Miksicek 1987

Formation Processes of the Archaeobotanical Record
Author(s): Charles H. Miksicek

Source: Advances in Archaeological Method and Theory, Vol. 10 (1987), pp. 211-247
Published by: Springer
Stable URL: http://www.jstor.org/stable/20210089
Accessed: 25-08-2015 13:41 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/20210089?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
Springer is collaborating with JSTOR to digitize, preserve and extend access to Advances in Archaeological Method and
Theory.
http://www.jstor.org
This content downloaded from 200.89.68.173 on Tue, 25 Aug 2015 13:41:27 UTC
All use subject to JSTOR Terms and Conditions
< 4
Formation Processes of the

Archaeobotanical Record
CHARLES H. MIKSICEK
Office of Arid Lands Studies
UniversityofArizona
Tucson, Arizona 85721
INTRODUCTION
Archaeobotany is the art and science of recovering, identifying, and in

terpreting plant remains from archaeological sites. Some authors (notably
Ford 1979;Renfrew1973) preferto call thedisciplinepaleoethnobotany,
stressing the relationships between past cultures and the plant kingdom, but
both termswill be used synonymously in the following discussion. This arti
cle will focus on the preserved traces of ancient food plants but other types
of botanical remains including fibers, wood, charcoal, pollen, and plant
crystals will also be mentioned briefly.
The formal study of plant remains from archaeological sites can be
traced back to at least 1826 when Kunth published an analysis of "mum
mified" cereals, fruits, and seeds from dry Egyptian tombs (Kunth 1826).
Other pioneering efforts in archaeobotany also focused on sites with excep
tional organic preservation. These included Heer's (1866) treatise on seeds
from waterlogged deposits associated with lake dwellings in Switzerland,
examinations of botanical materials inmummy bundles from the arid coast
of Peru by de Rochebrune (1879) and Wittmack (1888), and Harshberger's
211
ADVANCES INARCHAEOLOGICAL Copyright ? 1987 byAcademic Press, Inc.
METHOD AND THEORY, VOL. 10 All rightsof reproduction inany formreserved.
212 CHARLES H. MIKSICEK
(1896) analysis of plant remains preserved in dry rock shelters in

southwestern Colorado. A briefpaper byMills (1901) on carbonized seeds
fromBaum Village inQhio broadened the domain of archaeobotanical
researchby demonstratingthat even open siteswith less than optimal
preservationconditionscould yield suitablematerial forcarefulbotanical
analysis.During the firsthalf of thiscenturypaleoethnobotanicalreports
graduallybecame more common, and by the era of the large interdisci
plinaryprojectsof the1950s and 1960s,many archaelogicalresearchteams
includedone or more botanists in theirrosterof specialists. [A detailed
historyof thedevelopmentof paleoethnobotanyisbeyond thescope of this
article.For a more completediscussionof thedevelopmentof thediscipline
fromtheEuropean perspectiveseeRenfrew(1973:1-6) or fortheAmerican
viewpoint,see Ford (1979: 291-297).]
Within the last twodecades, archaeobatanyhas undergonea methodo
logicalrevolution with theadoptionof flotation(waterseparation)as a tool
for recoveringsmall charred plant remains.Although Struever (1968)
deservescreditforpopularizingflotation,thetechniqueitself was pioneered
over a centuryearlier. In 1860 ProfessorUnger, an Austrian botanist,
dissolvedancientEgyptianadobe bricks inwaterand examined the residue
for tracesof cerealsand other seeds (Wittmack1905: 6). Hendry (Hendry
and Kelly 1925) and V. Jones (Montgomeryet al. 1949: 88) used a similar
techniqueto recoverplant remainspreservedinmud bricks fromhistoric
mission buildings inCalifornia,Arizona, and northernMexico. Cutlerwas
one of the firstarchaeobotaniststo apply flotationto general site fill,
testingthemethod at Tularosa Cave, Higgins Flat Pueblo, and Point of
Pines in theearly1950s (Watson 1976:78-79).Matson (1955) experimented
withwater separation for concentratingcharcoal for radiocarbondating
and noted that it also yieldednumeroussmallcharredseeds and nutshell
fragments. (For a more completediscussionon various flotationtechniques
and theirdevelopmentseeWatson 1976.)
Concurrentwith thewider applicationof flotation,archaeobotany is
also undergoinga theoreticalrevolution.A commonthemeinmany articles
published over the last two decades is a careful reexamination of the nature
of the prehistoricbotanical record.Archaeological sites do not contain
pristine samples of all of the plants used by ancient peoples. A plant part
may undergo many transformations between the time itwas harvested by
someone in the past and the time when it is quantified, measured, and
reportedon inan archaeologicalmonograph.Organicpreservationvaries in
differenttypesof naturalenvironments. Not all typesof plantmaterials are
equallywell representedin a given site.Environmentalprocessesmay in
troducemore recentmaterial, mix deposits, or destroy fragile plant
fragments.Cultural factors such as site type,processingmethods, or
discard patternsaffect the archaeobotanical record.Even the sampling
FORMATION PROCESSES OF THE ARCHAEOBOTANICAL RECORD 213
strategiesand analytical techniquesemployedby the paleoethnobotanist

may alter thedata base. These natural,cultural,and analyticaltransforma
tions,henceforthcalled formation processesafterSchiffer(1976, 1983), are
the focusof thispaper.
PRESERVATIONENVIRONMENTS
Firstof all it is importantto consider thenatural conditionsthat favor

"the survivalof the evidence" (Renfrew1973). Archaeological siteshave
been discovered in environmentalsettingsrangingfromarctic tundra to
tropical rainforest. What kinds of sites inwhat localitiesoffer the best
potential for the preservationof ancient botanical remains?Table 4.1
presentsa hypotheticalcomparison of the preservationpotentialof dif
ferentcategoriesof biological remainsinvarious typesof sites.Tables 4.2
and 4.3 summarizesthedensitiesof seedsactually recoveredfromrepresen
tativesites.
Archaeologistshave tendedto overlook thecontributionof plants to the
subsistenceof huntersor pastoralistslivingat high latitudesor elevations.
Nevertheless,plants probablyprovided food foranimals, seasonal dietary
supplements,and rawmaterials for tools and shelters.Frozen sites,where
thedecompositionof organicmaterial is limitedby constanttemperatures
at or below freezing, offer great, almost untapped potential for
archaeobotanicalresearch.A cache of hemp seedswas found inassociation
with a copper censerand felttentina Scythiantomb in theAltai mountains
of Siberia (Rudenko 1970). Lowland tropicalplant remainswere identified
fromtheunlikely localityof dry, cold rock sheltersassociatedwith adze
quarries at an elevationof 4205meters on Mauna Kea inHawaii (M. S.
Allen, personal communication).
Peat bogs are a specializedclass of waterlogged sites inwhich thecon
stantlywet, anaerobic environment,combinedwith low pH fromhumic
and tannicacids, produce someof thebest conditions forthepreservation
of ancientplant remains.Bogs have longbeen favoredsamplinglocalities
forpalynologists.Unfortunatelyfor faunal specialists,the acidityof the
soil isverydeleteriousto thepreservationof bone and shell (hence the low
scores inTable 4.1). On theotherhand, hair, hide, and thechitinousex
oskeletonsof insectsmay survive in almost pristinecondition.The peat
bogs of northernEurope, especiallyDenmark, have yielded a numberof
remarkablypreservedhumanbodies (Glob 1971).Analyses of thestomach
contentsof two individualsfromTollund and Grauballe have been com
pleted by Helbaek (1950, 1958) and HiUman (1981). HiUman is currently
engaged in similar researchon a recentlydiscoveredEnglish "bog man"
fromCheshire.
+D + ++ ++ ++ ++ ++ ++ +
o +++++++++++++++
0 + | ++ + + + + + ++ 2C
E o_
0
go E o E + + ++ + + + + + + + +
E 0 2
26++ C +++ ++++ o

-
0 E+ +1
C0.0~~~~~~~~~~~~~0 I++++++++++I 0
0
L
C 0
o = 0= o 1++ ++ _
c
wii CDa
> 0 C C + + + + + + ++ + + + + + +
- cm
* 0 0c +
+ + ++ ++ + 0
+ +++ + + ID
002 's &-Z + + + + +
o 0 aoL ++I+++++I++++1++
zi
CD
l + + + + + + +| + + + + + + o~
x
g-
0+ co
U CoCD ++ +++++ ++++++0.
0 co
+ + +++++ ++ + ++ + +
++++++ + ++ + ++ + + +
_ 0 0 qu 0 ,O?OI
~~~00
c co0
0 0o -
* ~~~~~0 a
0 0~~~~E
0 0 00
c
i ( 0 - 0 .0 0..
I - CLE 00 -t=S- oc
00=> oc
0 CC m me coC
CCo 0C L0 ~0 00 c
214
C-4~~~~~~~~'
C)~~~~~~~~~~~~~~~~~~~~~~~~~~.
ci0 0P. -.
a:g C .i
_ ~ ~ ~ ~ ~ ~ c 200 c
C * 00 x
0 0 0 V f%
a~~~~~~~~~~~
E
c 0~~~c 0o - I-. 0 co
* 0~~~~~o oO to. co D C 0 =it

o C.) <~~~ ,~ oU,
~~~- ~ )C r- coCc
E CO)
a
.1'> co
e 4" 0. CC VD c
~~ 0 ci Co CC'~~~
0 - 0cm0
h. 0 -~~~ 0 E
0~~E
l0 o0 loC CC 0
cm -2~o~ - 0
2 ~~~~~ E l 0 B 0 0
=~~
0~~~~~~~~~
0 N C 0
a =~~~~~0C
C C - _ C 0
0 ci co go -C -~~~~~~
0 E 7~25 E 0c
o
0~~~~~~~,10 CM.
co V 'A 0 ,W
0 :~~ C
0.0~~~
0 0~~~
e 0 <
E0
.4-. NOO W~-Y c c

o 000c~~~~~ coc
%
CD C =.~ c~
o 07 go0 0)
* 0 =0ZO 0 It
0~~
0~~~~
00 0
0 C~~~ '
C~~~~~~C
o
0 _ 0
0 0 <
01 0O
~~~~ v N.
3 ..2.
=00 ,O 0e EO 00
7 >
EO* i~
C ?5~~ 10
c Oo oO 9
U -~~~~EU 2>0N C)cm
0 0 0 00
8
I*-.v aO.ZCD *
CD)
216
0 00~~
C~~ V0
I CD
0I 0~~~~~~~~~~~~~5
co
*
co 0
U N>
.0
N100 0 CO)'an
uor~~~~r~~~o a0'- v~ 0 co
Co~C
0 CNY2Y~
0.
O 0 0D
0CM
0
0 b0
10~~~~~~~~~
2 0 0
0. 0
~~~~0E_ 0 C
o a VC
? 0 4S 0 0
00~~~~~~o0
0 Co 'a 0 b-03
C
e 'FDE ID 0V
0. 0 ~~co coco co 0 a.
00 00 0.o
-L < 0. 0 1 0 co O o
Cl) Z Go00 0 Vc
0~~~is 0-.-
* 0 0060C 001 o
.5 ~~~~~ o~~C co. COC Zo EEL E '- ) 0
0 0~~O 0c 000 00 0c0C.) 0
co O> cis~0~~ .co
0 C Cs~~~~.
0 -a
* 0
0C C~ 0 O C Z E0oE VC o2*a
4.0 =cN o *
7i ~~ _ 0~ 3
C C 0
E ~<0C
V .
E~~~~
0 C00
o
C 0 0C~
>~~~ 0 co
o0c
0 CIS 100
Co
0 0
coc 0
b-
a.000 0 00 r
LI
> :5 co co 0=xC
c co0=c
coco x 217
The more generalizedclass of waterloggedsitessharesmany featuresin

common with peat bogs but lacks the acid conditions. The more neutral pH
is conducive to thepreservationof bone and shell(Table 4.1). Noteworthy
analyseshave been performedon plant remainsfromwaterloggeddeposits
includingSwiss and French lakeshoredwellings(Heer 1866; Lundstrom
Baudais 1984), theMesolithic villageof Star Carr inBritain (Clark 1954),
Roman timber-linedpits and middens below modern London (Willcox
1977), Carthage harbor (van Zeist and Bottema 1983), Neolithic Dutch
coastal villages (Pals 1984), German privypits (Knorzer 1984), and the
OlympicPeninsula villagesof Ozette andHoko River inWashington State
(Croes and Blinman 1980).
At theotherextremeof themoisturescale,a constantly dryenvironment,
suchas thatfoundinsomecavesand rockshelters,isconduciveto thepreser
vation of plant remains.Localities with abundant dehydratedbotanical
refuseincludedrypassages of Salts andMammoth caves inKentucky(Wat
son 1969, 1974;Yarnell 1974), rockshelters
in theTehuacanValley ofMexico
(Smith 1967; MacNeish 1967), cliffdwellingsin southwestern Colorado
(Harshberger1896),Egyptiantombs(Kunth1826),and thecoastal desertof
Peru (deRochebrune 1879;Wittmack 1888;Cohen 1975; Pozorski 1983).
Freezing,acidity,waterlogging,and aridityaremacroscale environmen
tal factorsthat favorbotanical preservation.In rare instancesuncharred
plant remainsmay persistinopen sitesdue tomicroscale chemicaleffects.
During recentexcavationsat theRoman siteofCuriumon Cyprus, inwhich
thisauthorparticipated,flaxfiberspreservedby copper salts leachingfrom
thealloywere discoveredduringthemicroscopicexaminationof bronzeob
jects. The copper salts mineralized the fibers and acted as a strong
fungicide.Wooden tools and basketsdatingat least to theRoman period
have been recoveredfromancientcoppermines inCyprus. In theeastern
United States,where pollen preservationis oftenmarginal, good recovery
was noted fromsoil adjacent to copper artifactsor barkwith high tannin
levels(Kinget al. 1975).Using scanningelectron
microscopy,Keepax (1975)
was able to identify wood fragments preservedby corrosionproductsfrom
iron artifactssuch as nails. In soilswith a highphosphate content,un
charredseedsmay be preservedbymineralization,especiallyincoprolitesor
ancient latrines(Green 1979). Fossilizationby castingormineral replace
mentmay also involvesilicates,carbonates,gypsum,or calcite.Seeds with
naturallyhighconcentrations of calciumcarbonatein theirseedcoats, such
as hackberry(Celtis spp.) or some sedgessuch as razorgrass(Scleria spp.),
will surviveformilleniawithoutcarbonization.Some typesofwood, such
as juniper, with a high resin and terpenoid content may persist for over a
thousandyears in relativelydry,open sites.
Another type of mineralization occurs as a natural result of transpira
tion. Inmost members of theplant kingdom,but especiallygrasses and
palms,hydratedsilica fromgroundwaterisprecipitatedinepidermaltissue
as silicabodies, plant opals, or phytoliths,as theyhave been referredto in
thepaleoecological literature(Rovner1983). These silicabodies are castsof
the epidermal cells, so they can be relativelydistinctive to taxonomic
groups.As leafyplantmaterial decays in soil, assemblagesof phytolithsare
leftbehind. Only stronglyalkaline soils seem to be relativelypoor en
vironmentsfor the preservationof silica bodies (Rovner 1983). Pearsall
(1978)utilizedphytolithanalysis to attemptto documentmaize agriculture
in earlydeposits fromReal Alto thatcontained fewotherpreservedplant
remains.Pearsall (1983) also suggestedearlycultivationof achira (Canna
edulis), a root crop, based on the identification of distinctivechains of
"phytoliths."These achira crystalsshouldmore properlybe referredto as
calciumoxalate druses,which have a different biochemicaloriginand com
position. These oxalate crystals are included in the "Other Crystal"
category in Table 4.1, along with calcium carbonate cystoliths,both of
which are adverselyaffectedby acidic environments.
PRESERVATIONBY CARBONIZATION
Frozen, acidic, waterlogged, permanentlydry, or chemicallyunique

localitiesaccount foronly a smallpercentageof thesitesthatarchaeologists
investigate.The most frequently encounteredsitesare inopen,well-drained
areaswith alternating wet and dryconditions.Fluctuatingmoisture levelsin
well-aeratedsoil are conditions favorableto thegrowthof bacteria, fungi,
insects,and other decomposers thatbreak down organicmatter. In these
open sites, plant remains are likely to be preserved for long periods of time
only if they have been charred.
Complete carbonization occurs when plantmaterials are subjected to
temperaturesbetween 250 and 500 ?C under low oxygen conditions
(Hillman 1981; Lopinot 1985). Rapid burningat high temperatures with
abundant oxygen reducesorganic remainstomineral ash. Carbonization
reducesa seed or fruitto 50 to 600o elementalcarbon (Meyer 1980: 403),
which is very resistantto furtherorganic decay. Mechanical damage is
about the only process that will destroy a completely charred seed.
Occasionally,whole sitesmay be destroyedby a catastrophicconflagra
tion. The eruption of Mount Vesuvius in A.D 79 froze the cities of Pompeii
and Herculaneum in a moment of time and preserved a fairly complete
cross sectionof Roman foodplantsby carbonization(Meyer1980).The ex
plosion of Ilopango Volcano in El Salvador in approximatelyA.D. 260
buried a Maya farmstead and an associated milpa (traditional field) with
germinatingcorn (Zea) seedlings in ash (Zier 1980). In theAmerican
Southwest,it isnot uncommon to find individualstructures
or occasionally
whole sitesdestroyedby fires.In theserarecircumstances, com

a relatively
plete inventory of plants available at one timemay be preserved.In sites
thathave not been destroyedby a catastrophicfire,only certain typesof
plant remainsare likelyto be preservedby charring.
The likelihoodthata given typeof plant remainwill be preservedin the
archaeological recordby carbonization is relatedtowhetheror not it is
directlyexposed to fireduring use or processing for consumption or
storage.
Munson et al. (1971: 427) dividedplant foods intothreecategoriesbased
on "preservability"and potentialvisibilityin thearchaeological record.
This approachhas been reiterated and expandedon bymany otherauthors,
for exampleDennell (1976),Minnis (1981), or Hammond and Miksicek
(1981).
The firstgroup includesfoodswith dense, inedibleparts such as nut
shells,maize cobs, or olive pits.After theediblepart is removed,thewaste
productsmay be recycledas a supplementalfuel.Massive quantitiesof
hickory(Carya) nutshellhave been reportedfrom many sitesin theeastem
United States. of
Charred fragments corncobs are perhaps themost ubiq
uitous remainsinSouthwesternsites.Ford andMiller (1978) suggestedthat
theabundance of charredolive (Olea) pits in flotationsamples fromCar
thageand otherNear Eastern andMediterraneansitesmay be due to theuse
of waste from olive oil presses, which included pits and was used as a fuel in
ancient limekilns, furnaces,ovens, and hearths.
The second group includesplant foods, such as edible seeds, thatwere
commonlyparchedbefore consumptionor storage.For example, toasting
mesquitepods (Prosopis) enhancestheirflavor, makes themeasier togrind,
much of theharvestduring
and also killsbruchidbeetles thatcould destroy
storage.Many small seeds, such as thoseof pigweed (Amaranthus)and
lamb's-quarters(Chenopodium),were commonlypopped inmuch thesame
way as popcorn.Dennell (1976)andHillman (1981,1984) suggestthatmany
of thehulledcerealsneed tobe toastedlightly to loosentheadherentglumes
beforepounding or threshing. Complete carbonizationof plant foods in
thissecondcategory would onlybe accidental,when a fewseeds spilled into
a hearth, or a batch of seeds was allowed to get too hot, during the parching
process.
The thirdcategoryincludesnondenseplant foodswith a highmoisture
content,such as leafygreens,pulpy fruits,or edible tubers.Sincemost of
theseare eithereaten fresh(and oftenaway fromsitesas snacks)or boiled
they are not very likely. to be preserved by carbonization. If by chance they
did get charred, the fragmentary remains of tubers or greens would be very
fragileand difficultto identifyexceptby carefulanatomical examination
under a scanningelectronmicroscope. This categoryis only likelyto be
representedarchaeologicallyby theoccasional pit froma fleshyfruitspat
into a hearth.Althoughmedicinal herbs are not technicallyfoods, they

could be includedwith thisgroup as theyare commonlypreparedas teasor
poultices and are thereforelikely to be very underrepresentedin ar
chaeological sites.
Even within theabove plant food categories(nuts,seeds, fruits),not all
taxa are equally likelytobe preserved.Lopinot (1985) experimentallycar
bonized nine differentspeciesof nuts commonly recoveredfromsites in
easternNorthAmerica at temperatures rangingfrom200 to 900 ?C. Thick
shellednuts such as hickoryand walnut (Juglans)survivedcharringbetter
than thin-shelledtypessuch as acorns (Quercus) and chestnuts(Castanea).
Lopinot concluded that because of differentialpreservationafter car
bonization and the greater edible meat to shell ratio of acorns and
chestnuts,thecontribution of thesetwo taxa toArchaic andWoodland sub
sistencein easternNorth America has been greatlyunderestimated.In a
similartestusing 12 common typesof European weed speciesand 12 dif
ferentexperimentaltreatments(varyingtime, temperature,and moisture
content),Wilson (1984) concluded thatseedswith oily ormucilaginous en
dospermwere less likelythanstarchyor proteinaceousones tobe preserved,
recovered,and identifiedfromarchaelogical sites.
APPROACHES TO UNDERSTANDING
DIFFERENTIAL PRESERVATION
Ethnoarchaeology
Laboratory experimentssuchas thoseconductedbyLopinot andWilson
are one approach to understandingdifferentialpreservation.Ethnoar
chaeology may offer other clues. During the summer of 1980, I had the op
portunity to observe traditional Papago saguaro fruit (Carnegiea)
harvestingand processing in the desertwest of Tucson, Arizona. The
followingspring,Fish and I returnedto thesaguaro camp to samplevarious
processingloci forpollen and charredseeds (Miksicekand Fish 1981).All of
theflotationsamplescollectedproduced saguaro seeds,but 44.4% of these
samples produced charred seeds. A total of 2088 saguaro seeds were
recovered,of which only 11.3% were carbonized.All of thecharredseeds
were recoveredfromhearthsor ash dumps.During theprocessof reducing
saguaro pulp to syrup or jam by boiling, foam rises to the top of the cook
ingvessel-and is continuouslyskimmedaway. This foam,which contains
seeds, isusually tossedintothefirepit,
where someof theseedsare charred.
Each saguaro fruit contains approximately 15 gm of edible pulp and 2000
seeds. During themorning that I observed saguaro processing,approx
imately 8 kg of pulp, containing over a million seeds, was reduced to syrup
and jam.The flotationsamplesyieldedevidence forapproximately0.2%oof
a morning'swork, of which only0.02% was carbonized.The problemof

underrepresentation is evenworse ifyou considerthatthesame campmay
have been used fairlycontinuouslyover a 5- to 6-week period during
saguaro season (JunethroughJuly).Our informant's familyhad been com
ing to the same localityforover 60 years.A singlecharred saguaro seed
could thereforerepresentthedistilledessenceof 0.03 kg (1 day), 0.9 kg (1
season),or 7.6 kg (60 years)of saguaropulp. In ordertobe able to estimate
thata givenplant food accounted fora certainpercentageof theprehistoric
diet,similarcorrectionfactors(transferfunctions) would have tobe derived
foreach typeof plant remainrecoveredfroman archaeological site.Even
so, therewould be no way to account for archaeologically invisible
categoriessuch as greensor tubers.
Significantamounts of saguaro pollen (higherthan levels in control
samplescollectedaway fromthecamp)were recoveredfrom35.7% of the
samplesanalyzedby Fish. In contrastto thedistribution of charredseeds,
saguaro pollen countswere lowestin firepits and ash piles and highestin
theareawhere theinitialopeningand processingof thefruitoccurred.The
high frequency of saguaro pollen in thisareamay be partiallyexplainedby
theuse of thedried saguaroblossom as a natural"can opener" forsplitting
open the fruit.
An interesting sidelightwas notedduringthisstudyof a saguaro camp.A
pack rat (Neotoma sp.) had moved into the roof of one of the ramadas and
built a nest after the saguaro harvestwas over. This nest contained
numerousseedsand otherplantparts (aswell as associatedpollen) thathad
been collecteld by the pack rat. If the ramada had been destroyed by a fire,
theonly clue thatsome of the seeds fromthisstructure were collectedby
pack ratsand not people would have been thepresenceof charred,rodent
fecalpellets. I have noticed similaroccurrencesof carbonizedfecalpellets in
a numberofHohokam structuresthathad burnedduringor afteruse. The
introduction of seedsby thepack rat isone exampleof faunalturbation, a
topic thatwill be discussed in a latersection.
Ethnoarchaeologicalexperimentssuch as thiswill become increasingly
more importantin understandingthearchaeobotanicalrecord. In an ex
aminationof plant remains fromtheHopi villageofWalpi, Gasser and
Adams (1981) noted that only 0.3 % of the seeds in deposits younger than 60
years were charred, whereas 8.6% of the seeds from rooms over 65 years old
were carbonized.Microbial, rodent,and insectactivityhad destroyedsome
of the unburned plant material in the older sample.. Rodents seemed to
preferoily seeds such,as squash (Cucurbita),melon (Citrullus), juniper
(Juniperus),pinyon (Pinus edulis), peach (Prunuspersica), and cherry
(Prunus cerasus). Insect damage was most severeon maize and beans
(Phaseolus). This studyalso suggestedan additionalexplanation for the
paucity of beans in the archaeobotanical record. Beans, which are usually
preparedby soakingand boiling,are rarelyevercharred.Gasser andAdams

(1981) noted that beans are more susceptible to fungaldamage during
storagethanmost other seed types.
In an ethnoarchaeologicalstudyofmodern cholla (Opuntia) roastingpits
byGreenhouse et al. (1981), fivecarbonized spine clusterswere theonly
charredevidenceforcholla buds, and only a fewclumpsofCylindropuntia
pollen were noted in pollen samples. Cheno-am (Chenopodiaceae and
Amaranthaceae) pollen fromseepweed (Suaeda) utilizedduringthecooking
processwas themost conspicuous evidence forcholla roasting.
Dennell (1974, 1976),Hubbard (1976),Hillman (1981, 1984), and Jones
(1984) have analyzed samples of modern seeds and chaff fromvarious
stagesof traditionalcrop processingmethods stillpracticed inGreece and
Turkey.They predictthattailings,theresidue leftafterwinnowingor siev
ing,should contain smallercereal grainsandmore weeds than thecleaned
end product ready for storage.Measurements on charredseeds from tail
smallerthansamplesof
ingstossed intoa fireforfuelwould be significantly
clean grain from a granary destroyed by a catastrophic fire. Cereals
harvestedby uprootingshould be mixed withmore weed seeds and stem
bases thanthosereapedwith sickles.Crops fromplowed fieldsshould have
different weedy assemblagesthanthosefromfieldstilledbyothermethods.
Hulled wheats (Triticum)or glumed barley (Hordeum),which are often
processedby parchingto loosen thechaff,aremore likelyto be preserved
thanbread wheats or naked barley.Based on theseethnographic models,
Hillman (1981) has suggestedthatthestomachcontentsof theTollund and
Grauballe Men may represent the tailings from a cereal crop, rather than
the prime harvest. Jones (et al. 1986) was able to identify several different
stages of processing in samples of grain from Mycenaean storerooms at
Assiros inGreece.
Comparison of Coprolites and Flotation Samples

The comparisonof the range and quantitiesof plant remains in open
sites to those recovered from contemporaneous and culturally related
depositswith betterpreservationmay offer another approach to under
Yarnell (1974) pioneered thismethod in
standingpreservationdifferences.
his comparison of taxa in flotation samples and human coprolites from
Salts Cave in Kentucky. Yarnell found a fairly close correspondence be
tween the two data sets. Nutshell fragments and Chenopodium seeds were
more abundant in the flotation samples (large quantities of nutshell were
probablynot consumed,and groundchenopod seedsmay have been almost
completelydigested).
Gasser (1982) conducteda similarcomparisonof theresultsof flotation
and coprolite analysis for the Anaszai area of the Southwest. In Gasser's
study,squash and beans, preparedby boiling,aregrosslyunderrepresented

in flotationsamples,whilemaize is only slightly underrepresented. Fresh
fruitsare generallyunderrepresented.Sincemany spinachlikegreensyielda
secondharvestof edible seeds, theymay be indirectlyrepresentedin the
flotationrecord.Gasser's data were assembled frommany differentsites
and probablycontainbiases derivedfromdifferencesinseasonality,age, or
site function.
Itmay be useful to compare coprolitesandmacrofossilsrecoveredfrom
a singlesite.Table 4.4 contains sucha comparisonforplant remainsfrom
theTehuacan Valley inMexico. In drycaves, rootcrops suchas pochoteor
freshfruitssuch as sapote, chile, avocado, hogplum, or various cactus
fruitsmay be well represented. The abundanceof plant remainsin thebulk
samplesreflectstheamountofwastematerialderivedfromfoodprocessing
thatisneverseen inopen sites.Earlymaize is almost synonymous with the
Tehuacan Valley, but it is interesting
tonote thatonly 9% of thecoprolites
produced evidence formaize consumption.Perhaps this is becausemaize
kernelsare not very recognizableaftertheyhave been ground and passed
throughthehuman digestivesystem.Difficultieswith identification may
also partiallyexplain theunderrepresentation incoprolitesofmany of the
Tehuacan plant remainsthatarewell represented as macrofossils.Future
comparisonsof plant remainsfromcoprolitesand flotationsamples,or be
tweenopen and protectedsites,may help identify many missingor under
representedtaxa.
CULTURALTRANSFORMATIONSOF
THE ARCHAEOBOTANICALRECORD
The effectsof processing

methodson thearchaeobotanicalrecord,which
has alreadybeen discussed, isone exampleof a culturaltransformation.A
combinationof many other cultural formationprocessesmay produce
patternsof theirown.
significant
At thispoint, itwould be usefultodefinedefacto, primary,and second
ary refuseas theyrelate to plant remains.De facto. refuse is defined as
usablematerial abandoned in an activitylocus (Schiffer1976;Rathje and
Schiffer1982).True examplesof defacto refuseare relativelyrarein thear
chaeobotanical record.Caches of seeds are occasionally found in sealed,
ceramicollas indrysitesin theSouthwest.These caches are so rarethatthey
have inspiredan elaborate oral mythologyconcerningviable seeds ger
minated fromthesedeposits (Nabhan 1977).Collections of plant remains
preservedby catastrophic firescould be consideredas de facto refuse,
although the term "usable" in the definitionno longer reallyapplies.
Primaryrefuseisdefinedas trashdiscardedat thelocationof use (Schiffer
TABLE 4.4
A Comparlson of Coprolltes and Plant Macrofosalla from the Tehuacan Valley:
The Problem of DIffeential Preseatlon'
Samples containing plant material (%)
Plantb Coprolites (N = 87) Bulk samples (N = 51)
Almost equal
Pochote (CelbaXb,r) 59 61
Millet (SetarlaXp) 59 45
Black sapote (DlospyrosXf) 26 31
Other grass (p) 25 27
Overrepresented Inbulk samples

Agave (AgaveXr) 50 92
Mesquite (ProsoplfXp) 16 53
Organ pipe (LemalreocereusXf) 16 24
Prickly pear cactus (OpuntlaXf) 11 61
Chile (CapsIcumXf) 11 31
Beans (PhaseolusXb) 11 25
Maize (ZeaXp,b,r) 9 72
Squash (CucurbitaXb) 3 47
Coyol palm (AcrocomlaXf) 0 33
Amaranth (AmaranthusXp) 0 35
Hogpium (SpondlasXf) 0 43
Avocado (PerseaXf) 0 45
Cotton (GossyplumXp) 0 47
Underrepresented Inbulk samples

Othercacti (f) 50 33
Groundcherry (PhysallsXf) 22 4
Manioc (ManihotXb,r) 11 0
"Data fromCallen 1967; Smith 1967.

bCommon method of preparation: b, boiled; f,fresh; p, parched; r,roasted.
1976;Rathje and Schiffer1982).The dcarred saguaro seeds recoveredfrom

thecookingpits in thesaguaro camp examplediscussedabove are botanical
examplesof primaryrefuse,as are thecholla spinesand pollen from the
roastingpit example (Greenhouseet al. 1981). Secondary refuseis defined
as trashdepositedat some locationother than the locationof use (Schiffer
1976; Rathje and Schiffer 1982). The vast majority of plant remains
recoveredfromarchaeological sitesshouldbe consideredsecondaryrefuse.
One of themost frustrating tasks routinelyassignedan archaeobotanist
is to figure out the function of a pit from a flotation sample, more often
An hypothetical
within thefillof thatfeature.
thannot, collectedsomewhere
Storagepit
Earlierarchaeological
Banialpit horizon
refuse
Primary
4# _ 7 | < g% _ VZ7 Secondaryrefuse
Grassmatting
Roasting
pit
Figure4.1. Three possibleuses of an hypothetical

pit, illustrating
thedepositionof primary
and secondaryrefuse.
example will illustrate some of the problems involved in this task. In Figure
4.1 a "generic" pitwith severalpossible functions
was excavated intosoil
which containedan earlierarchaeologicalstratum.In thefirstalternative,it
was filledwith grass matting and ears of corn and used as a storage pit. If it
was abandoned at this point, themaize ears and grass would be considered
defacto refuse, but since theywere not carbonized theywould soon decom
pose, leaving behind perhaps some phytoliths and a few grains of pollen. If
the ears of maize were removed before they decayed, fragments of grass
matting or cobs from which the kernels had been removed might have been
left behind as primary refuse. Once again they would probably not be
preserved unless theywere carbonized. In the end, the storage pit would be
filled with a mixture of soil from the initial excavation and any available
trash from around the site. This fillwould contain a mixture of secondary
refuse from both the earlier and current occupations. The fill of this
feature, which would probably be identified by an archaeologist as a trash
pit, would have no relation at all to the original (storage) function of the pit.
Based on experimental studies of replicated Iron Age storage pits at the
ButserExperimentalFarm inBritain,Reynolds (1979: 71-82) has suggested
that most types of grain do not need to be parched for storage in subter
ranean pits. Residue left after storage might only be burned if there was a
desire to sterilize the pit before the next usage.
In the second case, the pit was excavated for immediate use as a burial pit
and refilled with the original soil. Secondary refuse from the fill of this
burial pit would have little relationship to the burial itself and would actually
predateit.Perhaps a fewpollen grainsor phytolithsfromfoodor floralof

feringsincludedwith the body would be the only primarymaterial
associated with the burial. Leroi-Gourhan (1975) has suggested that
unusuallyhigh percentagesof polien frominsect-pollinatedflowersfound
associatedwith a Neanderthal grave fromShanidar Cave in Iraq reflect
floralofferingsintentionally depositedwith theburial.
In the thirdcase, thepitwas used for roastingagave. A large firewas
built in thepit toheat stones.After thefireburneddown and therockswere
heated,most of theasheswere removedand thepitwas refilled with alter
natinglayersof hot rock, leafy matting,and agave hearts.The pitwas then
sealedwith earth and a second firewas built on top to help complete the
roastingprocess.A few fragments ofmattingor agave leaves indirectcon
tactwith thehot rocksmight be preservedby charring.After thecooked
agave heartswere removed, some of thischarredmaterial and charcoal
fromthe firewould be leftbehind as primaryrefuse.Once again,most of
the fillof this roastingpitwould be secondaryrefuse,with only a small
amountof charredprimaryrefuseat theverybottomof thepit ormixed in
with the fill.In recentstudiesof roastingpits associatedwith dry farming
featuresnorthwestof Tucson, Arizona by P. Fish, S. Fish, and myself,
agave fibers,thorns,leafbases, and heart fragments have been identified
fromeverypit sampledby flotation.These agave remains,alongwithwood
charcoaland grass stems,seem to represent primaryrefusefromtheagave
roastingprocess.
How might de facto, primary,and secondaryrefuseappear in real ar
chaeobotanical assemblages? The "Mckellar Hypothesis" (discussed in
Rathje and Schiffer1982) suggeststhat inareas thatare regularlycleaned,
smallobjects aremore likelyto be leftbehind as primaryrefuse.This sug
geststhatflotationsamples fromhearthsthatare cleaned or floorsthatare
sweptare likelyto producemostly small seeds.Table 4.5 contains size and
densitydata forcarbonized seeds fromtheTanque VerdeWash site,a small
farming village in theeasternTucson Basin thatwas occupied betweenA.D.
1000 and 1100 (Miksicek 1986a). One of the catastrophicalyburned pit
houses was a storage structure with a floor assemblage consistingof 17
reconstructable vessels and over 180,000charredmaize kernels,beans, and
squash seeds.Concentrations of predominantly one taxonwere associated
with some of thevessels, suggestingstoragebehavior similarto thatfound
by Jones (et al. 1986) atAssiros. Two other structuresalso containedhigh
densitiesof charred seeds and relativelycomplete floor assemblages that
suggestedrapid abandonment.The plant remainsfromthe floorsof these
pit houses could be considereddefacto refuse.The seedswere fairlycom
plete (allowing for some damage during excavation and flotation).The
overalldensityof charredmaterialwas veryhigh (287.55 seeds/liter)and
most of themacrofossilswere smaUerthan3mm.Structuresthatwereburned
TABLE 4.5
Carbonized Plant Remains fromtheTanque Verde Wash Site: Size Sorting and Sed
Density
Density Large
Feature type (seeds/liter) seeds (%)
Castastrophically burned structures 287.55 36.4

Structures burned after abandonment 4.06 42.2
Extramural activity areas 3.41 30.7
Extramural pits 1.95 42.0
Unburned structures 0.69 70.0
Trash mounds 0.52 86.4
afterabandonmentyieldeda much lowerdensityof charredremains(4.06

seeds/liter)with 58% smaller than 3 mm. These small seeds, mostly
Amaranthus,were probablyprimaryrefusefromplant processingthatoc
currednear thehearth.Unburned structures yieldedvery fewcharred re
mains (0.69 seeds/liter)most of which were durablemesquite seeds and
maize cob fragments,that is, secondaryrefusedeposited in the structures
afterabandonment.This interpretation is strengthened by the similarity
with the trashmounds samples.Plant remainsfromtheactivityareas and
extramuralpitsweremost similarto thosefromthepit houses-thatburned
afterabandonment.Both thesecategories includedhearths and roasting
pits,which probablycontainedprimaryrefuse.
Hally (1981) discovered similarpatternsat the Little Egypt site in
Georgia. One burned structureyielded a pile of large hickorynutshell
fragments next toa hearth(probablytobe used as fuel),a clusterof persim
mom seeds next to a storagejar, and numerouscorncob fragments on the
floor(Hally 1981).This patternedfloorassemblage,combininga relatively
high densityof "seeds" (2.28/liter),at least forthat site, (seeTable 4.3)
with thepresenceof largenutshellfragments (82%were largerthan3mm),
suggestsmostly de facto refuse.One unburned structureyielded only
primaryrefuse:small cob and nutshellfragments.
Such factorsas the overall densityof preserved, plant remains, their
average size, and thedegree towhich theyare completemay help an ar
chaeobotanistrecognizethe typeof deposit fromwhich theyoriginated.
Both primaryand secondarydeposits are likelyto contain only limited
samplesof thepotentiallypreservedrangeof plantsfroma site.Secondary
trashis likelyto be biased towardthe largerandmore durable end of the
scalewhereas primaryrefusemay contain smallerseeds.By samplingboth
typesof contexts,or by takingfulladvantageof raredefacto depositswhen
theyare encountered,a botanical specialistismore likelyto identifythe
totalrangeof taxa preservedin a given site.
The "SchleppEffect" (discussedinRathje and Schiffer1982)was derived

fromtaphonomicstudiesand hasmostlybeen applied to faunalremains,but
itmay also have some relevancetoarchaeobotany.In itsoriginalzooarchaeo
logicalcontext,theSchleppEffectstatesthattheamountof butcheringper
formedon an animal is directlyrelated to the size of theanimal and the
distanceto theplace itwill be consumed.For a largeanimal,only themost
usable elementswill be broughtback to thehome base. In termsof plant re
mains, theSchleppEffectmay be interpreted as: themore primaryprocessing
of a plantproductthatoccursat a site,themore waste productswill be pro
duced.For example, inAnasazi sitesmaize seemstohave been storedon the
cob, but inHohokam sitesmaize seems to have been storedin the formof
shelledkemels in jars or baskets.Cob fragments are usuallyonly recovered
from Hohokam sitessituatedclose to canals or dry farmingfeatures.If cot
ton seedsor calyx fragments are recoveredfroma Hohokam site,itsuggests
thatcottonwas grownand processednearby.The recoveryof only cotton
fibersor completedtextilesfroma drycave, however,may be interpreted as
evidencethatthecottonwas produced elsewhereand importedto the site.
The "Clarke Effect" (discussedinSchiffer1983) statesthatthediversity
of artifactsrecoveredfroma site isdirectlyrelatedto the lengthof occupa
tionof a site.Thismay be directlyapplied to archaeobotanicalassemblages.
Figure4.2 presentsdata forthenumberof typesof "seeds" identifiedfrom
61Hohokam sites in southernArizona. The soild line representsthe ex
pectednumberof taxa fora givennumberof samplesbased on a regression
of the log1oof thenumberof samples analyzed against the log,. of the
numberof taxa identified. The taxon-samplerelationshipaccounts forap
proximately41%Oof thevariance in thedata set.As more samplesare ex
amined thenumberof rarertaxa encountered increases. It is possible to
predicttheexpectednumberof taxa froma givennumberof samplesusing
this log-log relationship.A single8-literflotationsample should yield an
averageof fivedistincttypesof seeds.Each additional fivesamples should
add another taxon.The dashed lines inFigure 4.2 indicatethe95% con
fidence intervalscalculated for 1 to 5, 5-10, 10-20, 20-50, 50-80, and
80-100 samples. Sites were classified as agriculturalfields, fieldhouses,
farmsteads, hamlets,villages,or primaryvillageswith ceremonialstructures
suchas ball courtsor platform mounds based on various archaeologicalcri
teriasuch as numberof structures, permancyof construction,amount of
accumulatedtrash,and presenceof hearths.Field houses (oneor two struc
tures)generially yielded fewerthan theexpectednumberof taxa (theytend
to fallbelow theregressionlineinFigure4.2). Field house sitesalso produced
lowerfrequenciesof maize and gatheredwild plants such as mesquite or
saguaro and higherproportionsof agriculturalweeds. Farmsteads (two to
fivestructures)tend to fallnear or below theregressionline inFigure 4.2.
Hamlets or villages (more thanfivecontemporaneousstructures), produced
35
30 A
A
~25
Nubr of A
-
A A
10 - -.
0 10 20 30 40 50 60 70 80 90 100
Number of samples
Figure 4.2. The relationshipbetween sample diversityand sample number for 61 sites in
southernArizona. Agriculturalfields(0)); fieldhouses (0); farmsteads(U); hamlets(-);
villages(O); villageswith ceremonialstructures mounds (*).
suchas ball courtsor platform
more maize and gathered

higher than expected levelsof plant diversity,
plants,and feweragricultural weeds. The villages tendedto fallabove the
regressionline inFigure4.2. These resultsseem to confirmtheClarke Ef
fect. The longer a site is occupied, the greater the range of activities thatwill
be carried on at that site, and themore diversity of plant remains that will
be preserved.
TRANSFORMATIONPROCESSES
ENVIRONMENTAL
Once charredbotanical remainsare deposited in an archaeologicalsite,

they are subject to any of a number of environmental factors. A very sim
ple relationship is evident from an examination of the data on seed density
for various types of open sites presented in Table 4.2. Older sites produce
fewercarbonizedseeds. Since charredseedshave been reducedto elemen
tal carbon,which is essentiallyimmuneto further organicdecomposition,
it is not age itselfthatproduces thisrelationship.Small, fragile,carbon
ized seedsmay be mechanicallydestroyed, mixed, or displaced in the soil
by any of the soil formation processes described inWood and Johnson
(1978).A second relationshipisevidentinTable 4.2. Sites inmore mesic en

vironmentsproduce fewercharredseeds thansitesinsemi-aridregions.This
suggeststhatthesesoil formationprocessesoccurmore rapidlyinmoist en
vironments.
Faunalturbatlon
In thediscussionof thesaguaro camp example, Imentionedone example
of faunalturbation, seeds transportedintoa siteby a pack rat.This isan ex
ampleof an additive,natural transformation; theadditionof newmaterial,
completelyunrelated to theoccupation of a site,by a nonhuman agent.
During theexcavationof StarCarr, a cache of hazelnutswas uncovered in
what appeared to be a goodMesolithic context,but thepresenceof rodent
incisormarks ledClark (1954: 60) to conclude that thesenuts had been
depositedby squirrelsaftertheoccupation of thesite.Granivorous rodents
or insectsmay createcachesof seeds in sitesthatcompletelyunrelatedto the
human occupationof thesame site.Sincemost archaeobotanists working in
open sitesuse thegeneral rule thatonly carbonized seeds should be con
sideredancient thisshouldnot be toomuch of problem. Intrusiveseeds are
not always so easilyrecognizedinwaterloggedsitesor drycaves. This prob
lem is not limitedto seeds. Since leaf harvesterants collect flowersor
vegetativematerial, intrusivemicrofossilsmay be an ever-present and not
in
easily recognizedproblem pollen and phytolithanalysis. Pulliam and
Brand (1975) estimatethat in the desertgrasslandsof southernArizona,
granivorous rodentsharvest an average of 3,000,000 seeds/hectareeach
year,while foragingantsmay harvestas many as 6,500,000/hectare.This
amountsto about 3% of thetotalannual seedcrop.Miller and Smart (1984)
describeda mechanism fortheintroduction of carbonizedseeds intositesby
a combinationof animal and human action. They suggestedthatcharred
seeds be incorporatedinto archaeological sites if the dung of domestic
aniimalswas used forfuel.
Faunalturbation is not only additive.The action of burrowinganimals
may mix previouslydepositedmaterial. Smallmammalsmay burrow to a
depthrangingfrom0.5 to 2.5m dependingon speciesand substrate(Kirmiz
1962; Schmidt-Nielsen1964;Wood and Johnson1978). Small rodentsmay
mix between2.5 and 18metric tonsof soilper hectareeach year (Wood and
Johnson1978).Tunnels inant nestshave been reportedas deep as 2 to 5m
below thesurface(Tevis 1958;Wood and Johnson1978).Earthwormsmay
burrowto a depthof 3m (Wood and Johnson1978) and theyfrequently use
seedsor small stonesto line theirtunnels(Keepax 1977).Earthwormsmay
mix between0.4 and 9metric tonsof soil per hectareeach year (Wood and
Johnson1978). Inmoist areaswith a shallowwater table,crayfish may bur
row to a depth of 5 to 8 m and bring a metric ton of soil per hectare to the
surfaceeach year (Wood and Johnson1978). Sinceburrowinganimals seem

to like the relativelyloose, organic, richsoils of archaeological sites, the
possibilityofmixingof prehistoric plant remainsby faunalturbation should
alwaysbe considered.
Floralturbatlon
Wood and Johnson(1978)describefloralturbation as themixingactivitiy
of soilby plants.As treerootsdecay theyleavebehindcavitiesthatmay fill
withorganicdebris.When treesblow over instormsthey may churnup and
mix largevolumesof soil or createdepressionsthattraporganicdebris. In
the tropicalregionsof CentralAmerica, cohunepalms (Orbignyacohune)
leavebehind largeconical depressions0.5m indiameterand up to Im deep
when theyareblown overor decay.Furley(1975)estimatesthatthisactivity
couldmix 5500m3 of soil per hectare(55% of theupper soil volume) each
milleniumin tropicalregionswhere thisspeciesgrows.Since slash-and-burn
farmingis almost synonymous with tropicalagriculture,and sincemodern
Maya farmers use cohunepalms as indicators of good soil formaize milpas,
thereisa veryrealdanger thatmodern seedscharredby fieldburningcould
go unrecognizedin shallow archaeologicaldeposits.The potential for the
mixingof soils by treefallsor infillingafterthedecay of stumpsor roots
shouldbe consideredforarchaeologicalsites inanywooded region.
Argililturbatlon
Argilliturbationis definedbyWood and Johnson(1978) as soil distur
bance caused by theshrinkingand swellingactivityof clays as theyabsorb
or losemoisture. The 1981 fleld season at PulltrouserSwamp (Miksicek
1983a)was one of thedriestperiods in recentclimaticrecords innorthern
Belize. Massive soil cracks that were 10 cm wide and over a meter in depth
formedin theupland vertisolsaway fromthePulltrouserBasin. This crack
ingactioncould havemixed earlierdepositsor trapped
modern seedschar
redby slash-and-burnland clearing.
Aeollan and Alluvial Processes
The action of wind (aeroturbation)or water (aquatubation)may alter

shallow archaeological deposits by mixing, covering,or eroding them
(Wood and Johnson 1978). The announcementof the discoveryof Late
Paleolithicbarley fromsitesnearWadi Kubbaniya inEgypt appeared to
shake establishedopinions on theantiquityof plantmanipulation in the
Near East (Wendorfet al. 1979). Grinding stonesand blade toolswere
collectedfromdeflateddepressionsinnearbydunes.The barleygrainswere
described as "carbonized" but "probably not burned" (Wendorfet al.
1979: 1345).Although the exact contextof thecereal grainswas not de
scribed, radiocarbon samples that seemed to be in association with the
barleydated to between 17,100 and 17,670 radiocarbonyearsbefore pre
sent. These were collected fromdepths rangingbetween 0 and 30 cm.
Several years laterHillman (et al. 1983) ran electronspin resonance spec
troscopictestson theWadi Kubbaniya barleygrains and determinedthat
themaximum temperaturetowhich theyhad been heated did not exceed
150'C, whichwould have been insufficient to char themenough to last the
presumed18millenia.More recently,thesebarleygrainswere dated at the
Universityof Arizona tandemaccelerator facilityand found to be a max
imumof 4850 yrold (Wendorfet al. 1984). Shiftingsands and othernatural
processes had mixed more recentbarley grainswith ancient charcoal.
Several factorsshouldhave alerted theexcavatorsto thepossibilityof con
tamination: the uncharred state of the grains, the shallowness of the
deposits, theunstablesitecontext(dunes), and the lackof clear association
with grindingstonesor similar tools.
Background Seed Rain and Fires

Soil acts as a "seed bank" for natural plant communities. Hopkins and
Graham (1983)have reportedseeddensitiesof 558 to 1068/Miof topsoilfor
lowlandrainforestinQueensland,Australia. Similardensitiesof 177 to 752
seeds m2 were noted for abandoned fieldsand adjacent rain forest in
Amazonian Venezuela (Uhl et al. 1982). Pulliam and Brand (1975)
estimatedthatannual seed production in thedesertgrasslandof southeast
ernArizona averaged 350,000,000 seeds/hectare.Lopinot and Brussell
(1982) recoveredan average of 38.6 uncharredseeds/literfrom flotation
samplescollectedat sites in southernIllinois.Keepax (1977) noted between
74 and 1506 uncharredseeds/kgof topsoil for IronAge sites inBritain.
Minnis (1981) identified100-2100modern seeds/literin control samples
collected fromoffsiteareas in southwestern New Mexico. Minnis (1981)
refersof theseuncharredseeds in archaeological sitesas the"modern seed
rain."With somany background seeds in soil it is conceivable thata few
could be carbonizedby natural orman-caused fires.
Sauer (1952) and Lewis (1972) have suggestedthatfirehas been an im
portant tool for hunting or clearing land for almost as long as humans have
existed.Slash-and-burnis essentiallysynonomouswith tropicalagriculture.
The Danish palynologistIversen(1941) cited thedeclineof elm pollen and
increases in herbaceous pollen in northernEuropean bog sequences as
evidencefor theuse of slash-and-burnclearingby earlyNeolithic farmers.
Recently theuse of a "controlledburn" strategyhas even been suggested
forprehistoricfarmersin theAmericanSouthwest.Sullivan (1982)propos

ed thatMogollon peoples could have intentionallyburnedsmallplots to in
crease the foragingand agriculturalpotentialof ponderosa pine forests.
To testforthepossibilityof a naturallycharredseed rain,Minnis (1981)
collectedsurfacesoil samplesfromoffsiteareas in southwesternNew Mex
ico.Although thousandsofmodern seedswere identified, not a singleone
was carbonized. Even if the real incidenceof naturallycharred seeds is
ratherlow, it is a factor that sould be consideredby everycautious ar
chaeobotanist.
Evaluating theOrigins of Seeds InSites

In the preceedingsections I have discussedmechanisms for thenon
human introduction of seeds intosites.An experimentbyMinnis (1981) il
magnitudeof someof thesefactors.In 1978he floated
lustratestherelative
floor sweepingsfroma chickencoop thathad servedas a dormitoryfor
membersof theMimbresArchaeologicalProject theprevious fieldseason.
He identified684 seeds of 19 differenttaxa fromthreeflotationsamples.
Only 0.7% could definitelybe attributedto human introduction(chile,
teparybeans, sunflower).Approximately11.2% of the seeds could have
been carried into the coop by rodents,but only 5.5% showed definite
evidenceforrodentgnawing.The remaining 88% couldhavebeen introduced
by the combined effectsof insects,shiftingsands, rodents,humans, or
otheranimals.
Various authors haves proposed criteriaforevaluating seeds fromar
chaeological sites (Keepax 1975;Minnis 1981; Lopinot and Brussell 1982;
Miksicek 1983b). When the is a doubt, the burden of proof is always on the
archaeobotanist.
1. What is the state of preservation of the seed? If it is from an open site

and it is not carbonized then itmay well be a recent introduction. If the seed
is not charred and it does not look old, then it probably is not. Black or
dark-colored seeds always present a special problem for the ar
chaeobotanist. If a taxon is abundant, it is always worthwhile to break a
fewseedsopen and determineif theyare completelycharredor iftheycon
endosperm.Although this techniqueis destructive(any
tain fresh-looking
criticalmeasurementsshould be takenfirst)itmay save a fewheadaches.
2. Is this species part of themodern local vegetation and seed rain? Con
trol samples collected away from a site will be useful for determining the
diversityand densityof backgroundseedsas well as indicatingthepossible
existence of a naturally charred seed rain.
3. Is thereany evidence fordisturbancein thesoil profile such as soil
cracks,animal burrows,plowing,or intrusivepits?
4. Even if theseeds are carbonized, is thereevidence forancientanimal

disturbancesuch as charred rodentor insectfecalpellets, teethmarks, or
charred insectbodies?
5. Is theregood ethnographicevidence for the species inquestion?
6. Is thereany size ormorphological featurethatwould distinguishthe
seed inquestion frommodem populations?
7. Will any of the uncharred seeds germinate? If a flotationsample
sproutswhile it isdrying,thenthoseseedsprobablyare not veryold. A sim
ple germinationtestmay resolvemany questions.
8. How abundant is the taxon in question? It is hard to argue with a
vessel fullof charredseedsor a storeroomfullof burned corn.One or two
individualsof a species froma largenumberof flotationsamplesmay not
be verysignificantin the long run.
9. How old is theseed itself?
A final,but ratherexpensiveway to resolve
any possibleproblemswith criticalmaterialwould be to date itdirectlyus
inga tandemaccelerator.
ANALYTICALTRANSFORMATIONSOF THE DATA BASE
In The structure
of scientificrevolutions
Kuhn (1970) discussed the rela
tionshipbetween"seeing" and theprocessof normal science.The theoretical
orientation of an investigator and themethods and instruments that he uses
influencethe interpretation
of theresultsof any research.To paraphrasea
contemporary adage, "How you see, affects what you get." In archaeobot
any, as well as any other research, how the data are collected influences the
finalanalysisand interpretation. are justas signifi
"Laboratorytransforms"
cant as "natural" or "cultural" transforms in paleoethnobotanical research.
Sampling
There is almost no such thing as a site with no preserved plant remains.
If enough samples of sufficient size are collected and analyzed just about
any site should yield some data. What is an adequate sample? This must be
answered both in terms of sample volume and sample number.
The data in Table 4.2 suggest that there is a very wide range in the abso
lute abundance of preserved plant remains in sites depending on age, preser
and sitehistory.In archaeobotany,"one sizedoes not
vationenvironment,
fit all." The sample volume must be adjusted according to the specific re
quirements of each site and project. It would be useful to experiment with
various sample volumes to determine the optimal size when working in a
new area. The choice of a given volume must be a compromise between ade
quate recovery and logistic problems involved in handling the samples.
100 ooo
80 ?
?0 -- Maize
10-~ ~ ~ ~~~~~~~~. *0
-.*-
60
40- __ Pigweed
U~~~~~~
Li. /~~~~~ /~~~
20-
Ben
0 i
0 10 20 30 40 50
Cumulative number of samples
FIgure4.3. A "species-area curve" fordeterminingtheadequacy of sampling.
Larger samplesproducemore plant remains,but theyaremore difficultto

transport,take longer to process and analyze,and requiremore storage
space. The collection of a standard volume sample makes statistical
manipulationsof thedata easier and more reliable.
The next question to consider is, "What is an adequate number of
samples?" To answer thisquestion I used data for 69 8-liter flotation
samples fromPueblo Las Fosas, a Classic Period Hohokam site near
Florence,Arizona. I shuffledtherawdata sheetsand calculated frequency
values for 1, 2, 3, 4, 5, 10, 15, 20, and so forth up to 50 samples at a time for
six taxa.The conceptof frequencywill be discussedfurtherina latersection
on quantificaton,but it isdefinedas "the percentageof analyzed samples
that contain a given taxon." The results for three species are plotted in
Figure 4.3. The curves for twoother taxa (Trianthemaor falsepurslane,
mesquite seeds) were essentiallyidenticalto those for beans and maize,
respectively.The sixthspecies (tobacco)was so rare that itwas not drawn
until the sixty-ninthtry.Figure 4.3 is similar to a species-area curve, a
techniquethat is commonlyused to testtheadequacy of a sample inquan

titativeecology. In Figure 4.3, it is possible to see thatbetween 5 and 10
samplesare necessaryto produce a reliableestimateof thepopulation fre
quency for a common taxon such as maize. On the other hand, 25-35
samples are required for rarer typesof plant remains such as beans or
pigweed.Fasham andMonk (1978)performedsimilartestsforcerealsfrom
IronAge pits inBritainwith essentiallythe same results.Five or six flota
tionsampleswould be an absoluteminimum fora feature,temporalphase,
or site to identifythemore common species preserved,but 30 samples
would be much more reliable.More sampleswill always add a fewaddi
tional typesof plant remains(see Figure 4.2).
The problem of where to sample depends on the specificgoals and
researchdesignof the individualarchaeologicalproject.General guidelines
have been presented inBohrer and Adams (1977) and Adams and Gasser
(1980). The developmentof a samplingdesign should involveclose col
laborationbetween the project director, the archaeobotanist,and other
analyticalspecialists.
Processing
Althoughmany differenttypesof flotationsystemsare inuse today,they
generally fall into twomajor types: tub flotationor continuous-flow
machines. These twomajor systemsare described inWatson (1976) along
withmany of thepossible variations.The choice of a systemdependson the
individualpreferencesof theprojectdirectorand analystas well as logistic
factorssuch as cost, availabilityofwater andmaterials, and theamountof
material to be processed.
When used by an experiencedtechnician,the resultsof either tub or
continuous-flowflotationshouldbe fairlyconsistent.To testthe recovery
efficiency of various flotationsystems, Kaplan andMaina (1977),Wagner
(1982), and Pendleton (1983) have suggestedadding charredmodern seeds
(ofvarious sizes) as tracersand calculatingthepercentagerecovered.This is
essentiallyidentical to the use of exotic pollen tracers in palynology to
calculate absolute pollen influx.Wagner (1982) reportedrecoveryratesof
84 to 98% for flotation machines and slightly lower values of 6 to 94% for a
tubsystem.The wide variance forthe tub systemcould be attributedto the
mesh sizes.When screenswithmesh openings smallerthan
use of different
0.59mm were used, recoveryrates increasedto 81 to 94%, comparable to
themachine system.Inmy ownwork, I tendtouse thetubsystemdescribed
inMinnis and LeBlanc (1976). I have testedrecoveryratesusing sample
volumesvaryingfrom 1 to 15 literswith resultsrangingfrom79 to 100%.
The lowestrecoverywas obtained froma 15-litersample,which suggests
thatsoil volumemay also affect recoveryefficiency.Wagner (1982) and
Pendelton (1983) have recommended usingvarious typesof tracersto com

pensate fordifferencesin seed buoyancy.The use of charred tracersmay
also helpmonitor cross-samplecontamination.
Preprocessingmay also affectrecoveryrates.Soil samples fromthemost
recentexcavationsat Snaketown inArizona were all screened in the field
before theywere given toBohrer (1970) forflotation. When the resultsof
the Snaketown analysis are compared to botanical data from other
Hohokam sites (Gasser anMiksicek 1985), severalstrikingdifferencesare
apparent.Maize was not recoveredfromtheSnaketown flotationsamples
even thoughit iswell representedat all otherHohokam sitesand despitethe
factthatitwas identifiedby Jonesfromearlierwork at Snaketown (Castet
terand Bell 1942: 31-32). In contrastto othersites along theGila River
(GasserandMiksicek 1985: Figure 1), cottonseemsconspicuouslyabsent in
theSnaketownsampleseven thoughit toowas reportedin theearlierstudy
(Castetterand Bell 1942: 32). Mesquite and saguaro,however seem to be
overrepresentedat Snaketown in comparison to other Hohokam sites
(Gasser and Miksicek 1985: Figure 1). I would stronglyadvise against
prescreeningflotationsamples.Charred seeds are fragileand it iswise to
minimizepossiblemechanical damage to theseeds.All of thesmallartifacts
or animal bones that are recovered in the heavy fraction may be retrieved by
the flotationanalystand sent to theappropriatespecialist.
Quantification
The topicof quantificationis one area of paleoethnobotanythat still
needs considerableexplorationand research.Various methods have been
utilized. In his analysisof plant remainsincoprolitesrecoveredfromSalts
Cave inKentucky,Yarnell (1969) used a relativescale rangingfromE
(trace) to A (abundant).Although this systemsomewhat limitsfurther
statisticalanalysis itmay be themost realisticapproach consideringthe
vagariesof differential preservationand recovery.Bohrer (1970) used a seed
concentrationindexdefinedas "the numberof seeds,dividedby thevolume
of charcoal recovered." I used a modified concentrationindexpreparing
Table 4.2 (seeds/liter of soil). Other authors(forexample,Renfrew 1973)
have used a relative abundance measure, defined as "the number of seeds of
one species divided by the total number of all seeds recovered." Relative
abundance is similar to the concepts of relative frequenciesused in
palynologyor relativedensitiesused inplant ecology. Perhaps themost
common statisticused today is frequency(definedpreviously),which has
also been referredto as presencevalue (Hubbard1980) or ubiquity (Gasser
1982).Frequencyor presencevalue isa statistic
borrowedfromquantitative
ecology.Relative abundance data use absolutecounts fordifferenttaxa,
while frenquencymeasures how commonlyrepresentatives of a taxonoccur
in independentsamples. Itwould seem useful to borrow another statistic

fromquantitativeecology and combine both typesof informationin a
summarystatisticsuch as importancevalue,which is an averageof twoor
more distinctmeasures (Miksicek 1983c). Importancevalue should be in
terpretedas importancein the archaeological recordand not importance
in thediet.
One of themost commonlystatedgoals of paleoethnobotanicalresearch
of past diets." In realitythisis impossiblebecause of
is the"reconstruction
problems with differentialpreservation, recovery,and many of the
aforementionedculturaland environmentaltransformation processes.An
archaeobotanical sample representsonly a small portion of theplant re
mains thatare preservedat a site,which are in turnonly a small fractionof
theplants thatwere actuallyutilizedby thepeople who lived there.
When Cohen (1975) analyzed plant remains in a Late Horizon midden
fromcoastal Peru, he identified over 40,000 itemsof vegetal refusefroma
4.5-m3volume.He estimatedthatthismiddenwas deposited inabout 70 yr.
Even with sucha large,well-preservedsamplehe feltthatitwas impossible
to calculate relativedietary proportions because certain items such as
squash, limabeans, root crops, and certain tropical fruitsthathad been
identifiedfromotherpartsof the siteor other sites in theregionwere con
spicuouslyunderrepresented. Several quantitativedietary reconstructions
have been attempted(forexampleMacNeish 1967; or Pozorski 1983)but in
thelongruntheseare probablypurelymathematicalexercises.It isdifficult,
ifnot impossible,to estimatethe relativedietary importanceof different
typesof plants.Chronological or spatial trendsfora single taxonmay be
farmore realistic.In the finalanalysis,general trendswill be much more
meaningful thanabsolute numbers.
CLOSING THOUGHTS
In the foregoingdiscussion I have triedto outline someof theprinciples

behind archaeobotanicalanalysis.Although I am farmore familiarwith
data fromtheAmerican Southwest,I have triedtobring togetherexamples
and ideas frombothEuropean andAmerican researchers.I have triedtobe
comprehensive but I certainlydo not assume thatthischapter is inanyway
exhaustive.
The concernsmentioned above are well understood by practicing
paleoethnobotanistsbut theyare often intuitiveand are not always stated
explicitlyineverypublishedreport.Almost all of theseconcernshave been
expressedsomewherein thepublished literature but I have triedto bringas
as
many possible tQgether inone place. I hope that in someway thisdiscus
sionwill proveuseful toboth specialistsand archaeologistsalike and that in
some smallway itmay contributeto an understandingof site formation

processesand archaeobotanical interpretation.
ACKNOWLEDGMENTS
The seedsofmany of the ideas discussed in thispaperwere planted almost 15yearsago in

discussionswithHugh C. Cutler and LeonardW. Blake, and I wish to dedicate thisarticleto
them.In someways, thispaper ismodeled aftera course inQuaternaryPalynology taughtby
Alan Solomon at theUniversityofArizona in thespringof 1976. I also benefitedgreatlyfrom
a workshop on archaeobotanyconducted in 1983 at theUniversityof Arizona by Vorsila
Bohrer and Karen Adams. I wish to thankBruce Benz, David Browman,Gary Crawford,
Owen Davis, Bill Doelle, Clark Erickson,Patricia Fall, Richard Felger, Paul Fish, Suzanne
Fish, Richard Ford, Robert Gasser, Norman Hammond, ChristineHastorf, Sissel Johan
nessen,ScottKwiatkowski,Neal Lopinot, ScottyMacNeish, Naomi Miller, Paul Minnis,Gary
Nabhan, Barbara Pickersgill,Virginia Popper,Michael Schiffer,C. Earle Smith Jr.,Alan
Sullivan, JillThompson,Molly Toll, B. L. Turne, Gail Wagner, Patty JoWatson, Fred
Wiseman, Richard Yarnell, and anyone I may have inadvertentlyforgottenformany
stimulatingdiscussions throughouttheyears.
REFERENCES
Adams, K. R.
1980 Pollen, parched seedsand prehistory: of prehistoricplant re
a pilot investigation
mains fromSalmon Ruin, a Chacoan pueblo in northwestern New Mexico.
EasternNew Mexdco University ContributionsinAnthropology 9.
Adams, K. R. and R. E. Gasser
1980 Plant microfossils fromarchaeological sites: researchquestions, and sampling
techniquesand approaches. TheKiva 45(4): 293-300.
Asch, N., R. I. Ford, and D. L. Asch
1972 Paleoethnobotanyof theKoster Site, thearchaichorizons. IllinoisStateMuseum
No. 24.
Reports of Investigations
Begler,E. S., and R. W. Keatinge
1979 Theoreticalgoals andmethodologicalrealities:problems in thereconstruction of
prehistoricsubsistenceeconomies.WorldArchaeology 11(2): 208-226.
Bohrer,V. L.
1970 Ethnobotanical aspects of Snaketown,a Hohokam village in southern Arizona.
AmericanAntiquity 35(4): 413-430.
Bohrer,V. L. and K. R. Adams
1977 Ethnobotanical techniquesand approachesat SalmonRuin, NewMexico. Eastern
New Mexico UniversityContributionsinAnthropology8.
Bottema,S.
1984 The compositionofmodern charredseedassemblages.InPlants and ancientman,
edited by W. van Zeist and W. A. Casparie. Boston: A. A. Balkema. Pp.
207-212.
Byers,D. S. (editor)
1967 The prehistoryof theTehuacan Valley,Vol. 1,Environmentand subsistence.
Austin: Universityof Texas Press.
Callen, E. 0.
1967 Analysisof Tehuacan coprolites. In The prehistoryof theTehuacan Valley,Vol.
1, Environmentand subsistence; edited by D. S. Byers. Austin: Universityof
Texas Press. Pp. 261-269.
Castetter,E. F., andW. H. Bell
1942 Pima and Papago Indian agriculture.Albuquerque: Universityof New Mexico
Press.
Clark, J.G. D.
1954 Excavations at Star Carr. Cambridge: CambridgeUniversityPress.
Cohen, M. N.
1975 Some problems in thequantitativeanalysisof vegetable refuse illustratedby a
Late Horizon siteon thePeruvian coast.Nawpa Pacha 10-12: 49-60.
Conrad, N., D. Asch, N. Asch, D. Elnore, H. Grove,M. Rubin, J. Brown,M. Wiant, K.
Farnsworth,and T. Cook
1984 Acceleratorradiocarbondatingof evidenceforprehistorichorticulturein Illinois.
Nature 308: 443-446.
Crawford,G. W.
1983 Paleoethnobotanyof theKameda Peninsula Jomon.Anthropological Papers,
Museum of Anthropology,UniversityofMichigan No. 73.
Croes, D. R., and E. Blinman (editors)
1980 Hoko River: A 2500 year old fishingcamp on the northwestcoast of North
America.WashingtonState University, LaboratoryofAnthropology,Reports of
No. 58.
Investigations
D'Altroy, T. N., and C. A. Hastorf
1984 The distributionand contentsof Inca state storehousesin theXauxa regionof
Peru. AmericanAntiquity49(2): 334-349.
Dennell, R. W.
1974 Botanical evidence for prehistoriccrop processing activities. Journal of Ar
chaeologicalScience 1: 274-284.
1976 The economic importanceof plant resourcesrepresented on archaeological sites.
Journalof Archaeological Science 3: 229-247.
de Rochebrune,A. T.
1879 Recherches d'ethnographiebotanique sur la flore des sepulturesp6ruviennes
d'Ancbn. Actes de la Societe Linn&ennede Bordeaux. 3: 343-358.
Doebley, J. F.
1981 Plant remainsrecoveredby flotationfromtrashat Salmon Ruin, New Mexico.
TheKiva 46:(3): 169-187.
Farnsworth,P., J.Brady,M. DeNiro, and R. S. MacNeish
1985 A re-evaluationof the isotopicand archaeological reconstructionsof diet in the
Tehuacan Valley. AmericanAntiquity 50(1): 102-116.
Fasham, P. J., and M. A. Monk
1978 Sampling forplant remainsfromIronAge pits: some resultsand implications.In
Sampling incontemporary Britisharchaeology,editedby J.F. Cherry,C. G., and
S. Shennan.BAR BritishSeries (50: 363-371.
Fish, P. R., S. K. Fish,A. Long, and C. Miksicek
1986 Early corn remainsfromTumamoc Hill, southern Arizona. American Antiquity
51(3): 563-572.
Ford, R. I.
1979 Paleoethnobotany in American archaeology. In Advances in archaeological
method and theory; Vol. 2, editedbyM. B. Schiffer.New York: Academic Press.
Pp. 285-336.
Ford, R. I., and N. Miller

1978 PaleoethnobotanyI. InExcavationsat Carthage,1976, conductedby theUniver
sity of Michigan Vol. IV, edited by J. H. Humphrey. Ann Arbor: Kelsey
Museum, The UniversityofMichigan. Pp. 181-187.
Furley,P. A.
1975 The significance of theCohune Palm, Orbignyacohune (Mart.)Dahlgren on the
natureand in thedevelopmentof thesoil profile.Biotropica 7(1): 32-36.
Gasser, R. E.
1982 The specialists' volume: biocultural analyses. The Coronado Project Ar
chaeological Invesitigations
No. 4,MNA ResearchPaper 23.
1985 Archaeological trashmounds and floorfeatures:don't believeeverything. Paper
presentedto theSociety forAmericanArchaeology,Denver, CO.
Gasser, R. E., and E. C. Adams
1981 Aspects of deteriorationof plant remainsin archaeologicalsites: theWalpi ar
chaeological project. Journalof Ethnobiology1(1): 182-192.
Gasser, R. E., and C. H. Miksicek
1985 The specialists:a reappraisalof Hokokam exchangeand the archaeobotanical
record. In Proceedingsof the 1983Hohokam Symposium,editedbyA. Dittert
and D. Dove. Phoenix Chapter, Arizona Archaeological Society Occasional
Paper (2): 483-498.
Glob, P. V.
1971 The bog people: IronAge man preserved.New York: BallantineBooks.
Green, F. J.
1979 Phosphatic mineralizationof seeds fromarchaeological sites. Journalof Ar
chaeological Science 6: 279-285.
Greenhouse,R., R. E. Gasser, and J.W. Gish
1981 Cholla bud roastingpits: an ethnoarchaeologicalexample. The Kiva 46(4):
227-242.
Hally, D. J.
1981 Plant preservationand the contentof paleobotanical samples: a case study.
AmericanAntiquity46(4): 723-742.
Hammond, N., and C. H. Miksicek
1981 Ecology and economyof a formative Maya siteatCuello, Belize. Journalof Field
Archaeology 8: 259-269.
Hargrave, L. L.
1938 A plea formore carefulpreservationof all biologicalmaterial fromprehistoric
sites.Southwestern Lore 4(3): 47-51.
Harshberger,J.W.
1896 The purposesof ethnobotany. AmericanAntiquarian 17: 73-81.
Hastorf,C., andM. DeNiro
1985 Reconstructionof prehistoricplant productionand cooking practicesby a new
isotopicmethod.Nature 315: 489-491.
Helbaek, H.
1950 Tollundmandens sidstemaaltid. Aarboger f.Nordisk oldkyndighedog historie
329-341.
1958 Grauballemandenssidstemaaltid. Kuml aarbog f. jyskArk. Selskab. 83-116.
Heer, 0.
1866 Treatise on theplants of theLake Dwellings. InLake Dwellings of Switzerland
and otherparts of Europe, edited by F. Keller, trans.by J. E. Lee, London:
Longman Green.
Hendry,G. W., andM. P. Kelley
1925 The plant content of adobe bricks. California Historical Society Quarterly
4:361-373.
Hillman, G.
1981 Reconstructingcrop husbandrypractices from charred remainsof crops. In
Farmingpractice inBritishprehistory,edited by R. Mercer. Edinburgh: Edin
burghUniversityPress. Pp. 123-162.
1984 Interpretation of archaeological plant remains: the application of ethnographic
models fromTurkey. InPlants and ancientman, editedbyW. Van Zeist andW.
A. Casparie. Boston: A. A. Balkema. Pp. 1-41.
Hillman,G., G. Robins, D. Oduwole, K. Sales, and D. McNeil
1983 Determinationof thermalhistoriesof archaeological seeds with electron spin
resonancespectroscopy.Science 222: 1235.
Hopkins,M. S., and A. W. Graham
1983 The speciescompositionof soil seedbanks beneath lowland tropicalrainforestsin
northQueensland, Australia. Biotropica 15(2): 90-99.
Hubbard, R. N. L. B.
1976 On thestrength of theevidenceforprehistoriccrop processingactivities.Journal
ofArchaeological Science 3: 257-265.
1980 Development of agricultureinEurope and theNear East: evidence fromquan
titativestudiesEconomic Botany 34(1): 51-67.
Iversen,J.
1941 Land occupation inDenmark's stoneage.Denmarks Geologiske Undersogelse
2(66): 1-67.
Johannessen,S.
1984 Paleoethnobotany. InAmerican bottomarchaeology,editedby C. J.Bareis and
J.W. Porter.Chicago: Universityof IllinoisPress. Pp. 197-214.
Jones,G.
1984 Interpretationof archaeological plant remains: ethnographicmodels from
Greece. In Plants and ancient man, edited by W. van Zeist and W. A. Casparie.
Boston: A. A. Balkema. Pp. 43-69.
Jones,G., K. Wardle, P. Halstead, and D. Wardle
1986 Crop storageat Assiros. ScientificAmerican 254(3): 96-103.
Kaplan, L. and S. L. Maina
1977 Archaeological botany of the Apple Creek Site, Illinois. Journal of Seed
Technology2(2): 40-53.
Keepax, C.
1975 Scanningelectronmicroscopyofwood replacedby ironcorrosionproducts.Jour
nal ofArchaeological Science 2:145-150.
1977 Contamination of archaeological deposits by seeds of modern originwith par
ticularreferenceto the use of flotationmachinery. Journalof Archaeological
Science. 4:221-229.
King, J.E., W. E. Klippel, and R. Duffield
1975 Pollen preservationand archaeology ineasternNorth America. AmericanAnti
quity40(2): 180-190.
Kirmiz, J. P.
1962 Adaptation to desertenvironment.London: Butterworth.
Knorzer,K. H.
1984 The prospectsof thepalaeoethnobotanicalexaminationof cesspits. InPlants and
ancientman, editedbyW. van Zeist andW. A. Casparie. Boston: A. A. Balkema.
Pp. 331-338.
Kuhn, T. S.
1970 The structure

of scientificrevolutions.Chicago: Universityof Chicago Press.
Kunth,C.
1826 Examen botanique. InCatalogue raisonneet historiquede antiquitksdkcouverte
en Egypte, editedby J. Passalaqua. Paris:Mus6es Nationaux.
Leroi-Gourhan,A.
1975 The flowersfoundwith Shanidar IV, a Neanderthalburial in Iraq. Science 190:
562-564.
Lewis,H. T.
1972 The roleof firein thedomesticationof plantsand animals inSouthwestAsia: a
hypothesis.Man 7:195-222.
Lopinot,N. H.
1985 Archaeobotanical formationprocesses and lateMiddle Archaic subsistence.
Ann Arbor: University Microfilms International, 85-10043.
Lopinot,N. H., and D. E. Brussell
1982 Assessing uncarbonizedseeds fromopen-airsitesinmesic environments:an ex
ample fromsouthernIllinois.JournalofArchaeological Science 9: 95-108.
Lundstrom-Baudais,K.
1984 Palaeo-ethnobotanicalinvestigation of plant remainsfroma Neolithic lakeshore
site inFrance: Clairvaux, Station III. InPlants and ancientman, edited byW.
van Zeist andW. A. Casparie. Boston: A. A. Balkema. Pp. 293-306.
MacNeish, R. S.
1967 A summaryof subsistence.In The prehistory of theTehuacan Valley,Vol. 1,
Environmentand subsistence,editedbyD. S. Byers.Austin:Universityof Texas
Press. Pp. 290-309.
Matson, F.
1955 Charcoal concentrationfromearlysitesforradiocarbondating.AmericanAnti
quity 21: 162-169.
Matthews,M., and B. Benz
1983 Developing an ecological-behavioral model for the interpretationof pioneer
plant remains in archaeological sites. Paper presentedbefore the Society for
AmericanArchaeology,Pittsburgh,PA.
Meyer, F. C.
1980 Carbonized food plants of Pompeii,Herculaneum,and thevilla at Torre An
nunziata.Economic Botany 34(4): 401437.
Miksicek, C. H.
1978 Basketmaker III-Pueblo I botanical remainsfromLupton, Arizona. In The
painted cliffsrestarea, editedbyA. Ferg.Arizona StateMuseum Contribution
toHighway SalvageArcaheology inArizona (50):143-153.
1983a Paleoecology and subsistenceat PulltrouserSwamp: theview from the float
tank. Report to the National Science Foundation on the 1981 Pulltrouser
Swamp Project, edited by P. D. Harrison and B. L. Turner, II.
1983b at Nolic, Papago
Plant remains.InArchaeological and historicalinvestigations
Indian Reservation, Arizona, edited by W. Harper Doelle. Institutefor
AmericanResearchAnthropologicalPapers (2): 147-156.
1983c Archaeobotanical aspects of Las Fosas: a statisticalapproach to prehistoric
plant remains,InHohokam archaeologyalong theSalt-Gila aqueduct,Central
Arizona Project, Vol. VI: Habitation siteson theGila River, edited by L. S.
Teague and P. L. Crown. Arizona StateMuseum Archaeological Series (150):
671-700.
1984a Historic desertification,prehistoricvegetationchange, and Hohokam sub
sistence in the Salt-Gila Basin. In Hohokam archaeology along the Salt-Gila
aqueduct, CentralArizopa Project, Vol. VII: Environmentand subsistence,
editedbyL. S. Teague and P. L. Crown. Arizona StateMuseum Archaeological

Series (150). Pp. 53-80.
1984b Hohokam adaptation to thechaparralzone: paleoecology and subsistencein the
Anamax-Rosemontarea of southern Arizona. Manuscript on file,Arizona State
Museum Archives.
1984c Iron age enviromment and diet.Muse 18: 51-53.
1985 Palaeoethnobotanyof a 14thcentury Hopi site:Homolovi II. InExcavation and
surfacecollectionofHomolovi II ruin,editedbyK. Hays and E. C. Adams, Ap
pendix II. Tucson: Archaeology Section,Arizona StateMuseum.
1986a Plant remainsfromtheTanque VerdeWash site. InArchaeological investiga
tionsat theTanque VerdeWash Site, editedbyM. Elson. InstituteforAmerican
ResearchAnthropologicalPapers Tucson 7: 371-394.
1986b Colonial and early classic plant remains from theDakota Wash site: AZ
AA:16:49. Manuscript on file,Pima CommunityCollege, Tucson.
Miksicek, C. H. and S. K. Fish
1981 Understandingprehistoricsubsistence activities in southernArizona. Paper
presentedto theFourthEthnobiologyConference,Columbia, Missouri.
Miller, N. F., and T. L. Smart
1984 Intentionalburningof dungas fuel:a mechanismfortheincorporation of charred
seeds into thearchaeological record.Journalof Ethnobiology 3(1): 15-28.
Mills, W. C.
1901 Plant remainsfromtheBaum Village site.Ohio Naturalist 1(5): 70-71.
Minnis, P. E.
1981 Seeds inarchaeologicalsites:sourcesand some interpretive
problems.American
Antiquity 46(1): 143-152.
Minnis, P. E., and S. LeBlanc
1976 An efficient,inexpensivearid lands flotationsystem.AmericanAntiquity 41:
491-493.
Montgomery,R. G., W. Smith, and J.0. Brew
1949 FranciscanAwatovi. Papers of thePeabodyMuseum ofAmericanArchaeology
and Ethnology, Harvard University, Vol. 26. Reports of the Awatovi Expedi
tion,PeabodyMuseum, Harvard UniversityReport (3).
Munson, P. J.
1981 Note on theuse andmisuse ofwater separation("flotation") fortherecoveryof
small scale botanical remains.Midcontinental Journal of Archaeology 6(1):
123-126.
Munson, P. J., P. W. Parmalee, and R. A. Yarnell
1971 Subsistenceecologyof Scovill, a terminal
Middle Woodland village.American
Antiquity 36(4): 410-431.
Nabhan, G. P.
1977 Viable seeds from prehistoric caches? Archaeobotanical remains in
Southwesternfolklore.TheKiva 43(2): 143-159.
Osborne, D.
1965 Contributionsof theWetherillMesa archaeologicalproject.AmericanAntiqui
ty31(2), Part (2).
Pals, J. P.
1984 Plant remainsfromAartswoud, a Neolithic settlementin the coastal area. In
Plants and ancient Man, edited by W. van Zeist and W. A. Casparie. Boston: A.
A. Balkema. Pp. 313-322.
Pearsall,D. M.
1978 Phytolithanalysis of archaeological soils: evidence formaize cultivation in
formativeEcuador. Science 199: 177-178.
1983 Evaluating the stabilityof subsistencestrategies

by use of paleoethnobotanical
data. Journalof Ethnobiology3(2): 121-137.
Pendleton,M. W.
1983 A commentconcerning"testingflotationrecoveryrates." American Antiquity
48(3): 615-616.
Pozorski, S.
1983 Changing subsistenceprioritiesand earlysettlementpatternsof thenorthcoast of
Peru. Journalof Ethnobiology3(1): 15-38.
Pulliam,H. R., andM. RileyBrand
1975 The production and utilizationof seeds in plains grassland of southeastern
Arizona. Ecology 56(4): 1158-1166.
Rathje,W. L., andM. B. Schiffer
1982 Archaeology.New York: Harcourt, Brace.
Renfrew,J.M.
1973 Palaeoethnobotany.New York: Columbia UniversityPress.
Reynolds,P. J.
1979 Iron-agefarm: theButser experiment. London: BritishMuseum Publications.
Rovner, I.
1983 Plant opal phytolithanalysis:major advances in archaeobotanical research. In
Advances inarchaeologicalmethod and theory Vol. 6, editedbyM. B. Schiffer.
New York: Academic Press. Pp. 225-266.
Rudenko, S. I.
1970 Frozen tombsof Siberia. Berkeley:UniversityofCalifornia Press.
Sauer, C. 0.
1952 Agricultural origins and dispersals.New York: The American Geographical
Society.
Schiffer,M. B.
1976 Behavioral archaeology.New York: Academic Press.
1983 Toward the identification of formationprocesses.American Antiquity 48(4):
675-706.
Schmidt-Nielson, K.
1964 Desert animals: physiologicalproblemsof heat and water. Oxford: University
Press.
Smith,C. E., Jr.
1967 Plant remains. In The prehistoryof theTehuacan Valley,Vol. 1,Environment
and subsistence,edited byD. S. Byers.Austin:Universityof Texas Press. Pp.
220-255.
1985 Recovery and processingof botanical remains.In The analysis of prehistoric
diets,editedbyR. Gilbertand J.Mielke. New York:Academic Press. Pp. 97-126.
Struever,S.
1968 Flotation techniques for the recoveryof small-scalearchaeological remains.
AmericanAntiquity 33: 353-362.
Sullivan,A. P.
1982 Mogollon agrarianecology. TheKiva 48(1-2): 1-15.
Taylor,W. W.
1957 The identificatipn
of non-artifactual materials.National Academy
archaeological
of Sciences,National Research Council, Publication 565.
Tevis, L., Jr.
1958 between theharvesterant Veromessorpergandei (Mayr) and some
Interrelations
desertephemerals.Ecology 39(4): 695-704.
Uhl, C., H. Clark, and K. Clark
1982 Successional patternsassociatedwith slash and burnagriculturein theupperRio

Negro regionof theAmazon Basin. Biotropica 14(4): 249-254.
van derVeen, M.
1984 Sampling forseeds. InPlants and ancientMan, editedbyW. van Zeist andW. A.
Casparie. Boston: A. A. Balkema. Pp. 193-200.
van Zeist,W., and S. Bottema
1983 Palaeobotanical studiesof Carthage. CEDAC Carthage,Bulletin 5.
Wagner, G. E.
1982 Testing flotationrecoveryrates.AmericanAntiquity 47(1): 127-132.
Watson, P. J.
1969 The prehistoryof Salts Cave, Kentucky. Illinois StateMuseum Reports of In
vestigations(16).
1974 Archaeology of theMammoth Cave area. New York: Academic Press.
1976 In pursuitof prehistoricsubsistence:a comparativeaccount of some contem
porary techniques.Mid-Continental Journalof Archaeology. 1(1): 77-100.
Wendorf, F., R. Schild,N. Hadidi, A. Close, M. Kobusiewicz,H. Wieckowska, B. Issawi,and
H. Haas
1979 Use of barley in theEgyptianLate Paleolithic. Science 205: 1341-1347.
Wendorf, F., R. Schild, A. Close, D. Donahue, A Jull,T. Zabel, H. Wieckowska, M.
Kobusiewicz, B. Issawi, and N. Hadidi
1984 New radiocarbondates on thecerealsofWadi Kubbaniya. Science 225: 645-646.
Willcox, G. H.
1977 Exotic plants from Roman waterlogged sites in London: Journal of Ar
chaeological Science. 4:269-282.
Wilson, D. G.
1984 The carbonisation of weed seeds and their representationin macrofossil
assemblages. In Plants and ancientman, edited byW. van Zeist and W. A.
Casparie. Boston: A. A. Balkema. Pp. 201-206.
Wittmack, L.
1888 Die Nutzpflanzen der alten Peruanet. Congris Internationalde Americanistes,
Berlin 7: 325-349.
1905 Our presentknowledgeof ancientplants. Transactionsof theAcademyof Science
of St. Louis XV (1).
Wood, W. R., and D. L. Johnson
1978 A surveyof disturbanceprocesses in archaeological site formation.InAdvances
inarchaeologicalmethod and theoryVol. 1, editedbyM. B. Schiffer.New York:
Academic Press. Pp. 315-381.
Yarnell, R. A.
1969 Contentsof humanpaleofeces. In The prehistoryof Salts Cave, Kentucky, edited
by P. J.Watson. IllinoisStateMuseum Report of Investigations16: 41-54.
1974 Plant foodcultivationof theSaltsCavers. InArchaeologyof the Mammoth Cave
area, editedby P. J.Watson. New York: Academic Press. Pp. 113-122.
1982 Problems of interpretationof archaeological plant remains of the Eastern
Woodlands. SoutheasternArchaeology 1(1):1-7.
Zier, C. J.
1980 A Classic-periodMaya agriculturalfield inwesternEl Salvador. Journalof Field
Archaeology 7: 65-74.

Miksicek 1987

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

Miksicek 1987

Загружено:

Авторское право:

Доступные форматы

Formation Processes of the Archaeobotanical Record

Author(s): Charles H. Miksicek

You may need to log in to JSTOR to access the linked references.

Formation Processes of the

Archaeobotany is the art and science of recovering, identifying, and in

(1896) analysis of plant remains preserved in dry rock shelters in

strategiesand analytical techniquesemployedby the paleoethnobotanist

Firstof all it is importantto consider thenatural conditionsthat favor

26++ C +++ ++++ o

* 0~~~~~o oO to. co D C 0 =it

.4-. NOO W~-Y c c

The more generalizedclass of waterloggedsitessharesmany featuresin

Frozen, acidic, waterlogged, permanentlydry, or chemicallyunique

whole sitesdestroyedby fires.In theserarecircumstances, com

into a hearth.Althoughmedicinal herbs are not technicallyfoods, they

a morning'swork, of which only0.02% was carbonized.The problemof

preparedby soakingand boiling,are rarelyevercharred.Gasser andAdams

Comparison of Coprolites and Flotation Samples

study,squash and beans, preparedby boiling,aregrosslyunderrepresented

The effectsof processing

Samples containing plant material (%)

Plantb Coprolites (N = 87) Bulk samples (N = 51)

Overrepresented Inbulk samples

Underrepresented Inbulk samples

"Data fromCallen 1967; Smith 1967.

1976;Rathje and Schiffer1982).The dcarred saguaro seeds recoveredfrom

Figure4.1. Three possibleuses of an hypothetical

predateit.Perhaps a fewpollen grainsor phytolithsfromfoodor floralof

Castastrophically burned structures 287.55 36.4

afterabandonmentyieldeda much lowerdensityof charredremains(4.06

The "SchleppEffect" (discussedinRathje and Schiffer1982)was derived

more maize and gathered

Once charredbotanical remainsare deposited in an archaeologicalsite,

(1978).A second relationshipisevidentinTable 4.2. Sites inmore mesic en

surfaceeach year (Wood and Johnson1978). Sinceburrowinganimals seem

Aeollan and Alluvial Processes

The action of wind (aeroturbation)or water (aquatubation)may alter

Background Seed Rain and Fires

forprehistoricfarmersin theAmericanSouthwest.Sullivan (1982)propos

Evaluating theOrigins of Seeds InSites

1. What is the state of preservation of the seed? If it is from an open site

4. Even if theseeds are carbonized, is thereevidence forancientanimal

ANALYTICALTRANSFORMATIONSOF THE DATA BASE

Cumulative number of samples

FIgure4.3. A "species-area curve" fordeterminingtheadequacy of sampling.

Larger samplesproducemore plant remains,but theyaremore difficultto

techniquethat is commonlyused to testtheadequacy of a sample inquan

Pendelton (1983) have recommended usingvarious typesof tracersto com

in independentsamples. Itwould seem useful to borrow another statistic

In the foregoingdiscussion I have triedto outline someof theprinciples

some smallway itmay contributeto an understandingof site formation

The seedsofmany of the ideas discussed in thispaperwere planted almost 15yearsago in

Ford, R. I., and N. Miller

1970 The structure

editedbyL. S. Teague and P. L. Crown. Arizona StateMuseum Archaeological

1983 Evaluating the stabilityof subsistencestrategies

1982 Successional patternsassociatedwith slash and burnagriculturein theupperRio

Вам также может понравиться