig. 22.3. Mineralogical classification of plutonic and phaneritic volcanic rocks (Jackson, 1970).
TABLE 22.3
Classifications of sediments and sedimentary rocks (after Folk, 1961; Dunham, 1962)
a. COMPONENTS OF
SEDIMENTS
- Terrigenous
- Allochemical
- Chemical
b. TERRIGENOUS
SEDIMENTS ROCKS
Grain Size (mm)
Rocks Soil
Gravel Conglomerate or
- 2.0 2.0 Breccia (angular frags.)
Sand Sandstone
- 0.0625 0.020
■a Silt Siltstone
- 0.0039 0.002
i Clay Shale (fissile)
or Mudstone
c. ALLOCHEMICAL; LIMESTONES
FOLK'S CLASSIFICATION
Over 2/3 lime mud matrix Subequal Over 2/3 spar cement
spar and Sorting
10-50% Over 50% Sorting Rounded,
0-1% 1-10% lime mud poor good abraded
Representative Micrite Fossil- Poorly

Sparce Packed Unsorted Sorted Rounded
Rock and iferous washed
Biomicrtte Biomicrtte Blospartte Blospartte Blospartte
Dismlcrtte Micrite Blospartte
Terminology ►
• Biomicrtte Biospartte
DUNHAM'S CLASSIFICATION
DEPOSITIONAL TEXTURE RECOGNIZABLE
Original components bound

together during deposition Original
DEPOSITIONAL
components
TEXTURE
Contains mud not bound
NOT
Lacks mud together
Mud-supported RECOGNIZABLE
and is during
Grain- deposition
Less than More than grain
supported supported
10% 10%
grains grains
WACKE- CRYSTALLINE
MUDSTONE
STONE PACKSTONE GRAINSTONE BOUNDSTONE
CARBONATE
Classification of earth matenals: a brief examination of examples 575
a. TERRIGENOUS — SANDSTONE
Maturity QUARTZ + CHERT

100 â" Quartz SS.
95 Mature
Moderately
85 mature Sublffliic SS.
75 4 Slightly
mature
Immature
FELDSPAR ROCK FRAGMENTS
NOTE: Matrix not considered here
b. ALLOCHEMICAL — LIMESTONES
(ignoring recrystailization)
ALLOCHEMICAL [
GRAINS
I = Sparry allochemical
limestone
II = Microcrystalline
allochemical limestone
III = Microcrystalline
limestone (mlcrfte)
10%
MICROCRYSTALLINE SPARRY
MATRIX CEMENT
Fig. 22.4. The three principal compositional components of sedimentary rocks (after Pettijohn, 1957).
576 W. Chesworth, G.A. Spiers, L.J. Evans and I.P. Martini
c. ALLOCHEMICAL — CARBONACEOUS SEDIMENTS
SAPROUTH
(petrotaim Mri··)
Fig. 22.4 (continued). The three principal compositional components of sedimentary rocks (after Petti-
john, 1957).
or chemistry of their component particles or crystals (Figure 22.4a; Table 22.4). In

the mineralogical classifications of clastic rocks the interstitial matrix may not be
considered (Figure 22.4a).
Textural crìterìa: grain size distribution of sediments and sedimentary rocks, and
prevalent shape of component particles (Table 22.3). This level of classification is
used primarily for clastic sediments and non-recrystallized carbonates.
Fades crìterìa: composition, texture, structure and fossils are considered. Charac
teristic sequences are sought which reflect recurring processes, which in turn could
indicate specific environments of sedimentation. Examples of well established facies
models are those of braided and meandering streams, barrier islands, tidal flats,
subaqueous fans and so on.
Field relations and age criteria: Lithostratigraphic units (Tkble 22.5a) are based on
one bed or sequence of strata containing one or more rock types and facies, formed
in a specified interval of time, that may or may not contain some characteristic
fossil assemblage. Each sequence is defined on the basis of a specific "type section"
described at one locality. Utilizing their fossil content or other dating methods, the
rock units can be placed in a temporal sequence, which defines a relative or an
absolute geological time scale (Table 22.4a and b).
Classification of earth matenah: a brief examination of examples 577
TABLE 22.4
Mineralogical characteristics of evaporite and other chemical sediments (after Boggs, 1987)
Mineral class Mineral name Chemical composition (Associated) rock name

(A) Evaporites.
Chloride Halite NaCl Halite; rock salt
Sylvite KC1
Carnal ite KMgCl3.6H2 Potash salts
Sulfates Langbeinite K 2 Mg 2 (S0 4 ) 3
Polyhalite K 2 Ca2Mg(S04)6.H 2 0
Kainite KMg(S0 4 )Cl.H 2 0
Anhydrite CaS0 4 Anhydrite: anhydrock
Gypsum CaS0 4 .2H 2 0
Kieserite MgS0 4 .H 2 0 Gypsum: gyprock
Carbonates Calcite CaC0 3 Limestone
Magnesite MgC0 3 -
Dolomite CaMg(C0 3 ) 2 Dolomite: dolostone
(B) Others
Silica (A mineraloid) Si02.nH20 Chert
Fe-oxides Hematite Fe203 Ironstone
Magnetite Fe304
Fe-hydroxides Goethite FeOOH
Fe-carbonates Siderite FeC0 3
Fe-sulphides Pyrite FeS 2
Pyrrhotite Fe^S
Silicates Glauconite K(Fe,Mg,Al)2(Si 4 Oio)(OH) 2
Zeolites Clinoptilolite CaAl 2 Si 7 0 1 8 .6H 2 0 Tliffaceous sandstone
Analcite NaAlSi 2 0 6 .H 2 0
Phosphates Fluorapatite Ca 5 (P0 4 ) 3 F Phosphorite
Mn oxides Manganite MnOOH Manganese nodules
Mn salts
Nomenclature of metamorphic rocks
Tectonic criteria: paired metamorphic belts of active continental margins; depth

zones of oceanic ridges.
Field relations and mode of occurrence: contact and regional metamorphism. Fa
cies series.
Age criteria: metamorphism associated with specific events: Alpine, Hercynian,
Caledonian metamorphisms, for example.
Textural criteria: characteristic metamorphic structures and textures are foliated
(schistose and gneissose), lineated, spotted and hornfelsic. Thefirsttwo are consid
ered to indicate regional metamorphism, the last two, contact metamorphism.
Compositional criteria: the ACF compositional triangle is a useful basis for dis
playing metamorphic nomenclature. Other componential systems used include the
AKF and AFM triangles.
578 W. Chesworth, G.A. Spiers, LJ. Evans and LR Martini
TABLE 22.5
Simplified stratigraphie classification and geologic time scale
a. STRATIGRAPHIC CLASSIFICATION
No time
significance Relative time significance
Rock units Time-rock units Time units
Era
Member
Formation
Bed
System Period
Series Epoch
Stage Age
b. GEOLOGIC TIME SCALE

Period Epoch Million
Era or System Years B.P.
Recent
0.0125
Quaternary
Pleistocene
2
Pliocene
12
N
O Miocene
c 26
a>
O Tertiary Oligocene
34
Eocene
55
Paleocene
71
u
Cretaceous 135
¼
N
O Jurassic
V) 180
2 Triassic
220
Permian
275
Pennsylvanian
330
o Mississippian 355
N
O
Devonian 410
σ Silurian
430
Ordovician 480
Cambrian
600
e _ 800
σ
Proterozoic _ 1600
E
o 2500
u
a)
_ 3000
a.
Arche a n
(AI2 O3 + FezOa - (NajjO + K2O)) Kyanite, an dal u site, sillimanite,

muscovite
Staurolite
Cordierite
Anorthite
Epidoto
Almandine garnet
Chlorite
Biotite
(MgO + FeO + MnO)

(CaO) C
Hypersthene
Fig. 22.5. Metamorphic rock and mineral compositions in the ACF diagram (from Read and Watson,
1962).
Figure 22.5 shows typical rock compositions displayed on the ACF triangle. Char-
acteristic metamorphic minerals are shown, and when tie lines are drawn between
coexisting minerals in such a diagram, the typical mineral facies diagram of the
metamorphic petrologist is the result.
Winkler has elaborated a simple double triangle based on quartz, feldspars, micas
and carbonates (Figure 22.6). His terminology also shows how typical rock names
can be modified by means of mineralogical prefixes, for example, albite quartz
phyllite.
Classification of soils
Systems of soil classification have been in existence since antiquity and have
traditionally been based upon criteria that are readily recognizable in the field. The
use of soil colour, texture and soil fertility, for example, were used by the Greeks
and Romans to establish the suitability of various soils for particular agricultural
crops (Strzemski, 1975). Marbut (1951), while developing an early version of the
U.S. system of soil classification, suggested the following criteria (in relative order
of importance):
(a) number of horizons;
582
CARBONATE CARBONATE
Marble -Marble
Cale-si licate-calcite
quartz tels
W. Chesworth, G.A. Spiers, LJ. Evans and LE Martini

QUARTZ SERICITE
Quartzite (BIOTITE,
Mica quartzite CHLORITE)
Abite (microcline)
quartzite
Quartz-albite
(microcline) fels Quartz-feldspar tels
Plagiodase fels
PLAGIOCLASE
(MICROCLINE)
Fig. 22.6. Nomenclature of metamorphic rocks: A. low grade; B. high grade (Winkler, 1974).
Classification of earth materìals: a brief examination of examples 581
(b) colour of horizons (with emphasis on the upper one or two);

(c) texture of horizons;
(d) structure of horizons;
(e) relative arrangement of horizons;
(f) chemical composition of horizons;
(g) thickness of horizons; and
(h) geology of soil material.
The authors of the modern American system of Soil Taxonomy essentially follow
this order of priority, with the concept of diagnostic horizon being central to their
classification (Tables 22.6 and 22.7). In certain soil orders however, mineralogy is
elevated to an important role (such as Oxisols, Andisols, Vertisols), whilst in the
Aridisols, for example, climate is considered to be an important differentiating
criterion.
Currently there are a number of soil classification systems in use throughout the
world, the majority being tailored to the nature and range of soils found within
a particular country or region. The classification of soils, much more than rocks,
has followed rather nationalistic lines, and Fitzpatrick (1980) gives useful tables of
comparison of the various national systems. However, two soil classification systems
do exist that are sufficiently comprehensive that they can be applied to the vast
majority of the world's soils. These are Soil Taxonomy, developed largely by U.S.
pedologists (Tkble 22.7) and the World System of Soil Classification, developed
by the Food and Agricultural Organization (FAO) of the United Nations. FAO
equivalents to the soil orders of Taxonomy are given in Table 22.8.
Fortunately most of the major systems of soil classification in use world-wide
are based on similar principles and contain a similar structure. It is relatively easy
therefore to move from one system to another once the basic principles of soil clas-
sification are understood. This is because central to all soil classification systems is
a hierarchical system in which soil properties that reflect processes of soil formation
are given greater priority than other soil properties. There is thus the concept of
genesis (soil development) embodied in most soil classification systems in which
the taxa at the highest level within the system (generally the Soil Order) are based
on properties that reflect the nature of the soil environment and the effects of the
dominant soil forming processes.
The highest level of generalization in most systems, the soil order, is generally
differentiated on the basis of the presence of certain horizons diagnostic of major
pedological processes. Nine soil orders are recognized in Soil Taxonomy, whereas in
the FAO system twenty-six orders are differentiated.
Discussion
Most classifications began as attempts to construct systems useful in recognis-

ing rocks or soils in the field. The practical purposes of geological or soil survey
TABLE 22.6
Generalised descriptions of diagnostic horizons
1. Surface horizons (epipedons)

Epipedon Description
Mollic Normally at least 25 cm thick with 1% organic matter; 50% saturated with
bases
Histic A thin organic layer overlying mineral soil; seasonally saturated with water <
30 cm if dry, or 45 cm if wet
Anthropic A mollic epipedon, with a high content of extractable phosphorus added as
fertilizer
Plaggen Produced by heaving manuring > 50 cm thick
Umbrie Surface mineral horizon; soil darked by O.M., but not meeting requirements
of mollic, anthropic or plaggen
Ochric Any surface horizon not meeting the requirements of the above
2. Sub-surface horizons
Horizon Description
Argillic Clay-rich horizon
Natric An argillic horizon with > 15% Na in total CEC; usually prismatic
Spodic Rich in illuvial humus and amorphous Fe and Al
Cambic Contains illuviated clay, humus, or Fe and Al, but insufficient for argillic or
spodic
Oxic Highly weathered horizon, high in 1:1 clays or gibbsite; low CEC
Agric Clay and humus rich horizon immediately below plow layer
3. Other horizons and features

Horizon or feature Description
Horizons in mineral soils
Albic Light "ashy" A horizon
Calcic Ca-rich layer
Duripan Layer cemented by silica
Fragipan Layer of brittle soil, commonly of a loam texture
Gypsic Gypsum-enriched layer
Petrocalcic Ca-cemented layer, Caliche is an example
Placic Layer cemented by iron (ironpan)
Plinthite Material cemented by precipitate sesquioxides; soft while moist, hardens irre
versibly on drying (laterite)
Sombric Subsurface horizon high in humus, base saturation < 50%
Sulfuric Horizon in any position with pH < 3.5 produced by the oxidation of sulphides
Horizons of organic soils a

Fibric Fibers compose over two-thirds of mass
Hemic Fibers compose between one-third and two-thirds of mass
Sapric Fibers compose less than one-third of the mass
Additional soil features

Lithic contact Boundary between soil and unweathered bedrock
Paralithic contact Boundary between soil and saprolite
Permafrost Permanently frozen soil
a
Organic deposits that exceed the limiting thickness of a histic epipedon.
Source: Hausenbueller, (1978).
TABLE 22.7
Brief characterization of the 11 soil orders of the United States

Soil order General features
Entisols No profile development
Inceptisols Weak to moderate horizon development
Andisols >60% volcanic éjecta, bulk density < 900 kg m~ 3
Histosols Organic soils (peat and mucks)
Andisols Dry climate soils, with Ca and alkali mineral concentrations
Mollisols Mostly grassland soils with deep dark A horizons
Vertisols High content swelling self-mulching
Alfisols Humid area soils with clay accumulation in the B horizon
Spodosols Acid soils with well leached A horizon with accumulation of humus, Fe and Al in
the B horizon
Ultisols Hot climate acid soils, reddish, with clay accumulation in the B horizon
Oxisols Hot climate acid soils with oxic horizon
Source: Miller and Donahue (1990).
TABLE 22.8
Correspondences between U.S. Soil Taxonomy groups and EA.O. groups
(A) Groups with one to one (B) U.S. groups with correspondences with more than
correspondence one F.A.O. group
U.S. F.A.O. U.S. F.A.O.
Andisol Andosol Entisol Arenosol, Fluvisol, Gleysol, Regosol
Spodosol Podzol
Oxisol Ferralsol Inceptisol Andosol, Cambisol, Fluvisol, Gleysol, Leptosol
Vertisol Vertisol
Histosol Histosol Aridisol Solonchak, Solonetz, Xerosol, Yermosol
Mollisol Chernozen, Gleysol, Greyzem, Kastanozem,
Phaeozem, Renzina, Solonchaks, Solonetz
Alfisol Luvisol, Planosol, Solonetz
Ultisol Acrisol, Nitosol, Planosol
Source: Fitzpatrick (1980).
were paramount. If however, the purpose of the classification is to discover some

inherent structure in nature, the survey type of classification may not be particularly
useful. For example, colour (easily recognized in the field) may be used as a cri
terion in classifying soils. Often this is a function of the mineralogy (especially the
Fe-mineralogy) of the soils, and information on soil mineralogy, rather than the sec
ondary property colour, would be much more useful in revealing genetic relation
ships amongst soils. Of course, classifications for field purposes are still required,
but the increasing convenience and rapidity of instrumental analytical techniques
has made chemical and mineralogical criteria much easier to obtain than formerly,
so that these criteria are becoming increasingly used by today's taxonomists. And
there is no doubt that for scientific purposes, for example in determining the genesis
584 W. Chesworth, G.A. Spiers, LJ. Evans and I.R Martini
and the genetic relationships between rocks or soils, such criteria are absolutely
necessary. Classifications useful in this regard are the natural classifications referred
to earlier. Unfortunately, many so-called natural classifications are based on pre-
vailing theories of formation for the objects classified. The theory then becomes
an axiomatic part of the classification which inevitably leads to predictions that are
related in circular fashion to the chosen theory.
However, it should be possible to choose criteria for distinguishing classes, that
are related to the processes that formed the objects being classified. The genetic or
interpretive content will then be an intrinsic part of the classification. Thus prop-
erties such as chemical and mineralogical composition, texture and structure, when
used in a classification, result in built-in genetic information, since they are sensitive
to the chemical and physical processes by which rocks and soils form. Most rock
classifications have always emphasized these properties, with chemistry and miner-
alogy playing pivotal roles. By contrast, most soil classifications have emphasized
structure and texture and consigned the details of chemistry and mineralogy to a
subordinate role in classification. The result has been that whereas great advances
have been made in this century in understanding the chemical processes of forma-
tion of igneous rocks, in soils the elaboration of physical processes of genesis has
outstripped the chemical.
An attempt to incorporate rock classifications into soil classification has been
made by Whiteside (1953) but only at the level of parent material (Figure 22.7).
Conclusion
The number of classifications that have been erected to cover the populations
of rocks and soils is too great to list here. Many, perhaps most were constructed
to help the practical surveyor to recognise rocks or soils in the field. They rightly
emphasized visible criteria as the bases of their classes. The scientific petrologist
or pedologist is more likely to be interested in working out problems of rock or
soil genesis and the surveyor's classification is not going to serve this purpose well.
In this case, what is required is a classification that utilises criteria responsive to
the physico-chemical processes of formation of a rock or soil. The ideal would be
something with the rigour of thermodynamics, whereby a class could be defined by
a minimum number of necessary criteria, the way that a thermodynamic state can
be completely defined in terms of a small number of state factors. Short of this a
great deal has been accomplished in terms of simple properties such as chemical
and mineralogical composition, texture and structure.
Only within the last generation or so, in both petrology and pedology, has the
use of multivariate statistical analysis become popular in taxonomy. No doubt this
will lead to less arbitrary classifications in the future, though so far no generally
used classification of rocks or soils has been constructed on this basis. Rather, the
Classification
of earth materials: a brief examination of examples
COARSE
CRYSTALLINE
LU PORPHYRES
FINE
CRYSTALLINE
FERRO-ktAQNESIAN MINERALS \
BtOTfTE "^ «
^ ^ HORNBLENDE PYROXENE ETC
ACCESSORY MINERALS — ·
(9.0. MAONETrrEAPATfTEETC)
58
CHEMICALANDMINERALOGICAL COMPOSITION
Fig, 22.7. Classification of soil parent material: igneous rocks and their metamorphic and glacial equivalents.
technique has been used to justify or to criticise existing classifications erected on

different grounds.
At a high, general level, and on tectonic grounds, it is possible to group all three
rock types into ensembles or super-classes that reflect the crustal regime where they
are found. Since the regime is driven by the endogenic cycle, and since soils and
many sediments are more closely related to the exogenic cycle, it is not feasible to
use tectonics as a basis for relating rocks and soils into a comprehensive natural
classification. If the latter is desirable, and there is no prima facie reason why it
should be, it would be best attempted in terms of the parent material-offspring
relationship that soils have with geological materials.
References
Brewer, R., 1964. Fabric and Mineral Analysis of Soil. Wiley, New York, N.Y., 470 pp.
Boggs, S., Jr., 1987. Principles of Sedimentology and Stratigraphy. Merrill, Columbus, Ohio, 784 pp.
Dunham, R.J., 1962. Classification of carbonate rocks according to depositional texture. Am. Assoc.
Pet. Geol., Mem., 1: pp. 108-121.
pp.
Folk, R.L., 1961. Petrology of Sedimentary Rocks. Hemphill, Austin, Texas, 153 pp.
Griffiths, J.C., 1967. Scientific Methods in the Analysis of Sediments. McGraw-Hill, New York, N.Y.,
508 pp.
Grout, F.F., 1932. Petrography and Petrology. McGraw-Hill, New York, N.Y., 522 pp.
Hausenbueller, R.L., 1978. Soil Science. Brown, Dubuque, Iowa, 611 pp.
Hennig, W, 1979. Phylogenetic Systematics. University of Illinois Press, Urbana, 111., 263 pp.
Jackson, K.C., 1970. Textbook of Lithology. McGraw-Hill, New York, N.Y., 552 pp.
Marbut, CF. 1951. Soils, their genesis and classification. In: The Life and Works of CF. Marbut, Soil
Science Society of America, Madison, Wise, 134 pp.
Middlemost, E.A.K., 1985. Magmas and Magmatic Rocks. Longman, London, 266 pp.
Miller, R.W. and Donahue, R.L., 1990. Soils. Prentice-Hall, Englewood Cliffs, N.J., 768 pp.
Murray, J.W., 1981. A Guide to Classification in Geology. Ellis Horwood, Chichester, 112 pp.
Perry, K., 1967. An application of linear algebra to petrologie problems: Part 1. Mineral classification.
Read, H.H. and Watson, J., 1962. Introduction to Geology, Vol. 1. Principles. Macmillan, London, 693
pp.
Soil Survey Staff, 1975. Soil Taxonomy. USDA, Washington, D.C, 574 pp.
Sokal, R.R., 1974. Classification: purposes, principles, progress, prospects. Science, 185: 1115-1123.
Strzemski, M. 1975. Ideas underlying soil systematics. U.S. Department of Commerce, National Techni
cal Information Service, Va., 542 pp.
Whiteside, E.P., 1953. Some relationships between the classification of rocks by geologists and the
classification of soils by soil scientists. Soil Sci. Soc. Am. Proc, 17: 138-142.
Winkler, H.G.F., 1979. Petrogenesis of Metamorphic Rocks. Student Edition. Springer, Berlin, 348 pp.
587
References Index
Aagaard, P., 122,123, 124 Altukhov, V.M., 220

Abbott, P.L., 527, 529, 540 Ambrosi, J.P., 383, 386, 387, 389, 401, 457, 467
Abraham, K., 563 Amiel, A.J., 527, 540
Abreu, M., 100, 104 Amit, R., 264, 278, 279, 490, 495, 496
Absolon, K., 152 Amouric, D., 428, 437
Abtahi, A., 254 Amouric, M., 85,100, 103
Adams, J.B., 248, 257 Amundson, R.G., 168,176, 315, 344
Adams, M.J., 124 Anderson, J., 204, 219
Adams, S.L., 541 Andrejko, M.J., 203, 219
Adams, W.A., 426, 437 Andreoli, C , 85, 87, 90,100, 104
Adler, L, 47, 68 Andrew, A.S., 374
Ague, J.J., 253 Andrew, R.L., 464, 467
Ahlbrandt, T.S., 527, 540 Andrews, H.N., 548, 561, 563
Ahlricks, J.L., 101 Andrews, J.E., 122
Ahn, J.H., 84,100 Antoine, H., 303
Aitken, E, 69 Anton, D., 528, 529, 540
Aiton, W, 204, 218, 219 Antonioletti, R., 178
Ajie, H.O., 496 Apollo Soil Survey, 50, 53, 68
Al-Mashhady, A.S., 253 Arakel, A.V, 241, 242, 246, 253, 310, 317, 319,
Al-Rubaii, M., 496 343, 344, 350, 374
Aleva, G.J.J., 434, 437 Arbey, E, 350, 374
Alexander, E.B., 272, 278 Archer, A.C., 161,176
Alexander, L.T., 404 Aristarain, L.F., 313, 344
Ali, O.M., 260 Armands, G., 216, 220
Allen, B.L., 181, 237, 238, 245, 253, 256 Arn, A., 499
Allen, C C , 248, 253 Atkinson, CD., 525, 536, 537, 540
Allen, J.R.L., 480, 484, 485, 495, 525, 535, 537, Auer, V, 205, 207, 208, 210, 211, 212, 213, 220
540, 550, 551, 560
Allen, T.F.H., 562 Baas-Becking, L.G.M., 31, 38
Allen, V.T., 522 Badarello L., 379, 394, 402
Almeida, E.B. de, 454, 467 Badaut, D., 116,123
Alonso, O.E., 118,122 Bailey, S.W., 108,122, 123
Alonso-Zarza, A., 486, 488, 499 Bain, O.C., 110, 122,124, 141,148,151, 302
Alpers, C.N., 253 Baker, W.E., 453, 467
AlSari, A.M., 541 Balkay, B., 435, 437
Alt, D., 542 Bambach, R.K., 563
588 References index
Bamforth, D.B., 254 Besley, B.M., 477, 486, 495

Banfield, J.F., 81, 84,100, 101 Bespalov, N.D., 232, 253
Banks, H.P., 548, 560 Best, M., 560
Barbaroux, L., 116,123, 124 Bettenay, E., 346, 374
Barber, K.E., 207, 220 Beutelspacher, H., 80,100
Bardossy, G.Y., 396, 397, 398, 401, 404, All, 430, Bevan, J.C., 123
435, 437, 442, 453, 467, 556, 560 Beveridge, T.J., 25, 38,102
Baronnet, A., 103, 437 Bgatov, B.I., 216, 220
Barros de Oliveira, S.M., 441, 469 Bidini, D., 173,179
Barry, R.G., 188, 200 Bigham, J.M., 245, 253
Barshad, I., 72,100, 103, 119,123 Bildgen, P., 480, 498
Barsotti, N.J., 235, 254 Bilzi, A.F., 538, 540
Bartoli, C, 167,178 Bird, M.I., 370, 374, 374, 480, 495
Bartoli, F., 169,172,276 Birkeland, P.W., 7, 12, 261, 262, 269, 270, 271,
Bashour, I.I., 252, 253 275, 278, 279, 280, 484, 495
Bates, R.L., 3, 4,15 Birse, E.L., 161,177
Batile, J., 255 Biscayne, P.E., 116,123
Bauer, R.A., 521 Bischof, G., 215, 220
Beadle, N.C.W., 557, 560 Bish, D.L., 467
Beam, W, 217, 220 Bishai, H.M., 217, 220
Beaudet, G., 401, 412, 437 Bjorkholm, P., 68
Beaufort, D., 124 Blain, CF., 464, 467
Beaumont, A., 104 Blair, T.C., 490, 495
Beauvais, A., 410, 437, 441, 458, 460, 467, 469, Blank, M., 236, 253
470 Blaser, P., 172,277
Beckmann, G.G., 475, 479, 495, 499 Blodgeth, H., 68
Behrensmeyer, A.K., 485, 495 Blodgett, R.H., 528, 540, 551, 560
Beisigel, W.R., 471 Blondel, J.L., 167,277
Beissner, H., 434, 443 Bloom, PR., 292, 306
Bell, L.C., 292, 303 Blot, A., 410, 437
Bellamy, D.J., 208, 220, 223 Blount, D.N., 226, 253
Bellamy, R., 208, 220 Boardman, J., 475, 495, 498
Ben Rhaïem, H., 88, 89, 90,100 Bockheim, J.C, 261, 276, 279
Benedict, J.B., 279 Bockheim, J.G., 185,192,193,199, 200, 226, 253
Benier, J., 123 Bocquier, G., 380, 382, 386, 389, 393, 401, 402,
Bennema, J., 395, 401 403, 404, 405, 441, 467
Bentley, CF., 556, 563 Boeglin, J.L., 410, 411, 415, 437, 442, 458, 463,
Benton, M.J., 543, 560 464, 467, 469
Beresovkaya, V.V., 468 Boggs, S., Jr., 577, 586
Bergenback, R.E., 523 Boguslawski, E., 105
Bernado, J.C., 544, 560 Boland, J.N., 80,101
Bernardelli, A , 471 Bonamo, P.M., 560
Berner, R.A., 31, 38, 71,100, All, 437, 445, 467 Bonatti, E., 257
Berrier, J., 40, 88,102, 104,180 Bonifas, M., 391, 403, 454, 456, 469
Berry, M.E., 270, 274, 277, 278 Bonneau, M., 178
Berthelin, J., 73, 76, 77, 78,102, 103 Boreman, J.A., 68
Bertrand-Ayrault, M., 376 Bos, R.H.G., 477, 495
Bertrand Ayrault, M., 376 Boss, G., 237, 253
Besler, H., 239, 253 Bottero, J.Y., 100,100
References index 589
Bottner, P., 169,177 Brown, G., 108, 109,110, 111, 123, 124
Boucot, A.J., 547, 548, 560, 563 Brown, I.W.M., 102
Boudeille, M., 96,100, 376 Brown, J.L., 80,100, 102,187,191, 201
Boulangé, B., 40, 289, 302, 391, 396, 397, 398, Brown, K.M., 123
401, 410, 421, 434, 437, 438, 453, 454, 455, Brown, L.F., 116,123
467, 470 Brown, R.J.E., 200
Boulet, R., 382, 387, 394, 401, 402, 408, 438 Brown, R.W, 69
Bourgeat, E, 396, 401, 404, 421, 438 Browne, DJ., 216, 220
Bourman, R.R, 441 Brownlee, D.E., 68
Bourrié, G., 288, 302, 422, 438, 457, 467 Bruand, A., 104
Bova, J.A., 498 Bnickert, S., 74,100,167,168,177, 179, 289, 306
Bovis, M J., 188, 200 Brunn, J.H., 15, 36, 38
Bowden, M.E., 102 Bryan, W.H., 6, 15,15, 31, 38
Bowen, H.J.M., 220 Bryant, E C , 181
Bowes, D.R., 560 Buchanan, E, 379, 402
Bowler, J.M., 344 Bûcher, A., 165,177
Bowles, W.B., 255 Buckley, H.A., I l l , 123
Bowman, G.M., 272, 281 Budnikoff, P.P., 235, 253
Bown, T.M., 480, 481, 483, 484, 492, 495, 497, Bull, W.B., 256, 346
525, 526, 527, 530, 531, 532, 536, 537, 538, Bunch, T.E., 67, 68
540, 541, 542 Buoi, S.W., 30, 38, 238, 245, 253, 286, 302
Bowser, C.J., 115,124 Bureau, P., 247, 253
Boyadgiev, T., 226, 227, 241, 245, 253 Burger, K., 503, 505, 510, 517, 518, 521
Boyer, D., 404 Burke, R.M., 270, 271, 278, 279
Boyer, Ph., 415, 416, 438 Burns, S.E, 161,177
Boyle, R.W, 216, 220 Burtin, G., 172,176
Braithwaite, C.J.R., 235, 253 Buseck, P.R., 80, 81,100, 101, 105
Brakke, D.E, 220, 221 Busenberg, E., 137,151
Brasher, B.R., 256 Butler, B.E., 476, 496
Braslau, D., 65, 68 Button, A., 149,151
Bravard, S., 103 Butzer, K.W., 244, 254
Brett, R., 69 Buurman, P., 297, 299, 302
Brewer, R., 164,179,189,190, 201, 258, 286, 287,
302, 349, 372, 374, 388, 391, 401, 411, 495, Cabidoche, Y.M., 161,167,171,177, 178, 180
499, 586 Cady, J.G., 375, 404
Bricker, O., 457, 467 Canili, R.A., 521
Bridge, J.S., 480, 483, 484, 495, 495, 496, 532, Canili, T.A., 254
535, 539, 540 Calas, G., 100, 462, 469
Bridges, E.M., 532, 540 Caldas, E.E, 173,177
Brime, C , 118,122 Callen, R.A., 344, 350, 351, 374, 377
Brimhall, G.H., 236, 246, 253 Callot, G., 78,100, 317, 322, 344
Brindley, G.W., 461, 462, 467 Calvo, R.M., 289, 292, 302
Briner, G.P, 119,123, 296, 302 Cameron, R.E., 188, 201
Brinkman, R., 31, 33, 38, 40, 74, 100, 245, 253, Campbel, I.B., 13
289, 302 Campbell, A.S., 273, 275, 279
Briot, P., 246, 253 Campbell, D.E., 223
Brookins, D.G., 31, 38 Campbell, F.E., 152
Brown, C.N., 314, 344 Campbell, LB., 194,195,198, 201, 297, 302
Brown, E., 216, 220 Cantagrel, J.M., 302
Cantinone, P., 456, 467 Cicerone, R J., 220

Cardenas, F., 68 Cintala, M J., 68
Carlisle, D., 246, 254 Ciolkosz, E.J., 538, 540
Carlson, T.B., 257 Clanton, U.S., 69
Carmouze, J.-R, 116,123 Claridge, G.G.C, 13,194,195,198, 201, 297, 302
Carozzi, A.V., 11,15 Clark, A.H., 464, 470
Carrier, W.D., 47, 68 Clark, D.R., 151
Carter, G.S., 217, 220 Clark, J.S., 495
Carvalho, A., 434, 439, 441 Clark, M.N., 280
Casas, J., 100, 255 Clarke, F.W., 130,151
Cathelineau, M., 120,123 Clemency, C.V., 137,151, 297, 302
Catt, J.A., 477, 496 Clement, P., 162,162,177
Cecil, C.B., 523 Clinard, C , 104
Cerri, C C , 304 Cline, L.M., 123
Chadwick, O.A., 176, 344, 349, 350, 372, 375 Cline, M.G., 404, 553, 560
Chaloner, W.G., 546, 558, 560 Clisham, T.J., 541
Chamayou, H., 164,177 Cloke, PL., 418, 438, 447, 468
Chamberlin, T.C, 505, 521 Cloud, L., 415, 438
Chamley, H., 114,115,116,118,123 Cluff, L.S., 281
Chan, R.A., 498 Cody, A.M., 242, 254
Chandra, D., 523 Cody, R.D., 242, 254
Chao, E.T.C., 55, 68 Coen, G.M., 180
Chapin, CE., 529, 540 Cohen, A.D., 219, 479, 498, 503, 523
Chapman, S.B., 214, 220 Coincon, R., 38
Chappell, B.W., 294, 306 Cojean, R., 376
Chartres, C.J., 162,173,177, 349, 375 Cole, M.M., 544, 557, 560
Chatelin, Y., 391, 402 Cole, T.G., 117,123
Chaumont, C , 303 Coleman, J.M., 554, 560
Chauvel, A., 103, 391, 393, 395, 396, 402, 404, Colin, F., 103, 297, 302, 304, 447, 449, 450, 461,
426, 438 462, 468, 469
Cherven, V.B., 529, 540 Collinson, J.D., 487, 496
Chesters, G., 104 Coiman, S.M., 33,39, 262, 274, 279, 289, 289,302
Chesworth, W, 4, 6, 13, 21, 24, 33, 38, 39, 119, Combes, J.M., 99,100
123, 285, 286, 288, 289, 290, 294, 295, 296, Combes, J.P., 398, 402, 421, 438
297, 298, 299, 302, 303, 304, 422, 430, 438, Conacher, A.J., 240, 247, 254
457, 465, 467, 468 Conel, J.E., 541
Chiarenzelli, J.R., 556, 560 Conrad, G., 349, 350, 375
Childs, C.W., 124, 262, 280 Conte, C.N., 404, 441
Chinn, T.J.H., 306 Contini, ¼.,1 77
Chivas, A.R., 374, 480, 495 Conway, T.J., 224
Choisnel,E., 160,177 Cooke, R.U., 235, 248, 254, 255, 258
Choquette, P.W., 310, 345 Cooper-Driver, G., 555, 563
Christ, C.L., 39, 422, 439, 456, 468 Cope, M.J., 546, 558, 560
Christenson, G.E., 488, 496 Coque, R., 244, 254
Christian, R.P., 523 Corliss, J.B., 39
Christiansen, G., 440 Coude-Gaussen, G., 168,177, 179
Chukhrov, F.V., 458, 468 Court, J.E., 115,123
Churchward, H.M., 317, 344, 374 Couston, J.F.C, 440
Cicerone, F.J., 223 Cowin, D.K., 522
Craig, D.C., 128,148,151 Dean, W.E., 115,123, 206, 207, 221

Craig, G. Y., 551, 560 Debrëbant, P., 177
Craig, J.R., 256 Decarreau A., 462, 468
Craig, M.A., 498 Degallier, R., 424, 438
Crawford, AJ., 224 Dejou, J., 24, 39, 72,100, 296, 302, 303
Crawford, A.R., 317, 345 Delage, P., 88,101
Crawford, RJ., 346 Delaunay, A., 376
Creach, M., 464, 466, 468 Delmas, A.B., 15, 27, 40, 350, 373, 373, 375
Cremaschi, M., 3,15 Delmas,M., 168,177
Crerar, D.A., 350, 373, 377 Delvaux, B., 75, 85,101
Criswell, D.R., 41, 68 Delvigne, J., 297, 304, 305, 382, 396, 402, 421,
Crittenden, R.C., 280 423, 438, 454, 468, 469
Crocker, R.L., 309, 317, 320, 340, 343, 345 DeMaris, P.J., 515, 517, 521, 522, 523
Crook, K.A.W., 495 Dembroiï, G.R., 279
Crum, R , 208, 220 DeNiro, M.J., 252, 254, 496
Cunningham, A.B., 253 Dennen, W.H., 434, 438
Curi, N., 438 Deprairies, A., 15
Curran, M.K., 293, 305 des Marais, D.J., 55, 68
Curtin, D., 293, 296, 303 Desborough, G.A., 68
Curtis, CD., 107,118,123, 125 Desprairies, A., 36,38
Curtis, M.T., 259, 492, 496, 499 Dethier, D.P., 33, 39, 274, 289, 302
Curtis, R.L., 435, 441 Dever, G.R., 560
Cutler, EJ.B., 161,176 Devi Prasad, J., 253
Dewey, J.F., 547, 560
D'Alessio, D., 177 Dewolf, Y., 376
D'Hoore, J., 393, 402, 407, 438 Didier, P.H., 389, 392, 402, 437, 456, 457, 467, 468
Dachnowski, A., 204, 205, 206, 207, 208, 211, 220 Dietrich, W.E., 246, 253
Dahlem, D.H., 69 Dillon, P.J., 217, 218, 222
Dahlgren, R., 291, 292, 306 DiMichele, W.A., 516, 519, 522, 549, 560
Dalrymple, J.B., 475, 476, 499 Dimmock, G.M., 124
Damberger, H.H., 503, 505, 507, 510, 517, 518, Dinel, H., 208, 222
521 Dineley, D.L., 560
Dambrine, E., 74,100, 103,161,170,177 Dixon, J.B., 92,101, 245, 258
Damman, A.W.H., 214, 220 Dixon, J.C, 320, 345
Dan, J., 251, 254, 263, 264, 279, 309, 310, 311, Dobremez, J.F., 160,177
316, 345 Dohrenwend, J.C, 254
Dancette, C , 443 Dokuchayev, V.V., 6,15, 177
Dandjinou, T., 442 Donahue, D.J., 254
Darragh, RJ., 350, 375 Donahue, R.L., 8,15, 583, 586
Dash, B., 556, 560 Donaldson, J.A., 152, 560
Dau, J.H.C, 209, 220 Doner, U.E., 176, 344
Daubrée, M.A., 215, 220 Doornkamp, J.C, 255
Davis, CA., 208, 210, 220 Dorioz, J.M., 166,177, 179, 180
Davy, R., 447, 468 Dorn, R., 315, 345
Dawson, J.E., 203, 220 Dorn, R.I., 248, 252, 254, 490, 496
de Angelis, D.L., 562 Dougedroit, A., 160,178
de Castro, C , 542 Dougherty, M.T., 235, 254
de Coninck, F., 177, 553, 560 Douglas, L.A., 102, 191,192, 201
de Kimpe, C , 302 Drees, L.R., 40
Dregne, H.E., 227, 229, 230, 232, 254 Engel, CG., 248, 254
Drever, J.F., 245, 254 Englund, J.O., 115,123
Drever, J.I., 31, 39, 294, 303 Epis, R.C., 529, 540
Drew, J.V., 201 Epstein, B., 62, 68
Droubi, A., 124 Epstein, E., 28, 39
Dubief, X, 177 Erhart, H., 15,15, 37, 39, 285, 303, 415, 438
Duchaufour, P., 6,166,167,169,178, 482, 496 Ericksen, G.E., 239, 241, 242, 244, 254, 258
Ducloux, J., 300, 303, 441, 469 Eschenbrenner, V, 388, 391, 394, 402, 415, 438
Dudas, M.J., 40 Escourrou, G., 156,157,178
Dufery, J.E., 101 Eslinger, E.V., 124, 125
Duke, M.B., 47, 68 Esteban, M., 309, 310, 311, 317, 344, 345
Dumanski, J., 3,15,162 Esu, I.E., 496
Dunham, R.J., 574, 586 Eswaran, H., 242, 254, 286, 302, 306
Dunn, P.A., 496 Ettensohn, FR., 550, 560
Duplay, J., 101, 305, 442, 470 Eugster, H.P., 30, 39, 254
Dupraz, C , 170,178 Evans, D.L.C., 418, 438, 447, 468
Dupuis, C , 376 Evans, L.J., 7, 297, 303, 304
Dupuis, M., 164,178 Everett, K.R., 191, 201, 203, 221
Duquet, B., 177 Evstifeev, Y.G., 232, 255
Durand, R., 161,178
Duthie, D.M., 151 Fairbridge, F.W., 480, 496
Fairbridge, R.W., 309, 317, 345
Eberhart, J.P., 80,101 Falla, WS., 523
Eberl, D.D., 92,104,107,118,123,125, 503, 509, Fanning, D.S., 237, 245, 253
512, 513, 523 Farmer, V.C., 74, 79,101,113,123,124,133,143,
Edelman, M.J., 561 151, 299, 301, 303
Edinger, S.E., 242, 254 Farnham, R.S., 203, 208, 221
Edlund, S.A., 187, 201 Farrell, K.M., 482, 496, 533, 534, 541
Edmond, J., 26, 39 Fastovsky, D.E., 525, 527, 541
Edou Minko, A., 468 Fauck, A., 394, 402
Edwards, D.S., 549, 560 Faure, P., 389, 402
Edzwald, J.K., 116,123 Feakes, C.R., 128, 151, 544, 545, 547, 557, 560,
Eggleton, R.A., 33, 39, 80, 81, 82, 84, 100, 101, 561, 562
114,123,133,151, 300, 303 Felix, C, 103
Ehhalt, D.H., 218, 221 Fernandez-Caldas, E., 306
Ehrenberg, CG., 215, 221 Fernandez-Marcos, M.L., 292, 293, 303, 304
Ehrenberg, S.N., 118,123 Fielding, C.R., 478, 495, 496
Ehrlich, H.L., 215, 221 Fields, R.W., 542
Eiselen, J.C., 204, 221 Fiene, F., 219
Eisenberg, J., 240, 254 Fiessinger, F., 100
Eisenreich, S.J., 221 Finkelman, R.B., 50, 68
El Ansary, M., 447, 468 Finkl, C.W., 479, 480, 496
Elders, W.A., 120,124 Finney, H.R., 203, 208, 221
Eller, E., 68 Fisher, G.C., 553, 561
Ellis, F., 349, 372, 375 Fisher, R.V., 70
Ellis, S., 270, 279 Fitzpatrick, E.A., 225, 228, 231, 255, 284, 303,
Elsass, F., 100, 103, 104 570, 581, 583, 586
Ely, R.W., 529, 540 Fitzpatrick, J.J., 523
Endell, K., 216, 221 Fitzpatrick, R.W., 114,123, 433, 435, 438, 441
Flach, K.W., 256, 349, 372, 375 Garcia-Hernandez, J.F., 375

Fletcher, R., 560 Garcia-Lopez, L., 306
Fletcher, W.K., 216, 222 Garcia-Rodeja, E., 298, 302, 303
Fleury,Ph., 161, i 75 Gardner, L.R., 313, 314, 320, 345, 422, 439
Folk, R.L., 255, 574, 586 Gardner, T.W., 478, 479, 496, 502, 512, 513, 514,
Folkoff, M.E., 238, 239, 255 518, 522
Ford, J., 221 Garfias, A., 123
Forrester, J.D., 216, 221 Garreis, R.M., 31, 32, 39, 422, 439, 456, 468
Foscolos, A.E., 193, 201 Garrison, R.E., 70
Foster, R.C., 346 Gary, M., 349, 375
Foster, R.W, 561 Gaskin, A.J., 375
Foth, H.D., 544, 561 Gasparini, E., 464, 470
Fouquet, CD., 103 Gasser, U., 180
Francis, J.E., 551, 561 Gaucher, G., 33, 39
Frankel, J.J., 350, 351, 375 Gault, D.E., 58, 63, 64, 65, 68
Franzmeier, D.P., 426, 438 Gaultier, J.P., 90,101
Frasche, D.F, 460, 468 Gavaud, M., 294, 303
Fraser, A.R., 124 Gay, A.L., 127,144,146,147,149,151
Fraser, D.C., 216, 221 Genre-Grandpierre, G., 178
Fraser, PJ., 224 Gensac, P., 166,775
Freckman, D.W., 234, 259 Gensei, P.G., 548, 561
Freeman, ST., 256, 346 Georgiev, A., 158,178
Freise, F.W, 447, 468 Gerard, J., 68
Frenzel, B., 210, 221 Gerasimov, I.P., 186, 201
Freyssinet, P., 402, 404, 410, 411, 415, 417, 418, Gerrard, A.J., 481, 496
438, 439, 442, 447, 449, 450, 468 Gerson, R., 264, 278, 279, 490, 495, 496
Freytet, P., 528, 541 Gerward, L., 440
Friedman, G.M., 39 Gibbons, R.V, 63, 68
Frink, C.R., 116,125 Gibbs, R.J., 23, 39,116,124
Fritsch E., 393, 402 Gibson, D.L., 498
Fritz, B., 124, 288, 303, 402, 408, 422, 439 Gilè, L.H., 226, 237, 238, 247, 251, 255, 262, 266,
Frostick, L.E., 233, 257 267, 279, 310, 312, 313, 314, 315, 319, 320,
Fruland, R.M., 69 339, 340, 343, 345, 375, 484, 488, 489, 493,
Fryberger, S.G., 527, 540, 541 496, 527, 541
Fryxell, R., 69 Gilkes, R.J., 96,101,119,124, 296, 303
Fulchignoni, M., 51, 68 Gill, E., 309, 345
Fung, I., 204, 223 Gillespie, R., 344
Funiciello, R., 68 Gillespie, W.H., 549, 561
Fyfe, WS., 81,102, 104, 224, 304, 440 Gillman, G.P., 292, 303
Fyson, W.K., 560 Gillott, J.E., 88,101
Gilot, J.C., 166,161,178
Gac, J.Y., 116, 124, 289, 292, 303, 404, 408, 419, Gjems, O., 74,101
439 Glaeser, R., 88,101
Gaiffe, M., 167,168,169,177, 178 Glaser, P.H., 221
Gallenne, B., 116,123 Glass, H.D., 510, 522
Gammon, R.H., 224 Glaubig, R.A., 255
Ganor, E., 236, 259 Gluskoter, H.J., 515, 523
Garcia, C , 104, 305 Godard,, A., 39
Garcia, P., 302 Goddard, E.N., 70
Godwin, H., 207, 210, 221 Grossman, S., 279

Goetz, A.F.H., 536, 541 Grotzinger, J.R, 498
Goldberg, E.D., 124 Grout, F.F., 568, 586
Goldberg, S., 239, 249, 255 Grow, J.S., 560
Goldbery, R., 480, 496, 527, 541 Grubb, P.L., 434, 439
Goldhaber, M.B., 523 Grüner, J.W., 216, 221
Goldhammer, R.K., 493, 496 Guendon, J.L., 467
Goldich, S.A., 285, 295, 303 Guitian, F., 303
Goldich, S.S., 33, 39, 461, 468 Gumuzzio, J., 239, 255
Goldman, C.R., 123 Gursky, H., 68
Golub, L., 68 Gustafson, LB., 438
Gonzalez, V, 279 Güven, N., 118,124
Goodchild, J.G., 216, 221 Guyon, A., 100, 344
Goodfellow, W.D., 222 Guyot, J., 100, 303
Gorbunova, Z.N., 125 Guzman, R., 442
Gordon, L.I., 39
Gordon, M., 454, 468 Hait, M.H., 69
Gore, A.J.R, 208, 221 Hall, G.F., 259, 484, 497
Gorenstien, R, 68 Hall, R.D., 289, 303
Gorham, E., 115,123, 214, 215, 221 Hall, R.L., 280
Gorshov, A.L., 468 Hallsworth, E.G., 258
Goudie, A.S., 234, 235, 247, 250, 255, 259, 309, Halmos, P.R., 62, 68
311, 345, 349, 375 Hamilton, R.D., 346
Gradusov, B.R, 114,124, 297, 305 Hamilton, W.B., 21, 39
Grandin, G., 40, 349,375, 399, 402, 408, 439, 457, Hampel, J., 253
458, 468, 470 Hancock, P.L., 259, 499
Grandstaif, D.E., 7,15, 33, 39,127,144,146,147, Hang, P.T., 461, 467
149,151, 551, 561 Hansen, C.L., 244, 254
Granlund, E., 205, 208, 224 Hansen, K., 205, 207, 221
Grant, T.A., 540 Harden, J.W., 263, 271, 274, 275, 276, 277, 279,
Grant, W.H., 128,133,135,138, 139,143,152 541
Grasse, P.P., 415, 439 Hardie, L.A., 30, 39, 242, 256, 496
Gray, J., 547, 548, 561 Hardy, M., 103
Greenwood, W.R., 69 Harnden, R.F., 68
Gresens, R.L., 529, 541 Harper, R.M., 208, 221
Grierson, J.D., 560 Harris, B., 68
Griffin, J.G., 560 Harris, T.M., 546, 561
Griffin, J.J., 116,124 Harriss, R.C., 193, 201
Griffith, G., 415, 439 Hart, D.M., 284, 303
Griffiths, J.C., 567, 570, 586 Hartmann, W.K., 43, 68
Grigor'eva, T.R., 220 Härtung, J.B., 68
Grim, R.E., 118,124, 503, 522 Harvey, A.M., 233, 247, 255, 487, 490, 497
Grimaldi, G, 104 Hathaway, J.C, 36, 39
Grimaldi, M., 103, 104 Hausenbueller, R.L., 582, 586
Grisoni, J.-C, 376 Hawley, J.W, 256, 279, 345, 496, 527, 541
Groke, M.C.T., 305, 434, 439 Hayes, J.M., 68
Grosse-Brauckmann, G., 205, 207, 221 Hayward, A.B., 527, 541
Grossman, R.B., 225, 237, 251,255, 279, 345, 484, Hayward, M., 481, 482, 497
488, 496 Hearn, P.P., 280
Heiken, G.H., 69 Hughes, R.E., 10, 501, 502, 503, 507, 509, 510,
Heinselman, M.L., 213, 221 511, 513, 516, 517, 518, 519, 522
Heitowit, E.D., 65, 68 Hulver, M.L., 543, 563
Hem, J.D., 456,468 Humphrey R., 102
Hemley, J.J., 26, 39 Humphreys, B.A., 118,124
Hendricks, D.M., 271, 280, 375 Humphreys, E.W., 216, 222
Henin, S., 85,101 Hunt, C.B., 242, 255
Henke, K.F., 235, 255 Hunt, G.R., 537, 541
Hennig, W, 568,586 Hunt, T.S., 215, 222
Henriksen, A., 220, 221 Hunter, R.E., 243, 256
Henwood, J.H., 216, 221 Hurst, A., 118,124
Herail, G., 418, 439 Hutton, J.T., 255, 309, 310, 311, 317, 320, 322,
Herbillon, A.J., 74,101, 102, 297, 303, 391, 401, 343, 345, 346, 351, 370, 375
403, 468, 470 Hyme, N.J., 123
Hervio, M., 104
Hesse, P.P., 303 Idnurm, M., 480, 497
Hickox, D.F., 560 Iijima, S., 80,100
Hieronymus, B., 434, 439, 454, 468 Ildefonse, P., 75,102, 401, 404
Hill, CM., 374 Her, R.K., 350, 375
Hill, D.E., 191, 201 Imbernon, J., 424, 439
Hill, G J.C., 558, 563 Irion, G., 393, 403
Hillel, D., 424, 426, 439 Irving, A., 215, 222
Hinners, N.W., 44, 68 Irwin, H., 118,124
Hinsinger, Ph., 78,101 Isaac, K.P., 351, 375
Hobbs, N.B., 203, 208, 222 Isbell, R.F., 236, 239, 240, 255, 256, 297, 304, 346
Hodge, T., 245, 255 Ito, T., 306
Holbrook, P.W., 496, 522 Iwai, S., 124
Holdren, G.R., 71,100
Hole, F.D., 38 Jackson, J.A., 4,15
Holland, H.D., 15, 111, 144, 152, 289, 304, 556, Jackson, K.C., 573, 586
558, 561 Jackson, M.L., 24, 28, 33, 39, 72, 75, 80, 96,100,
Holliday, V.T., 267, 279 102,119,123, 284, 285, 291, 302, 304
Holmen, H., 213, 222 Jacob, A.F., 529, 540
Holowaychuk, N., 272, 278 Jacobson, G., 374
Horen, A., 458, 468 Jahren, J.S., 122,124
Hornung, M., 40 Jaillard, B., 78, 79,101, 102
Horta, J.C. de O.S., 244, 255 Jalalian, A., 165,178
Horz, F, 55, 68, 70 James, N.P., 309, 310, 314, 345
Hosterman, J.W., 502, 512, 518, 522 James, O.B., 68
Hotz, P.E., 460, 468 Janot, C, 403, 404, 441
Housen, K., 70 Janssens, J.A., 221
Hower, J.H., 121,123, 124, 510, 522 Jaynes, W.F., 253
Hower, M.E., 121,124 Jeans, C.V., 122,124
Huang, P.M., 74, 79,102 Jenny, H., 19, 20, 39, 261, 279
Huang, W.H., 102 Jens, K., 248, 256
Hubbard, R., 69 Jercinovich, M.J., 256, 280
Hubble, G.D., 258, 346 Jeschke, L., 210, 224
Hubert, J.F, 527, 541, 551, 561 Jessup, R.W., 244, 247, 248, 249, 255, 351, 375
Huels, F.W., 208, 222 Johnes, L.H., 125
Johnson, C.W., 210, 222 Kittrick, J.A., 40, All, 440, 4SI, 469
Johnson, D.L., 5,15, 262, 279, 280, 309, 345, 477, Kivinen, E., 208, 222
497 Klappa, CF., 14,15, 310, 311, 316, 317, 340, 343,
Johnson, G.I., 556, 561 344, 345, 550, 561
Johnson, J.H., 124 Klecka, M., 152
Johnson, K.G., 549, 561 Knauss, K.G., 252, 256, 346
Johnson, L.R., 123 Knoll, A.H., 558, 561, 562
Johnson, M.E., 563 Knox, G.J., 309, 345
Johnson, R.G., 279, 280 Knuepfer, P.L.K., 306
Johnson, S.W., 204, 215, 222 Kobilsek, B., 433, 434, 439
Johnson, T.V., 70 Koch, C, 428, 440
Join, P., 402 Kocurek, G., 243, 256, 529, 530, 541
Jonasson, I.R., 222 Kodama, H., 193, 201
Jones, B.F., 96,104,115,116,124 Koerschner, WE, 498
Jongerius, A., 401 Kolker, O., 254
Jörgen sen, P., 123 Kolmogoroff, A.N., 62, 69
Juchler, S., 180 Komarneni, S., 85,102
Juchler, S.J., 167,176,275 Kontz, ER., 292, 304
Julien, A.A., 216, 222, 503, 522 Koroleva, G.P., 418, 440
Juli, A.J.T., 254 Korpijaakko, E.O., 222
Jurkovic, I., 460, 468 Korzhinskii, D.S., 24, 25, 39
Koster, H.M., 94,105
Kahalf, EL, 349, 350, 375 Kotilainen, M.J., 208, 222
Kahle, CF., 242, 256 Kotschoubey, B., 393, 405, 434, 443
Kakuto, Y., 75, 84, 85,105 Koud, J.M., 464, 468
Kalk, E., 257 Kounestron, O., 72, 75,102
Kämpf, N., 407, 426, 427, 439 Kovalev, V.A., 222
Kaplan, I.R., 38 Koyumdjisky, H., 254
Karathanasis, A.D., 130,152 Krassilov, V.A., 549, 561
Karlinger, M.R., 125 Kraus, M.J., 11, 480, 481, 483, 484, 485, 492, 493,
Karpoff, A.M., 403, 441 495, 497, 525, 526, 527, 530, 531, 532, 533,
Kassim, J.K., 426, 437 534, 535, 537, 538, 540, 541
Kaufman, A., 256 Krause, H.H., 190, 201
Keenan, A.C., 102 Krauskopf, K.B., 418, 440, 447, 452, 453, 469
Keller, E.A., 280 Kronberg, B.I., 304, 434, 440
Keller, W.D., 33, 39, 74,102, 518, 522 Krumbein, W C , 31, 32, 39
Kellog, CE., 379, 403 Krumbein, W.E., 39, 248, 256
Kelly, W C , 418, 438, 447, 468 Krumkachev, L.I., 233, 256
Kemp, R.A., 498 Ku, TL., 250, 252, 256, 315, 346
Kendall, CG.St.C, 241, 258 Kuhnel, R.A., 460, 469
Kent, L.E., 350, 375 Kuiper, G.P., 43, 69
Keys, D., 213, 222 Kupsch, W.O., 200
Khalil, M.A.K., 224 Kurtz, H.F., 441
Kidston, R., 549, 561 Kushnir, J., 242, 256
Kiehl, J.T., 223 Kwiatkowski, R.E., 223
Kimber, R.W, 346 Kyle, J.R., 447, 471
Kimberley, M., 75, 561
Kindle, E.M., 206, 222 Labroue, L., 166,178
King, L.C., 407, 408, 439 Lacroix, A., 454, 460, 469
Ladle, G.H., 69 Lett, R.E.W., 216, 222

Lamothe, R., 68 Lettis, W.R., 490, 497
Lamotte, M., 410, 440 Leveque, A., 394, 403
Lamplugh, G.W., 309, 346, 350, 375 Levesque, M., 208, 222
Landva, A.O., 205, 208, 222 Lewis, C.J., 253
Lang, H.R., 536, 541 Leyval, C , 102
Lang, W.H., 549, 561 Lie, O., 216, 223
Langford-Smith, T., 350, 375 Uebig, J., 215, 222
Langmuir, D., 427, 428, 440 Lieres, A., 105
Lappartient, J.R., 408, 441 Undsay, J.F., 13, 45, 47, 49, 50, 53, 55, 56, 57, 58,
Larriestra, C.N., 540 59, 60, 61, 63, 64, 65, 69
Larroque, P., 302 Lisitzyn, A.P., 125
Lasaga, A.C., 285, 304 Lister, J.S., 123
Laul, J.C., 55, 69 Litaor, M.I., 165,179
Lautridou, J.P., 161,178 Llorca, S., 469
Lavelle, P., 415, 440 Lobova, E.V., 225, 228, 231, 232, 238, 256
Lavkulich, L.M., 4, 5,14,15 Lock, D.E., 346, 374
Lawrance, L.M., 447, 449, 469 Locke, W.W., 289, 304
Lawrence, C.R., 317, 346 Lof, P., 225, 228, 229, 256
Lawton, K., 102 Logan, N.E., 114,124
LaZerte, B.D., 217, 218, 222 Logan, W.E., 501, 502, 503, 522
Le Conte, J., 215, 222 Long, A., 254
Leahy, G.D., 542 Loope, D.B., 528, 530, 541, 546, 561
Leamy, M.L., 298, 304 Lord, T., 222
Ledere, J., 443, 471 Lorphelin, L., 180
Lee, J., 216, 222 Lortie, G., 222
Lee, K.E., 415, 440 Loughnan, E C , 128,148,151
Lee, S.Y., 80,102 Lovelock, J.E., 558, 561
Lee, W.G., 273, 281 Lovering, T.S., 128, 138, 139, 141,152, 216, 222,
Leeder, M.R., 483, 495, 496, 497, 525, 532, 535, 519, 522
537, 539, 540} 541 Lowe, D.R., 561
Legros, J.P., 6, 7, 161, 162, 164, 166, 168, 171, Lowenstein, T.K., 242, 256
172,174,777, 178, 180 Lowman, P., 68
Lehman, T.M., 484, 497 Lowman, R., 68
Leinen, M., 253 Lucas, C , 177, 555, 561
Lelong, F., 40,178, 424, 440 Lucas, Y., 393, 396, 397, 403, 423, 440
Lemoine, J., 424, 440 Ludbrook, N.H., 320, 322, 323, 346
Lemos, R., 401 Lukashev, K.I., 216, 222
Lemos, V.P., 434, 440 Lulli, L., 173,179
Leneuf, N., 289, 304, 382, 383, 396, 403, 407, 408,
440 Mabbutt, J.A., 557, 563
Lenhoff, C.J., 541 Mabesoone, J.M., 529, 542
Lepage, M., 415, 440 Machairas, G., 418, 440
Lepiller, M., 164, ^ 79 Machette, M.N., 262, 266,267, 268, 277,279, 477,
LePoole, R.S., 69 478, 484, 497
Leprun, J.C., 383, 386, 387, 389, 391, 394, 403, Macias, F., 4, 33, 285, 290, 291, 298, 299, 300,302,
408, 412, 413, 428, 433, 437, 440 303, 304, 305
Lerman, A., 33, 40 Mackenzie, K.J.D., 100,102
Lesniak, P., 442 MacKinnon, B., 304
Macko, M.E., 254 McCarthy, P.P., 223

Mackowsky, M.-Th., 523, 563 McCleary, J.R., 540
Mackumbi, L., 74,102 McClennan, S.M., 21, 22, 40
Macphail, R.I., 475, 497 McConchie, D., 246, 253
Madsen M., 440 McCracken, RJ., 22, 38
Magaritz, M., 250, 256 McCraw, J.D., 486, 497
Magonthier, M.C., 103 McCrea, R.C, 218, 223
Mahadevan, R., 470 McDougall, I., 374
Mahaney, W.C., 269, 279, 475, 497 McDougall, J.D., 193, 201
Mah mood, A., 68 McDowell, S.D., 120,124
Maignien, R., 379, 393, 394, 403, 407, 408, 410, McFadden, L.D., 248, 256, 264, 266, 267, 271,
417, 433, 440 274, 275, 277, 280
Maksimovic, Z., 462, 469 McFarlane, MJ., 379, 386, 387, 390, 393, 395,
Malanin, A.N., 228, 232, 256 403, 407, 408, 410, 415, 417, 433, 435, 441,
Malia, P.B., 85, 94,102 480, 497
Mamy, J., 90,101 McGuffie, B.A., 541
Manœau, A., 100, 460, 462, 468, 469 McHardy, WJ., 101, 102
Manickam, S., 116,124 McKay, D.S., 49, 50, 54, 56, 57, 69
Manley, E.P., 300, 304 McKerrow, W.S., 560, 563
Mann, A.W., 418, 440, 443, 447, 448, 449, 450, Mckinnon, B.J., 440
451, 452, 453, 469, 471 McLaren, M., 386, 403
Mann, H., 78,102 McLean, S.A., 245, 256
Marbut, CF., 579, 586 McLennan, S.M., 132,152
Mareil Audiguier, M., 101 McPherson, J.G., 484, 497, 551, 561
Markewich, H.W., 272, 274, 276, 279, 280 McSaveney, M.J., 306
Markovics, G., 133,138,139,141,144,152 McSweeney, K., 525, 527, 541
Markowski, S., 206, 207, 222 Medvedev, Y.T., 421, 441
Marsh, R.D., 157,181 Meentemeyer, V, 238, 255
Martin, D., 393, 393, 394, 395, 403 Mehra, O.P., 96,102
Martin, R.E., 289, 303 Meilhac, A., 296, 304
Martins, VP., 551, 561 Meinhold, R.H., 102
Martyn, J.E., 556, 561 Meinschein, W.G., 68
Masarie, K.A., 224 Melfi, AJ., 296,304,404, 405, 433, 434, 434,439,
Mason, B., 27, 29, 40 441, 442, 460, 464, 467, 4ù, 470, 471
Matson, D., 70 Mellor, A., 162,179, 270, 280
Matson, P.A., 562 Mencher, E., 563
Matthews, E., 204, 223 Mengeilk, G., 105
Matti, J.C, 277, 279 Menillet, E, 363, 375, 376
Mattigod, S.V., 24, 40, 422, 440, 457, 469 Merifield, P.M., 254
Matukhina, V.G., 220 Mering, J., 88,101,102
Maucorps, J., 560 Merino, E., 467, 469
Mauguin, C, 79,102 Merriman, J.C., 218, 223
Mausbach, M.J., 279, 280 Mestdagh, M.M., 101
Mazaltarim, D., 410, 411, 415, 440, 442 Meunier, A., 24, 40, 75, 76, 102, 120, 124, 293,
Mazroa, M., 253 303, 304
Mazzullo, S J., 553, 563 Meunier, J.D., 467
McAfee, R., 375 Mew, G., 280
McBride, E.E, 525, 527, 542 Meybeck, M., 33, 40
McCalpin, J.P., 277, 280 Meyer, B., 257
Meyer, C , 53, 69 Mortensen, H., 235, 256

Michalet,R., 167,279 Mortland, M.M., 78,102
Michel, P., 397, 403, 407, 408, 409, 410, 412, 433, Morup, S., 440
434, 441 Moshe, R., 254
Middlemost, E.A.K., 572, 586 Mosser, C , 464, 469
Middleton, G.V., 11,15 Mountain, E.D., 350, 376
Middleton, L.T., 485, 497, 527', 541 Mousain, D., 100,344
Middleton, NJ., 236, 256 Muhs, D.R., 271, 272, 274, 276, 280
Mikhovich, A.I., 237, 258 Muir, I., 133,152
Miller, R.S., 125 Mulcahy, M J., 258
Miller, R.W., 8, 25, 583, 586 Muller, D., 401
Miller, T., 253 Müller, G., 115,124
Millot, G., 107,118,119,124, 125, 290, 296, 304, Muller, J.P., 96,100, 389, 392, 393, 394, 395, 401,
305, 357, 358, 360, 372, 376, 379, 382, 383, 403, 404, 426, 441
391, 396, 403, 404, 405, 407, 415, 417, 441, Muller, M., 172,279
442, 443, 445, 453, 454, 456, 463, 469, 470, Murray, J.W, 572,586
471 Muurman, R.P., 292, 304
Milnes, A.R., 11, 247, 256, 309, 310, 310, 311, Muysson, J.R., 152
317, 320, 322, 323, 331, 334, 340, 343, 346,
350, 351, 367, 369, 372, 374, 375, 376, 433, N'ziengui Mapangou, P., 469
441 Nadeau, P.H., 87, 94, 99,102, 110, 118,122,123,
Minarik, L., 142,152 124
Mingelgrin, U., 88,104 Nagel, J.S., 47, 68
Minkin, J.A., 68 Nahon, D.B., 3, 9, 24, 40, 80,103, 288, 289, 302,
Mishra, V.K., 181 304, 382, 383, 386, 387, 388, 389, 390, 391,
Mitchel, P.B., 303 392, 395, 401, 402, 403, 404, 405, 407, 408,
Mitchell, J.K., 68 410, 412, 421, 426, 428, 430, 431, 433, 435,
Mohr, E.C.J., 379, 395, 396, 403, 407, 441 437, 441, 442, 445, 446, 456, 457, 458, 460,
Moinereau, J., 173,179 461, 462, 467, 468, 469, 470
Mondain, PH., 1 6 M 79 Nair, N.G.K., 470
Mooney, H.A., 559, 562 Nanson, G.C., 481, 491, 497, 498
Mooney, R.W, 88,102 Nash, R.A., 440
Moore, A.W, 225, 228, 229, 256 National Wetlands Working Group, 208,209,223
Moore, C.B., 27, 29, 40 Naucke, W, 213, 223
Moore, C.H., 226, 253 Neeley, C.L., 441
Moore, D.M., 517, 522 Nelson, A.R., 277, 280
Moore, J.E., 115,124 Nelson, R.E., 256
Moore, J.G., 49, 69 Nelson, W.J., 523
Moore, O., 38 Nesbitt, H.W., 4, 128, 132, 133, 135, 137, 138,
Moore, P.D., 208, 223 139, 140, 141, 142, 143, 144, 145, 147, 148,
Morales, C , 236, 256 149,152, 224, 284, 304
Morel, G., 440 Netterberg, F., 310, 346
Morgan, D.J., 125 Nettleton, W.D., 226, 237, 238, 240, 245, 256, 375
Morgan, J.J., 290, 305 Newman, A.CD., 101,109, 111, 123,124
Morikawa, H., 124 Nicholas, J., 480, 498
Morris, R.C., 556, 562 Nicholls, G.D., 28, 40
Morris, R.V., 68 Nickel, R.E., 464, 470
Morrison, D.A., 69 Nieva, D., 123
Morrison, R.B., 477, 497, 527, 542 Nikiforoff, C.C., 5,15
Niklas, H., 216, 223 Ossaka, J., 124

Niklas, K.J., 543, 562 Ottman, E, 124
Niklas, K.T., 546, 561 Overbeck, E, 208, 223
Nilsen, T., 527, 542 Ozenda, P, 156,157,179
Ninov, N., 160,180
Nissim, S., 254 Paces, T., 289, 304
NkediKizza, P., 529,542 Page, W.D., 257
Noack, Y., 302, 468 Pain, CF., 162, 173,777
Nockolds, S.R., 129, 130,138,139,141,152 Palmieri, E, 124
Noller, J.S., 265, 280 Papike, J.J., 55, 69, 70
Norris, R.M., 249, 255, 351, 375 Paquet, H., 165,179, 297, 304, 305, 379, 403, 404,
Noms, T.A., 223 405, 441, 442, 461, 462, 467, 468, 469, 470
Norrish, K., 90,103 Parc, S., 458, 459, 460, 470
North American Commission on Stratigraphie Nomen Parfitt, R.L., 173,179, 262, 280
clature, 477, 498 Parham, W.E., 502, 503, 510, 513, 514, 522
North, C.P., 259, 499 Parisot, J.C, 300, 304, 464, 470, 471
Northcote, K.H., 232, 238, 252, 256, 257, 258, Parkinson, R.J., 191, 201
318, 346 Parks, G.A., 377
Norton, D.A., 434, 438 Parneix, J.-C, 124
Norton, S.A., 220, 221, 456, 470 Parrà, M.A., 442
Nosov, G.L., 125 Parrish, J.T., 433, 434, 435, 437, 441
Novikoif, A., 382, 392, 395, 397, 404, 405, 421, Parron, C , 85,100, 404, 467
422, 426, 435, 441, 442, 457, 470 Parsons, R.B., 477, 498
Nyamapfene, K.W., 247, 257 Party, J.P., 168,779
Nziengui-Mapangou, P., 441 Partyka, S., 443
Patin, C , 204, 223
O'Brien, N.R., 518, 523 Pattenden, G.E., 152
O'Melia, C.R., 116,123 Patterson, S.H., 435, 441, 502, 512, 518, 522
O'Neill, R.V., 559, 562 Pauling, L., 79,103
Oades, J.M., 255 Pautou, G., 156,179
Oberbeck, V.R., 46, 60, 61, 69 Pavich, M.J., 279, 280
Oberlander, T.M., 233, 248, 254, 257 Pavkulich, 3
Oberlin, T., 88,102 Pawluk, S., 40,164,179, 189,190, 201, 479, 498
Odora, I.E., 502, 513, 522 Paylor, E.D., 541
Ohira, Y., 124 Payne, J.N., 504, 523
Okada, K., 113,124 Peacor, D.R., 84,100
Okazaki, R., 164, 172,180 Peck, D.L., 49, 50, 69
Oliver, R., 123 Pedro, G., 3, 9, 75, 27, 28, 33, 35, 40, 71, 88, 707,
Olives Banos, J., 81,103 103, 104, 105, 123, 161, 779, 288, 290, 291,
Oilier, CD., 248, 257, 480, 498 294, 303, 305, 375, 380, 380, 395, 402, 404,
Olmsted, R.G., 560 All, 426, 438, 457, 467
Olsen, H., 493, 498 Peech, M., 287, 306
Olson, J.C., 441 Pellerin, J., 376
Orbell, G.E., 779 Pendleton, R.L·, 379, 386, 404
Oremland, R.S., 220 Peppers, R.A., 505, 519, 522
Orlova, M.A., 241, 257 Percival, C.J., 478, 479, 498, 553, 562
Orme, A.R., 254 Perret, D., 402
Ortlam, D., 484, 498 Perry, E.A., 124, 510, 522
Osborn, C.C., 208, 223 Perry, K., 570, 586
Perry, R.S., 248, 257 Rabenhorst, M.C., 314, 346

Pereeil, E.A., 457, 458, 468, 470 Radforth, N.W., 208, 223
Petersen, L., 292, 299, 305 Ragot, J.P., 104
Peterson, F.F., 226, 238, 240, 245, 256, 257, 345 Rainbird, R.H., 143, 144,145,152
Peterson, N.P., 464, 470 Rajan, R.S., 67, 68
Petit, M., 433, 442 Ramaekers, P., 562
Pettapiece, W.W., 190, 201 Ramanathan, V, 214, 223
Pettijohn, F.J., 65, 69,129,130,152, 575, 576, 586 Ramann, E., 204, 207, 223
Pettyjohn, W.A., 525, 528, 529, 542 Ranger, J., 74,100, 103,180, 305
Pfefferkorn, H.W., 513, 523, 551, 561 Rao, C.P., 515, 523
Pham, V.N., 387, 404 Rasmussen, R.A., 224
Pheeney, P.E., 208, 222 Rasplus, L., 376
Phillips, S.E., 310, 310, 320, 322, 331, 334, 340, Rateev, M.A., 116,125
343, 346 Rautureau, M., 104
Phillips, T.L., 516, 519, 522, 549, 560, 562 Raymond, A., 549, 562
Pierce, K.L., 274, 279, 289, 303 Rayner, J.H., 123
Pion, J.C., 387, 404, 410, 437, 442 Razzaghe-Karimi, M.H., 74,103,180, 305
Plassard, C , 100 Read, H.H., 579, 586
Plaziat, JC, 528, 541 Read, J.F, 309, 310, 317, 319, 320, 322, 339, 343,
Pluth, DJ., 180 346, 493, 498
Pochon, M , 164, 167,168,169,179 Reeve, R., 255
Pogregnyak, Y.E, 418, 440 Reeves, C.C., 309, 311, 313, 314, 320, 343, 344,
Poirier, J.E., 100 346
Polach, H.A., 250, 259, 315, 322, 344, 347 Reheis, M.C., 245, 251, 257, 265, 271, 274, 280
Pollard, L.D., 111, 125, 295, 306 Rehfuess, K.E., 181
Pomerol, C , 553, 562 Reid, A.M., 52, 69
Pons, C.H., 100, 105 Reid, I., 233, 257
Ponti, D.J., 490, 498 Reilly, G.A., 152
Potter, P.E., 522 Reimer, T.O., 556, 562
Potter, R.M., 248, 257 Reinhardt, J., 277, 280
Pouget, M., 243, 257 Reisenauer, H.M., 292, 305
Powers, D.W., 540 Rennie, R., 204, 218, 223
Prescott, S.A., 379, 386, 404 Resplendido, CI., 440
Previtali, F.,i77 Retallack, G.J., 4, 12, 13, 14, 15, 15, 128, 151,
Price, R.C., 152 278, 280, 476, 485, 498, 525, 527, 528, 530,
Price Williams, D., 259 537, 538, 542, 543, 544, 545, 546, 547, 548,
Prospero, J.M., 235, 236, 253, 257 550, 551, 553, 555, 557, 558, 559, 560, 562
Prost, R., 418, 442 Rex, G.M., 550, 563
Proust, D., 102, 293, 305 Reynolds, B., 33, 40
Purcell, C , 488, 496 Reynolds, R.C., Jr., 511, 523
Putzer, H., 393, 404 Rhoades, J.D., 260
Puustjävi, V, 213, 223 Rich, C.I., 73, 80,100, 103
Pye, K., 234, 235, 236, 239, 242, 257 Richard, L., 156,179
Richards, P.W, 554, 563
Quaide, W.L., 46, 60, 61, 69 Richardson, CT., 207, 223
Quakernaat, J., 115,124 Ridley, W.I., 69
Quantin, P., 177, 302 Rigg, G.B., 207, 213, 223
Quinif, Y., 162,179 Righi, D., 73, 74,103, 170,180, 560
Quirk, J.P, 90,103 Rimmer, S.M., 118,125, 503, 509, 512, 513, 523
Rimsaite, J.Y.H., 101 Sahu, K.N., 560

Risacher, E, 116,123, 240, 257 Saleh, A.M., 96,104
Ritchie, RE, 151 Salehi, M., 374
Rix, P., 253 Salisbury, J.W., 541
Roaldset, E., 123 Sallantaus, T., 217, 223
Robain, H., 94,103 Salop, L.J., 556, 563
Roberson, C.E., 456, 468 Samie, C , 160,180
Robert, M., 11, 25, 40, 72, 73, 74, 75, 76, 77, 78, Sanders, J.V, 375
79, 92, 95, 96, 97,100,101,102, 103,104,166, Sands, W.A., 415, 443
170,179, 180, 290, 291, 298, 303, 305 Sanford, R.L., 557, 563
Robertson, R.A, 213, 223 Santelmann, M.V., 221
Robin, A.M., 560 Santos-Ynigo, L., 460, 470
Robinson, D., 259, 484, 499, 503, 523, 525, 534, Santosh, M., 470
542 Sarna-Wojcicki, A.M., 279
Robinson, T.W., 255 Sawhney, B.L., 116,125
Rock, B.N., 541 Schaber, G.G., 70
Rockwell, T.K., 277, 280 Schafer, J.W., 544, 561
Roederer, P., 247, 253 Schattner, I., 234, 257
Rognon, P., 168,777, 179 Scheckler, S.E., 549, 561
Rolim, J.L., 542 Scheffer, F., 248, 257
Romero, R., 81, 83, 85, 86,104, 293, 293, 305 Schellmann, W, 379, 396, 404, 410, 442
Schleicter, D.L., 70
Ronov, A.B., 22, 40
Schlesinger, W.H., 250, 258
Roorda, HJ., 469
Schloms, B.H.A., 349, 372, 375
Roquin, C , 7, 410, 442
Schmadebeck, R., 68
Rose, J., 476, 498
Schmitt, A., 161,178
Rose, K.D., 525, 540
Schmitt, J.M., 376
Rosello, V, 393, 404
Scholz, G., 124
Ross, C.W., 273, 276, 280
Schonfeld, E., 69
Rosser, H., 375
Schopf, J.M., 546, 563
Rossman, G.R., 248, 257
Schreyer, W, 556, 563
Roth, E.S., 234, 257
Schubert, C , 257
Rothwell, G.W., 561
Schultz, L.G., 118,125
Rougerie, G., 410, 440 Schwartz, D.P., 281
Roulin, E, 372, 376 Schwebel, D., 346
Rowland, T.R., 280 Schwertmann, U., 9,15, 114, 115,125, 380, 396,
Roy, S., 457, 470 404, 426, 427, 428, 439, 442
Ruellan, A., 310, 316, 340, 346 Scotchman, I.C., 118,125
Ruhe, R.V., 3,15, 475, 491, 498 Scotese, C.R., 441, 550, 563
Runge, E.CA., 5,15 Scott, A.C., 563
Ruppert, L E , 515, 523 Scott, G.R., 529, 542
Rusanov, D.K., 220 Searle, P.L., 280
Rüssel, M., 179 Seddoh, F.K., 15
Russell, E.W., 415, 442 See, T.H., 68, 70
Russell, J.D., 77,101, 105, 296, 302 Sehgal, J.L., 172,180
Rust, B.R., 481, 491, 498 Seisser, W.G., 324, 346
Rust, U., 250, 259 Self, P.G., 322, 346
Rutter, N.W., 322, 346 Selmer-Olsen, A.R., 216, 223
Ruttner, F., 206, 223 Senior, B.R., 351, 376, 480, 497, 556, 563
Sepkoski, J.J., 543, 563 Smith, J., 296, 305

Serdyuchenko, D.P., 556, 563 Smith, K.L., 300, 303
Seret, G., 410, 442 Smith, R.M.H., 486, 498
Sevink, J., 477, 495 Smith, S.A., 124
Shadfan, H., 245, 258 Smith, S.M., 273, 281
Shaler, N.S., 208, 223 Smith, W.H., 518, 523
Sharp, R.R, 248, 254 Soepraptohardjo, M., 297, 302
Shaw, D.M., 132,132,137,152 Sofer, Z., 252, 258
Shaw, H.F., 117,123 Soil Conservation Service, 8
Shearman, D.J., 241, 258 Soil Survey Staff, 225, 226, 227, 228, 229, 248,
Shephard, J.A., 377 258, 297, 305, 586
Sherman, CD., 119,125 Sokal, R.R., 567, 586
Sherman, G.D., 24, 28, 39, 284, 295, 305 Sokolov, I.A., 297, 305
Shoemaker, E.M., 47, 47, 70 Sokolova, T.A., 291, 305
Shoji, S., 306 Solomon, J.E., 541
Shomer, I., 88,104 Sombroek, W.G., 393, 404
Shotyk, W, 10,11, 214, 215, 224 Sonko, S., 443
Shridar, K., 75,102 Soper, E.K., 208, 224
Shroba, R.R., 269, 270, 278, 280 Souchier, B., 158,164,172,180
Shumakov, VS., 237, 258 Southard, A.R., 165,178
Sieffermann, G., 28, 33, 40, 71,103,119,125, Sowers, J.M., 176, 313, 315, 344, 346
294, 296, 305, 396, 404 Spaltenstein, H., 165,180
Siever, R., 215, 224, 522 Speight, J.G., 495
Sigleo, W.R., 277, 280 Sperling, C.H.B., 235, 235, 257, 258
Sigolo, J.B., 454, 470 Spiers, G.A., 40,180
Sikora, W, 119,125 Spoon, M.D., 542
Sillitoe, R.H., 464, 470 Sposito, G., 301, 305
Silva, B., 303 Spurr, A.R., 80, 88,104
Simon, S.B., 55, 69, 70 Spyridakis, D.E., 78,104
Simon-Sylvestre, G., 74,104 Srodon, J., 92,104,122,125
Simonson, R.W., 262, 281 Stace, H.C.T., 232, 258
Simpson, E.S., 386, 404 Stach, E., 515, 523, 549, 563
Sinclair, I.G., 464, 470 Stainier, X., 244, 258
Singer, A., 116,119,125, 245, 258, 295, 305 Stanton, R.W., 523
Singer, M.J., 164,180, 529, 542 Staub, J.R., 479, 498, 503, 523
Singer, P.C., 235, 259 Stebbins, G.L., 558, 563
Singer, S., 316, 347 Steele, L.P., 218, 224
Singh, H.B., 223 Steensma, JJ.S., 469
Sivarajasingham, S., 379, 404 Steffens, D., 105
Sivtsov, A.V., 468 Steiner, A., 120,125
Sjörs, H., 224 Stephens, C.G., 351, 376
Sleeman, J.R., 258 Stevens, P.A., 40
Smale, D., 350, 351, 376 Stevens, PR., 261, 281
Smalley, I.J., 235, 254 Stevens, R.E., 293, 305
Smeck, N.E., 40, 259 Stevenson, FJ., 74,104
Smilie, G.W., 296, 303 Stewart, W.N., 514, 523
Smith, C.A.S., 164,180 Sticher, H., 176,180
Smith, C.L., 245, 254 Stilianidi, B., 440
Smith, D.G., 533, 542 Stoch, L., 119,125
Stocking, M.A., 249, 258 lavant, H., 177

Stoertz, G.E., 239, 258 Tavany, Y., 177
Stoffers, P., 116,125 Tavernier, R., 180, 286, 306
Stoffregen, R., 418, 442, 447, 470 Taylor, E.M., 263, 275, 276, 279
Stoops, G., 180, 254, 393, 404 Taylor, G.H., 498, 523, 534, 542
Stout, W, 503, 523 Taylor, R.M., 48, 69,133,152, 427, 442
Strasser, A., 528, 542 Taylor, S.R., 21, 22, 40, 70,132,152
Stremme, H., 216, 224 Tazaki, K., 81, 83,102, 104
Strom, R.G., 69 Tchoubar, C , 80,104
Strong, G.E., 124 Teague, T., 253
Strother, P.K., 549, 563 Teakle, L.J.H., 6,15,15, 31, 38
Strzemski, M., 19, 40, 579, 586 Tedrow, J.C.F., 186, 187,188,191,192,198, 201
Stuckey, B.N., 279 Teichert, C , 309, 317, 345
Stumm, W, 25, 40, 216, 224, 290, 305 Teichmuller, M., 523, 563
Suarez, D.L., 255 Teichmuller, R., 523, 563
Succow, M., 208, 210, 224 Tejedor-Salguero, M.L., 177, 298, 306
Suddhiprakan, A., 96,101, 296, 303 Teoharov, M., 160,180
Summerfield, M.A., 226, 258, 349, 350, 351, 376 Tercé, M., 96,103
Sutton, R.L., 70 Tessier, D., 11, 80, 88, 90, 98, 99, 100, 101, 102,
Swain, F.M., 218, 224 103, 104, 105
Swain, T., 555, 563 Thiry, M., 11, 349, 350, 351, 357, 358, 360, 361,
Swan, F.H., III, 277, 281 362, 363, 367, 369, 372, 373, 374, 375, 376
Swann, G.A., 41, 70 Thomas, D.S.G., 225, 233, 234, 236, 258
Swanson, D.K., 270, 274, 275, 281 Thomas, P., 70
Swardt, A.M.J. de, 415, 442 Thompson, C.H., 272, 281, 346
Swett, K., 226, 258 Thompson, G.R., 529, 542
Swindale, L.D., 119,125, 295, 305 Thompson, J.B., 24, 40
Syers, J.K., 262, 281 Thompson, R.M., 377
Sys, C , 180 Thompson, T.D., 68
Thoning, K.W., 224
Taboada, T., 305 Thony, J.L., 443
Taddeucci, A., 68 Thornber, M.R., 464, 471
Tait, J.M., 102 Thurman, E.M., 217, 224
Takeda, H., 69 Tiffney, B.H., 561, 562
Talbot, M.R., 487, 488, 490, 498, 529, 542 Tilton, D.L., 214, 221
Tailing, J.F., 217, 224 Tinsley, J.C, 266, 280
Tallis, J.H., 210, 224 Toledo-Groke, M.C., 464, 471
Talour, B., 164,179 Torrent, J., 426, 442
Tandy, J.C, 96, 97,104 Tosca, C, 166,178
Tardy, Y., 7, 9, 19, 24, 39, 40,103, 105, 124, 285, Toulemont, J., 424, 442
288, 291, 292, 294, 296, 303, 304, 305, 306, Touret, O., 90,105
382, 383, 385, 386, 387, 389, 392, 395, 397, Tourtelot, H.A., 518, 523
402, 403, 404, 405, 407, 408, 413, 415, 419, Tracey, J.I., 454, 468
421, 422, 425, 426, 428, 429, 430, 431, 432, Trauth-Badaud, D., 468
433, 435, 439, 441, 442, 443, 446, 453, 456, Trauth, N., 376
457, 467, 468, 470 Traverse, A., 549, 563
Targulyan, V.O., 192,192, 201, 290, 305 Tremblay, L.P., 556, 563
Tarnocai, C , 190, 201 Trescases, J.J., 40, 289, 302, 306, 382, 405, 434,
Tauxe, L., 485, 495 441, 442, 460, 461, 467, 468, 470, 471
Tributh, R , 78,105 Vdagawa, S., 523

THcart, X, 247, 258 Veblen, D.R., 80,105
THcart, L E , 393, 405 Vedy, J.C., 289, 306
Tìngila, R., 68 Velde, B., 24, 40, 75, 76,102, 122, Ì24, Ì25, 293,
Ttìki, R., 80,101 295, 300, 303, 304, 306
Trolard, E, 422, 426, 428, 429, 430, 431, 435, 442, Veneau, G., 40, 75,103, 104
443 Verger, J.P., 175,176,180, 181
Trombka, J., 68 Vermaat, J.G., 556, 563
Trosset, L , 172,180 Veverka, J., 42, 67, 70
Thic, G., 376 Viehaye, L , 101
Thickenbrodt, W, 393, 405, 434, 443 Vieillard, P., 470
Tsawlassou, G., 404 Vieillefon, J., 252, 258
Tucker, M.E., 244, 258, 492, 498, 499 Villadsen, J., 440
TUrchenek, L W , 255 Villas, R.N., 434, 440
Tbrkevich, A.L, 48, 49, 70 Vinsot, A., 376
TUrland, M., 363, 372, 376 Violente, A., 79,102
Turner, A.R., 460, 471 Vitousek, P.M., 562
Turner, R, 486, 495 Vletter, R. de, 460, 471
Turrin, B.D., 254 Vogg, R., 233, 258
TWidale, C.R., 247, 259, 349, 350, 351, 370, 375, Vogt, J., 410, 443
376, 377 Void, T., 157,181
Volkoff, B., 395, 396, 403, 405, 435, 443
Uehara, G., 102,119,125, 295, 305 Volobuyev, V.R., 6, 7,15
Ugolini, F.C., 7,15, 164,180, 198, 201, 269, 281, von Bùlow, K., 204, 206, 207, 224
288, 291, 292, 295, 298, 306 Von der Borch, C.C., 320, 346
Urey, H.C., 43, 70 von Post, H., 204, 207, 224
U.S. Department of Agriculture, 34, 40 von Post, L , 204, 205, 206, 207, 208, 211, 224
Vreeken, W.J. van, 261, 281
Vachaud, G., 424, 443 Waal, S.A. de, 460, 471
Vairinho, M., 100
Valarelli, J.V., 458, 471 Wackermann, J.M., 410, 443
Valentine, K.W.G., 475, 476, 499 Wada, K., 75, 84, 85,105, 306
Valeton, I., 396, 405, 421, 434, 443 Wahlstrom, E.E., 128,141,142,152
Vali, H., 94,105 Waide, J.B., 562
Vallès, V, 424, 443 Walker, J., 204, 224
Van Arsdale, R.B., 277, 280 Walker, P.H., 477, 499
Van Baren, F.A., 403, 441 Walker, R.E., 541
van Breemen, 40 Walker, T.R., 239, 258
van der Gaast, S.J., 104 Walker, T.W, 261, 262, 281
van der Marel, H.W., 80,100 Walther, J., 10,15, 386, 405
van Oort, E, 74,105,166,177, 179, 180 Walton, E.K., 496
Van Raij, B., 287, 306 Wan, H.M., 467
van Reuwijk, L.P., 299, 302 Wan, H.N., 462, 467
van Ryswyk, A.L., 164,172,180 Wanless, H.R., 503, 523, 525, 542
Van Schuylenborgh, J., 403, 441 Ward, J.D., 251, 260
Vane, G., 541 Warner, J., 69
Vasconselos, P., 447, 471 Warren, E.A., 118,125
Vauclin, M., 424, 443 Warren, J.K., 241, 258, 309, 323, 347
Vaudour,J., 162,777 Wash burn, A.L., 255
Wasmund, E., 205, 207, 224 Whittemore, D., 428, 440

Wasson, R.J., 344 Wieder, M., 245, 247, 259, 260, 488, 488, 499
Waters, A.C., 49, 70 Wieland, E., 224
Watson, A., 226, 235,239, 240, 241, 242, 243,244, Wieneke, F., 250, 259
245, 246, 248, 249, 250, 251, 255, 258, 259 Wìgley, T.M.L., 245, 259
Watson, J., 579, 586 Wilbert, X, 170,180
Wktson, J.A.L., 438 Wìlcox, B.P., 175,181
Witson, J.P., 415, 443 Wilde, S.A., 104,190, 201
Watson-Stegner, D., 4, 5, 13, 15, 262, 279, 411, Wilding, L.P., 37, 40, 230, 232, 259, 346
497 Wilke, B.M., 181
Watts, N.L., 244, 245, 259 Willard, B., 549, 563
Watts, S.H., 351, 377 Willard, D.A., 505, 523
Wax, D., 70 Williams, B.P.J., 485, 487, 495
Weaver, CE., I l l , 118, 125, 135, 152, 295, 306, Williams, E.G., 496, 502, 518, 522, 523
555, 563 Williams, G.E., 250, 259, 315, 347, 544, 563
Weaver, R.M., 292,306 Williams, LA., 350, 373, 377
Webb, T., 544, 560
Williams, M.A.J., 488, 490, 498
Weber, B.N., 434, 443
Williamson, W.O., 351, 377
Weber, F., 421, 443, 458, 471
Willman, H.B., 504, 523
Weber, N.A., 415, 443
Wilson, A.F., 418, 443, 447, 449, 450, 471
Websky, J., 218, 224
Wilson, A.M., 320, 331, 347
Webster, J.G., 418, 443, 447, 448, 449, 471
Wilson, A.T., 196, 201
Wedepohl, K.H., 128,132,137,152
Wilson, M.J., 77, 92,102, 105, 514, 523
Wehrli, B., 224
Wilson, R.E., 148,152
Weil, E.B., 254
Windom, H., 124
Weimer, R.J., 530, 542
Weir, A.H., 116,123, 124 Windom, H.L., 116,125
Weiss, E, 216, 224 Wing, S.L., 536, 537, 542
Weldon, R.J., II, 271, 277, 280 Winkler, E.M., 235, 259
Weller, J.M., 503, 523 Winkler, H.G.F., 580,586
Wellman, H.W, 196, 201 Winston, R.B., 517, 523
Wells, L.E., 280 Witty, J.E., 256
Wells, R.T., 350, 351, 377 Wnuk, C , 513, 523
Wells, S.C., 280 Wolf, C.L., 375
Wells, S.G., 256, 495 Wollast, R., 291, 306
Werding, G., 563 Wones, W.R., 26, 39
Westbroek, P., 78,105 Wood, L.A., 102
Weyant, WS., 188, 201 Wood, T.G., 415, 440, 443
Whalley, W.B., 351, 376 Woodford, N., 215, 224
Wheeler, G.A., 221 Woodyer, K.D., 534, 542
White, A.J.R., 294, 306 Woolnough, WG., 349, 377
White, E, 544, 563 Wopfner, H., 247, 259, 351, 377
White, W.A., 510, 516, 517, 518, 522 Working Group on the Origin and Nature of Pa-
Whitehouse, I.E., 289, 306 leosols, 475, 499
Whiteman, CA., 498 Wright, H.E., Jr., 261, 281
Whiteside, E.P., 584, 585, 586 Wright, M.J., 238, 257, 372, 375, 377
Whitford, W.G., 234, 259 Wright, VP, 12, 246, 259, 277, 281, 476, 484, 486,
Whitley, D.H., 254 488, 491, 499, 503, 523, 525, 534, 542, 550,
Whitney, G., 125 562, 563
References index
Yaalon, D.H., 236, 244, 245, 247, 251, 254, 256, Zabel, T.H, 254
259, 263, 279, 316, 347, 475, 488, 499 Zbinden, E.A., 561
Yair, A., 310, 316, 347, 499 Zdislava, K., 152
Yakushev V.M., 415, 443 Zebrowski, C, 173,181
Yam, O.L., 553, 561 Zeegere, H., 439, 464, 468, 469
Yareshevsky, A.A., 22, 40 Zeigler, A.M., 557, 560
Yatsu, E., 19, 40 Zeissink, H.E., 460, 471
Yin, L., 68 Zen, E.A., 456, 471
Young, G.M., 128, 132, 135, 137, 138, 139, 140, Zenger, D.H., 553, 563
141, 142, 143, 144, 145, 147, 148, 149, 152, Zhukhovitskaya, A.L., 222
285,304 Ziegler, A.M., 563
Young, R.W., 350,377 Zirgler, A.M., 441
Yousaf, M., 239, 260 Zoltai, S.C., 190, 201
609
Subject Index
Accumulation rates, 56, 60, 61, 65, 275, 526, 530, Antarctica, 188,194,196, 200, 234, 551
537, 538 Apatite, 262
Acidification, 26, 74,165-167,169,172,174,176, Aragonite, 22, 206, 328
217 Archean, 13,144,146
Acido-complexolosis, 290 Arctic brown soil, 188, 191,192,198, 199
Acidolysis, 290 Arene, 381, 383, 387
Active layer, 183,184,189 Argillic horizon, 251, 262, 551, 582
Adamellite, 129,138,139 Aridisol, 13, 226, 227, 229, 230, 230, 232, 238,
Aerosol, 23, 236, 237, 240, 289 248, 251, 484, 490, 550, 551, 553, 581, 583
Agglutinate glasses, 55 Asteroids, 66, 67
Agglutinates, 47, 50, 51, 53, 54, 54, 55-57,57, 58, Attapulgite, 165, 245
59, 62, 63-65 Authigenesis, 107,115, 237-239, 245
Agglutination, 54, 55, 59, 61, 63, 64 Autochthonous, 165,168, 205, 236, 408, 415
Aggradation, 477, 483, 486, 490, 493, 526-528, Avulsed, 483, 534, 535
536 Avulsion, 482, 483, 483, 484, 486, 487, 493, 494,
Aggradational, 527, 528, 535, 538 534, 535, 539
Aggrade, 476, 482, 530, 532
Aggrading, 12, 313, 314, 343, 525, 526, 539
Albedo, 42, 67 Badlands, 225, 249, 485, 528
Alfisol, 13, 74, 85, 175, 227, 229, 271, 297, 484, Basalt, 48-53, 114, 119, 129, 129, 140, 141, 148,
486, 529, 553, 583 148,149,149, 243, 289, 295, 296, 296, 300
Allochthonous, 4,164,165, 205-207, 236, 237 Bauxite, 115, 236, 379-381, 391, 396-398, 398,
Allophane, 81, 83, 87, 99,108,113,115,122,129, 399, 407, 409, 409, 417, 419, 421, 428, 430,
130,133,143,173,191, 288, 290, 296-301 432, 433, 434, 434, 435, 436, 437, 446, 453,
Alteration domain, 382 456, 464, 556, 558
Alterite, 391, 392 Bauxitic, 115, 397, 398, 408, 422, 435, 450, 453,
Alunite, 370, 374 455-457, 556
Amino acid, 322, 323 Bauxitization, 453, 454, 454, 456
Amphibole, 80, 81,118,132, 175, 286, 386, 570 Beidellite, 75, 92,112,130, 291, 295, 300
Amphibolite, 296, 299, 300, 389, 415 Biostasis, 37, 285
Analcite, 75, 577 Biotite, 72, 76, 81, 83, 84, 84, 96, 111, 118, 120,
Anatase, 117, 351, 352, 354, 358, 368, 509 129,129,130,132,143,147, 285, 286, 288,
Andésite, 21, 296, 547, 548 296, 386, 389, 454, 455
Andosolisation, 35, 36, 294, 298, 301 Bioturbation, 327, 339, 408, 417, 533, 547
Anhydrite, 235, 577 Birnessite, 112, 459, 459
Anoxie zone, 215, 218 Bleached zone, 390
610 Subject Index
Boehmite, 24, 113, 115, 299, 397-399, 421, 428, Cenozoic, 20,115, 294, 370, 536
429, 429, 430, 430, 431, 435, 455-457, 503 Chalcedans, 366
Bog, 10, 11, 161, 205, 206, 208-210, 211, 212, Chalcedony, 363, 366, 366, 368
213-216, 219 Chalcopyrite, 217, 464
Bog lakes, 206 Chalk, 118, 206, 309, 316
Boreal, 7, 9, 217, 269, 298, 299, 554 Chaparral, 546
Brakeland, 548, 558 Charcoal, 546
Brown water, 205, 206 Charnockite, 296, 299
Brucitization, 81 Chlorite, 7, 22, 36, 74, 75, 80, 81, 84, 108, 109,
Brunification, 198,199 111, 114-118,120-122,149,165,168,170,
Burrow, 327, 328, 528, 544-547, 550, 551, 553 175,193,198, 296, 297, 478, 509, 510-514,
518, 520, 521
Calcarenite, 309, 311, 317, 320, 323 Chronofunction, 261, 277
Calcareous crust, 311 Chronosequence, 7, 261-263, 264, 265, 268, 269,
Calcic, 35, 79, 170, 226, 231, 237, 268, 269, 271, 271-273, 275, 276, 276, 277, 277, 278, 481,
285, 310, 315, 316, 582 490
Calcretes, 226, 235, 241, 242, 244, 244, 245-247, Clay illuviation, 99,199, 237, 548
252, 309-317,318,319,319,320,321,322- Coal, 10, 118, 218, 477-479, 501-504, 504, 505-
324,324, 325, 325,326, 327, 328,328,329, 507, 510, 512-516,516, 517-521, 549, 550,
330, 330, 331, 331,332, 333,333, 334,335, 553, 573
336, 337,337, 338-340,341,342-344,380, Comminution, 55, 57, 61, 63, 235
478, 490, 527, 528, 531, 546, 548, 550, 552 Complexolysis, 28
Calcshale, 165-167,174 Concretion, 31, 393-395, 422, 506, 517, 519
Caliche, 309, 311, 313, 582 Cordierite, 300
Cambic, 159,172, 226, 231, 237, 238, 582 Corrosion, 162, 352, 360, 418
Cambisol, 158,159, 170, 172, 173, 175, 285, 292, Corundum, 115, 294, 421, 429-431, 456, 456, 556
583 Covellite, 217, 464
Carapace, 380, 386, 387, 390, 391, 409, 414 Cretaceous, 122,167,309, 314, 362,364,365,369,
Carbonate, 11, 22, 29, 30, 32, 33, 37, 77-79,145, 397, 408, 419, 434, 435,436, 437, 480, 528,
151,158,162,164,166-169,174,175,191, 549, 551, 556
199, 206, 215, 216, 231, 241, 245, 246, 262, Cristalliplasma, 398
263, 266, 267, 267, 272, 278, 309-312,312, Cryoturbation, 190,192
313-317, 319, 320, 322, 323, 325, 327, 328, Cryptomelane, 421, 459, 459, 460
334, 337-340, 342, 343, 344, 349, 446, 447, Crystallite, 90, 92, 94, 242, 342, 503, 513
457, 458, 462, 464, 466, 478, 484, 511, 517, Cuirasse, 380, 387, 389-391, 409
528, 530, 536, 550, 551, 553, 558, 569, 573, Cutan, 239,330,336,339, 352, 357, 366, 368, 369,
576, 577, 579 373
—, nodules, 478, 536, 551, 553 Cyclothem, 502-504, 504, 505, 507, 517, 518
Carbonic acid, 107,198,199
Carboniferous, 13, 141, 150, 218, 283, 477, 491, Decoloration, 392
501, 514, 534, 549, 550, 553, 555 Deferruginization, 392
Carr, 550 Dehydration, 23, 24,115, 238, 242, 421, 424, 431
Cassiterite, 415, 417 Desert, 9, 13, 116, 165, 185-189, 192-194, 194,
Catena, 10, 11, 24, 247, 270, 316, 476, 481, 482, 195,196,196,197-200, 225-227, 229, 231-
488 243, 243, 245, 246, 246, 247-252, 262, 264,
Catenary, 24, 310, 314, 316, 340, 481, 486, 488, 265, 311-313, 315, 543, 546, 550, 557
494 Desilicification, 71, 87, 501
Cation exchange, 71, 94,168, 213 Devonian, 13, 372, 485, 493, 548, 549, 551, 552,
Celadonite, 111 553, 558
Subject Index 611
Diagenesis, 3, 14, 17, 19, 85, 121, 127, 128, 143, Evaporate, 29, 244, 291
144,146-151, 245, 246, 278, 283, 320, 349, Evaporation, 9, 30, 188, 195, 210, 211, 212, 235,
501, 503, 507, 510, 511, 514, 517, 521, 555 237, 239, 240, 242, 243, 247, 251, 312, 313,
Diaspore, 113,115, 421, 429, 431, 456, 503 320, 340, 357, 369, 373, 374, 426, 427
Diatom, 116, 206 Evaporites, 30, 118, 237, 239-241, 493, 493, 573,
Dickite, 109,130 577
Diorite, 129, 139, 296 Evapotranspiration, 25,162,173, 226, 426
Disaggregation, 19,166 Exchange capacity, 174
Dissolution, 9, 10, 23, 71, 73, 77, 79, 81, 84, 85, Exogenic, 21, 41, 42, 586
122,132,133,133,134-136,140,162,164, Exsolution, 81, 85
166,175,176, 209, 214-216, 218, 219, 234,
242, 251, 252, 311, 313, 317, 319, 323, 324, Fades, 10, 11, 14, 15, 32, 92, 299, 310, 313, 314,
325, 321, 340, 342, 344, 353, 356-358, 360, 317, 319, 320, 321, 343, 352, 353, 358, 360,
361, 361, 363, 366, 367, 369, 373, 381, 383, 363, 370, 371, 371, 390, 392, 397, 407, 409,
391-393, 418, 427, 445, 447, 448, 450, 452, 410, 411, 412, 413, 492, 493, 503, 525, 528,
453, 456, 457, 460 538, 576, 577, 579
Dolomite, 22, 320, 327, 328, 330, 330, 331, 334, Fan, 247, 271, 315, 481, 486, 486, 487, 487, 488-
336, 337-339, 342-344, 492, 510, 511, 550, 490, 492, 494, 527, 529, 576
551, 552, 553, 577 Feldspar, 7, 50, 71, 72, 75, 80, 81, 83, 85, 99,118,
Dolomitization, 551 128,129,129,130,131,131,132,133,134-
Dopplerite, 205 137,139-141,141,142,142,143,146,147,
Dorbank, 349, 372 149,150,165,170,193, 238, 285, 288, 291,
Dune, 190,191, 225, 233, 237, 238, 251, 272, 273, 295, 299, 300, 327, 330, 339, 386, 454, 459,
309, 311, 317, 319, 320, 321, 322-324, 324, 507, 510-512, 514, 516, 518, 520, 521, 569,
325, 327, 328, 330, 330, 331, 339, 489, 528, 570, 572, 573, 573, 579
546 Feldspathoid, 75, 573
Dunite, 51, 296 Fellfield, 546
Duricrust, 4, 241, 309-311, 349, 350, 380, 390, Fen, 11, 208-210, 213-215, 550
399, 401, 409, 413, 421, 480 Ferrallitic soils, 37, 379, 380, 401, 426
Duripan, 226, 237, 241, 251, 349, 372, 476, 582 Ferrallitisation, 35, 36, 37, 71, 298, 299, 301
Dy, 159,172,173,176, 205, 206, 553 Ferricrete, 309, 380, 381, 381, 383, 384, 386, 387,
389-393, 395, 396, 399, 401, 407-409, 409,
Eclogite, 285, 299 410,411,412, 413, 413, 414-416,416, 417-
Electrostatic transport, 41, 42 419, 420, 421, 424, 427, 428, 430, 432, 433,
Eluvial, 225, 388, 479 433, 434, 435, 436, 437
Eluviated, 170, 239, 531 Ferrihydrite, 79, 99,108,114,115, 396, 426
Eluviation, 238, 248, 357, 358, 476, 478 Ferruginous soil, 380
Endogenic, 21, 41, 42, 67, 586 Ferruginous nodule, 449, 450, 453, 528
Entisols, 13, 225, 227, 229, 230, 232, 237, 251, Fireclay, 507, 510-513, 521
482, 486, 490, 492, 544-546, 583 Flint, 360, 362, 362, 363, 503, 505, 510, 512, 514,
Entropy, 5, 430 515-518, 521
Epitaxie, 96 Fluvisols, 225, 228, 229, 237, 251, 252, 583
Eocene, 323, 351, 352, 361, 362, 397, 408, 434, Flysch, 36,167
435, 436, 480, 483, 485, 526, 528-530, 535, Freeze and thaw, 162, 183,190
544, 551, 553, 555
Eolian dust, 165, 168, 176, 216, 234-236, 244, Gabbro, 53, 114, 129, 129, 138, 139, 176, 292,
247, 277, 311, 312, 314 296, 300, 415
Epipedon, 226, 229, 231, 531, 582 Gannister, 478, 479
Eutrophic, 207, 208, 213, 214 Garnet, 300, 460, 466
612 Subject Index
Garnierite, 460, 461 Halloysite, 81, 85, 86, 87, 90, 92, 93, 96,108,109,
Gibbsite, 7, 24, 31, 36, 108, 113, 114, 117, 129, 114,129,130,133,173, 298-301, 454
130, 238, 288, 290, 291, 294-297, 299-301, Hamada, 225
382, 383,394, 396-399,401, 408, 421-423, Haploidization, 5
423,428,429,429,430,430,431,435,454- Hardpan, 31, 309-311, 314, 316, 320, 333, 333,
457, 460, 582 334, 336, 344, 349, 372
Gilgai, 229, 248, 551 — calerete, 310, 314, 333, 334, 336
Glaciation, 20, 21, 37,185,192, 288, 479 — horizon, 313, 334
Glaebular, 351, 381, 381, 383, 390, 393, 395, 399, Heat energy, 65
415 Hematite, 9, 24, 108, 112, 114, 115, 238, 380,
— zone, 380, 381, 387, 391-393, 394 381, 383, 389-391, 393,394, 395, 396,396,
Glass, 33, 47-52, 52, 53, 54, 55, 57, 58, 61, 62, 397-399, 401,408,414, 415,418,421,426-
63-65, 75, 81,113,117,164, 298, 510, 518 428, 430-432, 435, 455-457, 537, 577
Glauconite, 31, 85, 92, 111, 577 Hisingerite, 462
Gley, 35, 189, 289, 477, 482, 507, 510, 511, 513, Histosol, 13, 549, 583
514, 583 Hollandite, 112
Gleying, 14, 30,190,191, 273, 476, 501, 544 Holocene, 237, 240, 261, 267, 269, 270, 273
Gneiss, 159,169, 286, 289, 296, 415 Homogenization, 5, 413, 415
Goethite, 24, 32, 72, 80, 81, 82, 96, 97, 108,113, Horizon, 5, 9,11, 30, 31,32, 36, 37, 74, 75, 79, 94,
115,189, 295, 297, 300, 301, 381, 383, 385, 113-115,144,161,164-168,170,172,175,
388, 390, 391, 393, 394, 395-399,401, 408, 176,185,189,190,192-196,196,197-200,
414, 415, 418, 421, 426-428, 430-432, 435, 204, 225-227, 229, 231, 233, 235, 237-242,
454-457, 460-463, 466, 537, 577 244-246, 248, 250, 251, 262, 263, 263, 264,
Goethitic cortex, 390, 395 265, 267, 269,269,270-273,276, 278,285-
Gold, 9, 416-419, 420, 446-448, 448, 449, 450, 287, 289-292, 297-301, 309, 310-312,312,
451, 452, 453, 466 313-317, 319, 320, 323-325, 327, 330,330,
Granite, 7&-78, 83, 119,129, 131, 138,138, 139, 331,332, 334, 336-340, 342-344, 351-353,
140,141,142,143, 144,145,146,146,147, 354,355, 356-358, 360, 363, 367-369, 370,
150,169, 235, 285, 286, 289, 292, 296, 297, 372-374, 379, 380, 382, 383, 384, 385-387,
299, 309, 331, 414, 415, 454-456 389, 390-394,394, 395,396, 397-399, 401,
Granitoid, 172, 294, 296, 389, 572 401, 407, 409, 410, 415, 417, 419, 421, 422,
Granodiorite, 129, 137, 138, 138, 139-141, 141, 424, 445-447, 449, 450, 453, 455, 457, 459,
142,142,143-145,146,147,150, 296 460, 462, 464, 477-479, 484, 492, 527, 530,
Grassland, 14, 298, 544, 551, 583 531, 546, 549, 551, 552, 553, 554,554, 555,
Gravel, 193, 264, 309, 319, 323, 327, 328, 330, 555, 556, 570, 579, 581-583
337, 338, 368 Horizonation, 5, 36, 229, 237, 477, 568
Greenhouse gasses, 10 Hornblende, 118,129,129, 289, 455, 572
Humification, 185, 205, 213
Greenstone, 175, 299
Hummocks, 190, 248
Groundwater silcrete, 361, 363, 364, 366, 369,
Humus, 159, 167, 168, 192, 198, 262, 295, 298,
370, 372-374
301, 553, 554, 555, 582, 583
Gypcrete, 241, 244, 247, 527
Huronian, 144
Gypsic, 226, 231, 237, 238, 241, 244, 245, 248,
Hydration, 23, 80, 88, 193, 198, 234, 235, 284,
265, 491, 582
421, 424
Gyttja, 205, 206
Hydrobiotite, 143
Hydrocarbon, 495, 526
Halite, 200, 239, 241-244, 246, 251, 262, 264, 265, Hydrolysate, 26, 27, 29, 290, 301
334, 577 Hydrolysis, 10, 25, 26, 28, 71, 74, 79, 85, 87, 115,
Halite pseudomorphs, 492 234, 283, 284, 290, 521
Subject Index 613
Hydromorphic, 74, 166, 170, 237, 243, 245, 250, Karst, 162,163,164,169, 313, 328, 333, 333, 334,
363, 432, 477-479, 488 340, 363, 398, 421
Hydromorphism, 476, 478, 486 Karstic, 311, 324, 325, 344, 383
Hydrothermal alteration, 107,118,120, 461 — topography, 324
Kerolite, 462, 463
Ice sheet, 168,184,185, 299 Kinetic energy, 58, 61, 64, 65
Iddingsite, 80, 462 Krummholz, 546
Illite, 75, 85,87, 88, 90, 92,94,95, 96, 98, 99,108,
109-112,114-118,120-122,130,133,135, Lapiaz, 162,163
136,143-145,149,151,165,168,192,198, Lapilli, 49
238, 295, 300, 334, 338, 459, 503, 507, 509, Laterite, 9, 379, 380, 380, 387, 393, 401, 407-409,
510, 511, 513, 514, 517, 518, 520, 521 409, 417, 418, 422, 428, 430, 432, 435, 448,
Illitization, 116 460, 462, 479, 480, 529, 556-558, 582
Illuvial, 225, 227, 229, 235, 237-239, 241-244, Laterization, 407
246, 248, 250, 251, 317, 327, 339, 340, 343, Latosol, 395, 396
352, 353, 356, 357, 363, 484, 582 Leaching, 25, 26, 30, 71, 75, 85, 135, 137, 140,
— deposits, 356-358, 368, 369 142,147,173,185,190, 227, 237, 240, 246,
Illuviation, 251, 340, 343, 344, 357,368, 369, 373, 250-252, 278, 291, 296, 317, 357, 360, 363,
476, 519 381, 383, 386, 387, 389, 391-393, 446, 478,
Ilmenite, 51, 408, 415 479, 488, 553, 554
Imogolite, 81, 99, 108, 113, 115, 129, 130, 133, Lepidocrocite, 78, 108,113,115
290, 298, 299, 301, 382 Leptosol, 4,158,159,169,172,175, 285, 583
Inceptisol, 13, 237, 426, 546, 548, 549, 553, 554, Leptynite, 172
583 Leucite, 75
Intraclast, 324 Ligands, 216
Iron nodule, 390 Limestone, 118,164,165,167-169,174,174,176,
Ironstone, 506, 577 216, 272, 283, 309, 313, 314, 323, 334, 350,
Isalteritic, 392 351, 361, 361, 363, 363, 371, 372, 504, 505,
511-513, 518, 550, 551, 573, 577
Jarosite, 193, 512 Lithiophorite, 460
Jurassic, 122, 167, 397, 408, 409, 432, 434, 435, Lithoclasts, 319, 323, 325, 325, 327, 328, 328, 333,
437, 480, 529, 546, 551 333, 336, 338, 339
Lithomarge, 379-381, 383, 384, 385-388, 391,
Kamacite, 51 395, 397, 399, 401, 408, 409, 416-419
Kandites, 143-145, 149 Lithorelict, 389, 389, 394
Kaolin, 84, 389, 558 Lithorelictual, 384, 388, 389, 390, 392, 395
Kaolinite, 7,10, 24, 31, 36, 37, 75, 80, 81, 83-85, Iithosols, 225, 227, 229, 237, 251
87, 88, 90, 93, 94, 96, 97, 98, 99, 108, 109, Loess, 117,130, 234, 236, 243, 251, 313, 317, 342,
114,116-121,121,122,129,130,133-136, 479
144,168,191-193,198, 238, 286, 290, 294, Lunar soil, 45, 46, 46, 47-51, 52, 53, 54, 55, 56,
295-297, 299-301, 314, 334, 338, 352, 379, 56, 57, 58, 59, 61, 62, 63-67
380,380, 381, 381, 382, 383, 385-394, 394,
395, 396, 396, 397, 399, 401, 401, 408, 415, Maghemite, 112,114
419, 421-423, 423, 426, 428, 431, 432, 435, Manganile, 421, 459, 459, 577
446, 454-457, 460, 466, 478, 503, 506, 507, Marcasite, 217, 505, 510, 511, 515
509, 510-521, 529, 570 Marl, 36,174
Kaolinitic, 118, 352, 362, 379, 380, 381, 387, 399, Marsh, 157, 205, 208, 549, 558
407, 421, 422, 435, 453 — gas, 218
Kaolinitization, 83, 237 Maturation rate, 538
614 Subject Index
Maturity index, 65 Neoformation, 21, 22, 85,117,118,122, 290, 299,

Mesotrophic, 207, 208 392, 418
Mesozoic, 14, 192, 475, 491, 514, 525-527, 536, Neogene, 75, 79, 85, 309, 311, 320, 321, 323, 331,
539, 556 339, 340, 374
Metamorphic processes, 21,120 Neogenesis, 75, 79, 85
Metamorphism, 127,127,150,150,151, 514, 577 Nitratine, 239
Metasomatism, 25, 26, 127, 143, 145, 147, 150, Nodule, 10,114, 312, 316,317, 320, 335, 338, 342,
150,150,151 352, 354, 356-358, 365, 366-368, 379, 381,
Meteoric waters, 370, 373, 464 385, 388-395, 397, 398, 401, 417, 421, 426,
Meteorite, 43, 44, 49, 58, 66, 67, 559 446, 492, 515, 538, 546, 549, 550, 551, 553
— flux, 41, 44, 58, 66 Nontronite, 80, 82, 90, 91, 92,112,116,198, 291,
Methane, 10, 218 300
Mica, 22, 71, 72, 72, 73, 74, 75, 76, 77, 78, 78, 80, Nsutite, 459, 460
81, 85, 86, 94, 95, 108-111, 118, 165, 170,
171,192,193,198, 238, 290, 510, 513, 579 Oligocene, 358, 409, 485, 529
micromica, 92 Oligotrophic, 207, 208, 213, 214
Micaschists, 167,169,172 Olivine, 22, 48, 51, 52, 80, 82,132, 285, 461, 462,
Micrite, 314, 319, 323, 325, 327, 339 570, 572
Microaggregates, 96, 97, 446 Ombrotrophic, 209, 210, 212, 213
Microfacies, 11 Ooid, 421
Microsite, 11, 71, 75, 78 Oolith, 319, 325, 327, 330, 339, 340, 342, 342
Microsystems, 11, 24, 75, 76, 77, 78, 85, 291, 293, Oolitic, 323, 325, 327, 328, 330, 330, 331, 331,
297,383 339, 340
Migmatite, 389, 392 Opal, 37, 113, 176, 299, 349, 352, 354, 355, 355,
Mires, 203, 205, 211 356-358, 364,364, 365, 366,366, 368-370,
Mississippian, 118, 546 372-374
Mollic, 226, 231, 287, 297, 582 Ordovician, 13, 544, 545, 546, 547, 548, 558
Mollisol, 226, 227, 229, 271, 544, 551, 583 Organic acid, 10, 33, 73, 74, 75, 78,107,198,199,
Montmorillonite, 89,89, 90,91, 92,112,130,238, 215, 216, 289, 453, 519
245, 314 Organic matter, 7, 11, 32, 36, 74, 79, 96, 99,
Moon, 13, 41, 42, 42, 43-45, 47, 60, 66, 67, 543 160,161,167,168,169,173,185,189-194,
Moraine, 161, 270, 271, 277 198-200, 203, 210, 214, 217, 219, 262, 290,
Mottle, 31, 309, 316, 319, 334, 342, 384, 386-388, 299, 301, 315, 317, 322, 394, 426, 427, 478,
388, 389, 391, 392, 394, 395, 401 510, 513, 545, 547, 582
— zone, 384, 385-393, 399, 416, 417 Orthoclase, 76, 285, 454, 510, 520
Mottled clays, 386, 446 Oxalate, 77, 78, 262
Mottling, 166, 476, 528, 536, 538 Oxalic acid, 77
Mud, 129,130, 206, 215, 267, 491, 516 Oxidate, 26, 29, 290, 301
Mudstone, 491, 491, 492 Oxidation, 23, 31, 73, 74,115,140,145, 195, 199,
Mukkara structure, 551 218, 234, 235, 289, 360, 369, 452, 457, 459,
Muscovite, 76, 111, 119, 121,121, 135, 147, 285, 505, 512, 515, 521, 556, 558, 582
455, 459 Oxisols, 74, 96, 226, 238, 272, 394, 395, 396, 396,
Mycorrhizae, 78, 317, 556 399, 401, 421, 426, 432, 556, 557, 581, 583
Oxygen, 3, 25, 28, 30, 66, 82, 107-109, 144, 145,
Nacrite, 109,130 210, 213, 453, 460, 480, 558, 570
Nanoscale, 75, 79, 85, 88, 96, 99 Oxygénation, 557
Nari, 309, 311, 316
Natric, 226, 271, 582 Paleocatena, 476
Natron, 239 Paleocene, 397, 529
Subject Index 615
Paleoclimate, 3, 401, 432, 435,436, 445, 446, 525, Phlogopite, 111

539 Phreatic, 14, 237, 239-243, 246, 250, 252, 349,
Paleodrainage, 350 369, 479
Paleogene, 529 Phyllite, 159,167, 579
Paleomagnetic, 315, 537 Phytolith, 37,113, 299, 553
Paleomagnetism, 480 Pigeonite, 198
Paleozoic, 10, 13, 116, 141, 476, 480, 492, 510, Pimelite, 297, 462, 463
514,528,543, 544,546-548, 550, 553,557- Pisoid, 421
559 Pisolite, 320, 380, 381, 384, 393, 395, 397, 398,
Pallid zone, 383, 386 435
Pallisadic quartz, 363, 366, 368 Pisolitic, 309, 317, 390, 391, 394, 433, 435, 446,
Palygorskite, 108,112,114,118, 245, 314, 320 455
Paragenesis, 75 — structures, 309
Parent material, 4-6, 9, 22, 26, 33, 36, 114, 115, Plagioclase, 22, 50, 76, 77, 129, 129, 131-133,
118,155,158,159,164,170,173,175,185, 133,134-140,143,144,147,150,170, 285,
191-193,195,198, 227, 231, 238, 240, 245, 296, 454, 509, 510, 511, 517, 520, 553, 558,
247, 261-264, 267, 269, 212-21 A, 274,275, 573
278, 286, 287, 287, 295, 299-301, 339, 395, Planosolisation, 99
407, 410, 481, 527, 530, 553, 557, 584, 585 Plasma, 11, 76, 287, 390, 397
Pavement, 192-195,196, 226, 236, 243, 244, 248, Playa, 231, 342, 488, 489, 491, 491, 493
251, 264, 315 Pleistocene, 165, 168, 227, 237, 238, 247, 249,
Peat, 9,11,13, 185, 189, 191, 203, 203, 203, 204- 250, 252, 270, 309, 314, 317, 334, 479, 489
209, 212, 213-216, 218, 219, 478, 479, 503, Plinthite, 379, 380, 395, 582
515-517, 519, 521, 549, 550, 558, 583 Plutonic, 36, 120, 128, 138, 234, 297, 571, 572,
Peatlands, 9-11, 203, 208, 209, 213, 214-219 573
Pedalfer, 268 Podzol, 33, 74, 159, 164, 166, 170, 172, 173, 176,
Pedocal, 268 189-192,198,199, 272, 273, 290-292, 298,
Pedodiagenesis, 14, 343 299-301, 399, 554, 583
Pedofacies, 481, 483, 483, 484, 486, 488, 491, 491, Podzolic, 114, 115, 158, 159, 173, 190, 192, 291,
492-494, 530, 532, 532, 533-535, 535 294, 298-301
Pedogenesis, 4, 5, 12, 14, 30, 36, 155, 225-227, Podzolisation, 35, 36, 71, 74, 99, 164, 170, 172,
232, 233, 237-239, 241, 245, 248, 249, 250, 176,192,198, 287, 290, 294, 298, 299, 301,
251, 264, 269, 278, 287, 322, 339, 340, 475, 553, 555
476, 480, 481, 483, 485, 487, 490, 492, 493, Polygenetic development, 480
512, 514, 521, 526, 527, 533, 534, 537 Prasinite, 175,176
Pedorelict, 389 Precambrian, 12-15, 31, 116, 127, 128, 144, 150,
Pedorelictual, 384, 388, 389, 390, 392 370, 544, 546, 556, 557, 557
Pedostratigraphic, 480, 481, 488, 490, 493 Precipitation, 6, 9, 22, 23, 30, 71, 77, 78, 81, 85,
Pedostratigraphy, 488, 494 107,114,117,119,134,156,173,183,185-
Peneplanation, 408, 409, 410 188,192, 200, 205, 210, 211, 212, 215, 219,
Pennsylvanian, 116, 143, 147,148, 479, 504, 507, 226, 235, 242, 262, 263, 263, 267, 283, 289,
510, 511, 515, 518, 521 291, 311, 313, 314, 316, 317, 319, 325, 334,
Perched water table, 391 339, 342, 343, 344, 357, 358, 369, 373, 389,
Peridotite, 296, 300, 572 393, 418, 427, 447, 449, 450, 452, 453, 456,
Permafrost, 35, 161, 183, 186-189,189, 190,194, 457, 463, 573
196, 200, 582 Principle of uniformitarianism, 3,14
Permian, 486, 546, 549, 550 Profile development index, 263, 270, 276
Pervection, 199 Proton donor, 7,198-200
Petrocalcic, 226, 251, 314, 582 Pseudo-gley, 482
616 Subject Index
Pseudoanticline, 244, 248 Sabkha, 234, 241, 493

Pyrite, 30, 165, 193, 217, 464, 505, 506, 510, 511, Salerete, 241, 242
515, 516, 549, 577 Salic, 226, 231, 237, 238, 240, 241, 244
Pyroclastic, 46, 52, 283, 294, 297, 298, 301 Saline lakes, 116
Pyrolusite, 108,112, 421, 460 Salinity, 188,194, 231, 239, 240, 291
Pyroxene, 22, 50, 80, 82, 118, 132, 175, 461, 462, Salinolysis, 291
466, 570, 572 Salt weathering, 196, 227, 234-236, 247, 251, 264
Sand, 11, 65, 170, 176, 184, 185, 189, 193, 194,
Quartz, 22, 48, 65, 76, 77, 78, 132, 144, 164- 196, 229, 233, 235, 237-240, 246, 251, 272,
168,170,193, 235, 272, 273, 285, 290, 295, 273, 309, 314, 323, 324, 324, 325, 325, 327,
297, 299-301, 314, 320, 325, 327, 328, 331, 328, 330, 330, 331,331, 333, 333, 334, 336,
332, 334, 338-340, 341, 342, 342, 343, 349, 336, 338-340, 351, 357, 358,359, 360, 361,
352-354, 354, 355-358, 360, 360, 365, 366, 361, 363-369, 372, 382, 416, 481, 491, 492,
366, 367-370, 373, 374, 379-381, 381, 383, 501, 505, 507, 509, 511, 512, 514, 520, 527,
385-393, 394, 395, 396, 396, 399, 401, 408, 528, 532, 538, 544, 552, 553, 554, 558
413-415, 417, 418, 418, 419, 420, 422, 423, Sandstone, 118,120,122,141,147,172,192,193,
428, 446, 447, 455, 456, 461, 462, 466, 503, 198, 358-360, 360, 364, 365, 365, 366, 367,
509, 510, 511, 513, 515, 520, 570, 572, 573, 369, 370, 372, 389, 390, 396, 399, 422, 480,
579 492-494, 507, 510, 512, 530, 532, 533, 535,
Quaternary, 7, 20, 198, 227, 246, 266, 275, 277, 539, 544, 550, 551, 552, 554, 555, 573, 577
317, 322, 331, 361, 362, 397, 408, 475, 476, Saponite, 112,116, 462
479, 480, 484, 488-490, 491, 494, 495, 528 Saprolite, 5, 75, 94, 238, 283, 284, 349, 379, 381-
383,384, 385-388, 391, 392,394, 395, 397,
Racemisation, 322, 323 398, 401, 409, 410, 417, 418, 422, 446, 447,
Radiocarbon, 315, 322 450, 454^56, 480, 582
Ranker, 158,159 Savanna, 544
Red bed, 9, 37, 239 Schreibersite, 51
Reddening, 9, 238, 239, 262, 270 Scree, 159,161,162,167,169,172
Redox, 25, 30, 31, 215, 217, 219 Sepiolite, 108,112, 114,118, 245, 320, 338
Reduction, 31, 74, 78, 121, 217, 234, 242, 448, Sericite, 76, 111, 147, 149, 296
514, 545, 547, 548 Serpentinite, 175, 292
Regolith, 5, 45, 46, 66, 67,119,127,185, 239,240, Shale, 35, 118, 120, 121,129, 130,131,131, 193,
349, 364, 364, 369, 373, 448 234, 365, 369, 479, 502, 503, 505, 506, 507,
Regosols, 225, 227, 229, 232, 237, 251, 583 510, 512, 513, 517, 518, 536, 551, 573
Remote sensing, 526, 536, 537 Siderite, 30, 216, 217, 506, 510, 511, 577
Rendzinas, 169,191 Silcrete, 11, 241, 244, 246, 247, 252, 309, 310, 349,
Residua system, 119, 286, 295, 298 350, 351,352-355, 357-359,359, 360,360,
Resistate, 26, 29,132, 290, 301 361,361, 362, 362, 363, 364, 364, 367, 367,
Rhizolith, 317, 323, 324, 339, 343, 368, 528 368, 368, 369-371, 371, 372-374, 380, 480
Rhizosphere, 78, 79 Silica, 31, 75, 108-110, 112, 113, 117, 170, 241,
Rhodochrosite, 457, 459, 459 245, 246, 300, 309, 310, 313, 349-353,353,
Rhyolite, 140, 296 355-358, 360, 361, 363, 366-368, 368, 369,
Rock cycle, 28, 78 370, 372-374, 387, 391, 422, 428, 431, 432,
Root, 11, 28, 78, 78, 79, 79, 189, 193, 198, 205, 456, 466, 516, 517, 555, 558, 569, 577, 582
316, 317, 326, 327,328, 340, 492, 502, 518, Silicification, 349-351, 360, 361, 362-364, 365,
519, 521, 528, 530, 544-551, 553, 554, 556, 369, 370, 372, 374
557 Silt, 11, 162, 166, 189, 190, 196, 199, 234, 236,
Rubéfaction zone, 380 262-266, 271, 277, 314, 323, 324, 363, 416,
Rubification, 9, 262, 270 417, 532
Subject Index 617
Siltstone, 492, 503, 533, 544, 548, 549, 573 Temperate zone, 9, 36, 37, 71, 87, 156, 158, 234,
Silurian, 13, 372, 545, 547-550, 558 298
Smectite, 7, 24, 31, 32, 36, 72, 74, 75, 80, 85, 87, Termite, 381, 388, 391, 393, 399, 401, 415, 416,
88, 90, 91, 92, 94, 95, 96, 98, 99, 108, 109, 416, 417
111, 112,114-122,130,170,173,191-193, Tertiary, 192, 309, 317, 323, 361, 362, 362, 363-
198, 238, 249, 249, 290, 291, 293, 295-297, 365,369,371,372,437, 493, 512, 514, 525-
299, 300, 382, 383, 387, 461-463, 466, 503, 528, 530, 539, 549, 556
507, 509, 510, 511, 514, 517-519, 521 Textural interstratification, 92, 95
Smectitic, 7, 72, 87, 92, 380, 386, 387, 391, 401, Thenardite, 200, 235
515 Thermodynamic, 5, 6, 24, 25, 99, 139, 293, 295,
Smithsonite, 216, 217 297, 298, 392, 397, 421-424, 427, 430, 431,
Soil maturity, 65, 478, 483 435, 456, 459, 463, 584
Solid solution, 31, 92,129, 295, 430, 431, 456 Till, 161,184,185,194,196, 262, 269, 270, 479
Solid state transformation, 71, 74, 75, 81, 85 Titania, 351-353, 354, 355, 355, 356-358, 361,
Solubilization, 170 367, 368, 368, 369, 372, 373
Solum, 25, 190, 236, 238-241, 245, 248, 250, 251, Tonalité, 129,138
395, 475 Tonstein, 10, 503, 505, 512, 517, 518, 521
Toposequence, 481
Solution, 23-25, 28, 30, 31, 32, 72, 73, 73, 74, 75,
Tourmaline, 408, 417
85, 88, 90, 98,107, 111, 131,133,134-137,
Translocalion, 183, 190-192, 200, 237, 240, 252,
138,139,145,147,162,163,163,164,170,
265, 270, 297, 336, 339, 389, 448, 548
171,173,174,193, 198, 206, 209, 215, 235,
Troctolite, 51
242, 245, 265, 283, 284, 288-292, 294, 298,
Troilite, 51
301, 312, 317, 319, 339, 342-344, 349, 358,
Trona, 239
369, 372-374, 382, 389, 408, 410, 416, 417,
Tuff, 297, 485
422, 428, 432, 446, 450, 452, 455, 456, 457,
Tundra, 185-188,190-192,198, 269, 546
459, 461, 462, 466, 519, 573
Spinel, 51 Ultisol, 74, 85, 297, 393-395, 401, 401, 426, 432,
Spodic, 299, 300, 553, 554, 582 529, 555, 556, 583
— horizon, 161, 226, 262, 298, 555, 570 Unconformity, 12, 141, 144, 277, 479, 480, 528,
Spodosol, 226, 269, 270, 272, 273, 484, 553-555, 529, 539
583 Underclay, 12, 87, 118, 478, 501-503, 505-507,
Stalactite, 354, 355 509, 510-521
Staurolite, 300
Stevensite, 112,116, 291 Vadose, 14, 239-244, 250, 252, 360
Stone-line, 391, 395 Variegated clay, 386
Subarctic, subpolar 185-189,189,190,198 Varnish, 236, 248, 251, 252, 315
Substitution, 109-111,121, 428-430, 456, 463 Vermiculite, 72, 74, 75, 78, 81, 84, 85, 88, 90, 93,
Supergene, 111, 183, 246, 349, 418, 446, 448, 450, 94, 108, 109, 111, 112, 114, 115, 118, 120,
464 130, 143, 168,170, 193, 198, 288, 290, 296,
Surface tension, 51, 52, 53, 64, 340 382, 454, 509, 511-513, 520, 521
Swamp, 205, 207-209, 210, 214-217, 479, 503, Vertisol, 7,13, 75, 96, 97,173, 227-229, 231, 271,
276, 292,380, 386, 387, 482, 486, 488, 491-
505, 514, 516-519, 550, 554, 558
493, 551, 581, 583
Swelling, 72, 74, 80, 85, 87, 89, 90, 92, 94, 99, 235,
Vesicular, 51, 53,195, 226, 241, 262, 351, 415
243, 248, 482, 551, 583
Vetusol, 3,12
Vivianite, 216, 217
Taenite, 51 Void cu tan, 366
Taiga, 186,187, 546 Volcanic rocks, 9, 99,115,117,118,120,128-130,
Talc, 110, 112,175, 295, 297, 461-463 140, 295, 415, 573
618 Subject Index
Volcanic ash, 10,116,117,164,191, 503, 507, 514, Wetlands, 10, 203, 204, 208, 209, 213, 217, 218,
515, 517, 518, 520, 521 550
Wurztite, 217
Walther's law, 10,11
Xerosol, 226-229, 231, 237, 241, 251, 583
Weathering profiles, 25, 30, 127, 128, 139, 141,
141,142,144,145,145,147-149, 238, 239, Yermosol, 226-229, 231, 237, 238, 241, 251, 583
251, 283, 284, 380, 417, 418, 425, 426, 445,
446, 455, 457, 460, 461, 464, 466, 479, 480 Zircon, 117, 408, 415, 417
Weathering rates, 32, 33, 285, 289, 419 Zonality, 6,155,157, 288

Weathering, Soils & Paleosols - Martini & Chesworth (1992) PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Weathering, Soils & Paleosols - Martini & Chesworth (1992) PDF

Загружено:

Авторское право:

Доступные форматы

Developments in Earth Surface Processes

1. PALEOKARST: A SYSTEMATIC AND REGIONAL REVIEW

LP. MARTINI and W. CHESWORTH

Library of Congress Cataloging-in-Publication Data

Weathering, soils & paleosols / edited by I.P. Martini and W.

© 1992 Elsevier Science Publishers B.V., All rights reserved.

This book is printed on acid-free paper.

Printed in The Netherlands

Preface and Acknowledgements

This book is an attempt to connect present happenings on the planetary land

LP. MARTINI and W. CHESWORTH (Editors)

P.W. BIRKELAND Department of Geological Sciences, University of Colorado at Boulder,

B. BOULANGE Laboratoire de Géosciences de l'Environnement, URA 132 CNRS

W. CHESWORTH Department of Land Resource Science, University of Guelph, Guelph,

F. COLIN Laboratoire de Géosciences de l'Environnement, URA 132 CNRS

L.J. EVANS Department of Land Resource Science, University of Guelph, Guelph,

M.J. KRAUS Dept. of Geological Sciences, University of Colorado at Boulder, Cam­

F. MACIAS Catedratico de Edafologia y Geologia, Facultad de Biologia, Universi-

LP. MARTINI Department of Land Resource Science, University of Guelph, Guelph,

D.B. NAHON Laboratoire de Géosciences de VEnvironnement, URA 132 CNRS

H.W. NESBITT Department of Geology, University of Western Ontario, London, Ont.

G.J. RETALLACK Department of Geology, University of Oregon, Eugene, OR 97403Ί272,

C. ROQUIN Departement de Géologie, Université de Louis Pasteur, 1 Rue Blessig

W SHOTYK Geological Institute, University of Berne, Baltzerstraße 1, Berne CH-

G. SPIERS Department of Land Resource Science, University of Guelph, Guelph,

Y. TARDY Departement de Géologie, Université de Louis Pasteur, 1 Rue Blessig

M. THIRY Ecole Nationale Supérieure de Mines de Paris, Centre de Géologie

A. WATSON 9 Hartley Terrace, Boston, MA 02134, U.SA.

Reflections on soils and paleosols

features of earlier pedogeneses, occurrence of allochthonous materials and effect of

Soil and energy

non-steady fashion. Without doubt, relatively time-independent states are achieved

Another aspect of the soil-energy relationship relates to the global distribution,

Ecosystem and energetic conditions at the Earth's surface

extreme semidesert dry savanna equatorial

0.2 0.4 0.6 0.8

Mineral concentrations in soils

Sedimentary facies and soil catenas

a reconstruction of paleogeography at any recognizable (datable) geological time

Paleosols and time

In spite of Einstein, geologists follow Aristotle in thinking of time as a one way,

Paleosols and life forms

Ultimately, the analysis of paleosols yields information on the biosphere. Retal-

The process of classifying is complex and never-endingly controversial. Ideally

of uniformitarianism and, where applicable, it transfers whatever environmental in­

Future research trends

Conceptual view of weathering systems

This chapter is concerned with chemical weathering, which is described in terms

Type ana density of x::::::: ATMOSPHERE

Water aval lability J ^ x x x ï l

The inorganic solids

Fore Fore 0ld

Chemical composition Mineralogical composition

Si0 2 66.0 Quartz 12

As phases inherited from parent materials, most earth-surface systems contain

The aqueous environment

The functions of water, an absolute pre-requisite if chemical weathering is to

Functions of water in the weathering zone

Major weathering trends in a pe-pH framework

The acid trend

20% Na. Al 2 0 3 + Fe 2 0 3 20% Na20 27 Si0 2 + Al ? 0 3 -e re ? 0 3

The alkaline trend

The reduced trend

M.J. KRAUS Dept. of Geological Sciences, University of Colorado at Boulder, Cam

of uniformitarianism and, where applicable, it transfers whatever environmental in

The fundamental characteristic of the reduced trend of weathering is the pres

Weathering systems can also be considered as a succession of macroscopic weath

Transmission electron microscopy techniques were developed in 1940 and ap

reprecipitation, but crystallographic orientation is maintained between clay, hydrox