Indicator and Methods For The Ecological Status Assessment Under The WFD Linkages Between Chemical and Biological Quality of Surface Waters

Institute for
Environment and
Sustainability
EUROPEAN COMMISSION
Joint Research Centre
DIRECTORATE-GENERAL
Indicators and methods for the
ecological status assessment
under the Water Framework Directive
2006 EUR 22314 EN
Publications Office
The mission of the JRC is to provide customer-driven scientic and technical support for the conception,
development, implementation and monitoring of EU policies. As a service of the European Commission,
the JRC functions as a reference centre of science and technology for the Union. Close to the policy-making
process, it serves the common interest of the Member States, while being independent of special interests,
whether private or national.
L
B
-
N
A
-
2
2
3
1
4
-
E
N
-
C
I
n
d
i
c
a
t
o
r
s

a
n
d

m
e
t
h
o
d
s

f
o
r

t
h
e

e
c
o
l
o
g
i
c
a
l

s
t
a
t
u
s

a
s
s
e
s
s
m
e
n
t

u
n
d
e
r

t
h
e

W
a
t
e
r

F
r
a
m
e
w
o
r
k

D
i
r
e
c
t
i
v
e
L
i
n
k
a
g
e
s

b
e
t
w
e
e
n

c
h
e
m
i
c
a
l

a
n
d

b
i
o
l
o
g
i
c
a
l

q
u
a
l
i
t
y

o
f

s
u
r
f
a
c
e

w
a
t
e
r
s
Linkages between chemical and biological quality of surface waters
Edited by
Angelo G. Solimini, Ana Cristina Cardoso and Anna-Stiina Heiskanen
EUROPEAN COMMISSION
DIRECTORATE-GENERAL
Cover photo by David Boorman (Lake Kukkia, Finland)
Institute for
Environment and
Sustainability
EUROPEAN COMMISSION
DIRECTORATE-GENERAL
Indicators and methods for the
ecological status assessment
under the Water Framework Directive
2006 EUR 22314 EN
Linkages between chemical and biological quality of surface waters
Edited by
Angelo G. Solimini, Ana Cristina Cardoso and Anna-Stiina Heiskanen
Institute for Environment and Sustainability
The mission of the Institute for Environment and Sustainability kmbkbkmnbIEShihikis to provide
scientic and technical support to the European Unions policies for protecting the environment
and the EU Strategy for Sustainable Development.
European Commission
Directorate-General Joint Research Centre
Institute for Environment and Sustainability
I-21020 Ispra (VA)
Contact information
Address: Anna-Stiina Heiskanen,
Rural, Water and Ecosystem Resources Unit, TP 290
E-mail: anna-stiina.heiskanen@ec.europa.eu
Tel.: +39-0332 785969
Fax: +39-0332-789352
http://ies.jrc.cec.eu.int
http://www.jrc.cec.eu.int
Legal Notice
Neither the European Commission nor any person acting on behalf of the Commission is
responsible for the use which might be made of this publication.
A great deal of additional information on the European Union is available on the Internet.
It can be accessed through the Europa server
http://europa.eu
EUR 22314EN
Luxembourg: Ofce for Ofcial Publications of the European Communities
ISBN 92-79-02646-1
ISSN 1018-5593
European Communities, 2006
Reproduction is authorised provided the source is acknowledged
Printed in Italy
The Water Framework Directive (WFD; 2000/60/EC) creates a new legislative
framework to manage, use, protect, and restore surface water and groundwater
resources within river basins and transitional (lagoons and estuaries) and coastal
waters in the European Union. The WFD aims to achieve sustainable management
of water resources, to reach good ecological quality and prevent further deterioration
of surface waters and groundwater, and to ensure sustainable functioning of
aquatic ecosystems by 2015. In the WFD the ecological status of surface water is
dened as an expression of the quality of the structure and functioning of aquatic
ecosystems associated with surface waters, classied in accordance with Annex V
(Article 2:21). This implies that the classication systems of the ecological status
of water bodies should reect changes in the structure of the biological communities
and in the overall ecosystem functioning as a response to anthropogenic pressures.
Such requirements are a novel development in European water policy, which has
been mostly based on the regulation of emissions at the source through the
establishment of emission limit values.
The WFD necessitates that the ecological quality status of surface water bodies is
assessed using biological quality elements such as phytoplankton, sh, and benthic
ora and fauna. Therefore, there is a need to identify biological indicators that
respond in a predictable manner to human disturbances, and allow classication
of ecological quality based on functional relationships between pressures and
indicators. Currently, intensive research is being carried out in Europe to develop
biological indicators and classication metrics that can be used in assessment,
comparison and harmonization of ecological water quality targets across Europe.
This work is naturally based on the strong research tradition and knowledge on the
ecological functioning of surface waters in Europe. Still the WFD has imposed
challenges for aquatic ecologists to develop scientically sound, high-precision,
and practical biological assessment tools for managers.
This book provides an overview on current knowledge and methods on the
relationship between chemical and biological indicators for the major pressures for
rivers, lakes and coastal waters. It is aimed at scientists working in environmental
agencies, managers working on WFD implementation, university lecturers, students
of applied ecology, water management and those interested in limnology and
coastal management. Each section is dedicated to one of the major pressure types:
eutrophication, organic loading, acidication, toxic, and combined pressures which
Preface
are currently most widely impacting surface waters in Europe. An overview on the
known relationships between the pressure type in question and the biological
responses is provided for different surface water categories (rivers, lakes, coastal
and transitional waters). We have compiled information on the applicability of the
current existing biological classication tools and metrics, as well as the empirical
and deterministic models which include biological quality elements as parameters,
and could be used for WFD classication and management purposes. Finally, we
also provide information on some new emerging approaches and methodologies
that could be applied in the future.
This book is a deliverable of the EU FP6 REBECCA
1
(Relationships between
ecological and chemical status of surface waters) project. The general objective of
the REBECCA project is to provide scientic support for the implementation of the
WFD. The two specic aims of the project are, rstly, to establish links between
ecological status of surface waters and physico-chemical quality elements and
pressures from different sources, and, secondly, to develop and validate tools that
member states can use in the process of classication, in the design of their
monitoring programs, and in the design of measures in accordance with the
requirements of the WFD.
We thank the Advisory Board of the REBECCA project (Ursula Schmedtje, Roger
Owen, Ivan Karottki, Joachim DEugenio, and Jorge Rodriguez-Romero) for constructive
comments and guidance on the contents and structure of this review. We are grateful
for a number of anonymous reviewers that provided suggestions for improvements.
We also wish to thank Gary Free for revising the language and providing comments
on a number of contributions. The opinions expressed in this book are those of the
authors, who are responsible for the contents of each chapter.
July 11, 2005, Ispra, Italy The editors,
Angelo G. Solimini,
Ana Cristina Cardoso,
Anna-Stiina Heiskanen
1
REBECCA project is jointly funded by the EC 6th Framework programme as a Specic Targeted
Research or Innovation Project (Contract number SSPI-CT-2003-502158) and the research programmes
of the collaborating organizations. More information about REBECCA is available through the project
web-site at http://www.environment./syke/rebecca.
IV | PREFACE
Contributors ............................................................................................................................. VII
Eutrophication and organic pollution
1. Nutrients and eutrophication in lakes ..................................................................... 3
Laurence Carvalho, Liisa Lepist, Johanna Rissanen, Olli-Pekka Pietilinen,
Seppo Rekolainen, Liliana Torok, Anne Lyche Solheim, Tuomo Saloranta,
Robert Ptacnik, Gianni Tartari, Ana Cristina Cardoso, Guido Premazzi, Iain Gunn,
Ellis Penning, Jenica Hanganu, Seppo Hellsten, Ibram Orhan, Ion Navodaru
2. Nutrients and eutrophication in coastal and transitional waters ................. 33
Pirjo Kuuppo, Anouk Blauw, Flemming Mhlenberg, Hanne Kaas,
Peter Henriksen, Dorte Krause-Jensen, Gunni rtebjerg, Saara Bck,
Paul Erftemeijer, Miguel Gaspar, Susana Carvalho, Anna-Stiina Heiskanen
3. Nutrients and eutrophication in rivers .................................................................... 81
Amelie Deandre and Helen Jarvie
4. Organic pollution in rivers ........................................................................................... 99
Maurizio Pettine, Antonio Ruggiero, Stefano Fazi, Andrea Buffagni,
Jens Mller Andersen, Peter Roncak and Nikolai Friberg
Acidication
5. Acidication in rivers and lakes ................................................................................ 117
Torleif Bkken, Frode Kroglund, Eli-Anne Lindstrm, and Laurence Carvalho
Hydromorphological pressures
6. Hydromorophological pressures in lakes ................................................................ 135
Seppo Hellsten, Bernard Dudley
7. Hydromorophological pressures in rivers ............................................................... 141
Bente Clausen, Michael Dunbar, Nikolai Friberg and Annette Baattrup-Pedersen
Toxic pressures
8. Impacts of selected hazardous and priority substances (WFD)
on freshwater and marine biota ................................................................................ 151
Lena Blom and Eva Brorstrm-Lundn
Contents
9. Applicability of QSAR models (Quantitative Structure Activity
Relationships) for assessing impact of chemical compounds ......................... 167
Erik Furusj
10. Advanced Molecular-Based Techniques for Environmental Toxicology ........ 193
Teresa Lettieri
Combined pressures
11. Large scale evaluations of pressures ........................................................................ 213
Ana Garcia, Bertrand Villeneuve, and Jean-Gabriel Wasson
12. Responses of aquatic biota to combined pressures from nutrients
and priority substances ................................................................................................ 225
Constanze OToole and Kenneth Irvine
VI | CONTENTS
Contributors
Andersen, Jens Mller. National Environmental
Research Institute (NERI), Vejlsvej 25 DK 8600
Silkeborg, Denmark
rtebjerg, Gunni. National Environmental Research
Institute (NERI), Vejlsvej 25 DK 8600 Silkeborg,
Denmark
Bck, Saara. Finnish Environment Institute (SYKE),
P.O. Box 140, FIN-00251 Helsinki, Finland
Bkken, Torleif. Norwegian Institute for Water
Research (NIVA), Brekkeveien P.O.Box 173, N-0411
Oslo, Norway
Baattrup-Pedersen, Annette. National Environmental
Silkeborg, Denmark
Blauw, Anouk. WL Delft Hydraulics (WL Delft), P.O.
Box 177 NL-2600 MH Delft, The Netherlands
Blom, Lena. Swedish Environmental Research
Institute (IVL), Box 47086, SE-402 58, Gothenburg,
Sweden
Brorstrm-Lundn, Eva. Swedish Environmental
Research Institute (IVL), Box 47086, SE-402 58
Gothenburg, Sweden
Buffagni, Andrea. CNR Istituto di Ricerca Sulle
Acque, Sezione Territoriale di Brugherio, Via della
Mornera, 25, 20047 Brugherio (MI) Italy
Clausen, Bente. National Environmental Research
Denmark
Cardoso, Ana Cristina. European Commission Joint
Research Centre (JRC), Institute for Environment
and Sustainability, TP290, 21020 Ispra, Italy
Carvalho, Laurence. Centre for Ecology and
Hydrology (CEH), Bush Estate, Penicuik EH26 0QB,
Scotland, UK
Carvalho, Susana. Instituto Nacional de
Investigao Agrria e das Pescas (IPIMAR), Avenida
de Brasilia 1449-006, Lisboa, Portugal.
Deandre, Amelie. Centre for Ecology and Hydrology
(CEH), Wallingford, Oxon, OX10 8BB, UK
Dudley, Bernard. Centre for Ecology and Hydrology
(CEH), Bush Estate, Penicuik EH26 0QB, Scotland,
UK
Dunbar, Michael J. Centre for Ecology and Hydrology
Erftemeijer, Paul. WL Delft Hydraulics (WL Delft),
P.O. Box 177 NL-2600 MH Delft, The Netherlands
Fazi, Stefano. Water Research Institute (IRSA), 1,
via Reno 00198 Rome, Italy
Friberg, Nikolai. National Environmental Research
Denmark
Furusj, Erik. Swedish Environmental Research
Institute (IVL), Box 21060, SE-100 31 Stockholm,
Sweden
Garcia, Ana. Centre National du Machinisme
Agricole, du Gnie Rural, des Eaux et des Fo
-
rets
(CEMAGREF), Groupement de Lyon, UR Biologie
des Ecosystemes Aquatiques, Laboratoire
dHydroecologie Quantitative, 3 bis quai Chauveau,
BP 220, 69336 Lyon Cedex 09, France
Gaspar, Miguel. Instituto Nacional de Investigao
Agrria e das Pescas (IPIMAR), Avenida de Brasilia
1449-006, Lisboa, Portugal
Gunn, Iain. Centre for Ecology and Hydrology (CEH),
Bush Estate, Penicuik EH26 0QB, Scotland, UK
Hanganu, Jenica. Danube Delta National Institute
(DDNI), Babadag Street, 165 Tulcea, 8800, Romania
Heiskanen, Anna-Stiina. European Commission
Joint Research Centre (JRC), Institute for Environment
Hellsten, Seppo. Finnish Environment Institute
(SYKE), P.O. Box 140, FIN-00251 Helsinki, Finland
Henriksen, Peter. National Environmental Research
Denmark
Irvine, Kenneth. Trinity College Dublin (TCD),
Department of Zoology, Trinity College, Dublin 2,
Ireland
Contributors
Jarvie, Helen. Centre for Ecology and Hydrology
Kaas, Hanne. DHI Water & Environment (DHI),
Agern All 5, DK-2970 Hrsholm, Denmark
Krause-Jensen, Dorte. National Environmental
Silkeborg, Denmark
Kroglund, Frode. Norwegian Institute for Water
Oslo, Norway
Kuuppo, Pirjo. Finnish Environment Institute
Lepist, Liisa. Finnish Environment Institute (SYKE),
P.O. Box 140, FIN-00251 Helsinki, Finland
Lettieri, Teresa. European Commission Joint
Lindstrm, Eli-Anne. Norwegian Institute for Water
Research (NIVA), Brekkeveien P.O. Box 173, N-0411
Oslo, Norway
Mhlenberg, Flemming. DHI Water & Environment
(DHI), Agern All 5, DK-2970 Hrsholm, Denmark
Navodaru, Ion. Danube Delta National Institute
OToole, Constanze. Trinity College Dublin (TCD),
Department of Zoology, Trinity College, Dublin 2,
Ireland
Orhan, Ibram. Danube Delta National Institute
Penning, Ellis. WL Delft Hydraulics (WL Delft), P.O.
Box 177 NL-2600 MH Delft, The Netherlands
Pettine, Maurizio. Water Research Institute (IRSA),
1, via Reno 00198 Rome, Italy
Pietilinen, Olli-Pekka. Finnish Environment Institute
Premazzi, Guido. European Commission Joint
Ptacnik, Robert. Norwegian Institute for Water
Research (NIVA), Brekkeveien P.O. Box 173, N-0411
Oslo, Norway
Rekolainen, Seppo. Finnish Environment Institute
Rissanen, Johanna. Finnish Environment Institute
Roncak, Peter. Slovak Hydrometeorological Institute
(SHMU), Jesniova 17, 833 15 Bratislava, Slovak
Republic
Ruggiero, Antonio. Water Research Institute (IRSA),
1, via Reno 00198 Rome, Italy
Saloranta, Tuomo. Norwegian Institute for Water
Oslo, Norway
Solheim, Anne Lyche. Norwegian Institute for
Water Research (NIVA), Brekkeveien P.O.Box 173, N-
0411 Oslo, Norway
Solimini, Angelo Giuseppe. European Commission
Joint Research Centre (JRC), Institute for
Environment and Sustainability, TP290, 21020
Ispra, Italy
Tartari, Gianni. CNR Istituto di Ricerca Sulle Acque,
Sezione Territoriale di Brugherio, Via della Mornera,
25, 20047 Brugherio (MI), Italy
Torok, Liliana. Danube Delta National Institute
(DDNI), Babadag Street, 165 Tulcea, 8800,
Romania
Villeneuve, Bertrand. Centre National du
Machinisme Agricole, du Gnie Rural, des Eaux et
des Fo
-
rets (CEMAGREF), Groupement de Lyon, UR
Biologie des Ecosystemes Aquatiques, Laboratoire
Wasson, Jean-Gabriel. Centre National du
Machinisme Agricole, du Gnie Rural, des Eaux et
des Fo
-
rets (CEMAGREF), Groupement de Lyon, UR
Biologie des Ecosystemes Aquatiques, Laboratoire
VIII | CONTRIBUTORS
Eutrophication and
organic pollution
1 - Gary Free. Batrachospermum from Lough Gill (Ireland)
2 - Michela Ghiani. Sampling in the Mediterranean Sea
3 - Laurence Carvalho. Algal bloom in Rostherne Mare (England)
4 - Bruno Rossano. Chironomus larvae (lake Varese, Italy)
5 - Pirkko Kokkonen. Spring bloom phytoplankton species from the Northern Baltic Sea
1 2 3 4 5
1 Nutrients and eutrophication in lakes
Laurence Carvalho, Liisa Lepist, Johanna Rissanen, Olli-Pekka Pietilinen, Seppo Rekolainen,
Liliana Torok, Anne Lyche Solheim, Tuomo Saloranta, Robert Ptacnik, Gianni Tartari,
Ana Cristina Cardoso, Guido Premazzi, Iain Gunn, Ellis Penning, Jenica Hanganu,
Seppo Hellsten, Ibram Orhan, Ion Navodaru
1.1 Introduction
Within Europe, eutrophication has had a high prole since the early 1980s when wide-spread
occurrence of blue-green algal blooms in standing and slow-owing freshwaters gave rise to
considerable interest and concern by the public, the media and by water industries. Of all surface
waters, the impact of nutrient enrichment on lakes and reservoirs has probably received the
greatest attention and is well understood in general qualitative terms. Nutrients tend to be a
secondary and indirect driver of community composition but are a signicant driver of productivity,
leading to increased growths of algae (phytoplankton and lamentous algae) and aquatic plants.
The increase in phytoplankton crops can lead to secondary problems, such as the loss of sensitive
macrophyte and sh species. These impacts, as well as the increasing frequency and intensity of
toxic cyanobacteria blooms, have become a widespread problem in European lakes and have
considerable consequent environmental, social and economic implications.
These general effects of increased nutrient loading, or eutrophication, have been incorporated
within the denition of eutrophication in recent European legislation - Urban Waste Water
Treatment Directive (UWWTD) and Nitrates Directive (ND) and agreements (Oslo - Paris Marine
Convention, OSPAR). These dene eutrophication as:
The enrichment of water by nutrients, especially compounds of phosphorus (P) and/or nitrogen
(N), causing an accelerated growth of algae and higher forms of plant life to produce an undesirable
disturbance to the balance of organisms present in the water and to the quality of the water
concerned
One problem with this denition of eutrophication is that it cannot easily be quantied.
Eutrophication is a process, not a state such as eutrophic, the latter being more readily dened
in terms of phytoplankton biomass (e.g. as a concentration of chlorophyll-a) or a concentration of
a nutrient such as P. In the WFD, this issue of how to assess the degree of change in freshwater
quality is tackled by relating (ecological) quality to a baseline or reference state under minimal
human inuence. The requirement within the WFD to dene nutrient conditions for high and
good ecological status requires a specic quantitative dose-response understanding of how
nutrient conditions relate to biological quality in individual lake ecotypes.
The impact of nutrient pressures in a particular lake ecotype or individual site can be dependent
on a number of sensitivity factors. Deep lakes are often more predictable in terms of nutrient
effects on phytoplankton abundance and the indirect impacts on profundal (deep-water)
invertebrate and sh communities. In shallow lakes, ecological thresholds in response to nutrient
conditions are much more difcult to dene due to the feedback effects of macrophytes and sh
NUTRIENTS AND EUTROPHICATION IN LAKES | 3
4 | LINKING ECOLOGICAL STATUS OF SURFACE WATERS WITH ANTHROPOGENIC PRESSURES
on lake community structure. In addition, nutrient enrichment is frequently accompanied by other
pressures and it is therefore often difcult to identify ecological impacts solely due to nutrient
conditions, let alone impacts due to particular nutrients (N or P).
This chapter considers both the known qualitative and quantitative relationships between nutrient
conditions and direct or indirect biological responses, in lake ecosystems. It considers effects on
phytoplankton, phytobenthos, macrophytes, benthic invertebrates and sh. The effects of nutrients
acting in combination with other pressures (hydromorphological, acidication and toxic) are
considered in chapters 5, 6 and 12.
1.2 Phytoplankton
Introduction
The phytoplankton community is widely considered the rst biological community to respond to
eutrophication pressures and is the most direct indicator of all the Biological Quality Elements
(BQEs) of nutrient concentrations in the water column. There are numerous socio-economic
problems associated with increases in phytoplankton abundance, particularly with increasing
frequencies and intensities of toxic cyanobacteria blooms. These include detrimental effects on
drinking water quality, ltration costs for water supply (industrial and domestic), water-based
activities and conservation status (e.g. sensitive pelagic sh species, such as coregonids). In some
contexts, however, increasing phytoplankton abundance can be considered as a positive feature,
for example, in increasing sheries productivity, and some phytoplankton taxa can be considered
indicative of high ecological status.
Annex V of the WFD outlines three features of the phytoplankton quality element that need to be
considered in the assessment of the ecological status of lakes and for which there is thus the need
to relate in quantitative terms to nutrient conditions, i.e. for which dose-response curves need to
be investigated. These three are:
Phytoplankton composition
Phytoplankton abundance and its effect on transparency conditions
Planktonic bloom frequency and intensity.
The WFD normative denition indicates that declining ecological quality is associated with
increasing phytoplankton abundance, greater proportions of cyanobacteria and more frequent
and intense phytoplankton blooms. These three features are considered further in the context of
their specic use as metrics of nutrient pressures.
Phytoplankton composition
Individual species or taxa can be considered as positive, negative or indifferent indicators in
relation to nutrient pressures. Negative indicators include species of green algae (e.g. Scenedesmus)
and diatoms (e.g. Stephanodiscus) and many groups of cyanobacteria, such as the large colonial
and lamentous genera Microcystis, Aphanizomenon and Anabaena. The latter are favoured by
Figure 1.1 Proportion of impact phytoplankton indicator Cyanobacteria as a function of total
phytoplankton biomass is shown for four different Norwegian lake types. Yellow lines and pink condence
bands are LOWESS-based trendlines for the single water type, while the trendline for the complete dataset
(not divided into lake type)s are shown in all graphs as red lines and yellow condence bands
(Lyche Solheim et al., 2004).
1
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
0.1 1 10 0.1 1 10
0.1 1 10 0.1 1 10
soft, clear hard, clear
soft, humic hard, humic
Total phytoplankton biomass (mg WW / L) Total phytoplankton biomass (mg WW / L)
F
r
a
c
t
i
o
n

c
y
a
n
o
*
F
r
a
c
t
i
o
n

c
y
a
n
o
*
relatively stable stratication and high alkalinity and can, therefore, form a signicant natural
component of the phytoplankton community in deep alkaline lakes. As nutrient concentrations
increase, however, the dominance and abundance of cyanobacteria, in particular, generally
increases often resulting in dense, mono-specic blooms during summer (Reynolds, 1984).
The proportion of cyanobacteria of the total phytoplankton biomass is a metric that is being
recommended in Norway with regard to WFD requirements and for which thresholds can be
generated (Figure 1.1; Lyche Solheim et al., 2004). Positive indicators include species of chrysophytes
(e.g. Dinobryon), desmids (e.g. Cosmarium) and some diatoms (e.g. Cyclotella comensis).
The ratio of positive to negative species can be used as a metric of ecological status and some Member
States (Denmark - Sondergaard et al., 2003; UK Carvalho et al., 2004) are developing metrics based
on the relative abundance of positive and negative phytoplankton functional groups (as outlined in
Reynolds et al., 2002), rather than taxonomic units. Both taxonomic (Figure 1.1) and functional group
(Figure 1.2) dose-response relationships need to be developed and validated further.
Phytoplankton abundance and its effect on transparency conditions
In general, as nutrient concentrations increase, phytoplankton abundance shows more frequent
and sustained peaks throughout summer, and consequently transparency declines. The abundance
of phytoplankton can be expressed in three ways:
Density of individual organisms (numbers of individual cells/laments/colonies per ml)
Volume of these cells as a fraction of the volume of water (biovolume in mm
3
m
-3
)
Chlorophyll-a concentration expressed as a concentration in water (g l
-1
).
All of these approaches have their pros and cons. For WFD monitoring purposes, many Member
States are choosing to use chlorophyll-a concentration, as it is a relatively robust and simple
measure widely used in all Member States (Sondergaard et al., 2003; Carvalho et al., 2004).
Transparency is widely used as an indirect or surrogate estimate of the amount of phytoplankton
or chlorophyll-a, and, therefore, as an indicator of eutrophication. The optical properties of lakes
are, however, not just controlled by the amount of phytoplankton, other factors such as dissolved
colour, suspended inorganic and organic particles may contribute signicantly in some lakes
(Tilzer, 1988).
Transparency is mentioned in Annex V of the WFD as a general physico-chemical factor supporting
the biological elements. In the normative denition of ecological status (Annex V, 1.2.2) it is
stated that, under high status conditions, the average phytoplankton biomass is consistent with
the type-specic physico-chemical conditions and is not such as to signicantly alter the type-
specic transparency conditions.
The transparency of water is commonly estimated using a Secchi disk, less commonly it is measured
using remote sensing methods. The Secchi disk measurement is obtained for a very restricted area
TP (g l
1
)
B
i
o
m
a
s
s

(
m
3

m
l
1
)
10000
1000
100
10
1
0.1
1 10 100 1000
Figure 1.2 Relationship between biomass of the impact phytoplankton functional group Lm
(c.f. Reynolds et al., 2002) and total phosphorus concentrations in Finnish lakes (courtesy by Lepist et al.,
unpublished).
(a lake or more specically discrete points in a lake) while remote sensing methods can cover a
larger area at once (up to hundreds of lakes). A Secchi disk is a circular, 20 cm diameter, black and
white disk. The Secchi depth is the average of the depths where the disk disappears when lowered
into water and reappears again when raised. The actual photosynthetic layer of water is
approximately three times the Secchi disk depth. The Secchi disk method is a very simple, useful
and cost-effective way to monitor and assess the status of surface waters. Furthermore, this
method is a useful tool for monitoring by citizens to enhance public participation in the WFD. The
use of satellite remote sensing is, however, a cost-effective method to assess the transparency of
surface waters (turbidity, chlorophyll-a) on a large scale.
Quantitative relationships have been developed relating total phosphorus (TP) concentrations
with phytoplankton biomass (chlorophyll-a) and water clarity (Secchi depth). The most widely
reported relationships are those developed by Dillon and Rigler (1974) and Vollenweider OECD
(OECD, 1982). The latter relationship (Figure 1.3) was developed for a set of, predominantly large,
temperate lakes, not ecotype specic. There is a great deal of scatter in all the published
relationships highlighting the fact that a number of sensitivity factors are involved, such as water
colour and ushing rate, of which the latter can be altered by hydromorphological pressures such
as ow regulation. There is, therefore, a clear need to develop more complex regression models
that incorporate these sensitivity factors. The importance of such sensitivity factors highlights the
need to incorporate lake typology factors (e.g. depth, altitude, colour) into models to understand
how nutrients are transformed into algal biomass, highlighting the value of lake-type specic
models as advocated by the WFD.
Planktonic bloom frequency and intensity
The term planktonic bloom refers to the phenomenon where phytoplankton populations greatly
increase in number over a limited period of time. Most concern focuses on periods when these
blooms are composed of toxic cyanobacteria that can form surface scums and accumulate on lake
100
10
1
0.1
1 10 100 1000
[P]
mg m
3
[
C
h
l
]

m
g

m
3
[Chl] 0.28 [P]
0.96
r = 0.88, SE = 0.251,
n = 77
}
excluded
from regression
Figure 1.3 Regression of annual mean phosphorus
concentration with phytoplankton biomass
(as annual mean chlorophyll-a; from OECD, 1982).
shores. The development of dense bloom populations of cyanobacteria is more predictable in
deep lakes, compared with shallow lakes, as bloom frequency and intensity is particularly affected
by physical factors such as stratication intensity and the depth of light penetration, however, no
quantied relationships have been described in the literature detailing how bloom frequency or
intensity is related to nutrient conditions. Furthermore, there is no agreed European quantitative
limit for dening simply when an algal bloom is present and different Member States use a variety
of thresholds based on measures of cyanobacteria abundance, chlorophyll-a concentrations or
amounts of toxins (e.g. microcystin). A more widely accepted denition of what constitutes a
phytoplankton bloom is needed before dose-response relationships can be developed.
Classication schemes
The most widely used classication scheme is that developed by the OECD (1982), which is
summarised in Table 1.1. Many other regression models or classication schemes have been
published which incorporate relationships between nutrients (exclusively phosphorus) and
phytoplankton (usually abundance) and transparency, summarised in Table 1.2.
Models
Models simulating the impact of nutrients on the phytoplankton community can be divided into
three main categories:
1. Empirical statistical models
2. Mechanistic/deterministic models
3. Theoretical food-web models.
Many empirical statistical models have been developed based on correlations and regressions of
observed data on nutrient and chlorophyll-a concentrations and Secchi disc depth. Only one
published model relates Secchi depth to phytoplankton biomass (Hillbricht-Illkowska, 1993).
Caneld and Hodgson (1983), besides chlorophyll and Secchi disk depth, included dissolved
organic matter concentration to account for water colour in their model.
Table 1.1 Lake classication scheme based on trophic status (OECD, 1982).
Annual mean
Total P
(g.l
-1
)
Annual mean
Chlorophyll-a (g.l
-1
)
Annual maximum
Chlorophyll-a
(g.l
-1
)
Annual mean
Secchi depth (m)
Ultra-oligotrophic <4 <1 <2.5 >12
Oligotrophic <10 <2.5 <8 >6
Mesotrophic 10-35 2.5-8 8-25 6-3
Eutrophic 35-100 8-25 25-75 3-1.5
Hyper-eutrophic >100 >25 >75 <1.5
The nutrient used in the published regressions is almost exclusively in-lake total phosphorus
concentration and the regressions are usually based on logarithmic transformations (Carlson,
1977; OECD, 1982; Mazumder and Haves, 1998; Portielje and van der Molen, 1999), only one is
based on a power function (Sas, 1989). In-lake TP concentrations can also be estimated using
data on lake depth, area, ushing rate and external TP loads using equations of Dillon and Rigler
(1974) and Kirchner and Dillon (1975). By linking the two, it is therefore possible to model how
changes in ushing rate and catchment land-use could affect phytoplankton abundance.
Deep lakes tend to show a much more predictable response than shallow lakes; it is therefore
possible that a relatively precise empirical model could be derived for deep European lakes to
predict impacts of changing nutrient conditions.
Mechanistic models which explore relationships between the physico-chemical environment and
the phytoplankton community are potentially useful for identifying reference conditions or
evaluating the effectiveness of restoration measures. A number of mechanistic models have been
Reference Region Notes
Dillon and Rigler (1974) North America Regression based on mean summer chlorophyll and spring TP
concentrations from 19 lakes in Canada combined with published data
from other North American lakes. Regression used to predict summer
chlorophyll from a single spring measurement of TP.
OECD (1982) Several Lake classication scheme based on expert evaluation of lake data.
Classication denes boundary values of total phosphorus, chlorophyll
and transparency (Secchi depth) between trophic classes from ultra-
oligotrophic to hyper-eutrophic. The classication is based on data from a
large number of lake types from regions across the globe, although there
was a predominance of large, deep lakes in the original dataset.
Gibson et al. (2000) based on
Carlson (1977; 1980)
USA EPA lake classication scheme which uses phytoplankton abundance,
chlorophyll concentrations and transparency (Secchi disc depth).
Premazzi and Chiaudani (1992) Europe Classication scheme based on annual mean and maximum chlorophyll
concentrations dened for complying (excellent and good status) and non-
complying waters (fair, poor, bad status).
Cardoso et al. (2001) Europe Threshold chlorophyll
-
a concentrations for lakes subject to eutrophication
set in response to the UWWTD.
Moss et al. (1996) Heinonen (2000)
Bratli (2000)
Swedish EPA (2000)
Newman (1988)
Great Britain
Finland
Norway
Sweden
Denmark
Transparency (Secchi depth) used as a parameter in classication of water
quality in many European countries.
Bennion (1994)
Wunsam and Schmidt (1995)
UK
European Alps
Quantitative direct relationships between phytoplankton composition
and nutrient conditions for planktonic diatoms, based on relative
abundances within surface sediment sub-fossil assemblages. Individual
species TP optima are dataset specic and cannot necessarily be
applied outside the region of development or for different lake types.
Moss et al. (2003)
Sondergaard et al. (2003)
Carvalho et al. (2004)
Lyche Solheim et al. (2004)
Europe
Denmark
UK
Norway
Preliminary phytoplankton classications (composition and abundance)
developed specically for the WFD.
Table 1.2 Published phytoplankton models or classication schemes for lakes.
developed which predict phytoplankton abundance and composition (DBS Model: PC Lake: Janse
and Van Liere, 1995; Los et al., 1997; Reynolds, 1999; PROTECH model: Elliott et al., 2001). Many
of these models incorporate mixing regimes and are capable of simulating multiple resource
limitation (phosphorus, nitrogen, etc.) and grazer effects.
Dynamic, deterministic models have also been developed to determine light attenuation in the
water column associated with phytoplankton (Kirk, 1994). Models such as CAEDYM, DYRESM,
SOBEK calculate light attenuation in the water column using Beers law with varying light
extinction coefcients for different algal groups, suspended sediments and dissolved organic
matter. From light attenuation, many models further calculate transparency. Theoretical models
examining lake ecosystem structure and functioning, in particular issues of stability and resistance
in shallow lakes, have been developed by Scheffer (1998).
1.3 Macrophytes
Introduction
The macrophyte community is generally regarded as a key indicator of the ecological status of
lakes as macrophytes provide habitat for many other aquatic biota (e.g. sh, macro-invertebrates,
wetland birds) to feed, seek refuge, or breed. Macrophytes are relatively long-lived organisms
(months to years), compared with phytoplankton and invertebrates. Macrophytes have a very
limited motility and are intrinsically linked to the prevailing environmental conditions in both the
surrounding lake water and sediments, through their roots and leaves. Individual species are
sensitive to physical and chemical changes in these media and hence are good indicators of both
current environmental conditions and longer-term environmental changes. Macrophytes, in terms
of assessing eutrophication pressure, can indicate enhanced nutrient concentrations through the
direct effects on species growth (biomass) and through indirect effects on species composition, as
may be caused by a reduction in transparency associated with nutrient-related increases in
phytoplankton and epiphyton.
Transparency is a key physico-chemical factor controlling the distribution and abundance of
submerged macrophytes in lakes. Changes occurring in the intensity and quality of light as it
passes down through the aquatic vegetation itself may exert a crucial effect on the development
and structure of aquatic macrophyte communities, as well as on the photosynthetic efciency and
productivity of the vegetation (Sculthorpe, 1967). Aquatic macrophytes are profoundly inuenced
by the transmission of light (the photosynthetically active radiation (PAR) through a water body,
which impacts upon photosynthetic processes. In turbid conditions the PAR passing through the
water can be attenuated by the scattering and absorption of light by suspended particles, both
organic and inorganic. Increasing turbidity can often result in a shift in community from submerged
to oating leaved or emergent species (Hough and Forwall, 1988). However, in terms of its impact
on aquatic macrophytes, turbidity can be difcult to separate from other properties of a lake, e.g.
colour and shading. Indeed, little is known of the precise inuence of underwater changes in light
intensity and quality on the distribution of life forms and species of aquatic macrophytes. It may
be that some plants of a particular habitat may tolerate, or even prefer sustained low intensities
and/or deciencies in certain wavelengths. Phenotypical variations due to abiotic circumstances
have been recorded for several species under changing inundation conditions e.g. Nabben et al.
1999; Van den Berg et al. 1999. Hyper-eutrophication may even lead to deterioration of emergent
and oating leaved communities due to reducing conditions in the sediment. These non-linear
responses of species composition and abundance to eutrophication pressures complicate the
indicator value of aquatic macrophytes.
Annex V of the WFD outlines two macrophyte-related quality elements that need consideration in
the assessment of the ecological status of lakes:
macrophyte community composition
macrophyte abundance.
Annex V of the WFD includes transparency as a physico-chemical element for supporting this
assessment of macrophyte communities. The WFD normative denitions for high, good and
moderate ecological status determine that ecological quality is considered as declining when the
composition of macrophytic taxa differ moderately from the type-specic communities and are
signicantly more distorted than those observed at good quality (slight changes in composition
and abundance of macrophytic taxa compared to the type-specic communities).
Metrics
The macrophyte community composition of lakes can be broadly related to a nutrient gradient, as
individual species appear to show differing tolerances to nutrient conditions, although research
suggests that other factors such as alkalinity, sediment type and water depth may be more
important in determining species composition (Penning, pers comm.). Some species such as Isoetes
lacustris, Potamogeton polygonifolius and Sparganium angustifolium tend to be restricted to low
alkalinity, oligotrophic lakes, other species, for example, Lemna spp, Myriophyllum spicatum and
Potamogeton pectinatus are typical of high alkalinity, (natural) eutrophic lakes. However, many
macrophyte species can be regarded as being indifferent to a change in trophic level, occurring
over a wider range of trophic conditions.
The relationship between the abundance of macrophytes and nutrients is complicated by internal
interactions between other ecological components of the ecosystem, particularly in shallow lakes
(Jeppesen et al., 2000; Moss et al., 2003). For example, Scheffer et al. (1993) produced an
alternative stable states model to describe shallow lake functioning. Rather than exhibiting
linear responses to changing nutrient concentrations, macrophyte species may undergo a threshold
response whereby they abruptly change from being abundant to a state where they are absent,
when phytoplankton dominate and vice versa. It appears that this can be due to initial spring
conditions, but also the characteristics of the sh community and management thereof can play
an important role (Van de Bund and Van Donk, 2002). Abundance can also be calculated as the
coverage of aquatic vegetation as a percentage of potential littoral area (USEPA, 1998;
Anonymous, 2003). Finnish case studies showed that helophytes and oating leaved vegetation
coverage could be related to an increase of phosphorus and nitrogen of lake water (Kanninen et al.,
2003; Leka et al., 2003). In shallow lakes macrophytes may even disappear when a hypereutrophic,
algal dominated state is reached (Blindow, 1992).
Although no comprehensive study has been undertaken to relate aquatic macrophytes with
turbidity/transparency measurements, a number of studies have examined the depth range of
individual species (e.g. Spence and Chrystal, 1970; Blindow, 1992; Wilby et al., 2000) or focused
on specic locations such as the Danube Delta, where Coops et al. (1999) classied lakes into
three different types that were related to turbidity ranges. In general, in relative terms, it appears
that the deeper a species grows, the less light the species requires and, therefore, the assumption
could be made that deeper growing species can compete better for light than other species and
will be more resistant to increased turbidity. However, not all species tolerant to low light and
growing at greater depths in clear lakes will successfully colonise lakes with high turbidity.
Exceptions may be species that are sensitive to factors such as eutrophication, substrate type,
siltation of leaves, or light quality, rather than just quantity of light (Davis and Brinson, 1980).
The sensitivity of North American submerged macrophytes to turbidity has been explored using the
Turbidity Tolerance Index, i.e. the ratio of the depth maxima of a species to the Secchi disc transparency
depth (Davis and Brinson, 1980). The Turbidity Tolerance Index provides an approximation of the
relative resilience of macrophyte species to turbidity stress and has been expanded to include more
species, including a number of which are known to occur in the UK (Adamus and Brandt, 1990). In
addition, species such as charophytes are thought to actively inuence the turbidity in their close
vicinity (Van den Berg et al., 1997; Van Donk and Van de Bund, 2002; Noges et al., 2003).
In the Danube Delta Lakes project the lake typology in the Danube delta was described using
phytoplankton, hydrology, chemistry, macrophytes and sh species based on PCA analysis (Oosterberg
et al., 2000). Here, three groups of lakes could be characterised by their macrophyte communities by
correlating lakes with macrophytes and environmental factors, including nutrients. Coops et al.
(1999) also performed a classication for the Danube delta based on aquatic vegetation and
turbidity. For the WFD a number of macrophyte classication schemes have recently been reported
for Denmark (Sndergaard et al., 2003), the Netherlands (Van den Berg et al., unpublished), Finland
(Kanninen et al., 2003), Germany (Schaumberg et al., 2004), Northern Ireland (Dodkins et al., 2003)
and the UK (Willby, pers. comm.). All these macrophyte classication schemes share the common
characteristic of not being exclusively related to nutrients. Instead, most macrophyte classication
schemes are indirectly related to nutrient conditions. Examples are given in Table 1.3.
As no true direct relationship between transparency and macrophytes has been established, no
classication schemes for this parameter have been developed. Most classication schemes have
concentrated on the parameters that cause changes in transparency, and so are indirectly related.
In the case of nutrient pressures these have typically focused on the association of macrophyte
species with variables that correlate with trophic status (e.g. Palmer et al., 1992 - TRS and
alkalinity). The general expected trend in European lakes is that transparency will decrease with
increasing productivity of the water (Ellenberg, 1988; Palmer et al., 1992). The expected turbidity
states for various lake trophic types are as follows (see also Figure 1.4):
Dystrophic lakes coloured water
Oligotrophic (lime decient) lakes generally clear water
Oligotrophic (lime rich(marl)) lakes very clear water
Mesotrophic (macrophyte dominated) lakes generally clear water although may have
seasonal turbidity due to phytoplankton
Mesotrophic (phytoplankton dominated) lakes turbid water due to phytoplankton
Eutrophic lakes turbid water due to phytoplankton and suspended solids.
In dystrophic lakes the clarity of water is most likely to be as a result of peat colouring rather than
as a result of turbidity arising from phytoplankton growth resulting from increased nutrient
concentrations. Shifts between these broad lake trophic types would indicate changing status but
there is insufcient data available to establish more explicit working ranges for transparency.
Palmer et al. (1992) UK The Trophic Ranking Scheme (TRS) used the macrophyte composition recorded in
1224 standing waters to classify UK lakes into 10 vegetation groups, which were
related to lake alkalinity, pH and conductivity. Each of these 10 groups were
allocated site types based on lake trophy. Individual macrophyte species were also
allocated a TRS based on the range of site types within which they were found.
The average site TRS can be used to infer whether eutrophication has occurred.
However, the classication scheme is not reference based and does not directly
relate macrophyte species to nutrient conditions.
Swedish EPA (2000) Sweden The Swedish Environmental Quality Criteria (SEQC) scheme assesses the state of
lakes using a variety of factors including nutrients and species richness of
macrophytes. Comparisons of the current condition with reference values are used
in appraisals. The SEQC scheme denes conditions for both nutrient loadings and
macrophytes at high to low ecological status as well as deviations from the high
reference state. The macrophyte scheme is based on the UK TRS (Palmer et al.,
1992) and does not, however, directly relate the aquatic plant communities to
nutrient concentration data.
US EPA (1998) USA The US EPA Lakes and Reservoir Bioassessment and Biocriteria uses submerged
macrophytes as one of 7 biological monitoring elements for assessing the condition
of US lakes. Lake condition is assessed using additive indices that integrate both
habitat and biological scores. The LRBB scheme provides reference values for
macrophyte metrics and nutrients but doesnt directly relate them together. This
multimetric index indicates the overall biological condition of a lake, however, it
cannot quantify the actual cause of degradation, although it does suggest where
eutrophication may be the cause.
Heegaard et al. (2001) Northern Ireland The Northern Ireland Lake Survey quantied macrophyte species environmental
relationships from over 500 lakes using Generalised Additive Models (GAM) and
Canonical Correspondence Analysis (CCA). Nutrient concentrations appeared
inuential in explaining species distribution, but were highly correlated with
alkalinity and altitude.
Copps et al., (1999) Romania A classication of Danube Delta lakes based on aquatic macrophytes and
turbidity.
Moss et al. (2003)
Sondergaard et al. (2003)
Kanninen et al. (2003)
Schaumberg et al. (2004)
Dodkins et al. (2003)
Europe
Denmark
Finland
Germany
Northern Ireland
Preliminary macrophyte classication schemes developed specically for WFD.
Table 1.3 Published examples of lake macrophyte models or classication schemes indirectly related to
nutrient conditions.
Models
Detailed modelling studies have been carried out examining macrophyte attributes in relation to
environmental factors (e.g. Rrslett, 1991; Scheffer et al., 1993; Weisner et al., 1997; Muhammetoglu
and Soyupak, 2000; Aseada et al., 2001; Murphy, 2002; Murphy and Hootsmans, 2002). Linear
regression models and Generalised Additive Models have been employed to try and model the
response of individual macrophyte species to environmental conditions (Heegaard et al., 2001;
Van den Berg et al., 2003), or the interaction between species (Van Nes, 2002). These models
may have relevance for WFD bioassessment in linking macrophyte community attributes to
nutrient conditions.
However, as yet, no reliable quantitative models relating transparency and lake macrophyte
composition have been developed. However, it should be noted that in the Netherlands Schaminee
et al. (1995) described the vegetation types in the aquatic environment and included the main
abiotic characteristics. Bal et al. (1995) also provided detailed information of species composition
in relation to abiotic conditions. These paved the way for the development for the Dutch WFD
macrophyte classication scheme.
1.4 Phytobenthos
Introduction
Phytobenthos are an important component of primary production in lakes, although its contribution
to overall primary production decreases with increasing lake depth. In deep lakes, the phytobenthos
is, however, still an important component of littoral food webs, and will inuence the structure of
littoral macrophyte, invertebrate and sh communities. For example, an increasing abundance of
epiphyton, in response to nutrient enrichment, has long been considered an important factor in the
loss of submerged macrophytes (Phillips et al., 1978; Sand-Jensen and Sndergaard, 1981) and the
phytobenthos itself constitutes a signicant source of energy for most littoral invertebrate grazers.
Figure 1.4 Relationship between lake trophic
category and transparency (Secchi disc dept;
OECD, 1982). 5 1 15 2 3 5 10 15 20 30 50 m
Probability Distribution for Trophic Categories
10
5
0
Hypertrophic Eutrophic Oligotrophic Ultra-Oligotrophic
Mesotrophic
Av. yearly
Secchi disc
Transparency
Despite its potential signicance, phytobenthos has received relatively little attention in terms of
its use as an indicator of lake quality. The fact, that phytobenthos respond to both water column
nutrient concentrations and habitat quality, is accessible from the lake shore, and is less dynamic
than the phytoplankton community has led to increasing interest in its use as a monitoring tool
for lakes (US EPA, 1998).
Annex V of the Water Framework Directive specically outlines phytobenthos composition and
abundance as two criteria that need dening for type-specic ecological assessment of lakes in
relation to undisturbed conditions. Eutrophication of the overlying water generally results in an
enhanced growth of attached algae and changes in community composition. Species associated
with low phosphorus (the diatoms Achnanthidium minutissimum and Gomphonema tenellum)
and low nitrogen (nitrogen-xing species, such as the diatoms Epithemia adnata and Rhopalodia
gibba and cyanobacteria such as Anabaena spp.) waters have been shown to disappear following
nutrient enrichment (Fairchild et al., 1985). If nutrient enrichment proceeds further, phytoplankton
can shade the phytobenthos, reducing abundance and shifting community composition to species
tolerant of low light levels. The detailed response of the phytobenthos to nutrient conditions is,
however, much more complex than for phytoplankton. Nutrients diffuse much more slowly into
attached communities, with strong gradients through boundary layers and the attached
community. For example, denitrication by bacteria within biolms can result in nitrogen limitation
of the phytobenthos irrespective of nutrient availability in the water column. Further difculties,
in relating water column nutrient concentrations to benthic algae, is that they can also obtain
nutrients from the substrates that they are attached to.
Phytobenthos abundance is generally measured as chlorophyll-a per unit area of substrate.
However, quantitative analysis of phytobenthos abundance is much more difcult than for
phytoplankton, due to problems in standardising sampling area and the greater proportion of
detrital material within phytobenthos communities.
Only recently have attempts been made to develop quantitative relationships between the
phytobenthos community and nutrient conditions. Danilov and Ekelund (2000) analysed epiphyton
and epilithon species diversity from seven Swedish lakes. They concluded that epiphyton diversity
showed little relationship with nutrient concentrations, but epilithon diversity was consistently related
and could be used as an indicator of nutrient status. King et al. (2000) examined distributions of 138
epilithic diatom species from 17 lakes in the English Lake District and showed that total phosphorus
and calcium concentrations were the most important variables explaining species distributions.
Classication and monitoring schemes
No lake classication schemes based on phytobenthos composition or abundance have been fully
developed. Schemes are under development for the WFD in several Member States, although
details have only been published for Germany (Schaumburg et al., 2004). Phytobenthos is
recommended as one of seven potential biological parameters in Tier 2 of the US EPA lake
monitoring and classication scheme (US EPA, 1998). The scheme highlights the potential of
phytobenthos, but does point out that responses to pollution or disturbance are not adequately
known and require further development.
Surface sediment diatom assemblages are important members of the phytobenthos community.
Direct, quantitative relationships between diatom species and total phosphorus concentrations
have been developed across Europe (Bennion, 1994; Wunsam and Schmidt, 1995). The combined
response of phytoplankton and phytobenthos communities is probably the most established
representation of lake ecosystem response to eutrophication and would be highly compatible with
palaeolimnological methods for the setting of reference conditions.
Models
No models have been established associating phytobenthos composition or abundance in lakes
to specic hydro-morphological or physico-chemical parameters.
1.5 Benthic invertebrates
Introduction
Eutrophication is associated with an increased mineral nutrient supply, mainly N and P, which
usually results in an increase in primary production (Sas, 1989). This eventually leads to an
increase in the sedimentation of dead plankton (Baines and Pace, 1994) with a resultant increase
in decomposition by micro-organisms. Increased decomposition is accompanied by enhanced
oxygen consumption by micro-organisms, which can result in oxygen depletion, particularly at the
bottom sediments (Dinsmore and Prepas, 1997) and in the profundal zone of deep lakes. This
indirect effect of eutrophication on oxygen conditions can lead to signicant changes in the
structure of oxygen dependant communities, either as a change in dominant species with mild
enrichment or severe shifts in species composition with more severe enrichment. The most sensitive
communities to changes in oxygen conditions are faunal communities.
The benthic invertebrates are frequently used as indicators of the biological integrity of ecosystems
because they are found in most aquatic habitats. They are a diverse and generally abundant
group with a wide range of environmental tolerances and preferences. They therefore serve as a
useful tool for detecting long-term environmental perturbation resulting from point and non-point
sources of pollution (Cairns and Pratt, 1993).
Benthic invertebrates play an essential role in key processes of lakes - in food chains, productivity,
nutrient cycling and decomposition. In principle, any environmental changes in lakes, for example
in nutrient concentrations, would be reected by changes in the structure of the benthic
invertebrate community. Profundal macroinvertebrates form an important link between detrital
deposits and higher trophic levels in aquatic food webs (Brinkhurst, 1974).
Food availability has been postulated as a key factor for benthic communities, and the coupling
of pelagic and benthic production has been demonstrated in lake Esrom (Johnasson, 1996) and
lake Erken (Goedkoop and Johnson, 1996). However, in other freshwater bodies oxygen content
has proved to be the limiting factor for the benthos, especially when the anoxic period lasts more
than 4 months (Heinis and Crommentuijn, 1992).
Historically most studies have focused on the changes of the profundal benthic invertebrate
communities of deep stratifying lakes (mean depth greater than 3 m). In these deeper waters,
below the photic and wave action zones, environmental conditions tend to be relatively uniform
and predictable. In contrast, a greater variety of invertebrates are found in the more complex and
diverse habitat conditions of smaller shallow (mean depth less than 3 m), non-stratied lakes or
littoral areas of deep lakes.
The larvae of the midge genus Chironomus (chironomids belonging to the family Chironomidae)
are common in the profundal zones of lakes and reservoirs around the world (Armitage et al.,
1995). The capacity of these insects to live at very low oxygen concentrations is well known.
Chironomids are probably the most useful profundal indicator group for the trophic state of a
lake, as they have high species richness compared to other benthic invertebrate groups.
They occur over the whole spectrum of nutrient conditions and, because individual species have
highly specic environmental tolerances, they change species composition in tandem to changing
lake trophic status (Rosenberg, 1992). Multiple regression analyses of data from Spanish lakes
and reservoirs, showed that Chironomus density decreased with depth and temperature and
increased with alkalinity (Real, 2000).
Species indicative of eutrophic conditions will gradually colonize a lake (in most cases species
belonging to class Oligochaeta and family Chironomidae). Although, freshwater worms,
oligochaets (Oligochaeta) are more sensitive to changes in oxygen conditions than chironomids,
because they are less mobile and they feed and reproduce deeper in the sediments than the latter
(Lang, 1999).
Annex V of the WFD outlines the following benthic invertebrate quality components that need
consideration in the assessment of the ecological status of lakes:
Composition and abundance of benthic invertebrate fauna
Ratio of disturbance sensitive taxa to insensitive invertebrate taxa
Diversity of invertebrate communities.
In the WFD, a declining ecological quality is associated with poor diversity and dominance of a few
taxa. To comply with good ecological status the composition and abundance of invertebrate taxa, the
ratio of disturbance sensitive taxa to insensitive taxa and levels of diversity should show only slight
signs of alteration compared to the type-specic reference communities (Annex V, Section 1.2.2).
Metrics
Taxonomic composition. The tolerances of profundal macroinvertebrates to low dissolved oxygen
(DO) concentrations vary widely among species (Davis, 1975; Herreid, 1980). Attempts to devise
a lake typology for macroinvertebrates in general and specically for benthic chironomid
assemblages, based on trophic status and DO concentration, have met with varying degrees of
success (Brinkhurst, 1974), although clear general patterns exist. A combination of both sediment
type and oxygen supply have been shown to control the distribution of chironomids (Francis and
Kane, 1995) and oligochaetes (Verdonschot, 1996). The Orthocladiinae are characteristic of
highly oxygenated waters and thus tend to reect high status as does a high ratio of the
oligochaete Naididae to chironomids (Armitage et al., 1995). Common compositional metrics
include the percentage of individuals of the family Orthocladiinae among chironomid larvae
present and the ratio between the number of individuals of the oligochaete Naididae to the sum
of all specimens of Naididae and chironomids (Moss et al., 2003). In contrast, particularly in
shallow lakes, other authors have demonstrated that biotic factors, rather than environmental
factors, can be responsible for the composition of the benthic macroinvertebrate community.
The biotic factors examined were mainly the effect of predation (Kornijow, 1997) but also intra-
and interspecic competition (Van de Bund and Groenendijk, 1994).
Barbour et al. (1995) recommend relative (percentage) rather than absolute abundance in
calculating metrics as they have been shown to be more robust and reliable and were more likely
to reect structural changes resulting from nutrients. An additional composition attribute is Bray-
Curtis dissimilarity, a coefcient shown to be a robust and ecologically interpretable index of
changes in taxonomic composition (Legendre and Legendre, 1998).
Benthic invertebrates abundance. Increased phytoplankton production associated with
eutrophication and the concurrent increase in organic sedimentation results in decreasing oxygen
concentration at the sediment/water interface. For benthic invertebrates, this means more food
but reduced oxygen availability. As a result, characteristic species indicative of eutrophic conditions
gradually colonise (in most cases oligochaetes and chironomids). Oligochaetes are more affected
by enrichment than chironomids as they are less mobile and they depend more on the inner
sediment for their food and reproduction than the latter (Lang, 1999).
In lakes with a strong oxygen gradient, profundal macroinvertebrate biomass (PMB) typically
declines from the upper profundal region to the maximum depth (Brinkhurst, 1974). Conversely,
the articial enhancement of hypolimnetic DO concentrations in productive lakes, through
aeration and oxygenation treatments, is often followed by a dramatic increase in profundal
macroinvertebrate abundance (Dinsmore and Prepas, 1997). Rasmussen and Kalff (1987)
hypothesised that the observed negative correlation of PMB with high phytoplankton biomass in
their regression models truly reected the deleterious effects of decreased hypolimnetic DO
concentration on profundal macroinvertebrates. In addition, Takamura and Iwakuma (1990)
suggested that extremely high primary productivity leading to anoxic conditions at the sediment/
water interface may depress PMB even in shallow polymictic lakes.
Benthic invertebrates diversity. Measures of diversity have only occasionally been used in studies
of the profundal benthos primarily due to the difculty in identifying taxa and because profundal
communities are difcult to characterise using diversity indices as they tend to be dominated by
comparatively few species (Wiederholm, 1980). Wiederholm (1980) suggested that species
richness or the total number of taxa is a much better measure of the diversity of the benthos by
demonstrating that a strong correlation could be made between species richness (once adjusted
to sample depth) and the trophic state (as indicated by chlorophyll-a concentrations). Richness
and diversity attributes includes the total number of taxa (richness per unit area) and numerical
richness (Larsen and Herlihy, 1998). Low diversity indices are associated with eutrophicated
ecosystems.
Benthic invertebrates are widely utilised as indicators of organic pollution in freshwaters,
particularly rivers due to their sensitivity to oxygen conditions (Hynes, 1960; Hellawell, 1986).
Early classication schemes in lakes, similar to the Saprobic systems in rivers, were based mainly
on the analysis of the species composition of the chironomid communities. See Table 1.4 for
examples.
The Benthic Quality Index (BQI) was developed for assessing trophic status of Palearctic lakes
(Wiederholm, 1980). In these ecosystems, a eutrophic lakes BQI value is considered to be 1 and
Chironomus plumosus the dominant taxon; a BQI value of 5 is characteristic of an oligotrophic
lake. If no indicator species are present then a value of 0 is recorded and is indicative of a
hypereutrophic lake (Wiederholm, 1980). Lang developed three indices of trophy for lake Geneva
based on the structure of the Tubicidae and Lumbriculidae communities (Lang, 1998).
Numerous other studies highlight oxygen concentrations and food availability (as indicated by
total phosphorus, chlorophyll-a and algal biovolume) to be the most important community-
structuring factors linking chironomid communities with lake trophic state (e.g. Saether, 1979;
Brodin, 1982). In northern temperate waters these models have proved to be useful (e.g. Kansanen
et al., 1984) as long as oxygen concentrations in the profundal zone remained high enough to
support the chironomid communities (Brodersen and Lindegaard, 1999).
Wiederholm (1980) Palearctic Benthic Quality Index (BQI) used to assess trophic status of
Palearctic lakes. In this system in eutrophic lakes the BQI value is
considered to be 1 and Chironomus plumosus the dominant taxon;
a BQI value of 5 is characteristic of oligotrophic lakes. If no
indicator species are present then a value of 0 is scored indicating
a hypereutrophic lake.
Lang (1998) Lake Geneva, Switzerland Indices of trophy were developed based on the structure of the
Tubicidae and Lumbriculidae (Oligochaeta) communities.
Brodin (1982)
Heinis and Davids (1993)
Saether (1979)
Kansanen et al. (1984)
Lindegaard (1995)
Brodersen and Lindegaard (1999)
Sweden
The Netherlands
Norway
Finland
Denmark
Studies linking chironomid communities with lake trophic state
(indicated by variables such as oxygen concentration, total
phosphorus, chlorophyll-a and algal biovolume).
Dinsmore et al. (1999) Alberta, Canada Empirical model linking profundal macroinvertebrate biomass
(PMB) and water chemistry and morphometric variables from 26
lakes located within the Boreal Mixedwood and Boreal Subarctic
ecoregions.
King and Richardson (2003) Florida, USA Five macroinvertebrate core metrics were linked to surface-water TP
from an observed P gradient and a P-dosing experiment in coastal
wetlands of the south Florida to estimate numerical water quality
criteria for TP.
Astrakhantsev et al. (2003) Lake Ladoga, NW Russia Model of the annual dynamics and distribution of zoobenthos
biomass with respect of lake phosphorus dynamics
Table 1.4 Examples of benthic invertebrate classication systems and models.
Models
Empirical models of profundal macroinvertebrate biomass using dissolved oxygen concentration
and water temperature as predictors have explained a signicant proportion of the variance
found in profundal macroinvertebrate biomass (PMB; Dinsmore et al., 1999). Residual variation
in this model is high suggesting that the model could be improved by obtaining more representative
estimates of PMB and water chemistry variables; including estimates of detrital production,
accumulation rate and composition; or by quantifying the effects of biotic factors such as
competition, predation, life history phenologies
1
, which have an important inuence on PMB in
habitats where dissolved oxygen and food materials are not limiting (Kajak, 1987).
Empirical models have established relationships between total phosphorus and the structure and
function of macroinvertebrate communities (e.g. King and Richardson, 2003). A model developed
for Lake Ladoga, North-Western Russia, reproduced the annual dynamics of zoobenthos biomass
and its distribution over the lake bed (Astrakhantsev et al., 2003).
1.6 Fish
Introduction
According to the cascading trophic interaction hypothesis a top predator such as piscivorous sh
can inuence the abundance, size-structure, and productivity of zooplankton and phytoplankton
(Carpenter and Kitchell, 1984; Carpenter et al., 1985). The bottom-up and top-down model
(McQueen et al., 1986; McQueen et al., 1989) combines the inuence of both predators (top-
down) and resources (bottom-up). It predicts that top-down forces should be strong at the top of
the food web and weaken towards the bottom, whereas bottom-up forces should be strong at the
lower trophic level (algae) and weaken towards the top of the food chain (little or no inuence of
sh). Top- down pressure can exert substantial inuence on phytoplankton composition and
biomass through grazing pressure, however the importance depends on the number of food web
links (2 to 5) and trophic status of the lakes (oligo- to hyper-trophic).
Due to their complex ecological requirements, sh are sensitive indicators for habitat quality at
various spatial scales. As consumers and/or top predators, they integrate information on trophic
conditions across the food chain. They also provide detailed information on the respective trophic
level. There is a long tradition of linking the health of sh populations to water quality.
Dissolved oxygen is a dening and limiting parameter for sh, affecting survival, growth, spawning,
swimming performance, larval development and migration behaviour (Doudoroff and Shumway,
1970). Since oxygen is vital to other aquatic biota (i.e. invertebrates) that are a key food source
for sh, then the species composition and biomass of sh communities can also be affected
indirectly by oxygen availability. The composition of the sh community at a specic site is the
result of various factors including environmental factors such as temperature, oxygen, ow and
nutrients and the anthropogenic movement of sh.
1
Periodic life history events, such as breeding, migration, etc., which occur in relation to seasonality and climate.
Fish species can be separated by their oxygen requirements. As well as having different requirements
between species, they have different needs throughout their life stages (eggs, larvae, juveniles
and adults). Generally, stenotherms
2
need higher oxygen concentrations than eurytherms
3
and
adult sh tolerate low oxygen concentrations better than juveniles. The temperature of the
surrounding water has a substantial inuence on the dissolved oxygen concentration within the
water column, as well as signicantly affecting the physiological demand by sh. Warmer water
temperatures reduce the oxygen saturation capacity, as well as increasing the physiological
demand by the sh. The development of anoxia in lakes is most pronounced in thermally stratied
systems in summer and under the ice in winter when the water mass is cut-off from the atmosphere.
Besides the direct effects on aerobic organisms, anoxia can lead to increased release of phosphorus
from sediments that can fuel algal blooms when mixed into the upper euphotic (illuminated)
zone. It also leads to the build-up of chemically reduced compounds such as ammonium and
hydrogen sulphide (H
2
S) which can be toxic to bottom dwelling organisms. In extreme cases,
sudden mixing of H
2
S into the upper water column can cause sh kills.
The Water Framework Directive requires information on sh communities to be used in the
ecological quality assessment of lakes. Annex V of the WFD outlines three sh-related quality
elements that need consideration:
sh composition
sh abundance
sh age structure.
The WFD normative denitions of ecological status classication, indicate that: 1) changes in
species composition and abundance; 2) decreases in type sensitive species; and 3) changes in age
structure, as a sign of disturbance indicating a failure of reproduction or development of a
particular species; will downgrade water quality classication.
Each sh species has characteristic requirements for water quality, habitat and environmental
conditions necessary to satisfy their need for breeding, feeding, growing, recruitment and survival.
These characteristics are used to classify sh species according to the concept of ecological/
functional guild, dened as a group of species that use the same class of environment conditions
in the same way (Root, 1967; Austen et al., 1994). Examples of classication schemes are
summarized in Table 1.5.
The European Fish-Index (EFI) developed by the FAME project
4
, is a multimetric sh-index for
classication of European rivers. The EFI is based on the same principle of the Index of Biotic
Integrity (IBI; Karr, 1981), where the fundamental assumption is that the composition and
structure of sh assemblages are impacted by human pressures in a predictable manner. In the
EFI the sh species were classied according to their zoogeographic status (Native, Introduced,
2
Organisms adapted** to only slight variations in temperature.
3
Organisms adapted*** to a wide range of temperature.
4
EVK1-CT-2001-00094; http://fame.boku.ac.at.
Endemic), trophic guild, reproductive guild, habitat guild (degree of rheophily
5
and habitat
preference in the water body), migratory behaviour, longevity and tolerance capacity (Schmutz et al.,
in prep.). Where possible, tolerance to specic stressors was assessed (temperature, acidication),
including habitat degradation and water quality. Where information was lacking an overall
assessment of general tolerance was made. A group of sentinel species, i.e. known to be common
in specic river-zones and providing good information on ecological quality, was identied for
Jeppesen et al. (2000) Denmark Fish species richness, biodiversity and trophic structure are
linked to a trophic gradient of TP
Moss et al. (2003) Europe Classication system developed for 48 European lake ecotypes
using 28 variables for water quality status with 3 sh
parameters: sh community, sh biomass and piscivorus-
zooplanctivorus biomass ratio
Regier et al. (1998)
Navodaru et al. (2003)
Oosterberg et al. (2000)
Europe
Danube Delta lakes
Classication of sh communities in relation with a lake
ecology guild, such as riverine/ white sh, euritopic/ grey sh
and limnophilic/ black sh.
Ligtvoet and Grimm (1992) The Netherlands Occurrence of three sh communities (perch-type, pike-perch-
type, and pike perch-bream-type) and the size of total sh
stock, perch stock, pike stock and cyprinid stock linked with
summer average TP concentration.
Gassner et al. (2003) Austria A relationship between TP (mg.m
-3
) and total sh biomass in
kg.ha
-1
were calculated for prealpine oligo- and oligo-mesotrophic
lakes using: BM = 3.8148 x TP
1.0940
, r
2
= 0.59, n = 10.
Carvalho et al. (2002)
EIFAC (1973)
US EPA (1986; 1997)
UK
Europe
USA
Classication criteria between threshold requirement for
oxygen concentrations and sh communities (e.g. in UK:
dissolved oxygen >8mg.l
-1
for salmonids, >6 for non-salmonids
and >1 for tolerant species of cyprinids). Similar or slightly
different thresholds have been forwarded by the European
Inland Fishery Advisory Commission (EIFAC), National
Research Council of Canada (NRCC), and US Environmental
Protection Agency (US EPA).
Swedish Environmental Protection Agency
(www.internat.naturvardsverket.se)
Sweden Method for classifying sh communities in rivers and lakes
using parameters from sh surveys and physical characteristics
of the water bodies. The classications are not directly related
to specic physico-chemical parameters, but aim to dene the
extent to which these communities may deviate from
undisturbed waters.
Busnita and Brezeanu (1967)
Nikolski (1962)
Romania Classication systems based on sh ecology and oxygen
sensitivity.
Diudea et al. (1986) Romania Direct nutrient stress on sh. Toxic effects of nitrite and nitrate,
depend on species, salinity, concentration and time of
exposure. Maximum level of Romanian STAS range of
10-45 mg.l
-1
for nitrate (-NO
3
) and 0-1 mg.l
-1
for nitrite (-NO
2
),
depending of water category use.
Table 1.5 Examples of Fish-based classication systems for lakes.
5
Degree of preference for running waters.
each ecoregion. For these species, information regarding size or age structure is required in the
multimetric assessment. The EFI indicates the deviation from predicted reference metrics, and
provides a probability for a site to represent reference conditions.
Final ecological status of a site is identied by the degree of deviation from reference metrics. The
EFI is able to detect impacts of both physical and chemical pressures. However more precise data
on pressures are required to distinguish between different types of pressure. The EFI index cannot
be applied directly for sh biota in lakes, but the same approach could be applied to develop a
sh-index for lakes.
A Fish-based classication system has been developed for Danish lakes (Jeppesen et al., 2000).
Changes in sh species richness, biodiversity and trophic structure were linked to a trophic gradient
which was divided into ve total phosphorus (TP) classes (class 1-5: <0.05, 0.05-0.1, 0.1-0.2. 0.2-
0.5, >0.4 mg P L
-1
). The sh species richness was unimodally related to TP, being highest at 0.1-0.4
mg P L
-1
. At low nutrient concentrations, piscivorous sh (particularly perch, Perca uviatilis) were
abundant and the biomass ratio of piscivores to plankti-benthivorous cyprinids was high and
density of cyprinids low. With increasing TP, a major shift in trophic structure occurred.
A classication system developed for 48 European lake ecotypes by the research project ECOFRAME
proposed 28 variables for water quality status from which three were sh parameters: sh
community, sh biomass and piscivorus-zooplanctivorus biomass ratio (Moss et al., 2003). The
advantages of the proposed classication system are that it applies variables that are inexpensive
to measure and ecologically relevant. The scheme was tested using data from 66 shallow lakes,
and may be adapted for more lakes after minor adjustments.
A classication of sh communities into lake ecology guilds, such as riverine/ white sh, eurytopic/
grey sh and limnophilic/ black sh has been proposed for Europe. White sh are Acipenseridae,
Salmonidae and Clupeidae, grey sh Percidae and Esocidae and black sh Umbridae and
Cobitidae; representatives of Cyprinidae turn up in all three classes (Regier et al., 1998). Except
for diadromous
6
migratory species, all these categories are found in lakes of the Danube delta
that are connected to a river with a network of natural or man-made canals (Navodaru et al.,
2002). Abundance, biomass and species guild was used to describe a trophic gradient of Danube
delta lakes. PCA analysis was carried out to separate lake types for the Danube delta using
comprehensive data sets of hydrology, chemistry, vegetation, phytoplankton, and sh (Oosterberg
et al., 2000). Based on this study three types of lake were identied: Type I, mostly eutrophic with
intermediate sh species (grey and black), Type II, an intermediate with eurytopic/grey sh and
Type III, mesotrophic with limnophilic/black sh (Figure 1.5).
Models
Some examples of models for sh and environmental quality variables for lakes are reported in
Table 1.6. A study of 466 lakes from temperate to artic zones described the impact of nutrients
and lake depth on top-down control in the pelagic zone of lakes (Jeppesen et al., 2003). The study
6
Fish species migrating between fresh and marine waters.
concluded that (a) sh predation pressure on large-bodied cladocerans is overall unimodally
related to TP in both shallow and deep lakes; (b) the effect generally cascades to the phytoplankton
biomass level in eutrophic lakes, but less signicantly in oligotrophic lakes; (c) predator control of
large-bodied zooplankton tends to be higher in shallow lakes than in deep lakes. A multiple
regression model revealed that the percentage abundance of Daphnia was signicantly inversely
related to Catch Per Unit Effort (CPUE) and positively related to lake water TP, suggesting a higher
predation risk at a given CPUE in oligotrophic lakes.
Fish yield potential (FYP) was estimated for 786 lakes in north-east Germany using data on
primary production of phytoplankton (PP) and total phosphorus (TP) as the main variables
(Brmick and Lemcke, 2003). The following equation between TP and PP was empirically derived
from 8 stratied German lakes:
PP (g C.m
-1
a
-1
) = 148 logTP(g.l
-1
) 39.6
A relationship between TP in stratied and shallow lakes, that had similar trophic status, gave the
following equation:
TP
1
= 1.48198 TP
2
.
1.2278
where TP
1
= TP in stratied lakes, and TP
2
= TP in shallow lakes. The trophic status for both types
of lakes was obtained from the German States Association for Water (1998), where the trophic
state of a lake is based on four parameters, including TP-value during the spring turnover. Based
1.0
0.5
0.0
-0.5
-1.0
1.0
0.5
0.0
-0.5
-1.0
-1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0
(A) (B)
Umbrei
TINC TINC
Cuibul cu lebede
Plin
Furtuna
Baclanesti
ABRA BRAM
CARA AURA
RUTI RUTI
Isac
ALBU ALBU
BLIC BJOE
Rosu
Raducu
Raduculet
Tataru
PERC FLUV
SCAR ERYT
ESOX LUCI
Serbata
TINC TINC
Organic
AqVeget
FilamAlg
Amplit
PotTrich
ABRA BRAM
CARA AURA
RUTI RUTI
Clay
Size
ALBU ALBU
Depth
Sand
BLIC BJOE
PERC FLUV
Phyto
Reed
SCAR ERYT
TravTim
NitelObt
ESOX LUCI
CARA CARA
Figure 1.5 PCA analysis of the relative biomass of 10 dominant sh species in 11 lakes in relation to trophic
variables, (A) species-site bi-plot of the rst two axes, and (B) species-environment bi-plot of the rst two axes.
Abbreviations of sh species are composed by the rst 4 letters of both genus and species names. Soil type
(ORGANIC, SAND or CLAY), morphometry (SIZE, DEPTH), hydrological distance (TRAVTIM) together with
subsequent residence time in the lake (CURESTM), water level uctuation (AMPLIT), chlorophyll a concentration
(PHYTO), lamentous algae (FILAMALG), aquatic vegetation coverage (AQVEGET), Potamogeton trichoides
(POTTRICH), Nitellopsis obtusa (NITELOBT) and reed: open water ratio (REED). (original from Navodaru et
al., 2002).
on these models, Brmik and Lemcke (2003) developed regional sh yield estimation models for
shallow and stratied lakes. The models take into account the ndings by Lang et al. (1981), who
state that the relation between PP and FYP changes from an exponential to a logarithmic function
with increasing PP values and approaches the upper asymptote. Accordingly, the following models
were obtained (Brmick and Lemcke, 2003):
FYP (for lakes with PP < 380): FYP(kg.ha
-1
.a
-1
) = 6.315e
0.0062.pp
, r = 0.47;
FPY (for lakes with PP > 380: FYP(kg.ha
-1
.a
-1
) = 57.937lnPP 278.09, r = 0.34
These models are capable of simulating the effects of a large variety of environmental conditions,
and can be used as dynamic tools for ecosystem risk assessment since they produce both qualitative
and quantitative results, allowing for comparisons of predictions with on-going observational
research and ecosystem monitoring.
1.7 Summary
Although the impact of nutrient pressures on biological quality is relatively well understood for
lakes in qualitative terms, there has been very limited development of quantitative dose-response
relationships, classication tools or models. This review identied a number of widely-recognised
strong relationships between physico-chemical measures of eutrophication and associated
biological responses. These can be split into two types of relationships: (1) primary responses
(phytoplankton, phytobenthos and macrophytes) to nutrient state, and (2), secondary responses
Jeppesen et al. (2003) Denmark, Norway
(temperate, boreal and artic zones)
Multiple regression model data from 466 lakes was used
to study impacts of lake trophic status on top-down control
by sh. The results indicated that planktivorous sh have a
more limited effect on the grazing control of phytoplankton
in oligotrophic lakes than in eutrophic lakes, despite similar
predator control of large-bodied zooplankton.
Hanson and Legged (1982) Canada Linear model for relationship between total phosphorus
(TP) and sh yield (Y) in two lakes:
LogY = 0.708 log TP + 0.774
Log Y = 0.072 TP + 0.792
Brmick and Lemcke (2003) NW Germany Data from 786 lakes was used to estimate sh yield
potential (FYP; kg.ha
-1
.a
-1
) for shallow and stratied lakes as
function of primary production of phytoplankton (PP) and
total phosphorus (TP)
lakes with PP<380: FYP= 6.315e
0.0062.pp
, r=0.47;
lakes with PP>380: FYP= 57.937 lnPP 278.09, r = 0.34
Lyytikainen and Jobling (1998) Northern Finland Study from an Artic lake Inari, indicated that an increase in
lake temperature (T) resulted in an immediate increase in
oxygen consumption (M-accl in mg kg
-1
h
-1
) tting to an
exponential model:
M-accl=46.53e
(0.086T)

Table 1.6 Examples of models for sh and environmental quality variables for lakes.
to primary production or production-related decreases in transparency and oxygen. The advantages
and disadvantages associated with developing these elements as indicators of eutrophication
pressures are briey summarised below.
Primary responses to nutrients. Phytoplankton abundance (most often measure of abundance
taken: chlorophyll-a) is highly sensitive to nutrients and integrates the response to a range of
possible limiting nutrients (P, N, Si). It is also a simple, practical measure of eutrophication impacts
that is widely adopted across EC Member States. Phytoplankton composition is sensitive to
nutrients and qualitatively well understood. The response to nutrients is, however, complicated by
alkalinity, stratication and lake morphometry. Macrophyte composition is sensitive to nutrients
and qualitatively well understood. Historical records are frequently available for dening reference
conditions. The response to nutrients is, however, affected by substrate quality and water level
uctuations (hydro-morphological pressures).
Secondary responses to nutrients. Fish biomass is sensitive to nutrient-related primary production,
but limited data are available. The relationship between phytoplankton abundance and
transparency is a simple, practical measure of eutrophication impacts that is widely adopted
across EC Member States, but it is affected by water colour and clay particles. The relationship
between macrophyte composition/abundance and transparency is qualitatively well understood,
but limited data are available. The relationship between benthic invertebrate composition and
oxygen is sensitive to eutrophication pressures but limited data are available. The relationship
between sh composition and oxygen is qualitatively well understood and sensitive to
eutrophication pressures but limited data are available.
All these primary and secondary responses are likely to have robust relationships with eutrophication
pressure, but data are limited for some of these parameters compared with others, which puts
practical limits on their selection as indicators for WFD ecological quality assessment. In particular,
data on phytobenthos, benthic invertebrates and oxygen conditions in lakes are very limited, as
are data on macrophyte and sh abundance. Although, some relationships, such as sh
composition oxygen condition, are relatively well understood, despite the data limitations. This
calls for an intensied effort to develop indicators in order to allow implementation of the WFD.
Availability of indicators and classication metrics is not only restricted by data, but also by the
limited research effort to develop Europe-wide bio-indicators for all the quality elements required
to be assessed by the WFD.
1.8 References
Adamus, P., and K. Brandt. 1990. Impacts on quality of inland wetlands of the United States: a survey of
indicators, techniques, and applications of community level biomonitoring data. EPA/600/3 - 90/073, US
EPA, pp. 406
Anonymous. 2003. Common Implementation Strategy for the Water Framework Directive (2000/60/EC).
Guidance No 10, River and lakes Typology, reference conditions and classication systems, pp. 94
Armitage, P.D., P.S. Cranston, and L.C.V. Pinder. 1995. The Chironomidae:Biology and Ecology of Non-Biting
Midges. Chapman and Hall, London
Asaeda, T., V.K. Trung, J. Manatunge, and T.V. Bon. 2001. Modelling macrophyte-nutrient-phytoplankton
interactions in shallow eutrophic lakes and the evaluation of environmental impacts. Ecological Engineering
16:341-357
Astrakhantsev, G.P., T.R. Minina, N.A. Petrova, V.N. Poloskov, and L.A. Rukhovets. 2003. Modeling Lake
Ladoga zoobenthos and studying its role in phosphorus exchange at the waterbed interface. Water
Resources 30:209220.
Austen, D.J., P.B. Bayley, and B.W. Menzel. 1994. Importance of the Guild Concept to Fisheries Research and
Management. Fisheries 19:12-20
Baines, S.T., and M. Pace. 1994. Relationships between suspended particulate matter and sinking ux along
a trophic gradient and implication for the fate of planktonic primary production. Canadian Journal of
Fisheries and Aquatic Sciences 51:25-36
Bal, D.H.M., H.M. Beije, Y.R. Hoogeveen, S.R.J. Jansen, and P.J. van der Reest. 1995. Handboek natuuraeltypen
in Nederland. IKC-rapport 11 IKC Natuubeheer, Wageningen
Barbour, M.T., J.B. Stribling, and J.R. Karr. 1995. Multimetric approach for establishing biocriteria and
measuring biological condition. In: W.S. Davis, and T.P. Simon (Eds.) Biological assessement and criteria:
Tools for water resource planning and decision-making, CRC Press, Boca Raton, pp. 63-77
Bennion, H. 1994. A Diatom-Phosphorus Transfer-Function for Shallow, Eutrophic Ponds in Southeast
England. Hydrobiologia, 276:391-410
Blindlow, I. 1992. Decline of charophytes during eutrophication comparison with angiosperms. Freshwater
Biology 28:9-14
Brmick, U., and R. Lemcke. 2003. Regional application of sh yeld estimation procedure to lakes in north-
east Germany. Limnologica 33:205-213
Bratli, J.L. 2000. Classication of the Environmental Quality of Freshwater in Norway. In: P. Heinonen, G.
Ziglio, and A. Van der Beken (Eds.), Hydrological and Limnological Aspects of Lake Monitoring, John Wiley
& Sons, Chichester
Brinkhurst, R.O. 1974. The benthos of lakes. St Martins Press, New York
Brodersen, K.P., and C. Lindegaard. 1999. Classication, assessment and trophic reconstruction of Danish
lakes using chironomids. Freshwater Biology 42:143-157
Brodin, Y.Y. 1982. Paleoecological studies of the recent developments of the lake Vaxjosjon: interpretetation
of eutrophication process through analysis of subfossil chironomids. Archiv Fur Hydrobiologie 93:313-326
Busnita, T., and G. Brezeanu. 1967. Ihtiofauna. In: Limnologia sectorului romanesc al Dunarii. A.R.S. Romania,
Bucuresti, pp. 325-370
Cairns, J. Jr., and J.R. Pratt. 1993. A history of biological monitoring using benthic macroinvertebrates. In:
D.M. Rosenberg, and V.H. Resh (Eds.), Freshwater Biomonitoring and Benthic Macroinvertebrates, Chapman
and Hall, pp. 10-27
Caneld, D.E. Jr., and L.M. Hodgson. 1983. Prediction of Secchi disc depths in Florida lakes: Impact of algal
biomass and organic color. Hydrobiologia 99:51-60
Cardoso, A.C., J. Duchemin, P. Magoarou, and G. Premazzi. 2001. Criteria for the Identication of Freshwaters
Subject to Eutrophication. Ofce for Ofcial Publications: European Commission, Luxembourg, pp. 90
Carlson, R.E. 1977. A trophic state index for lakes. Limnology and Oceanography 22:361-369
Carlson, R.E. 1980. More Complications in the Chlorophyll-Secchi Disk Relationship. Limnology and
Oceanography 25:379-382
Carpenter, S.R., and J.F. Kitchell. 1984. Plankton community structure and limnetic primary production.
American Naturalist 124:159-172
Carpenter, S.R., J.F. Kitchelland, and J.R. Hodgson. 1985. Cascading trophic interaction and lake productivity.
BioScience 35:634-639
Carvalho, L., H. Benninon, A. Darwell, I. Gunn, A. Lyle, D. Monteithand M. Wade. 2002. Physico-chemical
conditions for supporting different levels of biological quality for the Water Framework Directive for
freshwater. R&D Technical Report to the Environment Agency, December 2002, pp. 145
Carvalho, L., S. Maberly, L. May, C. Reynolds, M. Hughes, R. Brazier, L. Heathwaite, S. Liu, J. Hilton, D. Hornby,
H. Bennion, A. Elliott, N. Willby, R. Dils, L. Pope, G. Phillips, and I. Fozzard. 2004. Risk Assessment Methodology
for Determining Nutrient Impacts in Surface Freshwater Bodies. R&D Technical Report P2-260/9 to the
Environment Agency and Scottish Environment protection Agency, July 2004. pp. 171
Coops, H., J. Hanganu, M. Tudor, and W. Oosterberg. 1999. Classication of Danube Delta lakes based on
aquatic vegetation and turbidity. Hydrobiologia 415:187-191
Danilov, R.A., and N.G.A. Ekelund. 2000. The use of epiphyton and epilithon data as a base for calculating
ecological indices in monitoring of eutrophication in lakes in central Sweden. Science of the Total Environment
248:63-70
Davis, J.C. 1975. Minimal dissolved oxygen requirements of aquatic life with emphasis on Canadian species:
a review. Journal of the Fisheries Research Board of Canada 32: 2295-2332
Davis, G.J., and M.M. Brinson. 1980. Responses of submersed vascular plant communities to environmental
change. U.S. Department of the Interior Report FWS/OBS-79/33
Dillon, P.J., and F.H. Rigler. 1974. A test of a simple nutrient budget model predicting the phosphorus
concentration in lake water. Journal of the Fisheries Research Board of Canada 31:1771-1778
Dinsmore, W.P., and E.E. Prepas. 1997. Impact of hypolimnetic oxygenation on profundal macroinvertebrates
in a eutrophic lake in central Alberta. I. Changes in macroinvertebrate abundance and diversity. Canadian
Journal of Fisheries and Aquatic Sciences 54:21572169
Dinsmore, P.W., G.J. Scrimgeour, and E.E. Prepas. 1999. Empirical relationships between profundal
macroinvertebrate biomass and environmental variables in boreal lakes of Alberta, Canada. Freshwater
Biology 41:91-100
Diudea, M., S. Tudor, and A. Igna. 1986. Toxicologie acvatica. Cluj Napoca. pp. 183
Dodkins, I., B. Rippey, and P. Hale. 2003. The advantage of metrics for aquatic macrophyte assessment in
Northern Ireland. Temanord report 547:29-34.
Doudoroff, P., and D.L. Shumway. 1970. Dissolved oxygen requirements of freshwater shes. FAO Technical
Paper 86, pp. 291
EIFAC. 1973. Water Quality Criteria for European Freshwater Fish. Report on Dissolved Oxygen and Inland
Fisheries. EIFAC Tech. Paper 86, pp. 19
Ellenberg, H. 1988. Vegetation ecology of central Europe. Cambridge University Press, Cambridge
Elliott, J.A., C.S. Reynolds, and A.E. Irish. 2001. An investigation of dominance in phytoplankton using the
PROTECH model. Freshwater Biology 46:99-108
Fairchild, G.W., R.L. Lowe, and W.B. Richardson. 1985. Algal Periphyton Growth on Nutrient-Diffusing
Substrates - an In situ Bioassay. Ecology 66:465-472
Francis, D.R., and T.C. Kane. 1995. Effect of substrate on colonization of experimental ponds by Chironomidae
(Diptera). Journal of Freshwater Ecology 10:57-63
Gassner, H., G. Tischler, and J. Wanzenbck. 2003. Ecological integrity assessment of lakes using sh
communities - suggestions of new metrix development in two Austrian prealpine lakes. International Review
of Hydrobiology 88:635-652
German States Association for Water. 1998. Gewsserbewertung stehende Gewsser, Berlin, Kulturbuch
Verlag, pp. 74
Gibson, G., R. Carlson, J. Simpson, E. Smeltzer, J. Gerritson, S. Chapra, S. Heiskary, J. Jones, and R. Kennedy.
2000. Nutrient Criteria Technical Guidance Manual: Lakes and Reservoirs. US EPA, Washington DC, pp. 231
Goedkoop, W., and R.K. Johnson. 1996. Pelagic-benthic coupling:profundal benthic community response to
spring diatom deposition in mesotrophic Lake Erken. Limnology and Oceanography 41:636-647
Hanson, J.M., and W.C. Legget. 1982. Empirical prediction of sh biomass and yield. Canadian Journal of
Fisheries and Aquatic Sciences 39:257263
Heegaard, E., H.H. Birks, C.E. Gibson, S.J. Smithand S. Wolfe-Murphy. 2001. Species-environmental relationships
of aquatic macrophytes in Northern Ireland. Aquatic Botany 70:175-223
Heinis, F., and T. Crommentuijn. 1992. Behavioural response to changing oxygen concentration of deposit
feeding chironomid larvae (Diptera) of littoral and profundal habitats. Archiv Fur Hydrobiologie 124:173-185
Heinonen, P. 2000, Water quality classication in Finland. In: P. Heininen, G. Ziglio, and A. Van der Beken
(Eds.), Hydrological and limnological aspects of lake monitoring. John Wiley & Sons, Chichester, pp. 345-350
Hellawell, J.M. 1986. Biological indicators of freshwater pollution and environmental management. Elsevier,
London
Herreid, C.F. 1980. Hypoxia in invertebrates. Comparative Biochemistry and Physiology 67:311-320
Hillbricht-Ilkowska, A. 1993. Phosphorus loading to lakes of Suwalki Landscape Park (north-eastern Poland)
and its relation to lake trophy indices. Polish Journal of Ecology 41:221-235
Hough, R.A., and M.D. Forwall. 1988. Interactions of inorganic carbon and light availability as controlling
factors in aquatic macrophyte distribution and productivity. Limnology and Oceanography 33:1202-1208
Hynes, H.B.N. 1960. The biology of polluted waters. Liverpool University Press, Liverpool
Janse, J.H., and L. Van Liere. 1995. PC-Lake, a modelling tool for the evaluation of lake restoration scenarios.
Water Science Technology 31:371-374
Jeppesen, E., J.P. Jensen, M. Sondergaard, T. Lauridsenand F. Landkildehus. 2000. Trophic structure,
species richness and biodiversity in Danish lakes: changes along a phosphorus gradient. Freshwater biology
45:201-218
Jeppesen, E., J.P. Jensen, C. Jensen, B. Faafeng, D.O. Hessen, M. Sondergaard, T. Lauridsen, P. Brettumand K.
Christoffersen. 2003. The impact of nutrient state and lake depth on top-down control in the pelagic zone of
lakes: a study of 466 lakes from the temperate zone to the arctic. Ecosystems 6:313-325
Jonasson, P.M. 1996. Limits for the life in the lake ecosystem. Verhandlungen der internationale Vereiningung
fur Theoretische und Angewandte Limnologie 20:1-33
Kajak, Z. 1987. Determinants of maximum biomass of benthic Chironomidae (Diptera). Entomologica
Scandinavica Supplements 29:303-308
Kanninen, A., J. Leka, V.M. Vallinkosk, and R. Ilvonen. 2003. Aerial photograph interpretation and eld
surveys of aquatic macrophytes in assessing the ecological status of small boreal lakes preliminary results.
TemaNord 547:123-126.
Kansanen, P.H., J. Aho, and L. Paasivirta, 1984. Testing the benthic lake type concept based on chironomid
associations in some Finish lakes using multivariate statistical method. Annales Zoologici Fenici 21:55-76
Karr, J.R. 1981. Assessment of biotic integrity using sh communities. Fisheries 6:21-27
King, L., P. Barker, and R.I. Jones. 2000. Epilithic algal communities and their relationship to environmental
variables in lakes of the English Lake District. Freshwater Biology 45:425-442
King, R.S., and C.J. Richardson. 2003. Integrating bioassessment and ecological risk assessment: An approach
to developing numerical water-quality criteria. Environmental Management 31:795-809
Kirchner, W.R., and P.J. Dillon. 1975. An emperical method of estimating the retention of phosphorus in lakes.
Water Resources Research 11:182-183
Kirk, J.T.O. 1994. Light and photosynthesis in aquatic ecosystems. Cambridge University Press, New York
Kornijow, R. 1997. The impact of predation by perch on the size-structure of Chironomus larvae-the role of vertical
distribution of the prey in the bottom sediments, and habitat complexity. Hydrobiologia 342/343:207-213
Lang, C. 1998. Using oligochaetes to monitor the decrease of eutrophication: the 1982-1996 trend in Lake
Geneva. Archiv Fur Hydrobiologie 141:447-458
Lang, C. 1999. Contrasting responses of oligochaetes (Annelida) and chironomids (Diptera) to the abatement
of eutrophication in Lake Neuchatel. Aquatic Sciences 61:206-214
Lang, Y., J.M. Melackand J. Wang. 1981. Primary production and sh yield in Chinese ponds and lakes.
Transactions of the American Fisheries Society 110:346-350
Larsen, D.P., and A.T. Herlihy. 1998. The dilemma of sampling streams for macroinvertebrate richness. Journal
of the North American Benthological Society 17:359-366
Legendre, P., and L. Legendre. 1998. Numerical ecology. Elsevier, Amsterdam
Leka, J., K. Valta-Hulkkonen, A. Kanninen, S. Partanen, S. Hellsten, A. Ustinova, R. Ilvonen, and O. Airaksinen.
2003. Vesimakrofyytit jrvien ekologisen tilan arvioinnissa ja seurannassa. Maastomenetelmien ja
ilmakuvatulkinnan kyttkelpoisuuden arviointi Life Vuoksi-projektissa. Alueelliset ympristjulkaisut, pp. 312
Ligtvoet, W., and M.P. Grimm. 1992. Fish in clear water. Fish-stock development and management in Lake
Volkerak/Zoom. How an estuary changed into freshwater lake: The water management of Lake Volkerak/
Zoom. J.C. Hooghart, and C.W.S. Posthumus. Delft. Proceedings and Information, CHO-TNO, 46:69-84
Los, F.J., M. Ouboter, and R.J. Fokkink. 1997. Onderzoek naar de slibdynamiek van de Loosdrechtse Plassen
met behulp van mathematische modellen. WL Delft Hydraulics, Delft
Lyche Solheim, A.L., T. Andersen, P. Brettum, T. Bkken, T. Bongard, F. Moy, T. Kroglund, F. Olsgard, B. Rygg,
and E. Oug. 2004. BIOKLASS - Klassisering av kologiske status i norske vannforekomster: Forslag til
aktuelle kriterier og forelpige grenseverdier mellom god og moderat kologisk status for utvalgte elementer
og pvirkninger. Norsk institutt for vannforskning (NIVA), Oslo, Rapport LNR 4860-2004, pp. 67
Lyytikainen, T., and M. Jobling. 1998. The effect of temperature uctuations on oxygen consumption and
ammonia excretion of underyearling Lake Inari Arctic charr. Journal of Fish Biology 52:1186-1198
Mazumder, A., and K.E. Havens. 1998. Nutrient-chlorophyll-Secchi relationships under contrasting grazer
communities of temperate versus subtropical lakes. Canadian Journal of Fisheries and Aquatic Sciences
55:1652-1662
McQueen, D.J., J.R. Post, and E.L. Miles. 1986. Trophic relationship in freshwater pelagic ecosystems. Canadian
Journal of Fisheries and Aquatic Sciences 43:1571-1581
McQueen, D.J., M.R.S. Johannes, J.R. Post, T.J. Stewart, and D.R.S. Lean. 1989. Bottom-up and top-down
impacts on freshwater pelagic community structure. Ecological Monographs 59:289-309
Moss, B., P. Johnes, and G. Phillips. 1996. The monitoring of ecological quality and classication of standing
water in temperate regions: a review and proposal based on a worked scheme for British waters. Biological
Review 71:301-339
Moss, B., D., Stephen, C. Alvarez, E. Becares, W. Van De Bund, S.E. Collings, E. Van Donk, E. De Eyto, T.
Feldmann, C. Fernndez-Alez, M. Fernndez-Alez, R.J.M. Franken, F. Garca-Criado, E.M. Gross, M. Gyllstrm,
L.A. Hansson, K. Irvine, A. Jrvalt, J.P. Jensen, E. Jeppesen, T. Kairesalo, R. Kornijw, T. Krause, H. Knnap, A.
Laas, E. Lill, B. Lorens, H. Luup, M.R. Miracle, P. Nges, T. Nges, M. Nyknen, I. Ott, W. Peczula, E.T.H.M.
Peeters, G. Phillips, S. Romo, V. Russell, J. Saluje, M. Scheffer, K. Siewertsen, H. Smal, C. Tesch, H. Timm, L.
Tuvikene, I. Tonno, T. Virro, E. Vicente, and D. Wilson. 2003. The determination of ecological status in shallow
lakes - a tested system (ECOFRAME) for implementation of the European Water Framework Directive. Aquatic
Conservation: Marine and Freshwater Ecosystems 13:507-549
Muhammetoglu, A., and S. Soyupak. 2000. A three-dimensional water quality-macrophyte interaction model
for shallow lakes. Ecological Modeling 133:161-180
Murphy, K.J. 2002. Plant communities and plant diversity in softwater lakes of northern Europe. Aquatic
Botany 73:287-324
Murphy, K.J., and M.J.M. Hootsmans. 2002. Predictive modelling of aquatic community attributes: biomass,
biodiversity, biointegrity and biomonitoring. Acta Limnologica Brasiliensia 14:43-60
Nabben, R.H.M., C.W.P.M. Blom and L.A.C.J. Voesenek. 1999. Resistance to complete submergence in Rumex
species with different life histories: the inuence of plant size and light. The New Phytologist 144:313-321
Navodaru, I., A.D. Buijse and M. Staras. 2002. Effects of Hydrology and Water Quality on the Fish Community
in Danube Delta Lakes. International Review of Hydrobiology 87:329-348
Newman, P.J. 1998. Classication of surface water quality. Review of schemes used in EC member states.
Heinemann Professional Publishing Ltd: Oxford
Nikolski, G.V. 1962. Ecologia pestilor, Bucuresti, pp. 362
Noges, P., L. Tuvikene, T. Feldmann, I. Tnno, H. Knnap, H. Luup, J. Saluje, and T. Nges. 2003. The role of
charophytes in increasing water transparency: a case study of two shallow lakes in Estonia. Hydrobiologia
506:567-573
OECD. 1982. Eutrophication of waters: monitoring, assessment and control. Organisation of Economic Co-
operation and Development, Paris, pp. 154
Oosterberg, W., M. Sataras, A.D. Buijse, L. Bogdan, A. Constantinescu, H. Coops, J. Hanganu, B.W. Ibelings,
G.A.M. Menting, I. Navodaru, L. Trk. 2000. Ecological gradients in the Danube Delta lakes. Present state
and man-induced changes. Lelystad, RIZA report 2000.015
Palmer, M.A., S.L. Bell and I. Buttereld. 1992. A botanical classication of standing waters in Britain:
applications for conservation and monitoring. Aquatic Conservation: Marine and Freshwater Ecosystems
2:125-143
Phillips, G.L., D. Eminson, and B. Moss. 1978. A mechanism to account for macrophyte decline in progressively
eutrophicated freshwaters. Aquatic Botany 4:103-126
Portielje, R., and D.T. Van der Molen. 1999. Relationships between eutrophication variables: from nutrient
loading to transparency. Hydrobiologia 408:375-387
Premazzi, G., and G. Chiaudani. 1992. Ecological Quality of Surface Waters. Commission of the European
Communities, pp. 124
Rasmussen, J.B., and J. Kalff. 1987. Empirical models for zoobenthic biomass in lakes. Canadian Journal of
Fisheries and Aquatic Sciences 44:990-1001
Real, M. 2000. Chironomus species (Diptera, Chironomidae) in the profundal benthos of Spanish reservoirs
and lakes: factors affecting distribution patterns. Freshwater Biology 43:1-18
Regier, H.A., R.L. Welcomme, and R.J. Steedman. 1998. Rehabilitation of degraded river ecosystems.
Proceedings of the International Large River Symposium (LARS), Canadian Special Publication of Fisheries
and Aquatic Sciences 106:86-97
Reynolds, C.S. 1984. The ecology of freshwater phytoplankton. Cambridge University Press, Cambridge
Reynolds, C.S. 1999. Modelling phytoplankton dynamics and its application to lake management.
Hydrobiologia 396:123-131
Reynolds, C.S., V. Huszar, C. Kruk, L. Naselli-Flores, and S. Melo. 2002. Towards a functional classication of
the freshwater phytoplankton. Journal of Plankton Research 24:417-428
Root, R.B. 1967. The niche exploitation pattern of the blue-gray gnatcatcher. Ecological Monographs
37:317-350
Rrslett, B. 1991. Principle determinants of aquatic macrophyte richness in northern european lakes. Aquatic
Botany 39:173-193
Rosenberg, D.M. 1992. Freshwater biomonitoring and chironomidae. Netherlands Journal of Aquatic Ecology
26:101-122
Saether, O.A. 1979. Chironomid communities as water quality indicators. Holarctic Ecology, 2:65-74
Sand-Jensen, K., and M. Sndergaard. 1981. Phytoplankton and epiphyte development and their shading
effect on submerged macrophytes in lakes of different nutrient status. Internationale Revue der gesamten
Hydrobiologie 66: 529-552
Sas, H. 1989. Lake restoration by reduction of nutrient control. Academia Verlag Richartz, St. Augustin,
pp. 497
Schaminee, J.H.J., E.J. Weeda and V. Westhoff. 1995. Die vegetatie van Nederland. Deel 2. Plantengemeeschapen
van watersen moerassen en natte Leiden. Opulus Press, Uppsala, Leiden, pp. 357
Schaumberg, J., C. Scranz, G. Hofmann, D. Stelzer, and S. Schneider. 2004. Macrophytes and phytobenthos
as indicators for ecological status in German lakes - contributions to the implementation of the WFD.
Limnologica
Scheffer, M., S.H. Hosper, M.L. Meijer, and B. Moss. 1993. Alternative equilibria in shallow lakes. Trends in
Ecology and Evolution 8:275-279
Scheffer, M. 1998. Ecology of Shallow Lakes. Chapman and Hall, London
Sculthorpe, C.D. 1967. The Biology of Aquatic Vascular Plants. Edward Arnold Editor, London
Sondergaard M, E. Jeppesen, J.P. Jensen, E. Bradshaw, H. Skovgaard, and S. Grunfeld. 2003. Vandrammedirektivet
og danske soer Undertitel: Del 1: Sotyper, referencetilstand og okologiske kvalitetsklasser. Danmarks
Miljministeriet, Copenhagen
Spence, D.H.N., and J. Chrystal. 1970. Photosynthesis and zonation of freshwater macrophytes. I. Depth
distribution and shade tolerance. New Phytologist 69:205-215
Swedish Environmental Protection Agency. 2000. Lakes and Watercourses: Environmental Quality Criteria.
Report 5050, pp. 102
Takamura, K., and T. Iwakuma. 1990. Fate of macrozoobenthos in hypertrophic lakes: in situ rearing experiments
with the larvae of Tokunagayusurika akamusi (Chironomidae). Archiv Fur Hydrobiologie, 119:65-78
Tilzer, M.M. 1988. Secchi Disk-Chlorophyll Relationships in a Lake with Highly Variable Phytoplankton
Biomass. Hydrobiologia 162:163-171
US EPA. 1998. Lake and Reservoir Bioassessment and Biocriteria Technical Guidance Document (EPA 841-B-
98-007). Ofce of Water, Washington DC
US-EPA. 1986. Ambient Water Quality Criteria for Dissolved Oxygen (Freshwater). EPA 440/5/86-003, U.S.
Environmental Protection Agency, Washington, D.C.
US-EPA. 1997. Ambient Water Quality Criteria for Dissolved
Van de Bund, W., and D. Groenendij. 1994. Seasonal dynamics and burrowing of littoral larvae in relation to
competition and predation. Archiv Fur Hydrobiologie, 132:213-225
Van de Bund, W., and E. Van Donk. 2002. Short-term and long-term effects of zooplanktivorous sh removal
in a shallow lake: a synthesis of 15 years of data from Lake Zwemlust. Freshwater Biology 47:2380-2387
Van den Berg, M.S., H. Coops, M-L. Meijer, M. Scheffer, and J. Simons. 1997. Clear water associated with a
dense Chara vegetation in the shallow and turbid lake Veluwemeer. In: E. Jeppesen, M. Sndergaard,
M.Sndergaard, and K. Christoffersen (Eds.), The Structuring Role of Submerged Macrophytes in Lakes.
Springer, New York, pp. 339-352
Van den Berg, M.S., M. Scheffer, E. Van Nes, and H. Coops. 1999. Dynamics and stability of Chara sp. and
Potamogeton pectinatus in a shallow lake changing in eutrophication level. Hydrobiologia 409:335-342
Van den Berg, M.S., W. Joosse, and H. Coops. 2003. A statistical model predicting the occurrence and
dynamics of submerged macrophytes in shallow lakes in the Netherlands. Hydrobiologia 506-509: 611-623
Van Donk, E., and W. Van de Bund. 2002. Impact of submerged macrophytes including charophytes on
phyto- and zooplankton communities: allelopathy versus other mechanisms. Aquatic Botany 72: 261-274
Van Nes, E.K. 2002. Controlling complexity in individual-based models of aquatic vegetation and sh
communities. Wageningen University, pp. 200
Verdonschot, P.F.M. 1996. Oligochaetes and eutrophication; an experiment over four years in outdoor
mesocosms. Hydrobiologia 334:169-183
Weisner, S.E.B., J.A. Strand, and H. Sandsten. 1997. Mechanisms regulating abundance of submerged
vegetation in shallow eutrophic lakes. Oecologia 109:592-599
Wiederholm, T. 1980. Using benthos for lake montioring. Journal of the Water Pollution Control Federation,
52:537-547
Willby, N.J., V.J. Abernethy, and B.O.L. Demars. 2000. Attribute-based classication of European hydrophytes
and its relationship to habitat utilization. Freshwater Biology 43:43-74
Wunsam, S., and R. Schmidt. 1995. A diatom-phosphorus transfer function for Alpine and pre-alpine lakes.
Memorie del Istituto Italiano di Idrobiologia 53:85-99
2 Nutrients and eutrophication in coastal and
transitional waters
Pirjo Kuuppo, Anouk Blauw, Flemming Mhlenberg, Hanne Kaas, Peter Henriksen,
Dorte Krause-Jensen, Gunni rtebjerg, Saara Bck, Paul Erftemeijer, Miguel Gaspar,
Susana Carvalho, Anna-Stiina Heiskanen
2.1 Introduction
Coastal ecosystems receive anthropogenic nutrients from direct sources on the coastline or
indirectly from riverine export from catchments and from the atmosphere. Increasing and long-
term nutrient loading from anthropogenic sources has caused eutrophication of coastal ecosystems,
the symptoms of which are: excess accumulation of phytoplankton biomass in the water column
and bottom sediments, depletion of oxygen in bottom waters, increased frequency of noxious
algal blooms, increased turbidity, deterioration of coastal food webs and reduction of biodiversity
(Figure 2.1).
At present, many coastal and estuarine ecosystems receive high rates of nutrient loading. Estimates
of the increase of phosphorus and nitrogen loading to estuarine systems range between 2-6-fold
and 1.5-14-fold compared to the turn of the 20
th
century (Conley, 1999; Cloern, 2001; Carstensen
et al., 2004). Atmospheric loading of nitrogen to the north Atlantic area has increased 5-10-fold
since industrialisation (Paerl and Whitall, 1999), and atmospheric deposition of nitrogen may
represent 10-50% of the anthropogenic N ux to water surfaces (Pryor and Barthelmie, 2000).
Nutrient loading from diffuse sources (e.g. arable and pastoral agricultural sources and forestry)
to the coastal waters is more difcult to estimate.
When considering eutrophication in coastal areas, of equal relevance as the concentrations of the
principal nutrients: nitrogen (N), phosphorus (P) and silicate (Si), are their ratios. As nitrate
concentrations increase due to anthropogenic loading, systems are moving towards not only
higher N:P ratios, but also lower Si:N ratios. As a consequence, coastal areas are likely to become
P and Si limited worldwide (Justic et al., 1995a,b), including northern Europe (Conley et al., 1993;
Jickells, 1998; Kristiansen and Hoell, 2002; Sommer et al., 2002; Turner et al., 2003).
In addition, other human activities such as damming of rivers leads to increased silicate retention
and less silicon export to the sea (Humborg et al., 2000a). During the last decades, N: P ratios
have increased dramatically in Dutch coastal waters (deJonge et al., 2002), Danish coastal areas
(Jorgensen, 1996; Kaas et al., 1996), and in the Black Sea (Shtereva et al., 1999).
Ecological indicators
Dale and Beyeler (2001) suggest a set of criteria that ecological indicators should meet when used
for assessing the condition of the environment. According to their criteria, the indicators should
be straightforward and relatively inexpensive to measure;
be sensitive to stresses of the system;
respond to stress in a predictable manner;
NUTRIENTS AND EUTROPHICATION IN COASTAL AND TRANSITIONAL WATERS | 33
be anticipatory, meaning that a change in the indicator should be measurable before a
substantial change in ecosystem integrity occurs;
predict changes that can be averted by management actions;
be integrative;
have a known response to disturbances, anthropogenic stresses and change over time, and
exhibit low variability in response.
It is evident that not many candidates for an indicator full all these criteria. Therefore, it may be
necessary to apply a suite of indicators that together could full the criteria and represent complex
ecological systems.
Coastal eutrophication and the Water Framework Directive
The Water Framework Directive outlines biological quality elements that should be used in the
classication of the ecological status of coastal and transitional waters. These elements include
macrophytes (macroalgae and angiosperms), benthic invertebrate fauna, and phytoplankton for
Figure 2.1 Schematic
presentation of relationships
between nutrient pressures and
biological parameters in coastal
and transitional waters. Red
arrow indicate positive/
stimulating effect, and blue arrow
indicate negative/inhibiting
effect. Redrawn and modied
from the report of the OSPAR
Commission (2003).
Zoobenthos and sh kills
benthic community structure
Transboundary
nutrient uxes
Supporting environmental factors:
physical forcing, hydrodynamics,
climatic or weather conditions,
Transboundary transport
Nutrient input
N,P, Si
Increased concentrations of N, P, Si
N:P, N:Si ratios
Nuisance
algal
blooms
Increase in
phytoplankton
biomass
Increase in
turbidity
Decrease in
light regime
Shift from long-lived to
short-lived macrophyte
species and reduced
depth distribution
DIRECT EFFECTS
Macrophytobenthos biomass
and primary production
Toxins
Increase in
primary production
Organic
matter
Increase
bacteria
Oxygen
deciency
INDIRECT EFFECTS
both coastal and transitional waters, as well as sh for transitional waters only. In Tables 1.2.3.
and 1.2.4 of Annex V of the WFD, the normative denitions for high, good, and moderate status
are described in narrative terms. These tables provide more specic denitions on the quality
elements that should be used in the classication.
For the phytoplankton, the following quality elements need to be assessed:
Phytoplankton composition and abundance
Average phytoplankton biomass and water transparency
Frequency and intensity of phytoplankton blooms
According to the WFD, declining ecological quality of coastal and transitional waters is characterised
by a slight (good status) or moderate (moderate status) disturbance in the composition and
abundance of phytoplankton taxa, slight or moderate changes in the biomass, and slight or
moderate increases in the frequency and duration of phytoplankton blooms. Such disturbances are
evaluated against reference conditions set for each coastal or transitional water type.
For the macrophytes, the composition and abundance of macroalgae and angiosperms need to be
assessed. The WFD stipulates that at high ecological status all disturbance-sensitive macroalgal
and angiosperm taxa, which are normally associated with undisturbed conditions, must be present.
Also the levels of macroalgal cover and angiosperm abundance must be consistent with undisturbed
conditions. Similarly, to full good ecological status most disturbance-sensitive macroalgal and
angiosperm taxa associated with undisturbed conditions must be present and the levels of
macroalgal cover and angiosperm abundance must only show slight signs of disturbance.
Likewise, the composition and abundance of benthic invertebrate fauna needs to be assessed.
A decline in ecological quality is identied when the diversity of invertebrate taxa is outside the
range normally observed in high or reference conditions for each coastal type.
2.2 Phytoplankton
Phytoplankton biomass
Phytoplanktonic primary producers are the rst organisms to respond to elevated nutrient
concentrations in their environment. Most phytoplankton species, with the exception of
picoplanktonic algae, respond positively and predictably to nutrient enrichment in all European
coastal areas (Olsen et al., 2001). Phytoplankton react to increasing ambient nutrient availability
with increased growth i.e. primary production. Primary production is normally measured by the
incorporation of radioactively labelled carbon (
14
C) or by the O
2
technique, and it normally shows
a rapid and positive response to nutrient increase (Smayda, 2004).
While primary production is directly linked to nutrient availability, biomass and composition of
phytoplankton are determined by the balance between growth and loss processes in the ecosystem.
The most important loss processes are food web interactions (grazing by zooplankton and benthic
animals) and hydrodynamic dispersion. The top-down control of phytoplankton and dispersion are
signicant determinants of the fate of nutrient input and subsequent phytoplankton growth in
coastal areas (Monbet, 1991; Kivi et al., 1993; Le Pape et al., 1996; Thomann and Linker, 1998;
Olsen et al., 2001; Perez-Ruzafa et al., 2002). Mussel beds on the bottom of estuaries can act as
a strong buffer against biomass accumulation, since the extra phytoplankton biomass is effectively
grazed by these lter-feeders (Le Pape and Menesguen, 1997; Meeuwig et al., 1998).
High phytoplankton biomass results in an increased amount of organic matter to be degraded by
bacteria and benthic fauna, which may lead to hypoxia or anoxia of bottom waters (Richardson
and Jrgensen, 1996). Long-lasting eutrophication may result in recurrent or even permanent
oxygen decit on bottoms, and collapse of bottom fauna. In anoxic conditions the nutrients that
have accumulated in sediments of estuaries and enclosed seas, such as the Baltic Sea, are released
again into the water column thus contributing further to eutrophication. However, the development
of hypoxia depends largely on hydrography in the area. Sheltered, non-tidal estuaries with
restricted water exchange are more vulnerable to oxygen deciency than tidal open estuaries,
where water exchange is substantial.
Nutrient limitation of phytoplankton
When the relationship between phytoplankton biomass and nutrient status is analysed, it is
important to recognise gross limitation patterns. In turbid estuaries such as the Western Scheldt,
light limitation of primary production may prevail through all seasons (Cloern, 1999). In such
environments, nutrient concentrations or their ratios provide little information. The concept of
nutrient limitation (Liebigs law of the minimum) has been extrapolated from agricultural use to
the aquatic sciences (de Baar, 1994). According to the concept, the growth of algae or higher
plants is limited by the nutrient that is least available relative to those required. In phytoplankton,
the optimal cellular ratio of nitrogen and phosphorus (N:P) has been found to be on average
ca. 7 by weight (Redeld, 1958). Phytoplankton growth is generally considered to be limited by
one of the major nutrients: nitrogen (N), phosphorus (P), or silica (Si) in the case of diatoms for
which it is a macronutrient. In principle, if the molar ratio of nitrogen and phosphorus available
for primary production deviates from the N:P ratio16, primary production is limited either by
nitrogen (N:P < 16) or phosphorus (N:P >16).
The ratio of dissolved inorganic nitrogen and phosphorus (DIN:DIP ratio), can be used as an
indicative number of potential nutrient limitation. Its use should, however, be limited to N:P ratios
of winter concentrations due to the fact that during the growing season inorganic nutrient
concentrations are usually below the detection limit, as most of the nutrients are incorporated
into the particulate and dissolved organic nutrient pools. The concept and interpretation of
nutrient ratios differs among countries and research groups. For example, in Dutch literature the
following denitions have been used: winter DIN concentrations divided by winter DIP
concentrations (de Vries et al., 1998); time series of total nitrogen concentrations divided by total
phosphorus concentrations (Philipart and Herman, 1998), and constant DIN:DIP concentrations
of the nutrient additions in continuous phytoplankton cultures (Riegman et al., 1998). The ratio
of total nitrogen and phosphorus (TN:TP) has been used to indicate nutrient limitation, but it is
not necessarily a good indicator, because total nutrient pools include a large fraction of dissolved
organic compounds that cannot be directly utilised by phytoplankton.
Although ambient nutrient ratios provide an indication of nutrient limitation, experimental
studies such as nutrient addition bioassays and mesocosm experiments, reveal direct responses of
phytoplankton communities to nutrient additions (Lignell et al., 1992; Kivi et al., 1993; Downing
et al., 1999; Seppl et al., 1999; Duarte et al., 2000).
Marine coastal open waters are primarily nitrogen limited (Ryther and Dunstan, 1971) in contrast
to freshwaters where phosphorus is regarded as the main limiting nutrient. An exception to this
pattern is the Mediterranean, which has been found to be phosphorus-limited (Krom et al., 1991;
Thingstad et al., 1998; Diaz et al., 2001). In estuaries, coastal transition areas and river plumes,
the limitation pattern can shift from P to N limitation, or even co-limitation, towards the open sea,
as well as shift between N and P limitation seasonally (Kivi et al., 1993; Conley, 1999; Prins et al.,
1999). In addition, it has recently been shown, that iron (Fe) can limit phytoplankton growth in
eutrophic coastal waters (Stal et al., 1999; Zhang, 2000), as well as in open oceans. Heavily
nitrogen-impacted estuaries and coastal waters are moving from nitrogen limitation towards
secondary nutrient (P, Si, Fe) limitation (Paerl, 1998).
Nitrogen vs. phytoplankton biomass
Positive relationships between total N and phytoplankton biomass have been shown in previous
studies. One example where total nitrogen concentration (TN) and phytoplankton chlorophyll-a
(and transparency) show clear dose-response relationships is the long-term study in Himmerfjrden,
Baltic Sea, where nitrogen and phosphorus loading were altered in a large-scale experiment
(Figure 2.2; Elmgren and Larsson, 1997).
Analog examples of simple relationships between chlorophyll-a and total nitrogen concentration
include studies of small non-tidal Baltic estuaries (Meeuwig et al., 2000) and Scottish coastal
waters (Gowen et al., 1992). The relationship between nutrient load and the response of
phytoplankton biomass is not uniform in different European coastal areas. For example, a series of
Figure 2.2 Relationship between total nitrogen (TN) and (A) Secchi depth in Himmerfjrden, Baltic Sea
(Tett et al., 2003), and (B) Chlorophyll-a in small Finnish bayments, Baltic Sea (Meeuwig et al., 2000).
1.0
0.9
0.8
0.7
0.6
0.5
0.4
2
1
1.2 1.3 1.4 1.5 1.6 2.4 2.6 2.8 3.0 3.2 3.4
log
10
TN mol l
-1
log

(TN (mgm
-3
))
l
o
g

(
C
h
l

(
m
g
m
-
3
)
)
l
o
g
1
0

(
S
e
c
c
h
i
,

m
)
A
y = -1.32x + 2.5
R
2
= 0.89
B
lChl = -2.10 + 1.13 lTN
n = 19 p = 0.0004
r2 = 0.53 SEE = 0.213
coordinated mesocosm studies revealed a signicantly lower biomass yield per unit nutrient
addition for coastal communities in the Mediterranean compared to Atlantic or Baltic (Figure 2.3).
The differences between the systems were attributed to differences in top-down control by
zooplankton, and vertical export (Olsen et al., 2001; Duarte, 2004).
Monitoring phytoplankton biomass - metrics
Phytoplankton biomass is a quality element that has been routinely used in coastal monitoring
for decades. Therefore, long-term observations are available in many coastal areas, and these
observations are often comparable over time. Metrics generally used for phytoplankton biomass
include either microscopically determined abundance (i.e. cells per liter), biomass (fresh weight, or
carbon biomass; mg or g per liter), or chlorophyll-a (mg or g per liter).
The examination of samples by microscope allows analysis of the community structure of
phytoplankton, revealing possible shifts in species composition or functional groups as a response
to environmental changes that may be caused by pressures. On the other hand, such an approach
is time-consuming and requires specic expertise; especially, converting the cell numbers to
volume and fresh weight, which requires laborious measurement of individual cells. Moreover,
the conversion of biovolume to carbon biomass preferably requires species-specic carbon
conversion factors that are not always available, if ever even measured. The development of
taxonomic knowledge and new methodologies based on molecular biology has led to changes in
the taxonomy of phytoplankton, with new taxons found and old renamed. Therefore, changes in
phytoplankton taxonomy have to be kept in mind when long-term data is being analysed, and
when historical data is used to dene reference conditions.
The concentration of chlorophyll-a, i.e. the amount of primary photosynthetic pigment in
phytoplankton, is measured routinely in coastal monitoring programmes, since it reects total
phytoplankton biomass relatively well. Chlorophyll analysis is rapid, easy to perform and
Figure 2.3 Example of different dose response
relationship between nitrogen and phytoplankton
biomass in European coastal areas. MesoNor,
MesoMed and MesoBal are data from mesocosm
experiments of the project COMWEB in North
Atlantic, Mediterranean, and Baltic, respectively.
Lagoon data are from the experimental fertilization
of open North Atlantic coast; North Sea is open
North Sea coast (Olsen et al., 2001).
800
600
400
200
0
0 1 2 3 4 5
Mean N-dose, molN day-1 litre-1
MesoNor
MesoMed
MesoBal
Lagoon
NorthSea
Autotrophic biomass, g C/litre
economical. There are several long time-series of chlorophyll-a measurements available, although
differences in analytical procedures (mostly related to the extraction protocol), can hinder the
comparability of the results. A weak point in the use of total biomass or chlorophyll-a alone is that
these indicators hold virtually no information on phytoplankton community structure. A related
chemical approach: quantifying accessory algal pigments by HPLC, can provide information on
composition of communities (at group level) and still avoid the bias associated with microscopical
examination (Schlter et al., 2000).
Phytoplankton biomass shows considerable natural variability due to seasonal succession, tidal
mixing, and localised freshwater input. Therefore, the selection of time-windows from monitoring
data for analysis, as well as sampling frequency, can be crucial when using phytoplankton biomass
as an indicator (Johnson et al., 2000).
Transparency
Transparency is not specically mentioned as a quality element in the Water Framework Directive.
However, transparency is linked to normative denitions (WFD; Annex V, sections 1.2.3. and 1.2.4.)
of the phytoplankton quality element so that at high status phytoplankton should have a biomass
that does not signicantly alter the type specic transparency conditions. Therefore, transparency
is an important supporting element for assessing the ecological quality of a water body.
The transparency of the water column in coastal waters is an indirect result of nutrient loading
and nutrient status. Increased nutrient loading often increases phytoplankton biomass in the
water column, which in turn decreases the transparency. However, local discharges of suspended
solids, dissolved coloured substances, untreated sewage or industrial efuents, may directly affect
the transparency of the water. Low transparency conditions are also often found in large river
plumes and shallow tidal areas due to sediment resuspension.
Reduced transparency inhibits the penetration of light and decreases the depth of the euphotic
zone. As a consequence, there is insufcient light near the sediment for primary production, which
make the deep macrophyte populations (seagrasses and macroalgae) decline or disappear. Since
macrophyte species have different light demands, reduced transparency affects dominance
patterns of the vegetation as well. In addition, low transparency strongly decreases the recreational
value of coastal waters.
Monitoring of transparency - metrics
There are two common methods for monitoring transparency: 1) Secchi depth and 2) light
attenuation. Secchi depth (the depth in meters at which a white disk becomes obscured by
suspended material or coloured tannins present in the water) is commonly used as a descriptor of
transparency. Secchi depth is a low-cost measurement, routinely performed in monitoring
programmes, and there is data available for several decades.
Comparison of long-term Secchi depths measurements carried out during, 1919-1939 with those
from the period, 1969-1991 in the Baltic Proper indicated that Secchi depth had decreased
about 0.05 m y
-1
(Sanden and Hkansson, 1996). Likewise, in the Adriatic water transparency has
decreased since, 1960 due to increased phytoplankton biomass related to nutrient enrichment via
the Po river discharge (Justic, 1988; Degobbis et al., 2000). In the Dutch Wadden Sea (the
Marsdiep) Cade and Hegeman (2002) found a decrease in the annual mean Secchi depth from
1.20 m to 1.0 m from the early, 1970s to 2000.
In some areas it has been possible to establish a quantitative relationship between transparency
and nutrient loading, either through direct correlation, or the use of more or less complex models.
For instance, in Danish estuaries, a signicant relationship has been established between total
nitrogen and chlorophyll-a concentration, as well as with Secchi depth (Nielsen et al., 2002a).
The use of transparency for assessing trophic status in coastal waters is not always straightforward
because inorganic suspended matter, detritus and humic substances can contribute more to
turbidity than phytoplankton (Wasmund et al., 2001). In such cases the transparency is not or
only weakly coupled to the nutrient loading or nutrient status of the system. In naturally turbid
waters such as some tidal areas, the anthropogenic discharges of nutrients may have little or no
effects on the transparency if, for example, the phytoplankton production is light limited and not
nutrient limited (Sanders et al., 2001; Colijn and Cade, 2003). Also in shallow non-stratied
areas the transparency can be more inuenced by the presence of a high density of suspension
feeding bottom fauna (e.g. mussels), than nutrient loading.
In near-shore Dutch waters the transparency was mostly determined by resuspended sediment,
while the contribution of phytoplankton to light extinction was of minor importance (van Gils and
Tatman, 2002; Blauw et al., 2004a,b). On the other hand, the variability in transparency in lake
Veerse Meer, a former tidal inlet that is now separated from the sea by dams, was 92% determined
by phytoplankton biomass (Stronkhorst et al., 1985). The underwater light climate in this deep
saltwater lake is strongly affected by the mixing layer depth following seasonal stratication.
Provided that the contribution from suspended solids is low and/or large data sets are at hand
the extinction coefcient may be distributed among the various phytoplankton groups using a
regression approach (van Gils and Tatman, 2002; Blauw et al., 2004a,b). On the other hand,
phytoplankters are inuenced differentially by light availability. Such effects may be quantied in
terms of PI-curves (production/irradiance) for different groups of phytoplankton or evaluated
based on observations of species composition along gradients of light availability (e.g. depth
distribution, correlation to suspended sediments).
Phytoplankton composition and abundance of phytoplankton taxa
While evidence for eutrophication may be apparent through an increase in phytoplankton
production and biomass, changes in phytoplankton community structure are not so obvious, due
to the complexity and variability of planktonic ecosystems. Among the factors creating such
complexity are selective grazing by micro- and mesozooplankton and physical factors such as
stratication, turbulence, irradiation, and salinity (Wasmund, 1999). Although the relationship
may not be straightforward, anthropogenic nutrient increase is reected in the composition of the
phytoplankton community, which is caused not only by the increased ambient nutrient
concentrations but also by their ratios (N:P:Si).
One important factor in the response of phytoplankton species to nutrient enrichment is their
specic physiological characteristics, such as growth potential, nutrient afnity and capacity for
intracellular nutrient storage. Phytoplankton indicator species for eutrophication in coastal areas
have not yet been identied (Smayda, 2004). However, evidence suggests that invasive, colonist-
species, that have been recognised in anthropogenically disturbed water bodies are agellated,
small (high surface/volume ratio) and potentially form blooms, which are often toxic or otherwise
harmful (Sellner et al., 2001; Smayda and Reynolds, 2001). In contrast, it has also been found
that in eutrophic nitrogen-controlled environments large phytoplankton species dominate, whereas
the fraction of small-sized species increases towards oligotrophy (Kuosa, 1990; Jrgensen, 1996).
However, most toxic dinoagellate blooms occur along oligotrophic coasts in Europe, whereas
blooms of non-toxic Phaeocystis dominate in the eutrophicated southern North Sea.
Indicators for phytoplankton composition should ideally have the following requirements:
indicative and sensitive to eutrophication and nutrient loading pressures
widely applicable in European coastal areas
cost effective and easy to analyse
include key species or groups, which are important in production and fate of biomass
applicable for validation of ecosystem models.
Potential indicators
Diatoms/agellates shift of functional groups. Changes in ambient nutrient concentrations
and their ratios give a competitive advantage to some phytoplankton species over others (Ofcer
and Ryther, 1980; Egge and Aksnes, 1992). Silicate limitation, indicated by decreasing Si:N ratios
as a consequence of eutrophication may affect the phytoplankton community by decreasing the
relative biomass of diatoms and increasing the biomass of agellates, some of which may develop
into harmful algal blooms (Sommer, 1995; Escaravage et al., 1999).
It has been observed in the North Sea coastal area, that long-term eutrophication has not only led
to an increase in phytoplankton biomass (chlorophyll-a), but also changes in the phytoplankton
community structure with a shift towards larger diatoms, and Phaeocystis pouchetii (Haptophyta)
-dominated blooms (Philippart et al., 2000; Cade and Hegeman, 2002). A similar trend has been
found in the southern Baltic Sea (Wasmund and Uhlig, 2003), where long-term monitoring data
have showed a decline in diatoms and an increase in dinoagellates during the spring bloom.
Other factors that affect the relationship between anthropogenic nutrient loading and
phytoplankton abundance and composition include vertical mixing and underwater light climate.
When light is limiting phytoplankton growth, a change in nutrient loading is not likely to affect
phytoplankton growth. In turbid transitional waters with a tidal effect, benthic primary production
by microphytobenthos may account for a large part of local primary production. Also, vertical
mixing has a large inuence on the competition between agellates and diatoms. This effect is
usually not taken into account in experimental studies of competition between functional groups
of phytoplankton. Therefore the results of competition experiments do not necessarily match with
eld observations.
Proportion of cyanobacteria. While eutrophication in freshwaters tends to shift phytoplankton
communities towards cyanobacterial dominance, this is not evident in coastal areas and estuaries.
However, in low salinity environments, the relative proportion of total phytoplankton biomass
composed of cyanobacteria can be used as a potential indicator of eutrophication. In some coastal
areas such as in the Eastern Gulf of Finland, Baltic Sea, where salinity is below 2 (practical salinity
scale), the dominance of freshwater cyanobacterial species (e.g. Planktothrix agardhii, Microcystis
spp.) indicates increased eutrophication (Kauppila et al., 1995). Likewise, the increasing frequency
and intensity of late-summer cyanobacterial blooms in low-saline estuaries (e.g. Nodularia
spumigena in the Baltic Sea; Kahru et al., 1994; Finni et al., 2001) is proposed to be connected to
anthropogenic eutrophication, and is further aggravated by excess phosphorus from internal
loading from anoxic sediments (Sellner, 1997).
Proportion of picoplankton. Under oligotrophic conditions when ambient nutrient concentrations
are low, the fraction of small-sized phytoplankton: the picoplankton (<2 m in size), that are
ubiquitous in aquatic systems, often exceed 50% of the total phytoplankton biomass. There is
considerable experimental evidence that picoplankton, as a proportion of total autotrophic biomass,
decrease in predictable manner with increasing nutrient concentration (Stockner, 1988; Kuosa,
1990; Thingstad et al., 1998; Agawin et al., 2000; Gotsis-Skretas et al., 2000). In hypereutrophic
environments their proportion of autotrophic biomass is only a few per cent of total autotrophic
phytoplankton biomass (Sndergaard, 1991; Kuuppo and Uronen, unpublished data).
However, picoplanktonic algae are often overlooked in monitoring programmes even though their
contribution to the biomass would be easy to estimate, e.g. by measuring the share of <2-3 m
chlorophyll-a fraction and comparing it with total chlorophyll. Their biomass is also relatively
simple to estimate by epiuorescence microscopy (Figure 2.4). The estimation of the picoplanktonic
fraction of total phytoplankton biomass would provide a simple metric for the analysis of the
trophic status of a water body. Such metrics could be tested both experimentally and by analysing
water samples over seasonal cycles, wide trophic gradients, and coastal water-body types.
Shifts in the phytoplankton composition. In long-term studies it has been observed that changes
within functional phytoplankton groups occur during the course of eutrophication with concomitant
increases in total phytoplankton biomass (Olenina, 1998). Olenina (1998) has reported shifts in
both the spring bloom diatom community as well as the late summer cyanobacterial community
towards eutrophic species in the Kuriu Marios lagoon, the Baltic Sea. In her time series,
Stephanodiscus hantzschii became dominant in the spring bloom, and Planktonema lauterbornii
in the summer cyanobacterial assemblage.
One potential indicator species of eutrophication is Skeletonema costatum (Figure 2.5). This marine
diatom has a worldwide distribution, extending also to the low saline Baltic Sea (Hllfors, 2004).
Based on monitoring data from Danish coastal areas it has recently been demonstrated that the
summer biomass of S. costatum correlates with nitrogen loading (Henriksen et al., unpublished
data). If this relationship also applies to other coastal areas then S. costatum would be a useful
indicator species because it is easy to identify and enumerate microscopically and it is one of the
key species in spring blooms in coastal areas, in addition to its universal distribution.
Diversity indices
Diversity indices
1
have been developed to describe, quantify and simplify the species richness and
diversity of ecosystems. Their purpose is to combine information related to species abundance
and species richness into a single number that can be used to assess the state of the community
(Washington, 1984).
Different types of diversity indices have been developed to describe community structure (Pielou,
1975, Washington, 1984):
diversity indices that express the richness and variety of communities
evenness indices (similarity indices), that express the similarity (equitability) of communities
dominance indices, that emphasise the most important species in the community
biotic indices, that are used for estimating the degree of water pollution based on indicator
species.
Phytoplankton indices, to be useful in describing eutrophication pressure, must be sensitive to
changes in phytoplankton community structure, robust to the requirements of statistical analyses
and be effective in distinguishing different trophic conditions (Tsirtsis and Karydis, 1998). Various
indices have been tested in coastal areas, but many have not been proven to be consistent and
sufciently sensitive to be applicable for assessment of the impact of nutrient pressure on
phytoplankton.
A major problem in the use of phytoplankton diversity indices is that they are extremely dependent
on taxonomic resolution and consistency of microscopical analysis, as well as the level of
taxonomical knowledge. When using indices in which the number of species plays the central role,
exclusion of rare species, either by insufcient taxonomical expertise or by small sample size, may
seriously violate the reliability and sensitivity of an index (Cao et al., 1998).
Figure 2.4 Epiuorescence
microscopy image of prokaryotic
picocyanobacteria.
Photo: Pirjo Kuuppo, Finnish
Environment Institute, Helsinki.
Figure 2.5 Microscopic photograph of diatom Skeletonema
costatum. Photo: Pirkko Kokkonen, Finnish Environment Institute,
Helsinki.
1
See Annex 1 for a list of common diversity indices and their formulas.
Examples of the use of diversity indices
Tsirtsis and Karydis have tested the sensitivity of various diversity indices in connection with
cluster analysis along a trophic gradient in the Aegean Sea, the Mediterranean, to assess
eutrophication levels (Karydis and Tsirtsis, 1996). Of the indices tested, the species number
indices, Kothes species decit, Gleasons and Margalefs diversity indices (see annex), which are
based on the number of species, had the highest sensitivity to distinguish between eutrophic,
mesotrophic and oligotrophic conditions using phytoplankton data. In contrast, Shannons H
index or Simpsons index, which are widely used in environmental studies, were unable to
discriminate between eutrophic and oligotrophic conditions.
Arhondsitsis et al. (2003) have tested the sensitivity of similarity indices for distinguishing eutrophied
areas from pristine areas in Mediterranean coastal phytoplankton communities, using Gleasons
index, Kulczynskis index, Ruzickas index, Pandeyas index, Ellenbergs index, Whittakers index of
association, Canberra metric, Morisitas index, Bray-Curtis index (see annex) with subsequent cluster
analysis. In the analysis, opportunistic and rare phytoplankton species were distinguished using Box
and Whisker plots, which improved the sensitivity of the similarity indices, and gave better resolution
across the trophic gradient than diversity indices alone (Tsirtsis and Karydis, 1998).
Common ecological indices (Hurlberts, Margalefs, Menhinicks, Shannons, species number,
Jaccards and the Saprobic index; see annex) have been tested in small bays of the Bothnian Sea,
the Baltic, using phytoplankton data derived from standard databases (Danilov and Ekelund, 2001).
Their study concluded that diversity, biotic and similarity indices can lead to diverse conclusions on
the trophic status of the ecosystems studied. While Hurlberts, Margalefs, Menhinicks, Shannons,
Jaccards index and species number were useful in comparing the areas of similar trophic state, the
saprobic index was unsuitable. Furthermore, the simultaneous use of several indices is recommended
in order to avoid erroneous conclusions on the level of eutrophication.
A larger-scale attempt to evaluate the applicability of diversity indices from phytoplankton
monitoring data from the whole Baltic sea area, for Water Framework Directive purposes, has
been done by within the EU project CHARM
2
. It was specically tested which indices were sensitive
to the number of taxa taken into analysis. They found that indices that calculate species richness
(Menhinicks and Margalefs indices), as well as the Shannons and Evenness indices were sensitive
to the reduction of taxa, whereas the Hurlberts, Simpsons and McIntosh indices were less sensitive
to taxa reduction (i.e. rare or difcult to identify species). The Menhinicks and Margalefs indices
were able to distinguish oligotrophic systems, but none were sensitive enough to differentiate
hydro-morphological conditions. Furthermore, none of the indices tested were able to discriminate
between seasons, and the indices had a greater interannual than seasonal variation.
Frequency and intensity of phytoplankton blooms
What is a phytoplankton bloom? Traditionally, the term phytoplankton bloom refers to an
accelerated increase in phytoplankton biomass, either due to an increased growth rate or
decreased top-down control, leading to a high biomass, a decrease in water transparency, and
water discoloration.
2
Characterization of the Baltic Sea Ecosystem: Dynamics and Function of Coastal Types (EVK3-CT-2001-00065, 2001-2004).
In temperate and high latitudes, the annually occurring spring bloom, consisting usually of
diatoms, is a normal phenomenon and part of seasonal plankton succession (Smayda, 1997). The
spring bloom is essential in the seasonal succession, as it serves as a signicant source of food for
zooplankton and bottom fauna. However, with eutrophication, the intensity of the spring bloom
is augmented, and the increasing biomass leads to increased amounts of organic matter settling
to the bottom. Decomposition of organic matter by bacteria consumes oxygen, leading in some
areas to deoxygenated sediments and bottom waters.
In contrast, episodic phytoplankton summer blooms occur as a response to pulses of new
nutrients or a reduction in grazing pressure and can be classied as exceptional (Smayda, 1997).
Some of these blooms are harmful, i.e. they either produce high biomass leading to detrimental
ecosystem effects per se, or by being toxic, causing mortality of sh and other organisms. Moreover,
toxic bloom events pose a serious health risk for humans when algal toxins are concentrated in
shellsh cultivated for human consumption (Cloern, 2001; Granli, 2004). Toxic blooms cause
considerable economic harm on a global scale as well as in European coastal zones (Cloern,
2001). Most harmful algal blooms (HABs) in coastal areas are caused by microalgae (20-200 m
in size; Moestrup, 2004). HABs can be composed of different taxonomic groups, of which
dinoagellates are the most numerous (Dinophyta: 69 species). Other groups represented are
diatoms (Bacillariophyta: 12 species), haptophytes (Haptophyceae: 9 species) and rapidophyceans
(Rapidophyceae: 7 species; Moestrup, 2004). Although a direct causal connection between
eutrophication and increased frequency of harmful algal blooms has not been proven, it is
generally believed that their frequencies have increased worldwide due to increased nutrient
input (Hallegraeff, 1993; Anderson et al., 2002; Granli, 2004).
Bloom denition. When discussing episodic phytoplankton blooms, the term phytoplankton
bloom needs to be dened before it can be used in an operational, or even in a conceptual way.
Blooms can be considered to be exceptional, unusual and nuisance and can be dened in
many ways (Smayda, 1997). For example, a bloom is a deviation from the normal cycle of
phytoplankton biomass (Parker, 1987), or a period when chlorophyll-a concentration exceeds
100 mg m
-3
(Tett, 1987).
Carstensen et al. (2004) have developed an operational denition that is useful for assessing the
frequency of phytoplankton summer blooms, i.e. other than the spring bloom. Their numerical
denition is based on monitoring data of chlorophyll-a, accompanied with information on
phytoplankton composition. They dene a bloom statistically as percentage of observations
where a station-specic chlorophyll-a concentration is outside the Gaussian distribution of
observations. This approach is potentially useful in assessing the status of coastal waters based
on anomalies in bloom frequencies as required by the WFD.
However, in many harmful algal blooms, the abundance or biomass of the causative organism (often
dinoagellates or diatoms) is relatively low, but are considered harmful because of toxin accumulation
in sh or shellsh that lter the toxic cells for food (Smayda, 1997). On the other hand, extensive kills
of sh and bottom fauna are often connected to high biomasses of toxic haptophytes (Dahl et al.,
1989). But even so, harmful blooms are often temporally and spatially restricted, and their detection
using conventional monitoring methods is difcult as well as stochastic.
Examples of Harmful algal blooms in European coastal waters
Phaeocystis blooms. Marine prymnesiophyte Phaeocystis pouchetii forms recurrent blooms along
the North Sea coast. It lives in either agellated form, or forms gelatinous colonies. During mass
occurrences in spring and early summer, Phaeocystis produces foam that accumulates at the
shoreline, affecting the recreational value and ecosystem functioning of coastal lines. Long-term
monitoring has shown that the Phaeocystis blooms are connected to nutrient enrichment of
coastal areas (Lancelot et al., 1987; Cade and Hegeman, 2002).
An analysis of Dutch and English monitoring data to determine the parameters controlling the
bloom dynamics of Phaeocystis globosa (HABES project -Harmful Algal Blooms Expert System;
Blauw et al., 2004a,b; Iriarte et al., 2004) has shown that the long term median of the maximum
bloom intensity was correlated to salinity and winter concentration of DIN and DIP. It was not
clear which of these parameters was the most important as salinity and nutrient concentrations
are often strongly correlated (Lancelot et al., 1987). In English coastal waters, the interannual
variability of maximum bloom intensity could be explained by river runoff from the nearby river
Tamar and winter chlorophyll-a concentrations (Iriarte et al., 2004).
Haptophyte blooms. Blooms of haptophytes, especially Prymnesium parvum and Chrysochromulina
polylepis have been reported from around the world in brackish and marine waters, including
European coasts (Edvardsen and Paasche, 1998). These haptophyte agellates can form massive
blooms and can cause sh kills or even extensive ecosystem disasters along coastal areas, an
example being the disastrous Chrysochromulina polylepis bloom on the Norwegian coast and
Kattegat in 1988 (Dahl et al., 1989). There is evidence that this particular bloom was stimulated
by anthropogenic nutrients (Aksnes et al., 1995).
Cyanobacterial blooms. In the Baltic Sea, lamentous cyanobacteria regularly form late summer
blooms, which are reported to have intensied and increased in frequency and duration (Kahru
et al., 1994; Bianchi et al., 2000; Finni et al., 2001). These blooms are dominated by Aphanizomenon
os-aquae and Nodularia spumigena, which are both able to x atmospheric nitrogen. N. spumigena
blooms are frequently toxic (Sivonen et al., 1989).
The cyanobacterial blooms occur at a time when planktonic production relies largely on nitrogen
and phosphorus that is recycled within the planktonic food web. There are indications that blooms
are initiated in hydrographic fronts, and where inorganic phosphate is introduced to the
trophogenic layer from below the thermocline (Kononen et al., 1996), thus giving N-xing
cyanobacteria a competitive advantage in N-limited late-summer situations. A low DIN: DIP ratio
plays a key role, in addition to which climatic conditions (strong thermal stratication) during
summer act as the triggering factor (Grnlund et al., 1996).
Dinoagellate and diatom blooms. A considerable proportion of harmful algal blooms are
caused by dinoagellates. They can produce toxins causing paralytic shellsh poisoning (PSP),
diarrhetic shellsh poisoning (DSP), or compounds that cause mass mortalities of sh. The most
regularly occurring taxons in European coastal waters, which are also listed as indicator species
by the OSPAR Commission (Ospar, 2003) are Gymnodinum mikimotoi, Alexandrium spp.,
Dinophysis spp. and Prorocentrum spp. In addition, species of the diatom genus Pseudo-nitzschia
can cause shelsh poisoning in European coastal areas (Moestrup, 2004).
Models relating nutrients and phytoplankton
Different types of models relating phytoplankton abundance to nutrient availability have been
developed. The models can be divided into two main groups: 1) data-driven (empirical) models
(inductive) and 2) deterministic models (deductive). Both types attempt to explain the ecosystem
with a combination of observations and theoretical understanding.
The data-driven models build on observations, derive models from the data and explain their
structure with theoretical understanding. The deterministic models start with the theory, develop
models based on theoretical understanding and validate them with observed data. Deterministic
models take a variety of forms but the equations describing the relationships between nutrients
and phytoplankton are physiologically based, with the difference that carbon and nutrient uxes
are strictly coupled in some models and partly uncoupled in others.
The models are usually combined with hydrodynamical transport-dispersion models than can either
be 1-, 2- or 3-dimensional. Coupled physical-chemical-biological models (ecosystem models) have
several advantages (e.g. temporal and spatial resolution) compared to empirical models but require
specic skills from the user. Examples of both types of models are discussed below. Table 2.1
and 2.2 contain a listing of various models with a brief description of their main characteristics.
Table 2.1 Examples of empirical models on phytoplankton and nutrients, developed for European
coastal areas.
Brief model characteristics Area applied to Reference
Vollenweider-type model to predict eutrophication from regressions
between chlorophyll-a and TN and TP. Inuence of morphology and
residence time of the estuaries was taken into account.
Small Baltic non-tidal
estuaries (Finland)
(Meeuwig et al., 2000)
Stepwise regression analysis showed that chlorophyll a concentrations in
shallow estuaries could be described as a function of area specic N-loads,
average depth of the estuary and the biomass of benthic grazers (mussels).
Based on the established regressions, a 50% reduction in N-loads should
result in approximately a 25% reduction of summer chlorophyll a
concentrations.
Danish estuaries
(< 8.5 m depth)
(Conley et al., 2000)
Empirical models based on multiple regression analysis between either
primary production, chlorophyll-a or Secchi depth and nutrient loadings (N
or P) combined with climatic parameters (irradiance, temperature and
wind). The analysis showed a clear connection between nutrient loads and
the environmental condition of the area. The most pronounced effect was
found for primary production where ca. 1% change was found per
percentage change in N-loads. Phytoplankton biomass (chlorophyll-a)
changed by ca. 0.7% per percent change in N-loads. Secchi depth increased
when nutrient loads were reduced.
Kattegat and Belt Sea,
Denmark
(Markager and Storm, 2003)
Total dissolved nitrogen was found to be the best predictor of phytoplankton
biomass (chlorophyll-a) with regression analysis (r
2
= 60%). The regression
model was further complemented using Bayesian network model, which
enabled the analysis of system dynamics as well as comparison of different
eutrophication scenarios.
Gulf of Gera, eastern
Mediterranean Sea
(Arhonditsis et al., 2003).
Table 2.2 Examples of deterministic models on phytoplankton and nutrients, developed for
European coastal areas.
Model to quantify the inuence of hydrodynamics climate and biological
factors to explain the good resistance of the bay ecosystem to nutrient
loading.
Bay of Brest, France (Le Pape and Menesguen,
1997)
An ecological box-model BIOGEN was developed to assess the response
of eutrophication caused by the River Danube and veried using
monitoring data collected between, 1985 and, 1995. Model describes C,
N, P and Si ux through aggregated planktonic ecosystem compartments
and results suggests that especially phosphate reduction has contributed
to observed positive development of the Black Sea ecosystem.
NW Black Sea (Lancelot et al., 2002)
A 3D numerical model was developed for analysis of eutrophication (and
dispersion of pollutants). In addition to the advection and diffusion
modules, the model includes interactions between nutrients, phytoplankton,
zooplankton, detritus and dissolved oxygen.
Generic, Venice lagoon (Drago et al., 2001).
A layered box model was developed to evaluate the major effects of
estuarine eutrophication. State variables in the model included
phytoplankton (chlorophyll-a), labile and refractory detritus, dissolved
inorganic nitrogen and phosphorus, photosynthetically active radiation,
temperature and oxygen. The model succeeded in hindcasting the
eutrophication development in the lagoon from, 1950 to, 1996.
Szczecin lagoon,
Poland
(Humborg et al., 2000b).
The GEM model includes both the effect of phytoplankton on transparency
and the effect of transparency on phytoplankton. The underwater light
climate is simulated as a function of solar irradiance, mixing layer depth
and extinction by different substances in the water column. The effect of
transparency on phytoplankton is simulated as production-irradiance
relations for the different groups of phytoplankton species included in the
model. The extinction coefcient for algae is species (group) specic. The
combined effect of nutrient availability, light intensity and temperature on
competition between phytoplankton groups is simulated through the
optimisation method linear programming.
Dutch coastal waters (de Vries et al., 1998)
Model describes factors that control phytoplankton community structure,
more specically diatom versus non-diatom balance. In addition to the
diatom-non diatom compartments, the model includes meso- and
microzooplankton, nitrogen and silicate, as well as sediment dynamics.
The model concludes that in stratied and non-stratied areas zooplankton
grazing has different impacts on the diatom-non-diatom balance.
Western Irish Sea (Kelly-Gerreyn et al., 2004)
Model simulates the effect of nutrient availability and ratio to
phytoplankton community structure. Phosphate and silicate, diatom, and
agellate biomass, and detritus and benthos constituted the state
variables. The model showed that the combination of phosphate increase
and silicate decrease caused a decline in diatom biomass and an increase
in agellate biomass.
Southern Baltic Sea (Dippner, 1998)
A phytoplankton model was developed to describe nutrient supply,
turbidity, atmospheric forcing and topography on phytoplankton
production. The physical module includes 3D water movement, temperature
and salinity, whereas the biological module includes light/ turbidity,
concentrations of N, P, and Si, as well as diatom and agellate biomasses.
The model simulations support the general conception that agellates are
stimulated by anthropogenic nutrients. Furthermore, the model could
simulate that the strong bloom of Chrysochromulina polylepis in Kattegat
Skagerrak area in, 1988 was stimulated by anthropogenic nutrients.
North Sea (Aksnes et al., 1995)
Continued
The European Regional Seas Ecosystem Model (ERSEM) was conceived as
a generic model, which should be capable of correctly simulating the
spatial pattern of ecological uxes throughout the seasonal cycle and
across eutrophic to oligotrophic gradients, when coupled to a qualitatively
correct physical model. It is one of the most complex ecosystem models
currently in use. The philosophy is to include all groups/pools and
processes which may signicantly inuence ecosystem dynamics and to
resolve the ecosystem into a number of functional groups allowing sensible
denitions of those processes. A key feature of the model is that it
decouples carbon and nutrient dynamics. Not only does this give a far
better approximation to how nutrient limitation acts on cells but it allows
the production of signicant pools of DOM. Thus it can simulate both the
classical, dynamics of large cells and processes within the microbial loop.
The dynamics of biological functional groups are described by population
processes (growth, migration and mortality) and physiological processes
(ingestion, respiration, excretion and egestion).
Generic
North Sea
Mediterranean Arabian
Sea
(Baretta et al., 1995; Baretta-
Bekker et al., 1997; Allen et
al., 2002; Blackford, 2002;
Blackford and Burkill, 2002)
POLCOMS constitutes a merge between ERSEM and a sophisticated 3D
hydrodynamic model developed at Proudman Oceanographic Laboratory.
POLCOMS is a nite difference hydrodynamic model developed particularly
for the simulation of small scale coastal-ocean processes, but has been
applied from estuaries to ocean domains.
Generic
Estuaries
Oceans
(Allen et al., 2001; Proctor et
al., 2003)
The Belgian MIRO model has been implemented for describing and
predicting C, N, P and Si in the Phaeocystis-dominated ecosystem of the
North Sea in response to the physical and nutrient forcing. The current
model is a result of the assemblage of 4 modules describing the dynamics
of phytoplankton (3 key functional groups), zooplankton (2 key functional
groups), dissolved and particulate organic matter degradation and nutrient
(NO3, NH4, SiO4, PO4) regeneration in the water column and the sediment.
Interestingly there is an explicit description of the 2 life forms of Phaeocystis
- free-living cells and colony forms - which characterize the polymorphic life
cycle of Phaeocystis and have a different trophic fate (free-living cells
feeding microzooplankton while colonies resisting grazing)
MIRO has been calibrated in a simple multi-box frame and has shown its
ability to describe reasonably the interannual variability observed over the
last decade in Belgian coastal waters.
Southern North Sea
and English Channel
(Lancelot et al., 2005)
A global nutrient-phytoplankton model developed by the Coastal Ecology
Department at IFREMER/Brest has been applied to the Channel with a
3D loop on the Bay of Seine, and an extension to the French continental
shelf. Models are now in an extensive validation process, using chlorophyll
and SPM image banks derived from SeaWiFS satellite data.
English Channel
Bay of Biscay
(Hoch and Mnesguen,
1997; Mnesguen and Hoch,
1997)
The Generic Ecological Model (GEM) is part of the Delft3D-WAQ / ECO
system for water quality and ecology, which includes an extensive process
library and dynamic process modules for e.g. oxygen, nutrients,
phytoplankton, microphytobenthos, detritus, etc. The model builds on the
previous NZBLOOM model (Los and Bokhorst, 1997). Time and spatial-
varying suspended sediment concentrations can be used as input for
Delft3D-ECO, in order to determine the seasonal changes in the underwater
light climate, water quality, primary production, biomass of phytoplankton,
species composition, etc. Delft3D-ECO has been applied with the 14-day
spring / neap tidal cycle with varying wind condition and the suspended
sediment yearly-averaged output with the addition of a technique called
the Random Cosine Model for derivation of the weekly variation of the
sediment concentration. Nutrient loads from rivers are based on data,
which are averaged over the period, 1989 , 1998. Recent advances
include improvements in the light extinction modelling based on validation
studies, and improvements in phytoplankton coefcients based on
laboratory data and subsequent calibration runs.
Generic,
Dutch coastal waters
(Blauw et al., 2006)
Continued
OECD classication. The OECD has recommended a classication system for trophic conditions
in lakes, rivers and reservoirs (Vollenweider and Kerekes, 1982). The classication is based on
probability distributions of easily analysable variables describing pressures (concentrations of
total N and total P) and phytoplankton biomass (chlorophyll-a concentration, peaks in chl-a, and
transparency i.e. Secchi depth), and it sets boundaries to ve trophic categories from ultra-
oligotrophic to hypertrophic. Weaknesses in this classication scheme are: 1) There is considerable
overlap in the probability distributions, which makes the boundary setting difcult. On the other
hand, the probabilities illustrate the uncertainty of the classication; 2) The probabilities are set
by each variable individually, and these criteria may be controversial; 3) The scheme presumes
that the environment is homogenous and in a stable state.
The OECD classication scheme was further developed by Zurlini (1996) and applied to the
coastal waters of the Adriatic Sea, Mediterranean. The approach was based on combining the
distinct probabilities introduced in the OECD scheme (TN, TP, Chl-a, transparency). The work
showed that classication uncertainties, caused by the uncoupling of trophic variables are related
to environmental disturbances by discharges, inows and turbulence.
Trophic status index. A trophic status index (TSI), developed by Carlson (1977) for lakes, denes
each trophic state by a doubling of phytoplankton biomass, measured with Secchi disk transparency,
and relating that to total phosphorus concentration (Carlson, 1977).
The Carlson index was the basis for the TRIX index, developed for the coastal marine environment
by Vollenweider et al. (1998). The TRIX index includes chlorophyll-a and oxygen saturation (in
percentage; expressing phytoplankton productivity), concentrations of total N and total P,
available DIN and DIP (expressing nutritional factors), and transparency (as a supplementary
water quality factor). The latter gives the possibility to include turbidity caused by factors (minerals,
dissolved substances) other than phytoplankton in the calculation of the TRIX index. The TRIX
The NORWegian ECOlogical Model system is a coupled 3D physical-chemical-
biological model system applied to study primary production, dispersion of
particles (sh larvae, pollution) and eutrophication. The biological-chemical
module simulates subsurface light, two groups of phytoplankton (diatoms and
agellates), and their growth as affected by nutrient concentrations, light
intensity and temperature. The model does not include zooplankton as a state
variable but instead a constant death rate of phytoplankton (which is also
assumed to include grazing) is included.
North Sea
Skagerrak and
Kattegat.
(Skogen and Soiland, 1998)
The MIKE EU/MIKE EcoLab Eutrophication Template describes nutrient cycling,
phytoplankton and zooplankton growth, growth and distribution of rooted
vegetation and macroalgae in addition to simulating oxygen conditions. The
eutrophication module is seamlessly integrated with the (1-, 2-, 3D) advection-
dispersion modules, which describes the physical transport process at each grid-
point covering the area of interest. The EcoLab allows the user to add new state
variables and modify process descriptions.
Generic DHI:
download description in
http://www.dhisoftware.
com/ecolab/
Review of deterministic models covering the North Sea. (Moll and Radach, 2003)
index can be useful in classication, especially when complemented with the N:P ratio, or other
additional information (Vollenweider et al., 1998). The TRIX index has been tested by Moncheva
et al. (2002) for a coastal region of the Black Sea with promising results.
Vollenweider et al. (1998) note that when trying to transfer limnological models to marine
conditions, it is necessary to take into account some basic characteristics that differ between lake
and marine ecosystems. First of all, whereas limnological criteria characterise lakes as an entity
(i.e. the trophic status is assigned on a whole lake basis), this concept is not applicable in marine
environments (except possibly in enclosed bays and lagoons), where prevailing currents, inshore-
offshore exchange, tidal forcing, and upwellings continuously modify the physico-chemical and
biological patterns in time and space.
OSPAR classication. The OSPAR commission has agreed upon classication of NW Atlantic seas,
including the North Sea and the Celtic Sea, into three classes with respect to eutrophication:
1) problem areas, 2) potential problem areas and 3) non-problem areas (OSPAR, 2003).
For phytoplankton, the classication criteria set to reect direct effects of nutrient enrichment
include maximum and mean chlorophyll-a concentration (elevated level of chl-a concentration is
dened as >50% above offshore or historical concentrations), and region/area -specic
phytoplankton indicator species (elevated nuisance bloom or toxic assessment levels). The indicator
species are categorized as 1) nuisance species, which cause nuisance, foam or oxygen deciency
(e.g. Noctiluca scintillans and Phaeocystis spp. in colony form), and 2) toxin producing species,
which cause sh and benthos kills or PSP and DSP
3
contamination in mussels (e.g. Chrysochromulina
polylepis, Gymnodinum mikimotoi, Alexandrium spp., Dinophysis spp. and Prorocentrum spp.;
OSPAR, 2003). Indirect and other possible effects of nutrient enrichment include oxygen deciency
(<2 mg l
-1
acute toxicity, 2-6 mg l
-1
deciency), and algal toxins (DSP/PSP mussel infection events).
Other classication schemes. All estuaries draining from the UK mainland have been classied
according to their annual nutrient loads, water ow and tidal action by Nedwell et al. (2002).
They concluded that chlorophyll level does not directly indicate nutrient loading to the estuaries,
because it is also inuenced by physical constraints and turbidity caused by suspended solids.
A WFD-based classication of the ecological status of coastal waters has been done in Basque country
(Borja et al., 2004a,b) which proposes EQR thresholds to be used for each ecological status level.
In the Netherlands three classication methods have been developed. The schemes have been
developed for all aspects of coastal ecosystems. The classication systems for the phytoplankton
element are:
1) AMOEBE. The AMOEBE indicator system uses the percent abundance of indicator species
as a percentage of the abundance under reference conditions (1930) as an indicator for the
health of the North Sea ecosystem (Baptist and Jagtman, 1997). For phytoplankton two
indicators are used: averaged total phytoplankton biomass and Phaeocystis.
2) GONZ. In the GONZ system (Kabuta and Duijts, 2000) there are 3 phytoplankton-related
indicators: 1) phytoplankton species diversity (indicated by Shannon Wiener index:
monospecic bloom = 0, even distribution over many species = 1); 2) phytoplankton structure
3
PSP = paralytic shellsh poisoning, DSP = diarrhetic shellsh poisoning
(indicated by N:P ratio: (DIN:DIP winter average and summer average); and 3) primary
production (calculated with GEM model, corrected for natural variability in meteorological
conditions and river discharges).
3) Ecosystem targets North Sea. As part of the Ecosystem targets North Sea (Boon and
Wiersinga, 2002) 11 indicators related to phytoplankton are used to evaluate if target
values are met: DIN and DIP concentrations, N:P ratio, chlorophyll-a, oxygen concentration,
selected species as indicators: Karenia mikimotoi, Dinophysis spp., Alexandrium spp.,
Pseudo-nitzschia spp., Phaeocystis globosa, and state of ship ballast water.
2.3 Macrophytes
Introduction
Benthic macrophytes in coastal and transitional waters include macroalgae and seagrasses and,
in brackish areas a few genera of salinity-tolerant angiosperms other than seagrasses
(e.g. Potamogeton, Zannichellia, Ruppia). Benthic macrophytes have a lifetime of months to years
and most species grow at the same spot attached to the bottom substratum. Their growth pattern
therefore reects a time-integrated response to physical and chemical characteristics of the
habitat. Seagrasses and different algal species or functional algal groups have different nutrient
demands and changes in nutrient concentration therefore affect dominance patterns of the
benthic vegetation as well as dominance patterns between planktonic algae and benthic
vegetation (Sand-Jensen and Borum, 1991; Duarte, 1995; Pedersen, 1995).
Planktonic algae and opportunistic macroalgae are generally favoured by high nutrient
concentration. When phytoplankton biomass increases the depth distribution of seagrasses and
perennial macroalgae decreases due to shading and ultimately benthic plants may disappear
(Duarte, 1995). Nutrient and transparency conditions thereby often exert combined pressures on
the benthic vegetation. Other physico-chemical factors such as salinity, physical exposure and
substrate composition as well as biological factors like grazing or contagious diseases may also
affect growth- and loss processes and may complicate the interpretation of causes of changes in
benthic vegetation (Schramm, 1999). The analysis of cause- and effect relationships is rendered
even more complex because of possible negative feed-back effects of eutrophication which may
accelerate the loss of benthic macrophytes and delay restoration (Duarte, 1995; Figure 2.6).
Seagrass meadows and benthic algal communities such as perennial kelp forests are appreciated
mainly because of the services they provide to overall functioning of coastal ecosystems (see
details regarding seagrasses in Hemminga and Duarte (2000): 1) they enhance biodiversity and
habitat diversity by providing possibility for attachment, refuge and food resources; 2) they
improve water quality by reducing particle loads in the water and absorbing dissolved nutrients,
and by promoting sedimentation of suspended particles; 3) they stabilise sediments due to
underground networks of seagrass rhizomes and roots and due to a reduced ow regime inside
the canopies which diminishes sediment resuspension; 4) they play a signicant role in global
carbon and nutrient cycling. Benthic plants constitute an important sink for carbon in the oceans
(Smith, 1981) because of their high levels of biomass and production and because most of the
biomass of seagrasses ends up as refractory detritus. In an assessment of the economic value of
the worlds ecosystems, the value of the nutrient cycling function of seagrass beds and beds of
macroalgae ranked second among the listed ecosystem values (Costanza et al., 1997).
Changed dominance patterns of the coastal primary producers from benthic macrophytes to
planktonic algae or from long-lived seagrasses and macroalgae towards opportunistic algae as a
consequence of increased nutrient concentrations and reduced water transparency may affect the
above-mentioned functional attributes. Moreover, opportunistic algae grow and decompose faster
than perennial species and may thereby generate a temporal imbalance between oxygen
production and consumption increasing the likelihood of anoxia (Sand-Jensen and Borum, 1991).
Blooms of lamentous algae also reduce the recreational value of coastal waters as they
accumulate in the water column and on the beach and anoxic events associated with macroalgal
blooms have negative effects on the ecosystem (Valiela et al., 1997).
Relationships between pressures and biological responses
In order to assess water quality according to WFD demands, it is necessary to identify vegetation
indicators, which respond predictably to changes in water quality/pressure levels and for which
levels corresponding to undisturbed conditions can be identied. In the following discussion we
present a number of possible vegetation indicators, and describe how they respond to nutrient
and transparency pressures, and how various models describe these relationships.
Depth limits. Depth limits of coastal macrophytes are largely regulated by transparency and,
consequently, also by levels of nutrient concentration and load (Duarte, 1991; Abal and Dennison,
1996; Nielsen et al., 2002a; 2002b). There are several examples in the literature of relationships
between depth limits of seagrasses and macroalgae and water quality (Table 2.3). The strongest
among the models can explain up to 70-80% of variations in depth limit based on variations in
water clarity or nutrient concentration or transparency between areas.
Figure 2.6 A cartoon depicting a conceptual model of the effect of increased nutrient loading on
submerged vegetation, stressing the self-accelerating nature of the process and the buffer mechanisms
operating. Redrawn and modied after Duarte (1995).
Grazing
Nutrient
load Plankton &
opportunistic
algae
Turbidity
Benthic resp.
Oxygen
Seagrasses &
perennial
macroalgae
Erosion &
resuspension
Although highly signicant, the relationships typically show a wide scatter. The models are
therefore best suited for describing mean expected depth limits at given levels of nutrient
concentration/transparency rather than providing precise estimates for specic areas (Nielsen
et al., 2002a; Sagert et al., 2005; Greve and Krause-Jensen, 2005). The wide scatter in the
relationships indicates that factors other than light and nutrients contribute to regulating depth
limits. Sediment characteristics are, for example, likely to play a role in this regulation.
Recent scientic results suggest that oxygen deciency may harm seagrasses (Terrados et al.,
1999; Holmer and Bondgaard, 2001; Koch, 2001; Greve et al., 2003) and reduced, organic rich
sediments prone to oxygen deciency might therefore hinder seagrass growth. Sediment
characteristics can also affect algal depth limits since lack of hard substratum may limit depth
penetration in some areas.
Indicator Area applied to Independent variables Empirical relation Statistical data Reference
Eelgrass depth limit (Zc, m) Woods Hole, MA Light attenuation (K, m
-1
) Zc = 1.62/K (Dennison, 1987)
World wide Light attenuation (K, m
-1
) log Zc = 0.27-0.84log K R
2
= 0.40, N = 29, p < 0.001 (Duarte, 1991)
Danish coastal waters Total nitrogen (TN, g/l
-1
) ln Zc = 6.039 0.755ln(TN) R
2
= 0.55, N = 128, p < 0.0001 (Nielsen et al., 2002b)
Danish coastal waters Secchi depth (Zs, m) Zc = 0.339 + 0.786Zs R
2
= 0.61, N = 101, p < 0.0001 (Nielsen et al., 2002b)
Danish coastal waters (1994) Water depth (Zmax, m)
Secchi depth (Zs, m)
ln winter [NH
4
+
] (NH
4
+
, M)
Zc = 2.44 + 0.11Zmax 0.87NH
4
+
+ 0.26Zs-winter R
2
= 0.71, N = 46, p < 0.001
(manual backward elimination)
(Greve and Krause-Jensen, 2005)
Danish coastal waters (1989-1998) As above Zc = 1.27 + 0.06Zmax 0.15NH
4
+
+ 0.38Zs-summer R
2
= 0.42, N = 390, p < 0.001
Danish coastal waters (1989-2001) Secchi depth (Zs, m)
Total phosphorus (TP, M)
Total nitrogen (TN, M)
Zc = 0.97 + 1,06Zs
Zc = 5.361 * (TP)
-1.86
Zc = 659.4 * (TP)
-1.456
R
2
= 0.668
R
2
= 0.602
R
2
= 0.632
(Sagert et al., 2005)
Zostera capricorni depth limits (Zc, m) S. Moreton Bay, Queensland, Australia Secchi depth (Zs, m)
Light attenuation (K, m
-1
)
Total Nitrogen (TN, M)
Susp. solids (SS, mg l
-1
)
Chlorophyll (Chl, g l
-1
)
Not dened R
2
= 0.78, N = 8
R
2
= 0.87, N = 8
R
2
= 0.91, N = 8
R
2
= 0.93, N = 8
R
2
= 0.88, N = 8
(Abal and Dennison, 1996)
Seagrass depth limit (Zc, m) World wide Light attenuation (K, m
-1
) log Zc = 0.26 1.07log K Zc = 1.86/K R
2
= 0.77, N = 72, p < 0.001, SE
slope
= 0.07 (Duarte, 1991)
Depth limit of submerged aquatic vegetation (SAV) (Zc, m) General Light attenuation (K, m
-1
)
Required % surface light (Iz/Io)
Zc = ln(Iz/Io)/K
NB: To be measured at mean tidal level (MTL)
(Koch, 2001)
General Upper depth limit (Zmin, m)
Minimum SAV depth range (X, m)
Zc X + Zmin
In Long Island Sound X=1
(Koch, 2001)
Macroalgal (brown algae) depth limit (Zc, m) Danish coastal waters Secchi depth (Zs, m) Zc = 1,252 + 1,427Zs R
2
= 0.58, N = 84, p < 0.0001 (Nielsen et al., 2002b)
-1
) ln Zc = 9.717 1.329ln(TN) R
2
= 0.48, N = 111, p < 0.0001 (Nielsen et al., 2002a)
Macroalgal (excl. brown algae) depth limit (Zc, m) Danish coastal waters Secchi depth (Zs, m) Zc = 1,1+1,68Zs R
2
= 0.64, N = 119, p < 0.0001 (Nielsen et al., 2002a)
-1
) ln Zc = 9.736 1.308ln(TN) R
2
= 0.73, N = 132, p < 0.0001 (Nielsen et al., 2002a)
Depth limit of F. vesiculosus (Zc, m) Stockholm Archipelago Secchi depth (Zs, m) Zc = 0.717 + 0.814Zs R
2
= 0.38, N = 69, p < 0.0001 (Kautsky, 1999)
Baltic Sea entrance Secchi depth (Zs, m) Zc = 1,06 + 0.06Zs R
2
= 0.014, N = 804, p < 0.001 (Torn et al., 2006)
Baltic Sea Secchi depth (Zs, m) Zc = 0,84 + 0.48Zs R
2
= 0.16, N = 180, p < 0.001 (Torn et al., 2006)
Baltic Sea, (upper boundary) Secchi depth (Zs, m) Zc = 0.74 + 0.94Zs R
2
= 0.85, p < 0.001 (Torn et al., 2006)
Table 2.3 Relationships between depth limits of coastal macrophytes and water quality.
Regarding macroalgae, the best relations between depth limits and water quality have been
developed for the algal community as a whole and for larger algal groups (e.g. brown algae).
Models have also been attempted for depth limits of key algal species like Fucus vesiculosus in the
Baltic Sea (Table 2.3). The best correlation between F. vesiculosus depth limit and transparency
was found for a restricted area of low salinity (Kautsky, 1999), while only very weak relationships
were found for high salinity regions at the entrance of the Baltic Sea (Torn et al., 2006). These
divergences are partly explained by large salinity-related changes in algal depth limits (Pedersen
and Snoeijs, 2001): At high salinity, species diversity is high and inter-specic competition forces
F. vesiculosus to colonize shallow water where light is not a limiting factor. As salinity declines,
species diversity is reduced and F. vesiculosus can penetrate to deeper, more light limited waters.
The marked inuence of salinity implies that depth limits of individual algal species are highly
area specic indicators and indicator levels should accordingly be dened for specic areas.
Indicator Area applied to Independent variables Empirical relation Statistical data Reference
Eelgrass depth limit (Zc, m) Woods Hole, MA Light attenuation (K, m
-1
) Zc = 1.62/K (Dennison, 1987)
World wide Light attenuation (K, m
-1
) log Zc = 0.27-0.84log K R
2
= 0.40, N = 29, p < 0.001 (Duarte, 1991)
-1
) ln Zc = 6.039 0.755ln(TN) R
2
= 0.55, N = 128, p < 0.0001 (Nielsen et al., 2002b)
Danish coastal waters Secchi depth (Zs, m) Zc = 0.339 + 0.786Zs R
2
= 0.61, N = 101, p < 0.0001 (Nielsen et al., 2002b)
Danish coastal waters (1994) Water depth (Zmax, m)
ln winter [NH
4
+
] (NH
4
+
, M)
Zc = 2.44 + 0.11Zmax 0.87NH
4
+
+ 0.26Zs-winter R
2
= 0.71, N = 46, p < 0.001
Danish coastal waters (1989-1998) As above Zc = 1.27 + 0.06Zmax 0.15NH
4
+
+ 0.38Zs-summer R
2
= 0.42, N = 390, p < 0.001
Danish coastal waters (1989-2001) Secchi depth (Zs, m)
Total phosphorus (TP, M)
Total nitrogen (TN, M)
Zc = 0.97 + 1,06Zs
Zc = 5.361 * (TP)
-1.86
Zc = 659.4 * (TP)
-1.456
R
2
= 0.668
R
2
= 0.602
R
2
= 0.632
(Sagert et al., 2005)
Zostera capricorni depth limits (Zc, m) S. Moreton Bay, Queensland, Australia Secchi depth (Zs, m)
Light attenuation (K, m
-1
)
Total Nitrogen (TN, M)
Susp. solids (SS, mg l
-1
)
Chlorophyll (Chl, g l
-1
)
Not dened R
2
= 0.78, N = 8
R
2
= 0.87, N = 8
R
2
= 0.91, N = 8
R
2
= 0.93, N = 8
R
2
= 0.88, N = 8
(Abal and Dennison, 1996)
Seagrass depth limit (Zc, m) World wide Light attenuation (K, m
-1
) log Zc = 0.26 1.07log K Zc = 1.86/K R
2
= 0.77, N = 72, p < 0.001, SE
slope
= 0.07 (Duarte, 1991)
Depth limit of submerged aquatic vegetation (SAV) (Zc, m) General Light attenuation (K, m
-1
)
Required % surface light (Iz/Io)
Zc = ln(Iz/Io)/K
NB: To be measured at mean tidal level (MTL)
(Koch, 2001)
General Upper depth limit (Zmin, m)
Minimum SAV depth range (X, m)
Zc X + Zmin
In Long Island Sound X=1
(Koch, 2001)
Macroalgal (brown algae) depth limit (Zc, m) Danish coastal waters Secchi depth (Zs, m) Zc = 1,252 + 1,427Zs R
2
= 0.58, N = 84, p < 0.0001 (Nielsen et al., 2002b)
-1
) ln Zc = 9.717 1.329ln(TN) R
2
= 0.48, N = 111, p < 0.0001 (Nielsen et al., 2002a)
Macroalgal (excl. brown algae) depth limit (Zc, m) Danish coastal waters Secchi depth (Zs, m) Zc = 1,1+1,68Zs R
2
= 0.64, N = 119, p < 0.0001 (Nielsen et al., 2002a)
-1
) ln Zc = 9.736 1.308ln(TN) R
2
= 0.73, N = 132, p < 0.0001 (Nielsen et al., 2002a)
Depth limit of F. vesiculosus (Zc, m) Stockholm Archipelago Secchi depth (Zs, m) Zc = 0.717 + 0.814Zs R
2
= 0.38, N = 69, p < 0.0001 (Kautsky, 1999)
Baltic Sea entrance Secchi depth (Zs, m) Zc = 1,06 + 0.06Zs R
2
= 0.014, N = 804, p < 0.001 (Torn et al., 2006)
Baltic Sea Secchi depth (Zs, m) Zc = 0,84 + 0.48Zs R
2
= 0.16, N = 180, p < 0.001 (Torn et al., 2006)
Baltic Sea, (upper boundary) Secchi depth (Zs, m) Zc = 0.74 + 0.94Zs R
2
= 0.85, p < 0.001 (Torn et al., 2006)
Table 2.3 Relationships between depth limits of coastal macrophytes and water quality.
Abundance at specic depths. The abundance of seagrasses and macroalgae can be measured
in terms of cover, biomass and (for seagrasses also) shoot density at specic water depths.
Abundance is typically quite variable in shallow areas where exposure to wind and waves plays a
major regulating role, but from intermediate water depths towards deeper water abundance tends
to decline in parallel to reductions in light availability with depth (Duarte, 1991; Krause-Jensen
et al., 2000; Dahl et al., 2001; Krause-Jensen et al., 2003). Vegetation abundance at specic
depths therefore responds to changes in nutrient levels and transparency and this relationship has
been described in several models (Table 2.4).
Total cover of macroalgal communities on Danish stone reefs declines with depth and this decline
has been shown to be most marked in years with high nutrient load (Dahl et al., 2001). Cover of
macroalgal communities along the Danish coasts is also highest in the clearest waters with lowest
nutrient concentration and declines towards more nutrient rich areas. However, while coastal
macroalgal cover responds to large-scale differences in water quality between areas, the indicator
is not sufciently sensitive to reect smaller year-to-year changes in water quality (Krause-Jensen
et al., in press). More studies of factors regulating algal cover are needed to make dose-response
relationships generally applicable.
Community structure relative abundance of opportunistic algae. Increased nutrient load
often leads to increased abundance of opportunistic macroalgae and phytoplankton at the
expense of perennial species (Duarte, 1995; Morand and Briand, 1996; Viaroli et al., 1996; Valiela
et al., 1997). Empirical relations between nutrient load and algal biomass exist, e.g. from Waquoit
Bay (Lyons et al., 1995). More advanced growth models of opportunistic algae in relation to
nutrient level have also been developed and linked with hydro-dynamic models (de Vries et al.,
1996; Coffaro and Sfriso, 1997; Solidoro et al., 1997; Kiirikki et al., 1998).
Indicator Area
applied to
Independent variables Empirical relation Statistical
data
Reference
Seagrass abundance:
downward slope/
attenuation of
seagrass biomass
(Kbio, m
-1
)
Worldwide light attenuation
(K, m
-1
)
log Kbio = 0.19 + 1.4 log K R
2
=0.91,
N=9,
p<0.001,
SE
slope
= 0.17)
(Duarte, 1991)
Eelgrass abundance:
probability of eelgrass
cover
Danish coastal
waters: 0-1m,
1-2m, 2-4m,
4-8m
Light (% of surface PFD)
Exposure (REId)
Salinity (S)
See paper See paper (Krause-Jensen et
al., 2003)
Macroalgal
abundance: total
algal cover (modelled
average) (TC, %)
Danish stone
reefs
Water depth (Z
max
, m)
Insolation
(S, MJ m
-2
year
-1
)
N-input (TN, ton year
-1
)
TC=3.6898-0.5973Z
max
-
XxTNxlocality, where X
depends on locality
(see paper)
See paper (Dahl et al., 2001)
Danish coastal
waters
salinity (S)
Limfjord sites (L=1)
TC=-0.386+0.0407Zs+
0.0205S-0.419L
R
2
=0.79,
N=28,
p<0.05
(Krause-Jensen et
al., in press)
Table 2.4 Relations between abundance of coastal macrophytes at specic depths and water quality.
In Greek coastal waters, cover of opportunistic algae relative to cover of late successional species
including seagrasses and perennial macroalgae has been suggested as an indicator of water
quality (Orfanidis et al., 2001; 2003). This indicator was tested in Danish coastal waters, but here
the relative abundance of opportunistic algae was more closely related to differences in salinity
than to differences in water quality between areas (Table 2.5).
There are indications that the relative abundance of green opportunists alone may be a more
sensitive indicator than the entire group of opportunists. A recent mesocosm experiment showed
no overall increase in the abundance of opportunistic algae along an articial eutrophication
gradient, but instead demonstrated a change in the composition of opportunistic algae from
dominance of corticated lamentous red algae to dominance of thin foliose green algae (Karez
et al., 2004). Middelboe and Sand-Jensen (2000) also found an increase in the relative abundance
of opportunistic green algae and a decline in large perennial brown algae following increases in
nutrient concentrations in a Danish estuary over a 50-year period.
The abundance of opportunistic macroalgae is extremely dynamic. The algae have fast growth
and decay rates and as they often occur free-oating, their distribution and abundance vary
depending on wind events. Moreover, differences in grazing pressure may markedly affect their
abundance (Geertz-Hansen et al., 1993). Such a dynamic nature is a hindrance for precise
prediction of abundance levels and suggests that a dense net of sampling stations with frequent
sampling is necessary in order to assess abundance.
Changes in the composition and dominance patterns of benthic vegetation in response to changes
in nutrient/transparency levels have also been described at the species level. One example is from
brackish German boddens where the composition of macrophyte communities has been described
for different water quality regimes (Blmel et al., 2002; Blmel et al., 2004). Differences in species
composition of macrophytes as a function of nutrient level has also been used in a saprobia
classication of Finnish coastal waters (Blomster, 1996; Viitasalo et al., 2002; Hllfors et al.,
1987). In the UK, species composition of macroalgal communities is also used in the classication
of water quality (Wilkinson and Wood, 2003).
Area distribution of seagrasses
As the depth limit and abundance of seagrasses are related to water quality, the same must be true
for seagrass area distribution; but models describing this relationship quantitatively are scarce in
the literature (Table 2.6). Eelgrass area distribution in Waquoit Bay has been found to decline in
Area applied to Independent variables Empirical relation Statistical data Reference
Waquoit Bay, MA N-input (TN, ton year
-1
) See paper R
2
= 0.51, N = 5 (Lyons et al., 1995)
Danish coastal waters salinity (S) Op=1.06+0.032S
No relation to water quality found
R
2
= 0.73, N = 28,
p < 0.0001
Krause-Jensen et al.,
in press
Table 2.5 Relations between relative abundance of opportunistic macroalgae (Op, in percentage) and
water quality. Indicator: relative abundance of opportunistic macroalgae (modelled average).
response to increased nutrient load because nutrient addition stimulates growth and standing
stock of algae which shade the seagrasses (Short and Burdick, 1996; Hauxwell et al., 2001;
Hauxwell et al., 2003). There is also substantial evidence that a serious decline in the depth and
area distribution of seagrasses in Dutch coastal waters is partly related to the increased nutrient
load and reduced transparency (Philippart, 1994). The best relations between water quality and
area distribution of seagrasses are likely to exist for relatively deeply located populations whereas
shallow populations which are not strongly light regulated may not be tightly related to changes
in nutrient load and water quality. Analyses of long-term changes in area cover of shallow Danish
eelgrass populations thus showed no relation to changes in nutrient load, probably because
stochastic events like storms and ice scour play a major role (Frederiksen et al., 2004a; 2004b).
In the Dutch Wadden Sea habitat characteristics like exposure levels, substrate characteristics and
salinity have been used to model the potential areal distribution of seagrasses (Philippart et al.,
1992; deJonge et al., 1997; deJonge and deJong, 1999; RIKZ, 2004).
Long-term changes in coastal macrophytes
Studies of long-term changes in coastal macrophyte communities can also be used to assess how
macrophytes respond to changes in water quality. These studies generally conrm the relationships
described in the empirical models presented in the previous section. Several long-term studies thus
document a decline in the distribution of seagrasses and perennial macroalgae and/or an increase
in the abundance of opportunistic macroalgae along with increases in nutrient load and decreases
in transparency during the latter half of the 20th century (Schramm and Nienhuis, 1996; Short and
WyllieEcheverria, 1996; Green and Short, 2003). Long-term decreases in transparency have also been
documented (Justic et al., 1995a,b; Sanden and Hkansson, 1996; Cadee and Hegeman, 2002).
2.4 Benthic Invertebrates
Introduction
The scientic community generally supports the use of benthic macrofaunal communities in the
assessment of biological effects caused by human activities in marine and estuarine environments
(Long and Chapman, 1985; Radenac et al., 1997; Savage et al., 2001). This widespread support
Area applied to Independent variables Empirical relation Statistical data Reference
Waquoit Bay, MA Nitrogen load
(Nload, kg km
-2
y
-1
)
Log area = 1.648 Nload R
2
= 0.88, p < 0.05 (Short and Burdick, 1996)
Waquoit Bay, MA Nitrogen load
(Nload, kg ha
-1
y
-1
)
Area = 31(log Nload) + 54 R
2
= 0.88, p < 0.05,
n = 7
(Hauxwell et al., 2003)
Danish coastal waters Nload, insolation, wind No relation found (Frederiksen et al., 2004a)
Table 2.6 Relations between area distribution of eelgrass and water quality. Indicator: area distribution of
eelgrass (area, %).
results from particular characteristics of benthic fauna: 1) Animals are in contact with sediments
(which are recognised as a sink for many contaminants); 2) They play an important role in the
cycling of nutrients and certain contaminants; 3) Benthic animals are relatively sedentary and
long-lived and thus provide a measure of contaminant effects integrated over time; 4) They are
sensitive to anthropogenic activities and respond in a rather predictable way; 5) Benthic animals
are a commercially valuable resource or are relevant food sources for economically or recreationally
important species (Rees et al., 1990; Dauer, 1993). Therefore, especially concerning the latter
characteristic, the depletion of benthic communities following sediment contamination, dredging
or bottom trawling may represent an important socio-economic problem since they directly or
indirectly support several human activities (Rhoads et al., 1978; Chapman et al., 1987; Rees et al.,
1990; Kaiser et al., 2000).
Univariate variables (number of taxa, abundance and biomass)
These variables are used as measures of some attributes of community structure and are applied
in the computation of some diversity indices (Clarke and Warwick, 1994).
Shannon-Weaver diversity index (H) and Margalef index. These indices are commonly used in
community studies (Magurran, 1988; Clarke and Warwick, 1994). When used alone these
measures may provide erroneous results, as diversity does not behave in a consistent way after
disturbance (Clarke and Warwick, 1994). Usually, under stable conditions of infrequent disturbance,
macrobenthic communities are dominated by K-selected (or conservative) species. These species
usually have a long life span, large body size and are rarely numerically dominant. The r-selected
(or opportunistic) species present contrasting characteristics. In general, opportunistic species
benet from disturbance events and usually become numerically dominant (Warwick, 1993;
Warwick and Clarke, 1993). Therefore, the proportion of the relative abundance of opportunistic
groups to the total abundance could be used as a potential indicator of anthropogenic activities
such as dredging, bivalve sheries and sediment contamination on macrofaunal communities. A
change in the ratio between taxonomic or functional groups (e.g. polychaete:amphipod) is also
pointed out by Elliott (1994) as a technique for the interpretation of functional trends.
Sanders rarefaction technique. This method (Sanders, 1968) was developed in order to avoid
problems when comparing samples of different sizes. This technique determines the expected
number of species present in samples of different sizes (i.e. the expected number of species in a
sample containing 10, 25, 50, 100, etc. individuals). There are, however, some limitations, some
of them related to the type of gear used and the screen mesh size. Hurlbert (1971) has modied
Sanders original formula.
Multivariate analysis
Multivariate methods are considered to be more sensitive than univariate measures for detecting
differences in community structure between samples, both in space and time (Clarke and Warwick,
1994). The most commonly used methods in macrobenthic community studies are hierarchical
clustering, non-metric multi-dimensional scaling (MDS), principal components analysis (PCA), and
correspondence analysis (CA). Nevertheless, these methods only assess differences between
communities and cannot be used as measures of community stress (Clarke and Warwick, 1994).
Benthic communities show great variability at regional scales resulting from natural environmental
conditions that, however, can be overcome by working at taxonomic levels higher than species
(Clarke and Warwick, 1994). The phylum meta-analysis approach uses an approximated value of
production provided by both abundance and biomass data. This approach was applied to NE
Atlantic fauna in disturbed and undisturbed areas. Although it may not be highly sensitive to
detect low-level or subtle perturbations, this technique seems to retain much of the sensitivity of
multivariate methods (Clarke and Warwick, 1994), and it can be useful, especially concerning the
study of wide-ranging areas. Savage et al. (2001) suggested that the phylum-level meta-analysis
appears to be better in assessing the impact of organic and chemical pollution on an ocean-basin
scale than physical disturbance due to offshore mining.
Abundance/Biomass Comparison Curves ABC. ABC (Warwick, 1986) is a graphical method
that directly compares abundance and biomass K-dominance curves and has been applied
worldwide (Warwick et al., 1987; Weston, 1990; Simboura et al., 1995; Kaiser et al., 2000).
The method is based on the assumption that the distribution of numbers of individuals and
biomass among species within a community show a different response to disturbance (Warwick
et al., 1987). There are, however, some precautions to be followed when applying this method and
interpreting the results: 1) it should not be used in intertidal communities without reference to a
long-term and special series of control samples (Beukema, 1998); and 2) similar results may be
obtained for disturbed and undisturbed conditions e.g. after recruitment events (Warwick and
Ruswahyuni, 1987). With ABC Curves it is possible to recognise three different conditions
(undisturbed, moderately disturbed and grossly disturbed) in a community without reference to a
control area both in time and space. In undisturbed communities, the K-dominance curve for
biomass lies above the abundance curve for its entire length. Under moderate disturbance, both
curves are closely coincident and may cross each other one or more times. In grossly disturbed
communities the abundance curve lies above biomass throughout its length (Warwick, 1993).
Benthic Habitat Quality (BHQ) index. This index (Nilsson and Rosenberg, 1997) uses sediment
prole images to assign sediment quality and biogenic structures to 4 different successional
stages (SS 0 to SSIII). The method uses the redox potential discontinuity (RPD) layer and the
presence of burrows and animal tubes as an indicator of hypoxic conditions. Instead of the
separation into four successional stages, Rosenberg et al. (2004) suggested that this index should
be divided into 5 classes in order to be in agreement with the Water Framework Directive (WFD).
This index varies between 0 and 15 where high scores are assigned to mature successional stages
and low values with pioneering stages or azoic bottoms (Rosenberg et al., 2004).
Marine Biotic Index (BI). The marine biotic index (Borja et al., 2000) is derived from separating
the proportional abundance of individuals into ve ecological groups, which are related to the
degree of sensitivity/tolerance to an environmental stress gradient. In comparison to the biological
indices proposed previously (Hily et al., 1986; Grall and Glmarec, 1997), BI uses a mathematical
formula to obtain a continuous value the Biotic Coefcient (BC), which is more suitable for
statistical analysis. The BI value represents the quality of the bottom conditions in a discreet
range from 0 (unpolluted) to 7 (extremely polluted), according to the BC value.
100 / )) % 6 ( ) % 5 . 4 ( ) % 3 ( ) % 5 . 1 ( ) % 0 (( GV GIV GIII GII GI BC + + + + =
where GI comprises very sensitive species; GII indifferent species; GIII tolerant species; GIV
second-order opportunistic species; GV rst-order opportunistic species, according to their
sensitivity to increasing pollution gradients. Information about the sensitivity of more than 2700
taxa from European and Mediterranean soft-bottom environments is available in AZTIs web page
(http://www.azti.es).
Borja et al. (2004a, b) presented some of the problems associated with the AMBI tool. The validation
of model development showed that different anthropogenic disturbances in the environment can
be detected through the use of the BI, such as dredging works, sewage loading or mud disposal.
According to Borja et al. (2000), the BI can be applied when attempting to determine the ecological
status of European coastlines. Nevertheless Simbura and Zenetos (2002) have identied some
discrepancies when using this index on Mediterranean samples, and they suggest that the BI is
more appropriate for Atlantic ecosystems and estuarine areas (Simboura, 2004).
Benthic Index of Biotic Integrity (B-IBI). This index (Weisberg et al., 1997) compiles a number of
measures into a single score of benthic community health. Included in calculating the B-IBI are
measures of species diversity, species composition (e.g. % abundance and biomass, pollution
sensitive taxa), productivity, trophic composition (e.g. % deep deposit-feeders) and the depth
distribution of taxa and biomass in the sediment. The index was developed for the Chesapeake
Bay, USA using reference sites and ranges from 1 to 5 (in later versions the scale is extended to
between 1 (poor) and 10 (high quality), and has been used routinely in Chesapeake Bay monitoring
since 1999.
Benthic Quality Index (BQI)
Benthic Quality Index (BQI). This index (Rosenberg et al., 2004) was proposed for the assessment
of environmental quality at a particular station and uses the expected number of species (ES) in
samples where abundance is higher than 50 individuals (ES50).
) 1 log( )) 50 ( (
10
05 . 0
1
+ =

=
S ES
totA
A
BQI
i
n
i
i
,
where A is the mean relative abundance of a species i; S is the mean number of species at a
station.
Exergy index. The exergy index has been used in several studies as an ecological organizing
principle (Xu et al., 1999; Jrgensen and Marques, 2001; Marques and Jrgensen, 2002). Exergy
can be dened as the amount of work that can be extracted from a system when brought into
thermodynamic equilibrium with its environment. This includes both an energetic and information
content embedded in the biomass, which has been estimated by the complexity of the organisms
DNA. The total exergy is the sum of the compartmental biomasses multiplied by an organism
specic weighting factor. Standard values for these weighting factors are available in the literature
for a variety of organisms (Jrgensen and Marques, 2001; Marques and Jrgensen, 2002). Fabiano
et al. (2004) concluded that exergy appeared to be sensitive to changes in environmental
conditions and trophic state and may therefore be proposed as a useful tool for the health
assessment of marine benthic ecosystems.
The WFD requires that the environmental quality assessment of coastal ecosystems comprises 5
classes of quality status: Bad, Poor, Moderate, Good, and High. Three of the indices
discussed above (BQI, BHQ and BI) were adapted in order to be in agreement with this classication
and are summarised in Table 2.7.
Models
Models simulating the impact of different disturbance types on the macrofaunal community can
be divided into three main categories: 1) screening models, 2) empirical models (Pearson-
Rosenberg and bethic-TOC; see below), 3) ecological or food-web models.
Screening models. A Geographic Information System (GIS) is a powerful management tool that
allows the integration of all information for a given area. Due to the high complexity of marine
and estuarine areas, GIS offers a good way to visualise, organise, combine and analyse data.
Pearson-Rosenberg model. This model (Pearson and Rosenberg, 1978) is based on species
richness, abundance, and biomass (SAB). This is probably the most commonly cited model and
predicts a specic sequence of successional stages after disturbance. It predicts that an increase
in disturbance results in an increase in the number of species and biomass. After an initial increase,
HIGH GOOD MODERATE POOR BAD
WFD
All the
disturbance-
sensitive taxa
associated with
undisturbed
conditions are
present
Most of the
sensitive taxa of
the type-specic
communities are
present
Taxa indicative of
pollution are present.
Many of the sensitive
taxa of the type-specic
communities are present
biological communities
deviate substantially
from those normally
associated with the
surface water body type
under undisturbed
conditions
large portions of the
relevant biological
communities normally
associated with the
water body type under
undisturbed conditions
are absent
BQI 20 m 14.4 < 14.4 - 10.8 < 10.8 - 7.2 < 7.2 - 3.6 < 3.6
BQI > 20 m 16.0 < 16.0 - 12.0 <12.0 - 8.0 < 8.0 - 4.0 < 4.0
BHQ 20 m 15.0 - > 8.0 8.0 - > 6.0 6.0 - > 4.0 4.0 - > 2.0 2.0 - 0
BHQ > 20 m 15.0 - > 11.0 11.0 - > 7.0 7.0 - > 4.0 4.0 - > 2.0 2.0 - 0
BI 0 - 1.2 > 1.2 - 3.3 > 3.3 - 4.3 > 4.3 - 5.5 > 5.5 - 7
Dominant
Ecological
Groups (Borja
et al., 2004)
EGI
Species very
sensitive to
pollution
EGIII
Species tolerant
to an excess of
organic matter
EGIII, EGIV and EGV
Species tolerant to an
excess of organic matter
and opportunistic species
EGIV and EGV
Opportunistic species
EGV
First-order opportunistic
species
Table 2.7 Correspondence between Environmental quality status as described in WFD and three benthic
indices BQI, BHQ and BI values (see text).
both variables decrease to the ecotone point, followed by a secondary maximum of biomass and
abundance associated with a few opportunistic species. Under extreme conditions the community
may attain an azoic state. Although it is applied worldwide, results from other studies indicate
that predictions from this model can be quite variable; therefore it must be applied with caution
(Smith and Brumsickle, 1989; Thrush et al., 1996).
Benthic-TOC relationships. A model for the macrobenthic fauna and total organic carbon (TOC)
relationships (Hyland et al., 2000) has been produced based on the Pearson and Rosenberg
model (Pearson and Rosenberg, 1978). According to the benthic-TOC model, benthic variables
(number of species, abundance, and biomass) would increase up to a certain critical level in
relation to a sediments increasing organic carbon content (TOC) declining afterwards. Population
increases below this TOC critical level reect a combination of the nutritional value of moderate
levels of organic matter present in the sediment and other favourable environmental conditions,
such as sufciently high levels of dissolved oxygen and low sediment-associated stressors
(contaminants, ammonia, sulphide). As TOC exceeds a certain critical level, the benthos is exposed
to increasing amounts of physiological stress resulting from oxygen depletion and increasing
concentrations of contaminants or other sediment-associated stressors.
Ecological models. Several ecological models have been used to evaluate the impact of
anthropogenic activities in marine ecosystems, mainly sheries. Basically, the models can be split
into seven categories, depending both on the required input variable data and produced output
predictions (ICES, 2000). According to Robinson and Frid (2003), the categories are:
Habitat-based models including those that give an insight into how a population range will
t to the most suitable habitat as shing changes the total habitat size.
Models based on community metrics; these are models that reect how community-level
metrics change in response to shing disturbance.
Single-species models with variable prey or predators incorporating one-way trophic feedback
reactions on dynamic single-species models following a shery disturbance.
Multispecies production models are those that show how shery harvest of predator or prey
will affect the abundance of the other.
Dynamic multispecies models can incorporate spatially dynamic or age/size-structured
populations into changes in predatorprey interactions brought about by shing disturbance.
Aggregate ecosystem models derive from food webs and energy budgets, illustrating changes
in energy, carbon or biomass of aggregated functional species groups.
Ecosystem models with age/size structure, distinguishable from aggregate ecosystem models
in that the individual functional groups are generally less aggregated and there is greater
temporal resolution in their dynamics.
Among these categories, the ecological models that proved to yield good insights were those
categorised as aggregate ecosystem models and ecosystem age/size structures. Usually, these
kinds of models are derivates of the ECOPATH with Ecosim modelling approach.
The Ecopath with Ecosim (EwE). This modelling approach combines software for ecosystem
trophic mass balance analysis (Ecopath), with a dynamic modelling capability (Ecosim) for
exploring past and future impacts of shing and environmental disturbances as well as for
exploring optimal shing policies. Ecopath with Ecosim(EwE) is being widely used as a tool for
analysis of exploited aquatic ecosystems, but EwE can produce misleading predictions about the
direction of impacts of policy proposals. Erroneous predictions usually result from bad estimates
or errors of omission for a few key parameters, rather than diffuse effects of uncertainties in all
the input information (Christensen and Walters, 2004).
2.5 Summary
Phytoplankton. Coastal and transitional waters are the recipients of land-based nutrients, and
thus integrate nutrient loads from various sources. As the phytoplankton blooms have been
intensifying in many coastal areas, so too has the research effort to understand the factors
governing the initiation and fate of blooms during the last few decades. Despite the large amount
of results and literature available on phytoplankton and nutrients, region or coastal type-specic
dose-response relationships between nutrient concentrations and phytoplankton production have
only been described for some coastal areas. Such relationships, coupled with nutrient loading
models and appropriate phytoplankton status indicators are needed to allow WFD compliant
classication, and the setting of environmental objectives for different coastal areas, as a part of
river basin management programs and a program of measures.
WFD compliant classication systems are currently under development in many countries.
Currently, classication systems based on phytoplankton biomass and composition are only
developed for some parts of the NE Atlantic ecoregion. For most of the Baltic Sea and Mediterranean,
WFD compliant classication systems are not yet established.
Despite the increasing frequency, duration and coverage of algal blooms, little is known about the
factors that govern the initiation and taxonomic composition of blooms. It is not known what is
actually the role of nutrients in the blooms compared to other factors such as light, mixing,
nutrient composition, life cycles, etc. In addition, other agents such as dissolved organic matter
(e.g. humic substances) may be important for bloom initiation and success of toxic dinoagellates
(Doblin et al., 1999; Granli et al., 1999; Stolte et al., 2002).
Empirical, Vollenweider type models have not been developed for many coastal areas. Whereas in
lake ecosystems the Vollenweider model on the relationship between phosphorus loading and
phytoplankton biomass (chlorophyll-a) is applicable for 75% of lakes, it is not possible to develop
a common data-driven model for the interaction between one nutrient and phytoplankton
dynamics in coastal/estuarine ecosystems. This is mainly because nutrient limitation patterns are
variable across European coastal zones, as well as seasonally within a coastal area. Coastal areas
and estuaries respond to increased nutrient loading in different ways, depending on their
characteristic features that inuence their ability to buffer nutrient enrichments on the ecosystem
including different limitation patterns (N, P, Si, light, etc.). More information on the regional/
local nutrient limitation patterns and applicability of indicative nutrient ratios to predict the
changes in phytoplankton biomass and composition will be needed for development and
application of empirical dose-response models for European coastal areas.
There are a relatively large number of deterministic ecosystem models developed for European
coastal seas. Many of those simulate the development of phytoplankton biomass, and some are
also able to simulate changes in phytoplankton composition. Models that simulate both
parameters would be a helpful tool to develop WFD compliant classication systems, including
hind-casting or simulating reference conditions as well as setting the class boundaries with respect
to ecological discontinuities in the phytoplankton response curves. Such nonlinear points in dose-
response relationships that may lead to shifts between alternative stable states of an ecosystem
would be needed to set ecologically relevant class boundaries for WFD classication systems,
allowing scientic setting of the environmental objectives for good ecological quality. Identication
of such thresholds or point-of-no-return requires thorough understanding of the ecosystems
processes, bottlenecks in the resilience of the ecosystems and careful model formulation (Duarte,
2004). Furthermore, model testing and validation, with respect to monitoring alternative stages
in eutrophication in Europe, are restricted to the Baltic Sea area (Elmgren and Larsson, 1997; Rask
et al., 1999).
Macrophytes. This chapter demonstrates that coastal macrophytes do indeed respond to changes
in nutrient/transparency levels and therefore can be used as indicators of water quality. The best
empirical relations to water quality exist for deep communities because they are typically strongly
regulated by water transparency and only moderately by wave exposure. Existing relations do,
however, show considerable unexplained variation, which limit the predictive capacity of empirical
models. The variation can only be reduced by identifying its causes and including supplementary
regulating factors in future models. Several studies indicate that e.g. physical exposure, sediment
characteristics and occurrence of anoxic events may play signicant regulating roles. We encourage
further studies of such effects as well as analyses of possible negative feed-back effects of
eutrophication and identication of possible nutrient generated regime shifts in benthic
macrophyte communities.
Benthic invertebrates. Many of the techniques used in benthic ecology should be complementary,
as each only provides part of the complete information (Elliott, 1994). Macrofauna is recognised
as a proper tool for monitoring the environmental health of marine ecosystems, but studies on
this biotic component are highly labour intensive, both in terms of sampling effort and sample
processing (sorting and identication to species level). Therefore, several studies have focused on
the taxonomic resolution needed to assess environmental disturbances (Warwick, 1988; Somereld
and Clarke, 1995; Olsgard et al., 1997). According to these authors, for soft bottom macrobenthos
when using multivariate methods, disturbance effects are often detectable at high taxonomic
levels, such as the family level. Warwick (1993) also suggested that in environmental impact
studies, it would be more important to analyse a large number of stations/replicates at a higher
taxonomic level than a small number of stations at the species level.
Normally, for macrobenthos, there is a positive correlation between taxonomic groupings and
ecological functionality (Warwick, 1993). Several works have assessed macrobenthic impacts from
different sources of disturbance at high taxonomic levels (Bascom, 1982; Berge, 1990; Savage
et al., 2001). Nevertheless, Rosenberg et al. (2004) recommend the use of the highest possible
taxonomic level, as different species within the same genera may have different levels of tolerance
to disturbance. Independent of the taxonomic resolution applied, more information is needed
regarding the behaviour of benthic taxa, and not just whether they are sensitive or opportunistic.
2.6. References
Abal, E.G., and W.C. Dennison. 1996. Seagrass depth range and water quality in southern Moreton bay,
Queensland, Australia. Marine and Freshwater Research 47:763-771
Agawin, N.S.R., C.M. Duarte, and S. Agust. 2000. Nutrient and temperature control of the contribution of
picoplankton to phytoplankton biomass and production. Limnology and Oceanography 45:591-600
Aksnes, D.L., K.B. Ulvestad, B.M. Balino, J. Berntsen, J.K. Egge, and E. Svendsen. 1995. Ecological modelling
in coastal waters: Towards predictive physical-chemical-biological simulation models. Ophelia 41:5-36
Allen, J.I., J. Blackford, J. Holt, R. Proctor, M. Ashworth, and J. Siddorn. 2001. A highly spatially resolved
ecosystem model for the North West European Continental Shelf. Sarsia 86:423-440
Allen, J.I., P. Somereld, and J. Siddorn. 2002. Primary and Bacterial Production in the Mediterranean Sea:
A modelling study. Journal of Marine Systems 33-34:473-496
Anderson, D.M., P.M. Glibert, and J.M. Burkholder. 2002. Harmful algal blooms and eutrophication: Nutrient
sources, composition, and consequences. Estuaries 25:704-726
Arhonditsis, G., M. Eleftheriadou, M. Karydis, and G. Tsirtsis. 2003a. Eutrophication risk assessment in coastal
embayments using simple statistical models. Marine Pollution Bulletin 46:1174-1178
Arhonditsis, G., M. Karydis, and G. Tsirtsis. 2003b. Analysis of phytoplankton community structure using
similarity indices: A new methodology for discriminating among eutrophication levels in coastal marine
ecosystems. Environmental Management 31:619-632
Baptist, H.J.M., and E. Jagtman. 1997. De AMOEBES van zoute wateren. Water-. systeemverkenningen 1996.
Rapport RIKZ-97.027.
Baretta-Bekker, J.G., J.W. Baretta, and W. Ebenhoeh. 1997. Microbial dynamics in the marine ecosystem
model ERSEM II with decoupled carbon assimilation and nutrient uptake. Journal of Sea Research
38:195-211
Baretta, J.W., W. Ebenhoeh, and P. Ruardij. 1995. The European Regional Seas Ecosystem Model, a complex
marine ecosystem model. Netherlands Journal of Sea Research 33:233-246
Bascom, W. 1982. The effects of waste disposal on the coastal waters of southern California. Environmental
Science and Technology 16:226-236
Berge, J.A. 1990. Macrofauna recolonization of subtidal sediments. Experimental studies on defaunated
sediment contaminated with crude oil in two Norwegian fjords with unequal eutrophication status.
I. Community responses. Marine Ecology Progress Series 66:103-115
Beukema, J.J. 1998. An evaluation of ABC-method (abundance/biomass comparison) as applied to
macrozoobenthic communities living on tidal ats in Dutch Wadden Sea. Marine Biology 99:425-433
Bianchi, T.S., E. Engelhaupt, P. Westman, T. Andrn, C. Rolff, and R. Elmgren. 2000. Cyanobacterial blooms
in the Baltic Sea: Natural of human-induced? Limnology and Oceanography 45:716-726
Blackford, J.C. 2002. The inuence of microphytobenthos on the Northern Adriatic ecosystem. A modelling
study. Estuarine, Coastal and Shelf Science 55:109-123
Blackford, J.C., and P.H. Burkill. 2002. Planktonic community structure and carbon cycling in the Arabian
ea as a result of monsoonal forcing: The application of a generic model. Journal of Marine Systems
36:239-267
Blauw, A.N., L. Peperzak, F.J. Los, R.E. Uittenbogaard, L. Dardengo, M. Dedert, and M. Poelman. 2004a. Pilot
study Dutch coastal waters. In: HABES, Harmful Algal Blooms Expert System. Final report. WL Delft
Hydraulics, Delft.
Blauw, A.N., L. Peperzak, D. Purdie, A. Iriarte, and F.J. Los. 2004b. Integrated results on Phaeocystis globosa.
In: HABES, Harmful Algal Blooms Expert System. Final report. WL Delft Hydraulics, Delft.
Blauw, A.N., F.J. Los, and M. Bokhorst. 2006. GEM, a Dutch generic ecological model for estuaries. Submitted
to Hydrobiologia.
Blomster, J. 1996. Ravinnekuormituksen vaikutus rantavyhykkeen levyhteisihin ja vaikutusten arvioinnissa
kytetyt menetelmt. Report No. 5 Suomen Finland ympristkeskus, Helsinki.
Blmel, C., A. Domin, J.C. Krause, M. Schubert, U. Schiewer, and H. Schubert. 2002. Der historische
Makrophytenbewuchs der inneren Gewsser der deutschen Ostseekste. Rostock Meeresbiol Beitr 10:5-111
Blmel, C., A. Eggert, T. Rieling, M. Schubert, U. Selig, and H. Schubert. 2004. Entwicklung von
leitbildorientierten Bewertungsgrundlagen fr innere Kstengewsser der deutschen Ostseekste nach der
EU-WRRL Forschungsbericht. Report No. FKZ: 0330014, Bundesministerium fr Bildung und Forschung-
Landesamt fr Landesamt fr Umwelt, Naturschutz und Geologie Mecklenburg-Vorpommern
Boon, A.R., and W.A. Wiersinga. 2002. Parameters ecosysteemdoelen Noordzee: een analyse van mogelijke
parameters voor de concretisering van het natuurbeleid voor de Noordzee. Rapport Expertisecentrum LNV,
2002/116. Expertisecentrum Landbouw, Natuurbeheer en Visserij: Wageningen, The Netherlands. 69 pp.
Borja, A., J. Franco, and V. Prez. 2000. A marine biotic index to establish the ecological quality of soft-
bottom benthos within European estuarine and coastal environments. Marine Pollution Bulletin 40:1100-
1114
Borja, A., J. Franco, and I. Muxika. 2004a. The biotic indices and the Water Framework Di-rective: the required
consensus in the new benthic monitoring tools. Marine Pollution Bulletin 48:405-408
Borja, ., J. Franco, V. Valencia, J. Bald, I. Muxica, M.J. Belzunce, and O. Solaun. 2004b. Implementation of
the European water framework directive from the Basque country (northern Spain): a methodological
approach. Marine Pollution Bulletin 48:209-218
Cade, G.C., and J. Hegeman. 2002. Phytoplankton in the Marsdiep at the end of the 20th century; 30 years
monitoring biomass, primary production and Phaeocystes blooms. Journal of Sea Research 48:97-110
Cao, Y., D.D. Williams, and N.E. Williams. 1998. How important are rare species in aquatic community
ecology and bioassessment? Limnology and Oceanography 43:1403-1409
Carlson, R.E. 1977. A trophic state index for lakes. Limnology and Oceanography 22:361-369
Carstensen, J., D.J. Conley, and P. Henriksen. 2004. Frequency, composition, and causes of summer
phytoplankton blooms in a shallow coastal ecosystem, the Kattegat. Limnology and Oceanography 49:191-
201
Chapman, P.M., R.N. Dexter, and E.R. Long. 1987. Synoptic measures of sediment con-tamination, toxicity and
infaunal community composition (the Sediment Quality Triad) in San Francisco Bay. Marine Ecology Progress
Series 37:75-96
Christensen, V., and C.J. Walters. 2004. Ecopath with Ecosim: methods capabilities and limitations. Ecological
Modeling 172:109-139
Clarke, K.R., and R.M. Warwick. 1994. Change in marine communities: an approach to statistical analysis and
interpretation. 1st edition: Plymouth Marine Laboratory, Plymouth, UK, 144pp.
Cloern, J.E. 1999. The relative importance of light and nutrient limitation of phytoplankton growth: a simple
index of coastal ecosystem sensitivity to nutrient enrichment. Aquatic Ecology 33:3-16
Cloern, J.E. 2001. Our evolving conceptual model of the coastal eutrophication problem. Marine Ecology
Progress Series 210:223-253
Coffaro, G., and A. Sfriso. 1997. Simulation model of Ulva rigida growth in shallow water of the Lagoon of
Venice. Ecological Modeling 102:55-66
Colijn, F., and G.C. Cade. 2003. Is phytoplankton growth in the Wadden Sea light or nitrogen limited?
Journal of Sea Research 49:83-93
Conley, D.J., C.L. Schelske, and E.F. Stoermer. 1993. Modication of the biogeochemical cycle of silica with
eutrophication. Marine Ecology Progress Series 101:179-192
Conley, D.J. 1999. Biogeochemical nutrient cycles and nutrient management strategies. Hydrobiologia
410:87-96
Conley, D.J., H. Kaas, F. Mohlenberg, B. Rasmussen, and J. Windolf. 2000. Characteristics of Danish estuaries.
Estuaries 23:820-837
Costanza, R., R. dArge, R. deGroot, S. Farber, M. Grasso, B. Hannon, K. Limburg, S. Naeem, R.V. Oneill, J.
Paruelo, R.G. Raskin, P. Sutton, and M. vandenBelt. 1997. The value of the worlds ecosystem services and
natural capital. Nature 387:253-260
Dahl, E., O. Lindahl, E. Paasche, and J. Throndsen. 1989. The Chrysochromulina polylepis bloom in
Scandinavian waters during spring, 1988. In: E.M. Cosper, V.M. Bricelj, and E.J. Carpenter (Eds.), A Novel
Phytoplankton Bloom Causes and Impacts of Recurrent Brown Tides. Springer Verlag, New York, p 383-405
Dahl, K., J. Carstensen, and C.J.A. Lundsgaard. 2001. Hrdbundsvegetation som indikator og
mlstningsparameter. In: Henriksen, P. (Ed.), Marine Omrder, 2000: Miljtilstand og udvikling. Technical
report no. 375. NERI, Roskilde, Denmark.
Dale, V.H., and S.C. Beyeler. 2001. Challenges in the development and use of ecological indicators. Ecological
Indicators 1:3-10
Danilov, R.A., and N.G.A Ekelund. 2001. Comparative studies on the usefulness of seven ecological indices
for the marine coastal monitoring close to the shore on the Swedish East coast. Environmental Monitoring
and Assessment 66:265-279
Dauer, D.M. 1993. Biological criteria, environmental health and estuarine macrobenthic community
structure. Marine Pollution Bulletin 26:249-257
Degobbis, D., R. Precali, I. Ivancic, N. Smodlaka, D. Fuks, and S. Kveder. 2000. Long-term changes in the
northern Adriatic ecosystem related to anthropogenic eutrophication. International Journal of Environment
and Pollution 13:495-533
deBaar, H.J.H. 1994. Von Liebigs law of the minimum and plankton ecology (1899-1991). Progress in
deJonge, V.N., J. van deBergs, and D.J. deJong. 1997. Zeegras in de Waddenzee: een toekomstperspectief.
Beheersaanbevelingen voor het herstel van Groot en Klein zeegras (Zostera marina L. en Zostera noltii
Hornem). Report RIKZ-97.016, Rijskwaterstaat, Rijksinstituut voor Kust en Zee. RIKZ, The Hague.
deJonge, V.N., and D.J. deJong. 1999. Zeegras in de Nederlandse Waddenzee. Werkdocument RIKZ/OS-
99.808x, Rijksinstituut voor Kust en Zee / RIKZ, The Hague.
deJonge, V.N., M. Elliott, and E. Orive. 2002. Causes, historical development, effects and future challenges of
a common environmental problem: eutrophication. Hydrobiologia 475-476:1-19
deVries, I., C.J.M. de Philippart, E.G. de Groot, M.W.M. van der Tol. 1996. Coastal eutrophication and marine
benthic vegetation: a model analysis. In: W. Schramm, and P.H. Nienhuis (Eds.), Marine benthic vegetation -
recent changes and the effects of eutrophication. Ecological Studies, Springer-Verlag, Heidelberg, 123: p.
79-113
deVries, I., R.N.M. Duin, J.C.H. Peeters, F.J. Los, M. Bokhorst, and R. Laane. 1998. Patterns and trends in
nutrients and phytoplankton in Dutch coastal waters: comparison of time-series analysis, ecological model
simulation, and mesocosm experiments. Ices Journal of Marine Science 55:620-634
Dennison, W.C. 1987. Effects of light on seagrass photosynthesis, growth and depth distribution. Aquatic
Botany 27:15-26
Diaz, F., P. Raimbault, B. Boudjellal, N. Garcia, and T. Moutin. 2001. Early spring phosphorus limitation of
primary productivity in a NW Mediterranean coastal zone (Gulf of Lions). Marine Ecology-Progress Series
211:51-62
Dippner, J.W. 1998. Competition between different groups of phytoplankton for nutrients in the southern
North Sea. Journal of Marine Systems 14:181-198
Doblin, M.A., S.I. Blackburn, and G.M. Hallegraeff. 1999. Growth and biomass stimulation of the toxic
dinoagellate Gymnodinium catenatum (Graham) by dissolved organic substances. Journal of Experimental
Marine Biology and Ecology 236:33-47
Downing, J.A., C.W. Osenberg, and O. Sarnelle. 1999. Meta-analysis of marine nutrient-enrichment
experiments: Variation in the magnitude of nutrient limitation. Ecology 80:1157-1167
Drago, M., B. Cescon, and L. Iovenitti. 2001. A three-dimensional numerical model for eutrophication and
pollutant transport. Ecological Modelling 145:17-34
Duarte, C.M. 1991. Seagrass depth limits. Aquatic Botany 40:363-377
Duarte, C.M. 1995. Submerged aquatic vegetation in relation to different nutrient regimes. Ophelia 41:87-112
Duarte, C.M., S. Agusti, and N.S.R. Agawin. 2000. Response of a Mediterranean phytoplankton community
to increased nutrient inputs: a mesocosm experiment. Marine Ecology Progress Series, 195:61-70
Duarte, C.M. 2004. Eutrophication and dose-response relationships in European coastal waters. In: P.
Wassmann and K. Olli (Eds), Drainage basin nutrient inputs and eutrophication: an integrated approach. E-
book. www.ut.ee/
~
olli/eutr, p 139-145
Edvardsen, B., and E. Paasche. 1998. Bloom dynamics and physiology of Prymnesium and Chrysochromulina.
In: Anderson D.M., A.D. Cembella, and G.M. Hallegraeff (Eds.), Physiological Ecology of Harmful Algal
Blooms, Vol G 41. Springer-Verlag, Berlin Heidelberg, p, 193-208
Egge, J.K., and D.L. Aksnes. 1992. Silicate as regulating nutrientin phytoplankton competition. Marine
Ecology Progress Series 83:281-289
Elliott, M. 1994. The analysis of macrobenthic community data. Marine Pollution Bulletin 28:62-64
Elmgren, R., and U. Larsson. 1997. Himmerfjrden. Frndringar i ett nringsbelastet kustekosystem i
stersjn, Naturvrdsverket, Stockholm
Escaravage, V., T.C. Prins, C. Nijdam, A.C. Smaal, and J.C.H. Peeters. 1999. Response of phytoplankton communities
to nitrogen input reduction in mesocosm experiments. Marine Ecology Progress Series 179:187-199
Fabiano, M., P. Vassallo, L. Vezzulli, V.S. Salvo, and J.C. Marques. 2004. Temporal and spatial change of
exergy and ascendancy in different benthic marine ecosystems. Energy 29:1697-1712
Finni T., K. Kononen, R. Olsonen, and K. Wallstrm. 2001. The History of Cyanobacterial Blooms in the Baltic
Sea. Ambio 30:172-178
Frederiksen, M., D. Krause-Jensen, M. Holmer, and J.S. Laursen. 2004a. Long-term changes in area distribution
of eelgrass (Zostera marina) in Danish coastal waters. Aquatic Botany 78:167-181
Frederiksen, M., D. Krause-Jensen, M. Holmer, and J.S. Laursen. 2004b. Spatial and temporal variation in
eelgrass (Zostera marina) landscapes: inuence of physical setting. Aquatic Botany 78:147-165
Geertz-Hansen, O., K. Sand-Jensen, D.F. Hansen, and A. Christiansen. 1993. Growth and grazing control of
abundance of the marine macroalga Ulva lactuca L. in a eutrophic Danish estuary. Aquatic Botany 46:101-109
Gotsis-Skretas, O., U. Horstmann, and B. Wiryawan. 2000. Cell size structure of phytoplankton communities
in relation to physico-chemical parameters and zooplankton in a temperate coastal environment. Archive of
Fishery and Marine Research 48:265-282
Gowen, R.J., P. Tett, and K.J. Jones. 1992. Predicting marine eutrophication: The yield of chlorophyll from
nitrogen in Scottish coastal waters. Marine ecology progress series 85:153-161
Grall, J., and M. Glmarec. 1997. Using biotic indices to estimate macrobenthic commu-nity perturbations in
the Bay of Brest. Estuarine, Coastal and Shelf Science 44 (Suppl. A):43-53
Granli, E., P. Carlsson, and C. Legrand. 1999. The role of C, N and P in dissolved and particulate organic
matter as a nutrient source for phytoplankton growth, including toxic species. Aquatic Ecology 33:17-27
Granli, E. 2004. Eutrophication and harmful algal blooms. In: P. Wassmann and K. Olli (Eds), Drainage
basin nutrient inputs and eutrophication: an integrated approach. E-book. www.ut.ee/
~
olli/eutr, p 99-112
Green, E.P., and F.T. Short. 2003. World Atlas of Seagrasses, University of California Press, Berkely, USA
Greve, T.M., J. Borum, and O. Pedersen. 2003. Meristematic oxygen variability in eelgrass (Zostera marina).
Limnology and Oceanography 48:210-216
Greve, T.M., and D. Krause-Jensen. 2005. Predictive modelling of eelgrass (Zostera marina L.) depth limits.
Marine Biology 146:849-858
Grnlund, L., K. Kononen, E. Lahdes, and K. Mkel. 1996. Community development and modes of phosphorus
utilization in a late summer ecosystem in the central Gulf of Finland, the Baltic Sea. Hydrobiologia 331:97-108
Hallegraeff, G.M. 1993. A review of harmful algal blooms and their apparent global increase. Phycologia
32:79-99
Hllfors, G., I. Viitasalo, and . Niemi. 1987. Macrophyte vegetation and trophic status of the Gulf of Finland -
A review of Finnish investigations. Meri 13:111-158
Hllfors, G. 2004. Checklist of Baltic Sea Phytoplankton Species: including some heterotrophic protistan
groups. Baltic Marine Environment Protection Commission - HELCOM, Helsinki, 210 p.
Hauxwell, J., J. Cebrian, C. Furlong, and I. Valiela. 2001. Macroalgal canopies contribute to eelgrass (Zostera
marina) decline in temperate estuarine ecosystems. Ecology 82:1007-1022
Hauxwell, J., J. Cebrian, and I. Valiela. 2003. Eelgrass Zostera marina loss in temperate estuaries: relationship
to land-derived nitrogen loads and effect of light limitation imposed by algae. Marine Ecology Progress Series
247:59-73
Hemminga, M.A., and C.M. Duarte. 2000. Seagrass Ecology, Cambridge University Press, Cambridge.
Hily, C., H. Le Bris, and M. Glmarec. 1986. Impacts biologiques des emissaires urbains sur les ecosystemes
benthiques. Ocanis 12:419-426
Hoch, T., and A. Mnesguen. 1997. Modelling the biogeochemical cycles of elements limiting primary
production in the English Channel. II. Sensitivity analyses. Marine Ecology Progress Series 146:189-205
Holmer, M., and E.J. Bondgaard. 2001. Photosynthetic and growth response of eelgrass to low oxygen and
high sulphide concentrations during hypoxic events. Aquatic Botany 70:29-38
Humborg, C., D.J. Conley, L. Rahm, F. Wulff, A. Cociasu, and V. Ittekkot. 2000a. Silicon retention in river
basins: Far-reaching effects on biogeochemistry and aquatic food webs in coastal marine environments.
Ambio 29:45-50
Humborg, C., K. Fennel, M. Pastuszak, and W. Fennel. 2000b. A box model approach for a long-term assessment
of estuarine eutrophication, Szczecin Lagoon, southern Baltic. Journal of Marine Systems 25:387-403
Hurlbert, H. 1971. The nonconcept of species diversity: a critique and alternative parameters. Ecology 52:
577-586
Hyland, J., I. Karakassis, P. Magni, A. Petrov, and J. Shine. 2000. Ad hoc Benthic Indicator Group: results of
initial planning meeting, Paris, France, 6-9 December 1999. Unesco, IOC Technical Series N 57.
ICES. 2000. Report of the working group on ecosystem effects of shing activities, ICES, Copenhagen..
Iriarte, A., D. Purdie, and C. Holeton. 2004. Pilot study: English coastal waters. In: HABES, Harmful Algal
Blooms Expert System, nal report. WL Delft Hydraulics, Delft
Jickells, T.D. 1998. Nutrient biogeochemistry of the coastal zone. Science 281:217-222
Johnson, M.P., M.J. Costello, and D. ODonnell. 2000. Techniques for the objective detection of pattern in
chlorophyll samples at different temporal scales: an example using data from Lough Hyne, a seasonally
stratied inlet. Hydrobiologia 421:103-113
Jrgensen, B.B. 1996. Case study - Aarhus Bay. In: B.B. Jrgensen, and K. Richardson (Eds.). Eutrophication
in Coastal Marine Ecosystems, Vol 52. American Geophysical Union, Washington DC, p 137-154
Jrgensen, S.E., and J.C. Marques. 2001. Thermodynamics and ecosystem theory, case studies from
hydrobiology. Hydrobiologia 445:1-10
Justic, D. 1988. Trend in the transparency of the northern Adriatic Sea. Marine Pollution Bulletin, 19:32-35
Justic, D., N.N. Rabalais, and R.E. Turner. 1995a. Stoichiometric nutrient balance and origin of coastal
eutrophication. Marine Pollution Bullettin 30:41-46
Justic, D., N.N. Rabalais, R.E. Turner, and Q. Dortch. 1995b. Changes in nutrient structure of river-dominated
coastal waters: stoichiometric nutrient balance and its consequences. Estuarine, Coastal and Shelf Science
40:339-356
Kaas, H., F. Mohlenberg, A. Josefson, B. Rasmussen, D. Krause-Jensen, H.S. Jensen, L.M. Svendsen, J. Windolf,
A.L. Middelboe, and M.F. Pedersen. 1996. Marine omrder. Danske fjorde - status over miljotilstnd,
rsagssammenhaenge og udvikling. Vandmiljoplanens Overvgningsprogram, 1995. Danmarks
Miljoundersogelser/Miljo- og Energiministeriet, Roskilde/Copenhagen, Denmark
Kabuta, S.H., and H. Duijts. 2000. Graadmeters voor de Noordzee: eindrapport van het project
Graadmeterontwikkeling Noordzee (GONZ III). - Den Haag, The Netherlands: Ministerie van Verkeer en
Waterstaat, Directoraat-Generaal Rijkswaterstaat & Rijksinstituut voor Kust en Zee (RIKZ). - 104 p.
Kahru, M., U. Horsttmann, and O. Rud. 1994. Satellite detection of increased cyanobacteria blooms in the
Baltic Sea: natural uctuation or ecosystem change? Ambio 23:469-472
Kaiser, M.J., E.F. Pence, and P.J. Hart. 2000. Fishing-gear restrictions and conservation of benthic habitat
complexity. Conservation Biology 14:1512-1525
Karez, R., S. Engelbert, P. Kraufvelin, M.F. Pedersen, and U. Sommer. 2004. Biomass response and changes in
composition of ephemeral macroalgal assemblages along an experimental gradient of nutrient enrichment.
Aquatic Botany 78:103-117
Karydis, M., and G. Tsirtsis. 1996. Ecological indices: A biometric approach for assessing eutrophication levels
in the marine enironment. The Science of the Total Environment 186:209-219
Kauppila, P., P. Hallfors, P. Kokkonen, and S. Basova. 1995. Late summer phytoplankton species composition
and biomasses in the Eastern Gulf of Finland. Ophelia 42:179-191
Kautsky, H. 1999. Vegetationskldde bottnar. Dramatiske forndringar sedan, 1970-talet stersj 98 -
Swedish monitoring in the Baltic Sea, 1999, p 7-10
Kelly-Gerreyn, B.A., T.R. Anderson, J.T. Holt, R.J. Gowen, and R. Proctor. 2004. Phytoplankton community
structure at contrasting sites in the Irish Sea: a modelling investigation. Estuarine Coastal and Shelf Science
59:363-383
Kiirikki, M., J. Haapamaki, J. Koponen, A. Ruuskanen, and J. Sarkkula. 1998. Linking the growth of lamentous
algae to the 3D-ecohydrodynamic model of the Gulf of Finland. Environmental Modelling and Software
13:503-509
Kivi, K., S. Kaitala, H. Kuosa, J. Kuparinen, E. Leskinen, R. Lignell, B. Marcussen and T. Tamminen. 1993.
Nutrient limitation and grazing control of the Baltic plankton community during annual succession.
Koch, E.M. 2001. Beyond light: Physical, geological, and geochemical parameters as possible submersed
aquatic vegetation habitat requirements. Estuaries 24:1-17
Kononen, K., J. Kuparinen, K. Makela, J. Laanemets, J. Pavelson, and S. Nommann. 1996. Initiation of
cyanobacterial blooms in a frontal region at the entrance to the Gulf of Finland, Baltic Sea. Limnology and
Krause-Jensen, D., A.L. Middelboe, K. Sand-Jensen, and P.B. Christensen. 2000. Eelgrass, Zostera marina, growth
along depth gradients: upper boundaries of the variation as a powerful predictive tool. Oikos 91:233-244
Krause-Jensen, D., M.F. Pedersen, and C. Jensen. 2003. Regulation of eelgrass (Zostera marina) cover along
depth gradients in Danish coastal waters. Estuaries 26:866-877
Krause-Jensen, D., J. Carstensen, and K. Dahl. In press. Cover and composition of coastal macroalgae in
relation to water quality. Marine Pollution Bullettin.
Kristiansen, S., and E.E. Hoell. 2002. The importance of silicon for marine production. Hydrobiologia 484:21-31
Krom, M.D., N. Kress, and S. Brenner. 1991. Phosphorus limitation of primary productivity in the eastern
Mediterranean Sea. Limnology and Oceanography 36:424-432
Kuosa, H. 1990. Picoplanktonic cyanobacteria in the northern Baltic Sea: role in the phytoplankton
community. In: M. Barnes, and R.N. Gibson (Eds.), Proceedings of the 24
th
European Marine Biology
Symposium. Aberdeen University Press, p 11-17
Lancelot, C., G. Billen, A. Sournia, T. Weisse, F. Colijn, M.J.W. Veldhuis, A. Davies, and P. Wassman. 1987. Phaeocystis
blooms and nutrient enrichment in the continental coastal zones of the North Sea. Ambio 16:38-46
Lancelot, C., J. Staneva, D. VanEeckhout, J.M. Beckers, and E. Stanev. 2002. Modelling the Danube-inuenced
north-western continental shelf of the Black Sea. II: Ecosystem response to changes in nutrient delivery by
the Danube River after its damming in, 1972. Estuarine Coastal and Shelf Science 54:473-499
Lancelot, C., Y. Spitz, N. Gypens, K. Ruddick, S. Becquevort, V. Rousseau, G. Lacroix, and G. Billen. 2005.
Modelling diatom and Phaeocystis blooms and nutrient cycles in the Southern Bight of the North Sea: the
MIRO model. Marine Ecology Progress Series 289:63-78
Le Pape, O,Y. Del Amo, A.Menesguen, A. Aminot, B. Quequiner, and P. Treguer. 1996 Resistance of a coastal
ecosystem to increasing eutrophic conditions: the Bay of Brest (France), a semi-enclosed zone of Western
Europe. Continental Shelf Research 16:1885-1907
Le Pape, O., and A. Menesguen. 1997. Hydrodynamic prevention of eutrophication in the Bay of Brest
(France), a modelling approach. Journal of Marine Systems 12:171-186
Lignell, R., S. Kaitala, and H. Kuosa. 1992. Factors controlling phyto- and bacterioplankton in late spring on
a salinity gradient in the northern Baltic Sea. Marine Ecology Progress Series 84:121-131
Long, E.R., and P.M. Chapman. 1985. A sediment quality triad: measures of sediment con-tamination, toxicity
and infaunal community in Puget Sound. Marine Pollution Bulletin 16:405-415
Los, F.J., and M. Bokhorst-1997. Trend analysis Dutch coastal zone, 1975 -, 1994. Berichte aus dem Zentrum
fr Meers- und Klimaforschung Reihe Z: Interdisziplinre Zentrumsberichte 2:161-175
Lyons, J., J. Ahern, J. McClelland, and I. Valiela. 1995. Macrophyte Abundances in Waquoit Bay Estuaries
Subject to Different Nutrient Loads and the Potential Role of Fringing Salt-Marsh in Groundwater Nitrogen
Interception. Biological Bulletin 189:255-256
Magurran, A.E., 1988. Ecological diversity and its measurements. Princeton University Press, Princeton
Markager, S., and L. Storm. 2003. Assessment of environmental effects in the marine environment part 1:
Empirical modeling of environmental conditions in the open Inner Danish waters. Danish Environmental
Research Institute.
Marques, J.C., and S.E. Jrgensen. 2002. Three selected ecological observations inter-preted in terms of a
thermodynamic hypothesis. Contribution to a general theoretical framework. Ecological Modeling 158:213-221
Meeuwig, J.J., J. Rasmussen, and R.H. Peters. 1998. Turbid waters and clarifying mussels: their moderation of
Chl:nutrient relations in estuaries. Marine Ecology Progress Series 171:139-150
Meeuwig, J.J., P. Kauppila, and H. Pitkanen. 2000. Predicting coastal eutrophication in the Baltic: a
limnological approach. Canadian Journal of Fisheries and Aquatic Sciences 57:844-855
Mnesguen, A., and T. Hoch. 1997. Modelling the biogeochemical cycles of elements limiting primary
production in the English Channel. I. Role of thermohaline stratication. Marine Ecology Progress Series
146:173-188
Middelboe, A.L., and K. Sand-Jensen. 2000. Long-term changes in macroalgal communities in a Danish
estuary. Phycologia 39:245-257
Moestrup, . 2004. IOC Taxonomic Reference List of Toxic Algae, Intergovernmental Oceanographic
Commission of UNESCO.
Moll, A., and G. Radach. 2003. Review of three-dimensional ecological modelling related to the North Sea
shelf system - Part 1: Models and their results. Progress in Oceanography 57:175-217
Monbet, Y. 1991. Control of phytoplankton biomass in estuaries: A comparative analysis of microtidal and
macrotidal estuaries. Estuaries 15: 563-571
Moncheva, S., V. Dontcheva, G. Shtereva, L. Kamburska, A. Malej, and S. Gorinstein. 2002. Application of
eutrophication indices for assessment of the Bulgarian Black Sea coastal ecosystem ecological quality. Water,
Morand, P., and X. Briand. 1996. Excessive growth of macroalgae: A symptom of environmental disturbance.
Botanica Marina 39:491-516
Nedwell, D.B., L.F. Dong, A. Sage, and G.J.C. Underwood. 2002. Variations of the nutrients loads to the
mainland UK estuaries: Correlation with catchment areas, urbanization and coastal Eutrophication. Estuarine,
Coastal and Shelf Science 54:951-970
Nielsen, S.L., K. Sand-Jensen, J. Borum, and O. Geertz-Hansen. 2002a. Phytoplankton, Nutrients, and
Transparency in Danish Coastal Waters. Estuaries 25:930-937
Nielsen, S.L., K. Sand-Jensen, J. Borum, and O. Geertz-Hansen. 2002b. Depth colonization of eelgrass (Zostera
marina) and macroalgae as determined by water transparency in Danish coastal waters. Estuaries 25:
1025-1032
Nilsson, H.C., and R. Rosenberg. 1997. Benthic habitat quality assessment of an oxygen stressed fjord by
surface and sediment prole images. Journal of Marine Systems 11:249-264
Ofcer, C.B., and J.H. Ryther. 1980. The possible importance of silicon in marine eutrophication. Marine
Olenina, I., 1998. Long-term changes in the Kursiu Marios Lagoon: Eutrophication and phytoplankton
response. Ekologija 1:56-65
Olsen, Y., H. Reinertsen, O. Vadstein, T. Andersen, I. Gismervik, C.M. Duarte, S. Agusti, H. Stibor, U. Sommer,
R. Lignell, T. Tamminen, C. Lancelot, V. Rousseau, E. Hoell, and K.A. Sanderud. 2001. Comparative analysis of
food webs based on ow networks: effects of nutrient supply on structure and function of coastal plankton
communities. Continental Shelf Research 21:2043-2053
Olsgard, F., P.J. Somereld, and M.R. Carr. 1997. Relationships between taxonomic resolution, macrobenthic
community patterns and disturbance. Marine Ecology Progress Series 172:25-36
Orfanidis, S., P. Panayotidis, and N. Stamatis. 2001. Ecological evaluation of transitional and coastal waters:
a marine benthic macrophytes-based model. Mediterranean Marine Science 2:45-65
Orfanidis, S., P. Panayotidis, and N. Stamatis. 2003. An insight to the ecological evaluation index (EEI).
Ecological Indicators 3:27-33
Ospar. 2003. OSPAR integrated report 2003 on the eutrophication status of the OSPAR maritime area based
upon the rst application of the comprehensive procedure. OSPAR Commission, London.
Paerl, H.W. 1998. Structure and function of anthropogenically altered microbial communities in coastal
waters. Current Opinion in Microbiology 1:296-302
Paerl, H.W., and D.R. Whitall. 1999. Anthropogenically-derived atmospheric nitrogen deposition, marine
eutrophication and harmful algal bloom expansion: Is there a link? Ambio 28:307-311
Parker, M. 1987. Exceptional plankton blooms conclusion of discussions: Coveners report. Rapport et Proces-
verbaux des Reunions, Conseil international pour lExploration de la Mer, 187: 108-114.
Pearson, T.H., and R. Rosenberg. 1978. Macrobenthic succession in relation to organic enrichment and
pollution of the marine environment. Oceanography Marine Biology Annual Review 16:229-311
Pedersen, M.F. 1995. Nitrogen limitation of photosynthesis and growth: comparison across aquatic plant
communities in a Danish estuary (Roskilde Fjord). Ophelia 41:261-272
Pedersen, M., and P. Snoeijs. 2001. Patterns of macroalgal diversity, community composition and long-term
changes along the Swedish west coast. Hydrobiologia 459:83-102
Perez-Ruzafa, A., J. Gilabert, J.M. Gutierrez, A.I. Fernandez, C. Marcos, and S. Sabah. 2002. Evidence of a
planktonic food web response to changes in nutrient input dynamics in the Mar Menor coastal lagoon,
Spain. Hydrobiologia 475:359-369
Philipart, C.J.M., and P.M.J. Herman. 1998. Multivariate analyse van fytoplankton tijdseries, een pilot-study
naar de geschiktheid van de structuur van de algengemeenschap als graadmeter voor het ecologisch
functioneren van de Noordzee.
Philippart, C.J.M., K.S. Dijkema, and N. Dankers. 1992. De huidige verspreiding en de mogelijke toekomst van
het litoraal zeegras in de Nederlandse Waddenzee. Report No. 92/10, Instituut voor Bos- en Natuuronderzoek
(IBN-DLO), Texel
Philippart, C.J.M. 1994. Eutrophication as a possible cause of decline in the seagrass Zostera marina noltii
of the Dutch Wadden Sea. Ph.D. Thesis, Wageningen University
Philippart, C.J.M., G.C. Cadee, W. vanRaaphorst, and R. Riegman. 2000. Long-term phytoplankton-nutrient
interactions in a shallow coastal sea: Algal community structure, nutrient budgets, and denitrication
potential. Limnology and Oceanography 45:131-144
Pielou, E.C. 1975. Ecological diversity. Wiley, New York
Prins, T.C., V. Escaravage, L.P.M.J. Wetsteyn, J.H. Peeters, and A.C. Smaal. 1999. Effects of different N- and
P-loading on primary and secondary production in an experimental marine ecosystem. Aquatic Ecology
33:65-81
Proctor, R., J.T. Holt, J.I. Allen, and J.C. Blackford. 2003. Nutrient uxes and budgets for the North West
European Shelf from a 3-dimensional model. The Science of the Total Environment 314-316:769-785
Pryor, S.C., and R.J. Barthelmie. 2000. Particle dry deposition to water surfaces: Processes and consequences.
Marine Pollution Bulletin 41:220-231
Radenac, G., P. Miramand, and J. Tardy. 1997. Search for impact of a dredged material disposal site on growth
and metal contamination of Mytilus edulis (L.) in Charente-Maritime (France). Marine Pollution Bulletin
34:721-729
Rask, N., S.E. Pedersen, and M.H. Jensen. 1999. Response to lowered nutrient discharges in the coastal
waters around the island of Funen, Denmark. Hydrobiologia 393:69-81
Redeld, A.C. 1958. The biological control of chemical factors in the environment. American Scientist
46:205-222
Rees, H.L., D.C. Moore, T.H. Pearson, M. Elliott, M. Servile, J. Pomfret, and D. Johnson. 1990. Procedures for
the monitoring of marine benthic communities at UK sewage sludge disposal sites. Scottish Fisheries
Information Pamphlet 18:1-79
Rhoads, D.C., P.L. Mccall, and J.Y. Yingst-1978. Disturbance and production on the estuarine sea oor.
American Scientist 66:577-578
Richardson, K., and B.B. Jrgensen. 1996. Eutrophication: Denition, History and Effects. In: B.B. Jorgensen,
and K. Richardson (Eds.), Eutrophication in marine coastal ecosystems, Vol 52. American Geophysical Union,
Washington DC, p 1-20
Riegman, R., F.J. Los, and J.C.H Peeters. 1998. BEON-Garden: giftige algen en de reductie van de
nutrintenbelasting, eindrapport. BEON Rapport, 98(4). RIKZ, Den Haag, The Netherlands. 53 pp.
RIKZ. 2004. Groot Zeegras terug in de Waddenzee. Experiment in de westelijke Waddenzee., Brochure,
Rijksinstituut voor Kust en Zee. RIKZ, Den Haag, The Netherlands.
Robinson, L.A., and C.L.J Frid. 2003. Dynamics ecosystem models and the evaluation of eco-system effects of
shing: can we make meaningful predictions? Aquatic Conservation: Marine and Freshwater Ecosystems
13:5-20
Rosenberg, R., M. Blomqvist, HC. Nilsson, LH. Cederwal, and A. Dimming. 2004 Marine quality assessment
by use of benthic species-abundance distributions: a proposed new protocol within the European Union
Water Framework Directive. Marine Pollution Bulletin 49:728-739
Ryther, J.H., and W.M. Dunstan. 1971 Nitrogen, phosphorus, and eutrophication in the coastal marine
environment. Science 171:1008-1013
Sagert, S., D. KrauseJensen, P. Henriksen, T. Rieling, and H. Schubert. 2005 Integrated ecological assessment
of Danish Baltic Sea coastal areas by means of phytoplankton and macrophytobenthos. Estuarine, Coastal
and Shelf Science 63:109-118.
Sand-Jensen, K., and J. Borum. 1991. Interactions among Interactions among phytoplankton, periphyton, and
macrophytes in temperate freshwaters and estuaries. Aquatic Botany 41:137-175
Sanden P., and B. Hakansson. 1996. Long-term trends in Secchi depth in the Baltic Sea. Limnology and
Sanders, H.L. 1968 Marine benthic diversity: a comparative study. American Naturalist 102:243-282
Sanders, R., T. Jickells, and D. Mills. 2001. Nutrients and chlorophyll at two sites in the Thames plume and
southern North Sea. Journal of Sea Research 46:13-28
Savage, C., J.G. Field, and R. Warwick. 2001. Comparative meta-analysis of the impact of off-shore marine
mining on macrobenthic communities versus organic pollu-tion studies. Marine Ecology Progress Series
221:265-275
Schlter, L., F. Mhlenberg, and H. Havskum. 2000. The use of phytoplankton pigments for identifying and
quantifying phytoplanakton groups in coastal areas; testing the inu-ence of light and nutrients on
pigments/chlorophyll a-ratios. Marine Ecology Progress Series, 192:49-63
Schramm, W. 1999. Factors inuencing seaweed responses to eutrophication: some results from EU-project
EUMAC. Journal of Applied Phycology 11:69-78
Schramm, W., and P.H. Nienhuis. 1996. Marine benthic vegetation. Recent changes and the effects of
eutrophication. Springer, Berlin
Sellner, K. 1997. Physiology, ecology, and toxic properties of marine cyanobacterial blooms. Limnology and
Sellner, K.G., S.G. Sellner, R.V. Lacouture, and R.E. Magnien. 2001. Excessive nutrients select for dinoagellates
in the stratied Patapsco River estuary: Margalef reigns. Marine Ecology Progress Series 220:93-102
Seppl, J., T. Tamminen, and S. Kaitala. 1999. Experimental evaluation of nutrient limitation of phytoplankton
communities in the Gulf of Riga. Journal of Marine Systems 23:107-126
Short, F.T., and D.M. Burdick. 1996. Quantifying eelgrass habitat loss in relation to housing development and
nitrogen loading in Waquoit Bay, Massachusetts. Estuaries, 19:730-739
Short, F.T., and S. WyllieEcheverria. 1996. Natural and human-induced disturbance of seagrasses.
Environmental Conservation 23:17-27
Shtereva, G., S. Moncheva, V. Doncheva, O. Christova, and I. Shterev. 1999. Changes in chemical parameters
in the Bulgarian Black Sea coastal area as an indication of the ecological state of the environment. Water
Simboura, N., A. Zenetos, P. Panayotidis, and A. Makra. 1995. Changes in benthic community structure along
an environmental pollution gradient. Marine Pollution Bulletin 30:470-474
Simboura, N., and A. Zenetos. 2002. Benthic indicators to use in ecological quality classication of
Mediterranean soft bottom marine ecosystems including a new biotic index. Mediterranean Marine Science
3:77-111
Simboura, N., 2004. Bentix index vs. Biotic Index in monitoring: an answer to Borja et al., 2003. Marine
Pollution Bulletin 48:404-405
Sivonen, K., K. Kononen, W.W. Carmichael, A.M. Dahlem, K.L. Rinehart, J. Kiviranta, and S.I. Niemel. 1989.
Occurrence of the hepatotoxic cyanobacterium Nodularia spumigena in the Baltic Sea and structure of the
toxin. Applied and Environmental Microbiology 55:1990-1995
Skogen, M.D., and H. Soiland. 1998. A users guide to NORWECOM V2.0. The Norwegian Ecological Model
System. The Norwegian Ecological Model System. Fisken.
Smayda, J.J., and C.S. Reynolds. 2001. Community assembly in marine phytoplankton: application of recent
models to harmful dinoagellate blooms. Journal of Plankton Research 23:447-461
Smayda, T.J. 1997. What is a bloom? A commentary. Limnology and Oceanography 42:1132-1136
Smayda, T.J. 2004. Eutrophication and phytoplankton. In: P. Wassmann and K. Olli (Eds), Drainage basin
nutrient inputs and eutrophication: an integrated approach. E-book. www.ut.ee/
~
olli/eutr, p 89-98
Smith, C.R., and S.J. Brumsickle. 1989. The effect of patch size and substrate isolation on colonization modes
and rate in an intertidal sediment. Limnology and Oceanography 34:1263-1277
Smith, S.V. 1981. Marine Macrophytes as a Global Carbon Sink. Science 211:838-840
Solidoro, C., V.E. Brando, C. Dejak, D. Franco, R. Pastres, and G. Pecenik. 1997. Long term simulations of
population dynamics of Ulva r. in the lagoon of Venice. Ecological Modelling 102:259-272
Somereld, P.J., and K.R. Clarke. 1995. Taxonomic levels, in marine community studies, revisited. Marine
Sommer, U. 1995. Eutrophication related changes in phytoplankton species composition: Is there a role of
nutrient competition? ICES annual science conference, T:7, Aalborg, Denmark
Sommer, U., H. Stibor, A. Katechakis, F. Sommer, and T. Hansen. 2002. Pelagic food web congurations at
different levels of nutrient richness and their implications for the ratio sh production: primary production.
Sndergaard, M. 1991. Phototrophic picoplankton in temperate lakes: seasonal abundance and importance
along a trophic gradient. Internationale Revue der gesamten Hydrobiologie 76:505-522
Stal, L.J., M. Staal, and M. Villbrandt. 1999. Nutrient control of cyanobacterial blooms in the Baltic Sea.
Aquatic Microbial Ecology 18:165-173
Stockner, J.G. 1988. Phototrophic picoplankton: an overview from marine and freshwater ecosystems.
Stolte, W., R. Panosso, L.A. Gisselson, and E. Graneli. 2002. Utilization efciency of nitrogen associated with
riverine dissolved organic carbon (>1 kDa) by two toxin-producing phytoplankton species. Aquatic Microbial
Ecology 29:97-105
Stronkhorst, J., R.N.D. Duin, and H.A. Haas. 1985. Primaire produktie onderzoek in het Veerse Meer. 1982-
1983). Nota delta dienst Milieu onderzoek, 85.01, Middelburg, p. 52
Terrados, J., C.M. Duarte, L. Kamp-Nielsen, N.S.R. Agawin, E. Gacia, D. Lacap, M.D. Fortes, J. Borum, M.
Lubanski, and T. Greve. 1999. Are seagrass growth and survival constrained by the reducing conditions of the
sediment? Aquatic Botany 65:175-197
Tett, P. 1987. The ecophysiology of exceptional blooms. Rapport et Proces-verbaux des Reunions, Conseil
international pour lExploration de la Mer, 187:47-60
Tett, P., L. Gilpin, H. Svendsen, C.P. Erlandsson, U. Larsson, S. Kratzer, E. Fouilland, C. Janzen, J.Y. Lee, C. Grenz,
A. Newton, J.G. Ferreira, T. Fernandes, and S. Scory. 2003. Eutrophication and some European waters of
restricted exchange. Continental Shelf Research 23:1635-1671
Thingstad, T.F., U.L. Zweifel, and F. Rassoulzadegan. 1998. P limitation of heterotrophic bacteria and
phytoplankton in the Northwest Mediterranean. Limnology and Oceanography 43:88-94
Thomann, R.V., and L.C. Linker. 1998. Contemporary issues in watershed and water quality modeling for
eutrophication control. Water Science and Technology 37:93-102
Thrush, S.F., R.B. Whitlatch, R.D. Pridmore, J.E. Hewitt, V.J. Cummings, and M.R. Wilkinson. 1996. Scale-
dependent recolonization: the role of sediment stability in a dynamic sandat habitat. Ecology 77:2471-
2487
Torn, K., D. Krause-Jensen, and G. Martin. 2006. Present and past depth distribution of bladderwrack (Fucus
vesiculosus) in the Baltic Sea. Aquatic Botany 84:53-62
Tsirtsis, G., and M. Karydis. 1998. Evaluation of phytoplankton community indices for detecting eutrophic
trends in the marine environment. Environmental Monitoring and Assessment 50:255-269
Turner, R.E., N.N. Nancy, D. Justic, and Q. Dortch. 2003. Future aquatic nutrient limitations. Marine Pollution
Bulletin 46:1032-1034
Valiela, I., J. McClelland, J. Hauxwell, P.J. Behr, D. Hersh, and K. Foreman. 1997. Macroalgal blooms in shallow
estuaries: Controls and ecophysiological and ecosystem consequences. Limnology and Oceanography
42:1105-1118
van Gils, J., S. and Tatman. 2002. Light penetration in the water column. Report No. Z3379, WL Delft,
Delft
Viaroli, P., M. Naldi, C. Bondavalli, and S. Bencivelli. 1996. Growth of the seaweed Ulva rigida C. Agardh in
relation to biomass densities, internal nutrient pools and external nutrient supply in the Sacca di Goro
lagoon (Northern Italy). Hydrobiologia 329:93-103
Viitasalo, I., U-M. Hyytiinen, S. Pekuri, S-P. Saarnio, and H. Toppinen. 2002. Rantavyhykkeen
uposvesikasvillisuuden tila Helsingin ja Espoon merialueilla vuosina, 1998-99. Helsingin kaupungin
ympristkeskuksen julkaisuja 4
Vollenweider, R.A., and J.J. Kerekes. 1982. OECD. Organization for Economic Cooperation and Development).
Eutrophication of waters. Monitoring, assessment and control. OECD 164, Paris
Vollenweider, R.A., F. Giovanardi, G. Montanari, and A. Rinaldi. 1998. Characterization of the trophic
conditions of marine coastal waters with special reference to the NW Adriatic Sea: Proposal for a trophic
scale, turbidity and generalized water quality index. Environmetrics 9:329-357
Warwick, R.M. 1986. A new method for detecting pollution effects on marine macroben-thic communities.
Marine Biology 92:557-562
Warwick, R.M., and Ruswahyuni. 1987. Comparative study of the structure of some tropi-cal and temperate
marine soft-bottom macrobenthic communities. Marine Biology 95:641-649
Warwick, R.M. 1988. The level of taxonomic discrimination required to detect pollution effects on marine
benthic communities. Marine Pollution Bulletin, 196:259-268
Warwick, R.M. 1993. Environmental impact studies on marine communities: pragmatical considerations.
Australian Journal of Ecology 18:63-80
Warwick, R.M., and K.R. Clarke. 1993. Comparing the severity of disturbance: a meta-analysis of marine
macrobenthic community data. Marine Ecology Progress Series 92:221-231
Warwick, R.M., T.M. Pearson, and Ruswahyuni. 1987. Detection of pollution effects on marine macrobenthos:
further evaluation of the species abun-dance/biomass method. Marine Biology 95:193-200
Washington, H.G. 1984. Diversity, biotic and similarity indices. Water Research 18:653-694
Wasmund, N., M. Zalewski, and S. Busch. 1999. Phytoplankton in large river plumes in the Baltic Sea. ICES
Journal of Marine Science Suppl. 56:23-32
Wasmund, N., A. Andrushaitis, E. Lysiak-Pastuszak, B. Muller-Karulis, G. Nausch, T. Neumann, H. Ojaveer, I.
Olenina, L. Postel, and Z. Witek. 2001. Trophic status of the south-eastern Baltic Sea: A comparison of coastal
and open areas. Estuarine Coastal and Shelf Science 53:849-864
Wasmund, N., and S. Uhlig. 2003. Phytoplankton trends in the Baltic Sea. ICES Journal of Marine Science
60:177-186
Weisberg, S.B., J.A. Ranasinghe, D.M. Dauer, L.C. Schaffner, R.J. Diaz, and J.B. Frithsen. 1997. An estuarine
benthic index of biotic integrity (B-IBI) for Chesapeake Bay. Estuaries 20:149-158
Weston, D.P. 1990. Quantitative examination of macrobenthic community changes along an organic
enrichment gradient. Marine Ecology Progress Series 61:233-244
Wilkinson, M., and P. Wood. 2003. Type-specic reference conditions for macroalgal and angiosperms in
Scottish transitional and coastal waters. Final report. Sepa project reference 230/4136. Heriot-Watt
University, Edinburgh. 114 pp
Xu. F-L, S.E. Jrgensen, and S. Tao. 1999. Ecological indicators for assessing freshwater eco-system health.
Ecological Modeling 116:77-106
Zhang, J. 2000. Evidence of trace metal limited photosynthesis in eutrophic estuarine and coastal waters.
Zurlini, G. 1996. Multiparametric classication of trophic conditions. The OECD methodology extended:
combined probabilities and uncertainties -- application to the North Adriatic sea. The Science of the Total
Environment 182:169-185
Annex to chapter2
Diversity indices used for phytoplankton community assessment with respect to coastal eutrophication
(Karydis and Tsirtsis, 1996; Tsirtsis and Karydis, 1998; Danilov and Ekelund 1999; Arhonditsis et al.
2003b). Original references are cited in these publications. See also (Washington 1984).
1) Diversity indices
Simpsons D
) (
) (
( )

( )
Kothes species decit

) (
) (
( )

( )
Odums species per 1000 individuals

) (
) (
( )

( )
Gleasons D
) (
) (
( )

( )
Margalefs D
) (
) (
( )

( )
Menhinicks D
) (
) (
( )

( )
Shannons H
) (
) (
( )

( )
Brillouins H
) (
) (
( )

( )
Hills N 0 S
Hills N 1
) (
) (
( )

( )
Hills N 2
) (
) (
( )

( )
Hurlberts PIE
) (
) (
( )

( )
McIntoshs M

Keefes TU

2) Evenness indices
Evenness E1

Evenness E2


Evenness E3

Evenness E4

Evenness E5

Redundancy R

Jaccards J

3) Dominance indices
Berger-Parker

McNaughtons

4) Saprobic indices
Saprobic index

5) Similarity indices
Gleasons SI
GL

Kulczynskis SI
KU

Ruzickas SI
RU

Pandeyas SI
PA

Ellenbergs SI
EL

Whittakers index of association SI
WH

Canberra metric SI
CA

Morisitas index SI
MO

Bray-Curtis index SI
BC

S = the number of species in a sample or a population
N = the number of individuals in a population or a community
N
i
= the number of individuals in a species i of a population or community
n = the number of individuals in a sample from a population
n
i
= the number of individuals in a species i of a sample from a population
p
i
= n
i
/n = the fraction of a sample of individuals belonging to species I
S
max
= the maximum number of species in a sample
n
1
, n
2
= the number of individuals in the two most abundant species
H
max
= ln S (Pielou 1975)
H
min
= (1/N) * ln[N!/(N S + 1)] (Pielou 1975)
t
i
= the number of species in the population i
t
j
= the number of species in the population j
t
c
= the number of species common in the populations I and j
K = saprobic index
k
i
= saprobic value of indicator species i
x
i
= abundance of the ith phytoplankton species at site x
y
i
= abundance of the ith phytoplankton species at site y
T = subset of species occurring in both samples
U, V = subsets of species occurring only in one sample or the other
D
1
, D
2
= Simpsons indices of the two samples
3 Nutrients and eutrophication in rivers
Amelie Deandre and Helen Jarvie
3.1 Introduction
The impacts of inorganic nutrient enrichment on river ecosystems are seen in many rivers and
have been intensively studied. For the implementation of the WFD it is equally important to be
able to also describe quantitatively the ecological consequences of reduced nutrient concentrations
in the different types of rivers to be able to decide on nutrient reductions necessary to meet WFD
quality objectives. Enrichment in inorganic phosphorus and nitrogen concentrations in rivers can
cause the following impacts, as summarised in Nijboer and Verdonschot (2004):
Increase in primary production rate and mass growth of algae and plants. Increase in
nutrients can potentially increase the growth of algae and macrophytes if these nutrients
were initially limiting and if the other in-stream conditions allow more growth (e.g. light).
Change in communities. A change in nutrients can lead to a shift from nutrient-sensitive
slow-growing species to nutrient-tolerant fast-growing species. This may result in competition
for light between epiphytes, macrophytes and phytoplankton, modifying the balance of these
algae and plants. This may have impacts on macroinvertebrates and sh through changes
to habitats and the food web.
Reduction in oxygen. High nutrient levels favour a high biomass of algae and plants.
Consequently, larger amounts of biomass are decomposed under oxygen consumption at a
later stage. A high autotrophic production of oxygen during the day is usually followed by a
high rate of respiration and often low oxygen concentration during the night. These processes
can lead to oxygen depletion in the water or in the sediment.
Nitrogen and phosphorus are naturally present in streams through transport of terrestrial organic
matter, leaching and runoff of terrestrial inorganic nutrients and deposition from the atmosphere.
Mainstone and Parr (2002) estimated that the background export rates of total phosphorus in
four UK rivers would be between 0.1 and 0.2 kg TP ha
-1
year
-1
, which means a total phosphorus
concentration level of 5-25 g l
-1
. All anthropogenic activities that release organic matter and
inorganic nutrients in the environment may cause in-stream enrichment in inorganic nutrients.
The predominant sources are population, industry, agriculture, transport and power plants.
In western countries the wastewater contribution from each person amounts to 1-1.5 kg P
capita
-1
.year
-1
and 2.2 to 5.5 kg N capita
-1
.year
-1
. The main sources are excretion from the human
body and use of detergents (European Environment Agency, 1999). The wastewater nutrient
loadings to rivers depend mainly on the population density in the river catchment and the level
of nutrient removal at wastewater treatment plants. Depending on the treatment level, up to 95%
of the phosphorus can be removed from wastewaters (Brett et al., 1997).
Industries, primarily the food and drink industry, the industries producing phosphorus-based
fertilisers, cleaning materials, pesticides and/or metal nishing (Morse et al., 1993), can discharge
signicant amounts of phosphorus to surface waters. Emissions from industries are driven by the
technology used, the level of wastewater treatment demanded (EC Urban Waste Water Treatment)
NUTRIENTS AND EUTROPHICATION IN RIVERS | 81
and regulations on discharges (discharge consents) and emissions (UNECE Gothenburg Protocol
to the Convention on the Long Range Transboundary Emissions, EU National Emission Ceiling
Directive and EC Integrated Pollution, Prevention and Control Directive).
Agriculture releases nutrients to the environment through different pathways: emissions of ammonia
to the air and leaching and runoff of nitrogen and phosphorus that reach the river. The export rate
of phosphorus and nitrogen are affected by the amount of fertiliser applied, cultivation practice,
climate, geology, topography, soils hydrology, land management and the spatial location of nutrient
sources in the catchment (Johnes et al., 1996). The transport and power production sectors combust
fossil fuels. By-products of these combustions are nitrogen oxides. Once deposited, they can be
transported to the rivers. 60% of the nitrogen oxides in 1995 came from transport, 21% from the
energy sector and 14% from other industries (European Environment Agency, 1999).
Relative importance of the different sources
Nitrogen load. Nitrogen within rivers is derived predominantly from agricultural sources (e.g.
Jarvie et al., 1998). In Europe, agricultural sources can represent between 46 and 87% of the
total nitrogen load. Agricultural N losses depend on a range of factors, including the area of land
in agricultural use, intensity of agricultural production and fertilizer application rates. Point
sources (principally sewage discharges) can also play an important role accounting for up to 35
to 43% of N loads in some catchments (e.g. Boorman, 2003). Nitrogen from the atmosphere can
also represent an important source of N, for example up to 35% for a Swedish catchment between
1982 and 1987 (European Environment Agency, 1999) whereas the atmospheric deposition of N
in the Humber catchment (UK) reached values comparable with river output (Smith et al., 1997).
Phosphorus load. Source apportionment of phosphorus can highly vary between catchments. It is
usually proportional to population density. Phosphorus present naturally would, in many catchments,
represent only 3 to 15% of the current levels (Morse et al., 1993; Farmer, 2004). In densely populated
catchments, 50 to 76% of the phosphorus load was found to come from point sources, mainly from
sewage discharges in highly industrial catchments, and 20 to 40% from agricultural diffuse sources
(European Environment Agency, 1999). The percentage contribution of non-point sources of
phosphorus could vary between 2 and 60% for European rivers (Farmer, 2004).
Quantication of diffuse losses of nitrogen N and phosphorus P. Several quantication tools
exist to estimate the diffuse nutrient losses to surface freshwater systems. An EU research project
EuroHARP
1
is evaluating nine different methodologies for quantifying diffuse losses of nitrogen
and phosphorus (AMINO, REALTA, N-LES, MONERIS, TRK, SWAT, EVENFLOW, Nopolu and Source
apportionment methodology) on 17 catchments.
Nutrient concentrations in rivers
The level and uctuation of nutrient concentrations in rivers is determined mainly by natural and
anthropogenic inputs. Catchment behaviour contrasts between urban/industrial catchments and
1
EuroHARP is a research project under the 5
th
Framework program. It started on the 1
st
of January 2002 and is running
for 4 years. For information see www.euroharp.org
agricultural catchments. In the urban/industrial catchments, dilution effects can mean that
nutrients peak at low ows and are potentially nitrogen limited. In agricultural catchments, runoff
can lead to a nutrient peak at high ow with potentially phosphorus limitation (Jarvie et al.,
1998; Mainstone and Parr, 2002). Variation of the ratio between nitrogen and phosphorus in
time also leads to a variation in potential nutrient limitation. However the concentration levels
are also inuenced by plant assimilation and sedimentation (Young et al., 1999).
Phosphorus can be adsorbed or released from sediments. Interactions between river water and
sediments are particularly important for phosphorus (House, 2003). Soluble reactive phosphorus
can be adsorbed or released from sediments as a result of a combination of physico-chemical and
biochemical processes as well as hydrodynamic effects and bioturbation (House and Denison,
1998; House and Warwick, 1999; House, 2003; Jarvie et al., 2005). Particulate phosphorus is
also subject to deposition on the river bed and resuspension according to hydrodynamic factors
and the inuence of biolms in stabilizing the sediment surface (Battin and Sengshmitt, 1999).
The biogeochemical cycling of phosphorus in rivers is therefore complex and inuenced by a wide
range of factors such as the composition and reactivity of the sediments, sediment redox conditions,
growth of biolms, hydrology and antecedent conditions (House, 2003). Nitrogen transformations
include e.g. nitrication of ammonium into nitrate and denitrication of nitrate into atmospheric
nitrogen also depending on the redox potential (Hamilton et al., 2001; Peterson et al., 2001).
Decomposition of organic matter releases plant-available inorganic nutrients to the water
column.
A high population of macrophytes, phytobenthos and phytoplankton can strip soluble reactive
phosphorus (Mainstone and Parr, 2002) and ammonium (Balbi, 2000) out of the water column,
even at high nutrient levels. Effects of nutrients on biota also vary in time: greater impacts by
nutrients during the active season (light and temperature driven) and time to uptake the nutrients
(water residence times, ow driven). Increase in biomass leads to an increase in organic matter.
This organic matter can either decompose in situ (very slow owing rivers) or be carried away,
increasing the nutrient levels of the river either locally or downstream when organic matter is
mineralised.
Models of water quality
A wide range of models has been developed. They all include a hydrological compartment and
can represent the in-stream processes and/or the land processes with varying degrees of
complexity. Benchmarking Models for the Water Framework Directive (BMW) created a database
of water quality models that include their main characteristics and their domain of applicability.
The BMW database
2
includes models of in-stream processes for the river domain, for example
QUESTOR, QUAL-2E, RIVQM, SOBEK and HERMES, and models of terrestrial processes for nutrients,
for the diffuse domain, for example CREAMS, CASCADE, MIKE SHE, and SWAT.
2
This database is available on Internet, after a free registration to the BMW website: http://www.rbm-toolbox.net/
toolbox3/index.php.
Effects of reduction in nutrient inputs
It is expected that a reduction in nutrient inputs to a river leads to a reduction in nutrient
concentrations. However, few studies have explored the effect of reduction in nutrient inputs.
There is a potential for a release of phosphorus stored in soft river sediments and thus delaying
the reduction in river concentration (House, 2003). The catchment as a whole can also buffer the
effect of a reduction in nutrient use: in Denmark and Estonia, no evidence of reduction in river
nutrient concentration was found after 15 to 20 years of reduction in fertilizer use (Stalnacke et
al., 2003; Stalnacke et al., 2004; Kronvang et al., 2005).
As for impacts on the biota, Jarvie et al. (2002) indicate that implementation of phosphorus-
stripping before the discharge of wastewater did not lead to an improvement of macrophytes and
phytobenthos. Young et al. (1999) examine this situation resulting from an insufcient reduction
of in-stream nutrients that can maintain a eutrophic status. On the other hand, Sosiak (2002)
found a decrease in phytobenthic biomass after reduction of phosphorus in a wastewater discharge
(even if not directly downstream) and in macrophyte biomass following reduction in nitrogen.
These differing conclusions indicate the lack of knowledge of the effect of nutrient stripping and
the recovery that can be expected from the biota.
3.2 Macrophytes
Relationships between nutrients and macrophytes
Macrophytes are macroscopic aquatic plants. They are important primary producers in rivers. Their
growth is partly conditioned by the uptake of inorganic nutrients by their roots and their shoots.
Inorganic nutrients can be taken up either from the water column or the sediments, or both in
varying proportion (Clarke and Wharton, 2001). In nutrient-limited rivers enrichment in inorganic
nutrients may lead to an increase in biomass and species richness up to a point when competitive
exclusion occurs (Mesters, 1995).
However, these effects of nutrients are contested in several studies. Demars and Harper (1998)
underlined the difculty of identifying the specic effect of nutrients because of the synergistic
effect of a number of physical factors. Holmes et al. (1999) indicated that species diversity in
itself is not an indicator of trophic status of rivers as a wide range of diversity can occur at any
phosphorus concentration and low to intermediate diversity occur at any phosphorus concentration
as well. The Environment Agency (2002) also pointed out that macrophytes were not an effective
indicator of the level of nutrients in lowland rivers and Madsen and Cedergreen (2002) did not
nd an effect of enrichment on the growth rate of two macrophyte species, suggesting that
nutrients were not limiting for their growth.
Other studies conrmed the prominence of physical characteristics in determining macrophyte
biomass and species composition (Demars and Harper, 1998; Sand-Jensen, 1998; Schorer et al.,
2000; Baattrup-Pedersen et al., 2003; Barendregt and Bio, 2003). They underlined the effects of
ow and its variation, light (water depth, shading by riparian vegetation, suspended solids,
phytoplankton and epiphytes), substrata and weed cutting.
Existing methods to investigate the relationships
Macrophytes communities can be assessed by their biomass and their composition. Direct
measures of macrophyte biomass are not widely used. Macrophyte biomass is more often assessed
using measures of coverage or abundance. Macrophytes species composition can be assessed by
recording the species/taxa present at a site and possibly the species abundance. Several indices
for species richness and evenness have been developed. Average Score Per Taxon (H.M.S.O., 1987)
based on number of taxa, surface cover class and biomass, distribution factor based on species
abundance per section, Shannon-Wiener functions (Shannon and Weawer, 1963), Sorensen index
and the Simpson index (Simpson, 1952). The relationships between inorganic nutrients and
macrophyte biomass and composition can potentially be studied through 1) laboratory bioassays,
2) in-situ enrichment, and 3) analysis of monitoring data.
Monitoring data have been far more widely used than laboratory or in-situ experiments. Analysis
of the data can vary from simple relations to more complex analyses. Regression analysis allows
the assessment of linkages between species composition (as indices), abundance or biomass with
nutrients or other factors (Flynn et al., 2002; Carr et al., 2003). Correlations are used to analyse
interactions with a wider range of variables, including nutrients, as in Mesters (1995), Szoszkiewicz
et al. (2002) and Thiebaut et al. (2002). Examples are shown in Table 3.1.
Macrophyte models and indices
Several dynamic models have been developed; for example Wright et al. (1986) presented a
model of macrophyte biomass in shallow waters as a function of light and nutrient limitation. A
model of macrophyte and epiphyte biomass has been developed for the River Kennet (Southern
England), including SRP from sediment pore water (Wade et al., 2002a, b).
The equation of macrophyte biomass in the Kennet model (Wade et al., 2002b) is:

+ +
=

where x
7
macrophyte biomass at time t (g C m
-2
); c
10
macrophyte growth rate (day
-1
);
M
(T-20)

macrophyte temperature dependency; x
12
SRP in pore water at time t (mg P L
-1
);

R solar radiation
at time t (normalised to 0-1 from W.m
-2
); c
12
self shading (g C m
-2
); c
11
half saturation of P for
macrophyte growth (mg P L
-1
); c
14
macrophyte death rate (s g C
-1
day
1
); x
8
epiphyte biomass at
time t (g C m
-2
); x
1
ow out of reach at time t (m
3
.s
-1
);
An example of a macrophyte index used for classication, the Mean Trophic Rank index (Holmes
et al., 1999) is presented in Box 3.1.
The MTR index has been developed for England and Wales to implement the EC Urban Waste
Water Directive. It is used to assess the impact of point sources on rivers (Kelly, 1998). It is based
on the combination of species at a site and, for each species, its indicator value and its abundance.
128 species have been identied for this index. They can be split into 4 broad groups and 10 River
Community Types. These River Community Types are physically described and the mean MTR
value for each of these groups can be used as a reference for physically similar sites. The MTR has
been adapted to several countries, including Poland (Szoszkiewicz et al., 2002).
Other macrophytes indices based on nutrients have been developed in Europe, such as the GIS
Macrophytes in France (Haury et al., 1996), the Trophic Makrophyten Index (Schneider and Melzer,
2003) and the Reference Index (Meilinger et al., 2005) in Germany, and the Canonical-Based
Assessment System for Northern Ireland (Dodkins et al., 2005).
Biological
element
Water quality
element
Relation Sites Nutrient range References
Total %
macrophyte
cover
SRP Regression:
y = 120.37 + 77.1 . log SRP
(R
2
= 0.45. p = 0.002)
River Kennet, UK
n = 20
Not indicated (Flynn et al.,
2002)
MTR Soluble NO
3
Soluble PO
4
Correlation:
0.43 (p < 0.042)
0.58 (p < 0.003)
River Welland, UK Not indicated (Demars and
Harper, 1998)
Shift in
species
PO
4
3-
1
st
ordination axis correlated
among other things with:
PO
4
3-
(R = 0.74, p = 0.0015,
n = 17)
28 streams,
The Netherlands
Not indicated (Mesters, 1995)
Species
richness
NH
4
-N
PO
4
-P
Coefcient of rank
Spearman correlation
0.457 (p < 0.05)
0.611 (p < 0.001)
30 sites of Vosges/
Alsace oodplain,
France
Mean NH
4
-N = 31422 g.l
-1
PO
4
-P = 6-413 g/l
-1
(Thiebaut et al.,
2002)
Abundance NH
4
-N
PO
4
-P
Coefcient of rank
0.600 (p < 0.001)
0.707 (p < 0.001)
30 sites of Vosges/
Alsace oodplain,
France
Mean NH
4
-N = 31422 g.l
-1
PO
4
-P = 6-413 g/l
-1
(Thiebaut et al.,
2002)
Species
eveness
NH
4
-N
PO
4
-P
Coefcient of rank
0.238 (non signicant)
0.780 (p < 0.0001)
30 sites of Vosges/
Alsace oodplain,
France
Mean NH
4
-N = 31422 g.l
-1
PO
4
-P = 6-413 g.l
-1
(Thiebaut et al.,
2002)
MTR NO
3
-N
NH
4
-N
N total
SRP
P total
Pearson linear correlation
coefcient
0.156
0.299 (p < 0.05)
0.368 (p < 0.05)
0.260
0.294 (p < 0.05)
48 sites (19 rivers)
in lowland Poland
Not indicated (Szoszkiewicz
et al., 2002)
6 phyto-
sociological
groups
NH
4
-N
NO
3
-N
PO
4
-P
0.68
+ 0.81 (contextual)
0.92
29 sites NH
4
-N = 090 g.l
-1
PO
4
-P = 142 g.l
-1
NO
3
-N = 0.2-7.5 mg.l
-1
(Carbiener
et al., 1990)
Table 3.1 Summary of some of the relations found between macrophytes and inorganic nutrients.
3.3 Phytoplankton
Phytoplankton are microscopic plants suspended in the water column. They can include both
true phytoplankton species and phytobenthic species detached by ow, especially after spates
(Descy, 1993; Reynolds, 2003). The generation time of phytoplankton is typically several days.
This makes them highly reactive to changes in the condition of water and light. It also underlines
that a signicant phytoplankton biomass cannot be reached unless the residence time of water in
the river is greater than the generation time (large rivers, slow-moving rivers, canals and on-river
reservoirs).
Nutrient enrichment can lead to an increase in biomass and/or a change in species composition
(e.g. Vanni and Findlay, 1990). Algal blooms have been found to occur principally in spring in rivers
when increasing light and temperature and decreasing ow are favourable to phytoplankton
growth (Marker and Collett, 1997; Balbi, 2000). Species composition is also affected by the way
nutrient enrichment occurs: nutrient pulses allow the persistence of species preferring low and high
nutrient concentrations, thus leading to increased species richness (Hecky and Kilham, 1988).
The effect of nutrient enrichment will, however, be modulated by the initial nutrient concentrations
in the water column. Thus, Mallin et al. (2004) showed no impact of phosphorus enrichment on
phytoplankton biomass (0-5 mg P/l, stream concentration: 3-143 g P/l
-1
) for 2 streams. In this
study nitrogen enrichment over 200 g N/l in nitrate (stream concentration: 5-930 g N/l
-1
) led
to an increase in phytoplankton, whatever the form of nitrogen. Young et al. (1999) also showed
that there was little correlation (r = 0.29) between chlorophyll a and phosphorus concentrations
(0.1-6.7 mg PO
4
-P /l
-1
) for 3 lowland rivers of the Thames catchment. These studies demonstrate
that the relationships between nutrients and phytoplankton in rivers are not simple.

where CVS = SCV x STR

and where SCV = Species Cover Value (cover class per taxon, on a scale 1-9)
STR = Species Trophic Rank (score assigned to species on a scale 1-10: the higher the score, the lower the tolerance to nutrient
enrichment)
MTR score Interpretation
MTR > 65 Unlikely to be eutrophic*
25 - 65 Site likely to be either eutrophic or at risk of becoming eutrophic**
MTR < 25 Site badly damaged by eutrophication, organic pollution, toxicity or is
physically damaged
*If the score is less than might be expected for a site of this physical type, the site may be at risk of becoming eutrophic.
**If the score is less than might be expected for a site of this physical type, the site is eutrophic or at risk of becoming eutrophic.
Box 3.1 Brief description of Mean Trophic Rank (derived from Holmes et al., 1999).
Phytoplankton-nutrient relationships
Commonly-used models are mainly simple regression models (Table 3.2). Hakanson et al. (2003)
concluded that it is unlikely that a regression model for mean monthly chlorophyll concentrations
in rivers could yield R
2
higher than about 0.6 (for all n). Other more complex and dynamic models
exist and are described in Recknagel et al. (1997).
The dynamic models describe the phytoplankton development and need very substantial data
input e.g. algal growth and respiration rates, light attenuation, nutrient concentrations and
growth rates as a function of the nutrient concentrations. Some of these models are described
below. SWAT2000 (Neitsch et al., 2002): models algal biomass through chlorophyll a, depending
on light, nitrogen and phosphorus. AQUATOX (Park and Clough, 2004): models phytoplankton,
periphyton, macrophytes, invertebrates and sh depending on nutrients, organic matter, organic
toxicant, oxygen, suspended sediments, sediments, temperature and discharges. It is applicable to
streams, lakes and reservoirs.
RIVERSTRAHLER (Billen et al., 1994): combines a hydrological component based on the stream
order and the AQUAPHY model. The AQUAPHY model models the chlorophyll a concentration
and biomass evolution in each stream order depending on light, temperature, ow, ammonium,
nitrate and phosphate concentrations. In Garnier et al. (2002), this model is further rened to
distinguish between diatoms and chlorophycae, to include predation by zooplankton and mussels
and to model the interactions of nutrients across the sediment-water interface.
Biological
element
Water quality
element
Chlorophyll-a,
January to
spring
chlorophyll
maximum
Silicate
NH
3
NO
3
-N
TN
SRP
N/P ratio
Regression
0.71 (p < 0.05)
0.43 (p < 0.05)
0.25 (p < 0.05)
0.20 (p < 0.05)
0.09 (p < 0.05)
0.15 (p < 0.05)
River Nene (small
lowland river), UK
11 sites, up to over
22 years
NH
4
N = 0-5.3 mg.L
-1
NO
3
N = 0.4-20 mg.L
-1
NO
2
N = 0-0.4 mg.L
-1
SRP = 0.1-4.4 mg.L
-1
Balbi, 2000
Chlorophyll-a,
summer
period
TP Quadratic model
log(Chl) = 1.65 + 1.99
(log TP) 0.28(log TP)
2

(r
2
= 0.67)
Chl in mg.m
-3
and TP in mg.m
-3
Temperate streams
(mainly North
America, but also
Europe, Korea, Iraq
and Australia)
N = 292
TP = 5 1030 mg P.m
-3
Van Nieuwenhuyse
and Jones, 1996
Chlorophyll-a NO
3
SiO
2
Pearson correlation coefcient:
0.91 (p = 0.003)
0.87 (p = 0.006)
Canojoharie Creek,
Hudson River
Basin, US
NO
3
N = 0.05-3.3 mg.L
-1
SiO
2
= 15 mg.L
-1
Wall et al., 1998
Table 3.2 Summary of some relations found between phytoplankton and inorganic nutrients.
3.4 Phytobenthos
Whereas phytoplankton is considered to be a good indicator of eutrophication in rivers with a
residence time longer than the generation time of phytoplankton, phytobenthos is a better one
for smaller rivers (Piirso et al., 1997). Phytobenthos is a highly diversied biota (Stevenson, 1996)
with different types of organisms (diatoms, lamentous algae, blue-green etc), growing on different
substrates (rocks, soft riverbed, macrophytes etc) and developing different forms (lamentous,
thin or thick mat etc). Most of the studies undertaken have focused on diatoms. Phytobenthos is
present in every type of river where light reaches the riverbed or surface of macrophytes.
As for all autotrophs, phytobenthos development depends on the availability of inorganic
nutrients. An increase in inorganic nutrients (P and N) potentially leads to an increase in
phytobenthos biomass. Simultaneously, a shift in species composition occurs from slow-growing,
oligotrophic species at very low nutrient concentrations to fast-growing species with competitive
advantages at high nutrient concentrations (Biggs et al., 1998).
The response of phytobenthos to nutrient enrichment depends on whether the nutrient was
limiting and the initial nutrient level. Indeed, Perrin and Richardson (1997) obtained little
phytobenthic response to P addition, a moderate one when adding N, and an even higher one
when adding both N and P (P became limiting once N was added). Stanley et al. (1990)
demonstrated a rapid response of phytobenthic biomass when the P level was initially low (<0.010
mg.l
-1
) and only a moderate one when ambient P was higher (0.015-0.025 mg.l
-1
).
Flow and its variation can also play an important role in determining the response of phytobenthos
to nutrient concentrations. The discharge stability over periods of less than a year may govern the
average phytobenthic biomass, and nutrients will control the overall biomass only over prolonged
periods of stability with moderate to low ow (Biggs, 1996). So, a small increase in dissolved
nutrients may greatly increase the frequency of high biomass events if there are infrequent oods
and accrual periods over 100 days (Biggs, 2000).
In term of species richness, nitrogen and phosphorus enrichment have been found to increase the
species richness in oligotrophic streams up to a point where competitive exclusion occurs and
species richness decreases (Marcus, 1980; Snyder et al., 2002). As for phytoplankton (Hecky and
Kiham, 1988), the way nutrients are delivered affects how the community assemblage will
respond. Nutrient pulses can lead to higher richness, with a mix of species of different nutrient
preferences, as found in Finnish rivers (Soininen, 2002).
Kjeldsen (1994) underlined a different response of phytobenthos peak biomass to phosphorus
enrichment depending on the riverbed substrate: on ne-grained sediments phosphorus was the
main driver of the biomass peak, whereas on stony substrate P concentrations are only linked to
the maximum potential biomass and the actual peak biomass is determined by a range of other
factors.
Estimations of biomass levels that hampers the aesthetic quality of rivers have been developed. In
Welch et al. (1988) 100 mg chlorophyll-a per square metre is considered to be the threshold over
which development of phytobenthos, often as lamentous algae, is perceived unfavourably.
Dodds et al. (1998) reviewed several studies of such thresholds. They range from 50 to 200 mg
chlorophyll-a per square metre, and mainly over 100 mg chlorophyll-a per square metre.
Kiffney and Bull (2000) showed that light was the single best indicator explaining the variation
in peak chlorophyll-a biomass. Light varies seasonally and the quantity that reaches the riverbed
depends on the water depth and a thick carpet of diatoms covering the sediment is often seen
during spring where irradiance is high (Sand-Jensen, 1983). Dense growth of phytobenthos can
cause a high diurnal variations in oxygen concentrations (higher during the day, lower during the
night) unfavourable for other organisms like macroinvertebrates and sh (ten Cate et al., 1991).
Anderson et al. (1999) showed that grazing controlled biomass more than enrichment in soluble
reactive phosphate whereas Walton et al. (1995) pointed out that grazing lead to a shift in
species whatever the enrichment was and grazing is the third dimension in the habitat matrix
developed by Biggs et al. (1998) with an effect similar to spates.
Methods to investigate the phytobenthos-nutrient relationships
Chlorophyll-a per square metre is the most commonly used parameter describing phytobenthic
biomass as it is easy to use and not as labour-intensive as species composition surveys. The dry
mass, ash-free dry mass, cell density and biovolume can also be used to estimate the biomass of
phytobenthic communities (Stevenson, 1996). Species composition is assessed by recording the
number of species/taxa present in the community and their relative abundance. Different indices
can be calculated from these measures to reect the species richness, the species evenness and
the similarity of species among communities (Zelinka and Marvan, 1961).
The relationships between inorganic nutrients and phytobenthos can be studied via three different
methods: laboratory bioassays (Horner et al., 1990; Humphrey and Stevenson, 1992), in-situ
enrichment experiments and empirical methods from monitoring data. Whole river fertilisation has
mainly been used in several artic streams to study the long-term effect of fertilisation on phytobenthos
and other biota (Harvey et al., 1998, Slavik et al., 2004) and to other streams (Biggs et al., 2000;
Bernhardt and Likens, 2004). Monitoring of both phytobenthos and water quality and then using
empirical methods to form relationships is the most commonly used method to analyse the link
between phytobenthos biomass and composition and environmental conditions (including nutrient
levels). Once the phytobenthos and the water quality have been measured, similar techniques as for
phytoplankton and macrophytes can be used, from the simplest to the most complex.
Simple regressions have been used in Biggs (2000) to link the mean monthly chlorophyll a and
nutrients and days of accrual (reects the effect of ow variation). Multiple regressions, sometimes
with stepwise selection, have been used as well in Biggs (2000) and Kiffney and Bull (2000).
Regressions are usually used to link biomass with nutrients.
Relationships, models and indices
Regression models have been developed between phytobenthic biomass and nutrient concentration,
for example in Biggs (2000) and Winter and Duthie (2000) and examples given in Table 3.3.
The peak biomass of benthic algae in 10 Danish lowland streams followed the equation:
Y=929.3.x/(49.2+x) (r
2
=0.61), with Y the maximum biomass (mg chlorophyll.m
-2
) and x the
dissolved inorganic phosphorus (g P l
-1
, 1 - 273 g P l
-1
; Kjeldsen, 1994).
Indices and classication
A large number of phytobenthos indices (mainly diatoms) have been proposed, e.g.:
Specic Pollution Sensitivity Index (Cemagref, 1982)
Generic Diatom Index (Coste and Ayphasshorho, 1991)
Descys Index (Descy, 1979)
Sldec
^
eks index (Sldec
^
ek, 1986)
Leclerq & Maquets index (Leclercq and Maquet, 1987)
Trophic Diatom Index (Kelly and Whitton, 1998; Holmes et al., 1999; Kelly et al., 2001)
EPI-D (Dell'Uomo, 1997; Dell'Uomo et al., 1999).
Biological
element
Water quality
element
Chlorophyll-a
(in mg.m
-2
)
TP, in g.l
-1
TN, in g.l
-1
log Chl a = 0.905 log TP + 0.490
(r
2
= 0.56, p < 0.001)
log Chl a = 0.984 log TN 0.935
(r
2
= 0.50, p < 0.001)
33 sites, 13
rivers of Ontario
and Quebec
(Canada)
TP = 6 - 130 g.l
-1
TN = 179 - 2873 g.l
-1
(Chetelat
et al., 1999)
Mean
monthly
benthic algal
biomass
In mg/m
2

chlorophyll-a
SIN
SRP
Log
10
(chla) = 0.109 + 0.483.log
10
SIN (r
2
= 0.122, p = 0.057)
Log
10
(chla) = 0.468 + 0.697 log
10

SRP (r
2
= 0.226, p = 0.008)
30 sites, 25
streams,
New Zealand
SIN = 6.2 232 mg.m
-3
SRP = 1.3 31.6 mg.m
-3
(Biggs,
2000)
Max monthly
benthic algal
biomass
In mg/m
2

chlorophyll-a
SIN
SRP
Log
10
(chla) = 0.711 + 0.688.log
10
SIN (r
2
= 0.325, p = 0.001)
Log
10
(chla) = 1.400 + 0.797 log
10

SRP (r
2
= 0.295, p = 0.002)
30 sites, 25
streams,
New Zealand
SIN = 6.2 232 mg.m
-3
SRP = 1.3 31.6 mg.m
-3
(Biggs,
2000)
Diatom taxa
richness
TN Spearman correlation:
0.28 (p < 0.01, diatoms at
genus level)
0.42 (p < 0.001, diatoms at
species level)
199 streams,
Mid-Appalachian
streams, USA
Not indicated (Hill et al.,
2001)
% dominance
(diatoms)
TP Spearman correlation:
0.41 (p < 0.01, diatoms at
genus level)
0.45 (p < 0.001, diatoms at
species level)
199 streams,
Mid-Appalachian
streams, USA
Not indicated (Hill et al.,
2001)
TDI NH
4
-N Pearson correlation coefcient
0.430 (p < 0.001)
9 sites, for
Vistula river,
Poland
NH
4
-N = 0.2 - 3.0 mg.l
-1
NO
3
-N = 0.2 3.0 mg.l
-1
PO
4
-P = 0.05-0.55 mg.l
-1
(Kwandrans,
2002)
Diatom
richness
NH
4
-N
TP
Spearman correlation coefcient
0.55
0.64
11 rivers in
Idaho, USA
NO
3
N + NO
2
-N = 0.002-
1.61 mg.l
-1
NH
4
N = 0.005-0.037 mg.l
-1
(Snyder
et al., 2002)
Table 3.3 Some examples of relations found between phytbenthos and inorganic nutrients.
A classication of river sites has been developed for the UK based on the Trophic Diatom Index
(TDI) as an indicator of eutrophication (Kelly, 1998). In Italy, the EPI-D is considered suitable to
diagnose inorganic nutrients, organic matter and minerals impacts on phytobenthos (DellUomo
et al., 1999). Lastly, Biggs (2000) developed a classication depending on soluble reactive
phosphorus and soluble inorganic nitrogen, and the number of days of accrual in gravel/cobble-
bed streams, the boundaries being determined by 2 nuisance levels of chlorophyll a: 60 and 200
mg chlorophyll a per square metre (Figure 3.1).
3.5 Summary
An increase in plant available phosphorus and nitrogen concentrations in rivers potentially leads
to an overall increase in the biomass of phytobenthos, phytoplankton and macrophytes and to
changes in species composition, favouring nutrient tolerant species. Large-scale phytoplankton
growth is generally limited to slow-owing rivers and macrophytes and phytobenthos tend to
occur in rivers where light penetration reaches the river bed.
Nutrient enrichment often leads to an increase in biomass and a change in species from nutrient-
sensitive toward more nutrient-tolerant species. In term of species richness, enrichment of
oligotrophic streams can lead to an increase in species richness until the enrichment reaches a
threshold after which competitive exclusion of nutrient-sensitive species decreases the overall
species richness. The effects of nutrient enrichment depend on the initial nutrient level. In
oligotrophic streams, even a slight enrichment can lead to a high increase in biomass and alter
the species assemblage, whereas it may not cause any change at higher nutrient levels. However,
no normative thresholds for inorganic nutrients, applicable beyond the individual rivers of the
studies, have been found in the literature.
Inorganic nutrients usually do not directly inuence heterotrophic organisms such as invertebrates
and sh, but it is important to consider also the indirect effects from the changes in the plant
communities when assessing the overall impacts of nutrient in stream ecology. Further, the
autotrophic communities are inuenced by heterotrophic activities (grazing) and the quality of
the river sediment, which must therefore be considered in the assessment of river eutrophication.
Figure 3.1 Simplied schematic classication
as developed by Biggs (2000). The boundary
between oligotrophic and mesotrophic is set
at 60 mg m
-2
chlorophyll-a and the boundary
between mesotrophic and eutrophic is set at
200 mg m
-2
chlorophyll-a.
1000
1
0 100
S
I
N

(
m
g
/
m
-
3
)
Eutrophic
Days of accrual
100.0
0 1
S
I
R
P

(
m
g
/
m
-
3
)
Mesotrophic
Oligotrophic
Numerous studies have been undertaken to examine the relationships between macrophytes and
benthic diatoms and nutrient status in rivers. However, the relationships identied are not
sufciently detailed or targeted to meet the needs in the implementation of the WFD. For setting
the criteria for ecological classication and the environmental objectives, and for planning the
program of measures thereafter, it is necessary to know quantitative relationships between
nutrient concentrations and biological quality elements (periphyton, macrophytes and
phytoplankton) in the different types of rivers. On-going work in REBECCA and other projects such
as DARES (http://craticula.ncl.ac.uk/Dares/index.htm) are exploring these relationships.
Major needs for the implementation of the WFD include clarication of the conditions where the
models and relationships established are operational and valid. The conditions to be claried are
related to the type of river and sediment, the type of vegetation and the concentration level of the
nutrients. Few studies have focussed on the recovery of a river after reduction in either phosphorus
or nitrogen (or both) sources. This aspect should be considered fundamental in determining the
measures needed to be implemented to achieve good chemical and ecological status.
3.6 References
Anderson, E.L., E.B. Welch, J.M. Jacoby, G.M. Schimek, and R.R. Horner. 1999. Periphyton removal related to
phosphorus and grazer biomass level. Freshwater Biology 41:633-651
Baattrup-Pedersen, A., S.E. Larsen, and T. Riis. 2003. Composition and richness of macrophyte communities
in small Danish streams - inuence of environmental factors and weed cutting. Hydrobiologia 495:171-179
Balbi, D.M. 2000. Suspended chlorophyll in the River Nene, a small nutrient-rich river in eastern England:
long-term and spatial trends. The Science of the Total Environment 251-252:401-421
Barendregt, A., and A.M.F. Bio. 2003. Relevant variables to predict macrophyte communities in running
waters. Ecological Modelling 160:205-217
Battin, T.J., and Sengschmitt. 1999. Linking sediment biolms, hydrodynamics and river bed clogging:
Evidence from a large river. Microbial Ecology 37:185-196
Bernhardt, E.S., and G.E. Likens. 2004. Controls on periphyton biomass in heterotrophic streams. Freshwater
Biology 49:14-27
Biggs, B.J.F. 1996. Hydraulic habitat of plants in streams. Regulated Rivers: Research and Management
12:131-144
Biggs, B.J.F., R.J. Stevenson, and R.L. Lowe. 1998. A habitat matrix conceptual model for stream periphyton.
Archiv fur Hydrobiologie 143:21-56
Biggs, B.J.F. 2000. Eutrophication of streams and rivers: dissolved nutrient-chlorophyll relationships for
benthic algae. Journal of the North American Benthological Society 19:17-31
Biggs, B.J.F., S.N. Francoeur, A.D. Huryn, R. Young, C.J. Arbuckle, and C.R. Townsend. 2000. Trophic cascades
in streams: effects of nutrient enrichment on autotrophic and consumer benthic communities under two
different sh predation regimes. Canadian Journal of Fisheries and Aquatic Sciences 57:1380-1394
Billen, G., J. Garnier, and P. Hanset. 1994. Modelling Phytoplankton Development in Whole Drainage
Networks - the Riverstrahler Model Applied to the Seine River System. Hydrobiologia 289:119-137
Brett, S., J. Guy, J.K. Morse, and J.N. Lester. 1997. Phosphorus removal and recovery technologies. Centre
Europen dEtudes des Polyphosphates, pp.142
Boorman, D.B. 2003. LOIS in-stream water quality modelling. Part 2. Results and scenarios. Science of the
Total Environment 314:397-409
Carr, G.M., S.A.E. Bod, H.C. Duthie, and W.D. Taylor. 2003. Macrophyte biomass and water quality in Ontario
rivers. Journal of the North American Benthological Society 22:182-193
Carbiener, R., M. Tremolieres, J.L. Mercier, and A. Ortscheid. 1990. Aquatic macrophyte communities as
bioindicators of eutrophication in calcareous oligosaprobe stream waters (Upper Rhine plain, Alsace).
Vegetatio 86:71-88
Cemagref. 1982. Etude des mthodes biologiques quantitatives dapprciation de la qualit des eaux.
Agence de lEau Rhne-Mditerrane-Corse, pp. 28
Chetelat, J., F.R. Pick, A. Morin, and P.B. Hamilton. 1999. Periphyton biomass and community composition in
rivers of different nutrient status. Canadian Journal of Fisheries and Aquatic Sciences 56:560-569
Clarke, S.J., and G. Wharton. 2001. Sediment nutrient characteristics and aquatic macrophytes in lowland
English rivers. The Science of the Total Environment 266:103-112
Coste, M., and H. Ayphasshorho. 1991. Etude de la qualit des eaux du Bassin Artois-Picardie laide des
communauts de diatomes benthiques (Applications des indices diatomiques). Cemagref, Bordeaux -
Agence de leau Artois-Picardie, Douai, pp. 227
DellUomo, A. 1997. Use of algae for monitoring rivers in Italy: current situation and perspectives. In: J.
Prygiel, B.A. Whitton, and J. Bukowska (Eds.), 3rd Sypmosium Use of Algae for Monitoring Rivers, Douai,
France, pp.17-25
DellUomo, A., A. Pensieri, and D. Corradetti. 1999. Epilithic diatoms from the Esino river (central Italy) and
their use for the evaluation of the biological quality of the water. Cryptogamie Algologie 20:253-269
Demars, B.O.L., and D.M. Harper. 1998. The aquatic macrophytes of an English lowland river system:
assessing response to nutrient enrichment. Hydrobiologia 384:75-88
Descy, J.P. 1979. A new approach to water quality estimation using diatoms. Nova Hedwigia 64:305-323
Descy, J.P. 1993. Ecology of the phytoplankton of the River Moselle: effects of disturbance on community
structure and diversity. Hydrobiologia 249:111-116
Dodds, W.K., J.R. Jones, and E.B. Welch. 1998. Suggested classication of stream trophic state: distributions
of temperate stream types by chlorophyll, total nitrogen, and phosphorus. Water Research 32:1455-1462
Dodkins, I., B. Rippey, and P. Hale. 2005. An application of canonical correspondence analysis for developing
ecological quality assessment metrics for river macrophytes. Freshwater Biology 50:891-904
Environment Agency. 2002. Assessment of the impact of nutrient removal on eutrophic rivers. R&D Technical
Report P2-127/TR, Environment Agency, Bristol, pp. 110
European Environment Agency. 1999. Nutrients in European ecosystems. European Environment Agency,
Copenhagen, pp. 156
Farmer, A.M. 2004. Phosphate pollution: a global overview of the problem. Phosphorus in Environmental
Technologies - Principles and Applications. IWA Publishing. pp. 174-191
Flynn, N.J., D.L. Snook, A.J. Wade, and H.P. Jarvie. 2002. Macrophytes and periphyton dynamics in a UK
Cretaceous chalk stream: the River Kennet, a tributary of the Thames. The Science of the Total Environment.
282-283:143-157
Garnier, J., G. Billen, E. Hannon, S. Fonbonne, Y. Videnina, and M. Soulie. 2002. Modelling the Transfer and
Retention of Nutrients in the Drainage Network of the Danube River. Estuarine, Coastal and Shelf Science
54:285-308
HMSO (Her Majestys Stationery Service). 1987. Methods for the use of Aquatic Macrophytes for Assessing
Water Quality 1985-86: methods for the examination of waters and associated materials. HMSO, London,
pp. 176
Hakanson, L., J.M. Malmaeus, U. Bodemer, and V. Gerhardt. 2003. Coefcients of variation for chlorophyll,
green algae, diatoms, cryptophytes and blue-greens in rivers as a basis for predictive modelling and aquatic
management. Ecological Modelling 169:179-196
Hamilton, S.K., J.L. Tank, D.F. Raikow, W.M. Wollheim, B.J. Peterson, and J.R. Webster. 2001. Nitrogen uptake and
transformation in a midwestern US stream: A stable isotope enrichment study. Biogeochemistry 54:297-340
Harvey, C.J., B.J. Peterson, W.B. Bowden, A.E. Hershey, M.C. Miller, L.A. Deegan, and J.C. Finlay. 1998.
Biological responses to fertilization of Oksrukuyik Creek, a tundra stream. Journal of the North American
Benthological Society 17:190-209
Haury, J., M.C. Peltre, S. Muller, M. Tremolieres, J. Barbe, A. Dutartre, and M. Guerlesquin. 1996. Des indices
macrophytiques pour estimer la qualit des cours deau franais: premires propositions. Ecologie 27:233-244
Hecky, R.E., and P. Kilham. 1988. Nutrient limitation of phytoplankton in freshwater and marine environments:
A review of recent evidence on the effects of enrichment. Limnology and Oceanography 33:796-822
Hill, B.T., R.J. Stevenson, Y.D. Pan, A.T. Herlihy, P.R. Kaufmann, and C.B. Johnson. 2001. Comparison of
correlations between environmental characteristics and stream diatom assemblages characterized at genus
and species levels. Journal of the North American Benthological Society 20:299-310
Holmes, N.T.H., J.R. Newman, S. Chadd, K.J. Rouen, L. Sharp, and F.H. Dawson. 1999. Mean Trophic Rank: A
users manual. E38, Environment Agency, pp. 129
Horner, R.R., E.B. Welch, M.R. Seeley, and J.M. Jacoby. 1990. Responses of periphyton to changes in current
velocity, suspended sediment and phosphorus concentration. Freshwater Biology 24:215-232
House, W.A., and F.H. Denison. 1998. P dynamics in a lowland river. Water research 32:1819-1830
House, W.A., and M.S. Warwick. 1999. Interactions of phosphorus with sediments in the River Swale,
Yorkshire, UK. Hydrological Processes 13:1103-1115
House, W.A. 2003. Geochemical cycling of phosphorus in rivers. Applied Geochemistry 18:739-745
Humphrey, K.P., and R.J. Stevenson. 1992. Responses of Benthic Algae to Pulses in Current and Nutrients
During Simulations of Subscouring Spates. Journal of the North American Benthological Society 11:37-48
Jarvie, H.P., B.A. Whitton, and C. Neal. 1998. Nitrogen and phosphorus in east coast British rivers: Speciation,
sources and biological signicance. Science of the Total Environment 210/211:79-109
Jarvie, H.P., E. Lycett, C. Neal, and A. Love. 2002. Patterns in nutrient concentrations and biological quality
indices across the upper Thames river basin, UK. Science of the Total Environment 282:263-294
Jarvie, H.P., M.D. Juergems, R.J. Williams, C. Neal, J.J.L. Davies, C. Barrett, J. White. 2005. Role of river bed
sediments as sources and sinks of phosphorus across two major eutrophic UK river basins: the Hampshire
Avon and Herefordshire Wye. Journal of Hydrology 304:51-74
Johnes, P., B. Moss, and G. Phillips. 1996. The determination of total nitrogen and total phosphorus
concentrations in freshwaters from land use, stock headage and population data: Testing of a model for use
in conservation and water quality management. Freshwater Biology 36:451-473
Kelly, M.G. 1998. Use of community-based indices to monitor eutrophication in European rivers. Environmental
Conservation 25:22-29
Kelly, M.G., and B.A. Whitton. 1998. Biological monitoring of eutrophication in rivers. Hydrobiologia 384:55-67
Kelly, M.G., C. Adams, A.C. Graves, J. Jamieson, J. Krokowski, E. Lycett, J. Murray-Bligh, S. Pritchard, and C.
Wilkins. 2001. The Trophic Diatom Index: a users manual. E2/TR2, Environment Agency, Bristol
Kiffney, P.M., and J.P. Bull. 2000. Factors controlling periphyton accrual during summer in headwater streams
of southwestern British Columbia, Canada. Journal of Freshwater Ecology 15:339-351
Kjeldsen, K. 1994. The relationship between phosphorus and peak biomass of benthic algae in small lowland
streams. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte Limnologie
25:1530-1533
Kronvang, B., E. Jeppesen, D.J. Conley, M. Sondergaard, S.E. Larsen, N.B. Ovesen, and J. Carstensen. 2005.
Nutrient pressures and ecological responses to nutrient loading reductions in Danish streams, lakes and
coastal waters. Journal of Hydrology 304:274-288
Kwandrans, J. 2002. Upper Vistula River: Response of aquatic communities to pollution and impoundment.
IX. Benthic diatom communities. Polish Journal of Ecology 50:223-236
Leclercq, L., and B. Maquet. 1987. Deux nouveaux indices chimiques et diatomiques de qualit de leau
courante. Application au Samson et ses afuents (bassin de la Meuse belge). Comparaison avec dautres
indices chimiques, biocniques et diatomiques. Document de travail 28, Institut Royal des Sciences Naturelles
de Belgique, pp. 113
Madsen, T.V., and N. Cedergreen. 2002. Sources of nutrients to rooted submerged macrophytes growing in a
nutrient-rich stream. Freshwater Biology 47:283-291
Mainstone, C.P., and W. Parr. 2002. Phosphorus in rivers ecology and management. The Science of the Total
Environment 282-283:25-47
Mallin, M.A., M.R. McIver, S.H. Ensign, and L.B. Cahoon. 2004. Photosynthetic and heterotrophic impacts of
nutrient loading to blackwater streams. Ecological Applications 14:823-838
Marcus, M. D. 1980. Periphytic community response to chronic nutrient enrichment by a reservoir discharge.
Ecology 61:387-399
Marker, A.F.H., and G.D. Collett. 1997. Spatial and temporal characteristics of algae in the River Great Ouse.
I. Phytoplankton. Regulated Rivers: Research and Management 13:219-233
Mesters, C.M.L. 1995. Shifts in macrophyte species composition as a result of eutrophication and pollution
in Dutch transboundary streams over the past decades. Journal of Aquatic Ecosystem Health 4:295-305
Meilinger, P., S. Schneider, and A. Melzer. 2005. The reference index method for the macrophyte-based
assessment of rivers - a contribution to the implementation of the European Water Framework Directive in
Germany. International Review of Hydrobiology 90:322-342
Morse, G.K., J.N. Lester, and R. Perry. 1993. The economic and environmental impact of phosphorus removal
from wastewater in the European community. Selper Publications, London, pp. 92
Neitsch, S.L., J.G. Arnold, J.R. Kiniry, J.R. Williams, and K.W. King. 2002. Soil and Water Assessment Tool
Theorical Documentation. TWRI Report TR-191, Texas Water Resources Institute, College Station, Texas
Nijboer, R.C., and P.F.M. Verdonschot. 2004. Variable selection for modelling effects of eutrophication on
stream and river ecosystems. Ecological Modelling 177:17-39
Park, R.A., and J.S. Clough. 2004. AQUATOX (RELEASE 2) Modelling environmental fate and ecological
effects in aquatic ecosystems. Volume2: technical documentation. U.S. Environmental Protection Agency,
Washington DC, pp. 210
Perrin, C.J., and J.S. Richardson. 1997. N and P limitation of benthos abundance in the Nechako River, British
Columbia. Canadian Journal of Fisheries and Aquatic Sciences 54:2574-2583
Peterson, C.G., M.A. Horton, M.C. Marshall, H.M. Valett, and C.N. Dahm. 2001. Spatial and temporal variation
in the inuence of grazing macroinvertebrates on epilithic algae in a montane stream. Archiv Fur Hydrobiologie
153:29-54
Piirso, K., T. Trei, and R. Laugaste. 1997. Use of algae for monitoring rivers in Estonia. In Use of Algae for
Monitoring Rivers III. Agence de lEau Artois-Picardie. pp. 66-71
Recknagel, F., M. French, P. Harkonen, and K.I. Yabunaka. 1997. Articial neural network approach for
modelling and prediction of algal blooms. Ecological Modelling 96:11-28
Sand-Jensen, K. 1983. Chapter 10: Physical and chemical parameters regulating growth of periphytic communities.
In: R.G. Wetzel (Ed.), Periphyton of freshwater ecosystems. Dr W. Junk Publishers, The Hague, pp. 63-71
Sand-Jensen, K. 1998. Inuence of submerged macrophytes on sediment composition and near-bed ow in
lowland streams. Freshwater Biology 39:663-679
Reynolds, C.S. 2003. Planktic community assembly in owing water and the ecosystem health of rivers.
Ecological Modelling 160:191-203
Schneider, S., and A. Melzer. 2003. The trophic index of macrophytes (TIM) - a new tool for indicating the
trophic state of running waters. International Review of Hydrology 88:49-67
Schorer, A., S. Schneider, and A. Melzer. 2000. The importance of submerged macrophytes as indicators for
the nutrient concentration in a small stream (Rotbach, Bavaria). Limnologica 30:351-358
Shannon, C.E., and W. Weaver. 1963. The mathematical theory of communication. University Illinois Press, Urbana
Simpson, G.G. 1952. How many species? Evolution 6:342
Sladecek, V. 1986. Diatoms as Indicators of Organic Pollution. Acta Hydrochimica Et Hydrobiologica 14:555-566
Slavik, K., B.J. Peterson, L.A. Deegan, W.B. Bowden, A.E. Hershey, and J.E. Hobbie. 2004. Long-term responses
of the Kuparuk River ecosystem to phosphorus fertilization. Ecology 85:939-954
Smith, R.I., J.N. Cape, J. Binnie, T.D. Murray, M.Young, and D. Fowler. 1997. Deposition of atmospheric
pollutants to the LOIS area. The Science of the Total Environment 194:71-85
Snyder, E.B., C.T. Robinson, G.W. Minshall, and S.R. Rushforth. 2002. Regional patterns in periphyton accrual
and diatom assemblage structure in a heterogeneous nutrient landscape. Canadian Journal of Fisheries and
Aquatic Sciences 59:564-577
Soininen, J. 2002. Responses of epilithic diatom communities to environmental gradients in some Finnish
rivers. International Review of Hydrobiology 87:11-24
Sosiak, A. 2002. Long-term response of periphyton and macrophytes to reduced municipal nutrient loading
to the Bow River (Alberta, Canada). Canadian Journal of Fisheries and Aquatic Sciences 59:987-1001
Stalnacke, P., A. Grimvall, C. Libiseller, M. Laznik, and I. Kokorite. 2003. Trends in nutrient concentrations in
Latvian rivers and the response to the dramatic change in agriculture. Journal of Hydrology 283:184-205
Stalnacke, P., S.M. Vandsemb, A. Vassiljev, A. Grimvall, and G. Jolankai. 2004. Changes in nutrient levels in
some Eastern European rivers in response to large-scale changes in agriculture. Water Science and Technology
49:29-36
Stanley, E.H., R.A. Short, J.W. Harrison, R. Hall, and R.C. Wiedenfeld. 1990. Variation in nutrient limitation of
lotic and lentic algal communities in a Texas (USA) river. Hydrobiologia 206:61-71
Stevenson, R.J. 1996. An introduction to algal ecology in freshwater benthic habitats. In: R.J. Stevenson,
M.L. Bothwell, and R.L. Lowe (Eds.), Algal ecology, freshwater benthic ecosystems. Academic Press, San Diego,
pp. 3-30
Szoszkiewicz, K., K. Karolewicz, A. Lawniczak, and F.H. Dawson. 2002. An assessment of the MTR aquatic
plant bioindication system for determining the trophic status of Polish rivers. Polish Journal of Environmental
Studies 11:421-427
ten Cate, J. H., J. Simons, and H. Schreurs. 1991. Periphytic macro- and microalgae in dutch freshwater ditches
and their signicance as ecological indicators of water quality. Archiv fur Hydrobiologie 122:275-296
Thiebaut, G., F. Guerold, and S. Muller. 2002. Are trophic and diversity indices based on macrophyte
communities pertinent tools to monitor water quality? Water Research 36:3602-3610
Van Nieuwenhuyse, E.E., and J.R. Jones. 1996. Phosphorus-chlorophyll relationship in temperate streams and
its variation with stream catchment area. Canadian Journal of Fisheries and Aquatic Sciences 53:99-105
Vanni, M.J., and D.L. Findlay. 1990. Trophic cascades and phytoplankton community structure. Ecology
71:921-937
Wade, A.J., P.G. Whitehead, G.M. Hornberger, H.P. Jarvie, and N. Flynn. 2002a. On modelling the impacts of
phosphorus stripping at sewage works on in-stream phosphorus dynamics: a case study for the River Kennet.
The Science of the Total Environment 282-283:395-415
Wade, A.J., P.G. Whitehead, G.M. Hornberger, and D.L. Snook. 2002b. On modelling the ow controls on
macrophyte and epiphyte dynamics in a lowland permeable catchment: the River Kennet, southern England.
The Science of the Total Environment 282:375-393
Wall, G.R., P.J. Phillips, and K. Riva-Murray. 1998. Seasonal and spatial patterns of nitrate and silica
concentrations in Canajoharie Creek, New York. Journal of Environmental Quality 27:381-389
Walton, S.P., E.B. Welch, and R.R. Horner. 1995. Stream Periphyton Response to Grazing and Changes in
Phosphorus Concentration. Hydrobiologia 302:31-46
Welch, E.B., J.M. Jacoby, R.R. Horner, and M.R. Seeley. 1988. Nuisance biomass levels of periphytic algae in
streams. Hydrobiologia 157:161-168
Winter, J.G., and H.C. Duthie. 2000. Epilithic diatoms as indicators of stream total N and total P concentration.
Journal of the North American Benthological Society 19:32-49
Wright, R.M., and A.J. Mc Donnell. 1986. Macrophyte growth in shallow streams: biomass model. Journal of
Environmental Engineering 112:967-982
Young, K., G.K. Morse, M.D. Scrimshaw, J.H. Kinniburgh, C.L. MacLeod, and J.N. Lester. 1999. The relation
between phosphorus and eutrophication in the Thames catchment, UK. Science of the Total Environment
228:157-183
Zelinka, M., and P. Marvan 1961. Zur Przisierung der biologischen Klassikation des Rheinheit Fliessender
Gewsser. Archiv fur Hydrobiologie 57:389-407
4 Organic pollution in rivers
Maurizio Pettine, Antonio Ruggiero, Stefano Fazi, Andrea Buffagni, Jens Mller Andersen,
Peter Roncak and Nikolai Friberg
4.1 Introduction
The European Environment Agency (EEA), in the Dobris Report and in the Report on European
Rivers and Lakes (EEA 1991, 1994), underlined the importance of organic pollution in European
rivers. Although, more recently, the EEA has stated (EEA, 2003a and 2003b) that the amount of
degradable organic matter (OM) discharged with wastewaters has been signicantly reduced in
Europe since the 1990s. The objective of this chapter is to review the existing knowledge on the
relationships between the content of organic matter in rivers and the WFD relevant biological
quality elements, mainly focusing on the impacts of OM discharged with wastewaters, one of the
main causes of pollution in rivers.
Role of organic matter in rivers
The sources of organic matter in rivers are mainly allochtonous inputs from the catchment
(especially from the riparian areas) and autochthonous production by primary producers. When
an excess of organic wastewaters are discharged into natural waters the ecological balance of the
system is disturbed. Organic pollution can affect the natural longitudinal trends of respiration and
photosynthesis rates. Respiration is enhanced downstream of an input of organic waste, whereas,
further downstream photosynthesis becomes more important, where organic wastes are partially
mineralized and suspended solids have settled out (Figure 4.1). The inputs of degradable organic
matter have led to dramatic deteriorations of river ecosystems. This is caused by the increased
availability of trophic resources for heterotrophic organisms, by covering the riverbed habitats
with sludge and, in extreme cases, by deoxygenating the river water.
Figure 4.1 Longitudinal distributions of
photosynthesis (green line) and respiration (blue
line) in a river downstream of a wastewater
discharge containing a high concentration of
organic matter
Distance
Organic Waste
High Respiration
High Production
ORGANIC POLLUTION IN RIVERS | 99
The major sources of organic matter pollution are domestic wastewater, industrial wastewater
from the food, pulp and paper processing industries and agricultural discharges from livestock
production and sh farming. In addition, occasional discharges, such as storm-water run-off from
urban areas, can also be important sources of pollution.
The specic impacts mainly depend on two factors: the amount and the quality of the organic
matter discharged. Furthermore, the extent of the ecological damage depends on the
hydromorphological characteristics of the river and the presence of other concurrent pressures
affecting the system. Figure 4.2 shows changes in water quality and biota downstream of
wastewater discharges. It is worth noting that the benthic fauna can be more heavily affected in
areas next to the discharge point, rather than at sites where oxygen is at it its lowest, due to
deposits of organic matter or bacterial growths on the riverbed. These patterns are typical for
large rivers while in small streams chemical and biological characteristics downstream of a
wastewater input show a much higher variability.
Figure 4.2 Typical changes of chemical and
biological characteristics downstream of a wastewater
input to a river (modied from Hynes, 1960).
A
Distance downstream
Outfall
B
C
D
BOD
Oxygen
Bacteria Algae
Sewage fungus
Asellus
Tubicidae
Clean water fauna
Chironomidae
NH
4
NO
3 PO
4
Characteristics of organic matter in wastewaters and rivers
Human activities alter the natural levels of organic matter in rivers mainly by (i) modications of
the river watershed (urban and agricultural land use), and (ii) direct organic matter release into
rivers (domestic and industrial wastewater, livestock and sh farming). Impacts on chemical water
quality from organic discharges have been described and modelled during the last hundred years
(Streeter and Phelps, 1925; Hynes, 1960). A key research eld focused on the fate of the organic
matter and on the development of models able to describe the impact of organic matter on the
concentration of dissolved oxygen.
Decomposition and dilution processes reduce the concentration of organic matter downstream of
the discharge point and, as the distance from a sewage efuent increases, organic matter
concentration may possibly reach the same values detected immediately upstream of the discharge
point. The rst models (Streeter and Phelps, 1925) were based on oxygen consumption described
by a rst order mineralisation of the organic matter and a re-aeration proportional to the oxygen
decit. The implementation of this model during the last decades has led to the development of
more comprehensive models describing the temporal and spatial dynamics of the oxygen
concentrations in rivers. Examples of such models are the river models established by the IWA
(International Water Association) Task Group on River Water Quality Modelling (Reichert et al.,
2001; Shanahan et al., 2001; Vanrolleghem et al., 2001) and the modelling framework for
simulation of river and stream water quality available at US EPA homepage (Chapra and Pelletier,
2003). These recent river water quality models give a dynamic description of chemical variables
such as dissolved oxygen (DO, mg O
2
/l), biological oxygen demand (BOD) and ammonium.
Domestic and industrial wastewater discharges are the most important sources of organic
pollution in rivers (Table 4.1). If not properly treated, such discharges greatly affect the quantity
and quality of organic matter in rivers. Since organic matter represents a trophic resource for many
organisms, these changes will directly affect the river biota.
Parameter Food processing
(IRSA, 1980)
Urban Runoff
(Gray, 1989)
Paper industry* Domestic
(Henze et al., 1995)
Suspended Solids 4.2-13000 2-11300 <5-70 120-450
BOD
5
5.2-25000 1-700 90-1150 100-350
COD 2-22400 5-3100 235-2000 210-740
N total 0.05-1313 1.1-6.2 3.2-51.2 20-80
N organic 0.3-510 0.1 16 8-30
N-NH
4
0.1-66 0.1-14 12-50
P total 0.02-520 0.09-4.4 0.1-0.3 4-23
Proteins 210-560 7-25
Fats 1.2-3000 30-100
Carbohydrates 100-1500 10-40
* Report by Burgo paper industry (2002)
Table 4.1 Concentration (mg/l) of various compounds in wastewaters.
4.2 Parameters describing organic pollution
As the decomposition of organic matter requires oxygen, severe organic pollution leads to
deoxygenation of river water. Therefore, dissolved oxygen (DO, mg O
2
/l) is probably one of the
most important tracers of organic pollution. However, DO varies diurnally, especially in small
rivers, making its use in traditional monitoring programmes problematic. Photosynthesis during
daylight hours often leads to oxygen supersaturation, but the high respiration rates during the
night may lead to critically low levels of dissolved oxygen.
The parameters generally used to describe the levels of organic pollution in both wastewater and
rivers are:
Biological Oxygen Demand (BOD, mg O
2
/l). The BOD-analysis is performed by measuring the
oxygen consumption (by micro-organisms) for the oxidation of organic matter and ammonia
in a sample during a predetermined period (5, 7 or 25 days). The method consists of lling to
overowing, an airtight sample-bottle of the specied size and incubating it at 20C in the
dark to prevent any possibility of photosynthetic production of DO. Dissolved Oxygen is
measured initially and after incubation, and BOD is computed from the difference between
initial and nal DO (APHA, 1998). The addition of thiourea to the sample prevents the
oxidation of ammonia, thus making possible to measure BOD due to organic matter only.
Chemical Oxygen Demand (COD, mg O
2
/l). The COD analysis is performed by a chemical
oxidation with dichromate or permanganate. COD-values determined by oxidation with
dichromate give a good picture of the total content of organic matter in the sample as almost
all organic matter is oxidised, while permanganate gives an incomplete oxidation.
Total Oxygen Demand (TOD, mg O
2
/l). The TOD analysis is performed by oxidation at high
temperature in the presence of a suitable catalyst. All organic compounds are oxidised by
TOD analysis.
Total, particulate and dissolved organic matter (TOM, POM and DOM, mg/l). The separation
between dissolved and particulate matter is traditionally achieved by ltering through 0.2-0.7 m
lters. These different fractions are generally determined by instrumental analysis and expressed
as total, particulate and dissolved organic carbon (TOC, POC and DOC, mg C/l). Carbon is
approximately 50% of the organic fraction in domestic wastewaters (Tchobanoglous et al., 2002).
Levels of BOD in rivers and wastewater
The monitoring of BOD is probably the oldest and most widespread method to quantify organic pollution.
Hynes (1960) states that in 1898 the British Royal Commission on Sewage Disposal investigated a large
number of British rivers and found that a 5 day BOD (BOD
5
) of river water gave a very fair measure of
their quality, and classied the rivers according to the BOD
5
as reported in Table 4.2.
The European Environment Agency (EEA, 1991; EEA, 1994) compiled monitoring results from
European rivers and found average BOD
5
levels for near pristine rivers of about 1.6 mg/l. EEA
concluded that river reaches slightly affected by human activities generally have a BOD below 2
mg/l whereas a BOD exceeding 5 mg/l generally indicates pollution.
River class BOD
5
(mg O
2
/l)
Very clean 1
Clean 2
Fairly clean 3
Doubtful 5
Bad 10
Table 4.2 River classication based on BOD
5
.
Typical BOD levels in untreated and standard biologically treated urban wastewater are given in
Table 4.3. An extended biological treatment is likely to produce a wastewater with a BOD
5
of
about 5 mg/l which is far below the limits of 20 mg BOD/l and 75 mg COD/l required by the
Urban Waste Water Directive (91/271/EEC).
Element Name Raw Wastewater
(g O
2
/m
3
)
Biologically treated wastewater
without nitrication
(g O
2
/m
3
)
C
BOD5
5-daybiochemical oxygen demand 280 25
C
BOD7
7-daybiochemical oxygen demand 320 30
C
BOD
Total biochemical oxygen demand 400 35
C
COD
Chemical oxygen demand 600 100
S
S,COD
Easily degradable matter, measured as COD 60 5
X
S,COD
Slowly biodegradable matter 200 10
Table 4.3 Values for the different measuring methods for two different types of wastewater (Henze et al., 1995).
Total organic carbon in rivers
Organic matter in rivers has a continuous size spectrum, from dissolved small molecules to
macromolecules, and from aggregates to organisms. An arbitrary denition of particulate organic
matter (POM) refers to the fraction of material retained on a lter with a pore size from 0.2 to
0.7 m. POM can originate both from autochthonous production (phytoplankton and periphyton)
and from allochthonous sources (soils, riparian vegetation and wastewater). It can also be
produced in situ by coagulation-occulation processes induced by chemico-physical variations of
pH and ionic strength, which often happens in the tidal zone of rivers. In situ production of POM
may be also mediated by biological activities through the excretion in the water column of
esopolymeric substances which are able to give rise to colloids and particles, depending on the
water turbulence conditions of the system.
Dissolved organic matter (DOM) may originate from allochthonous discharges or be produced in
situ during the mineralization of organic matter. Most DOM in rivers is very stable and, generally,
the largest fraction of organic matter in rivers consists only of slowly degradable compounds.
There is generally 10 times more DOM than POM in river water (Meybeck, 1982). However, the
POM fraction can be considerably higher in rivers with high contents of suspended matter (Ittekkot
and Laane, 1991).
Dissolved and particulate organic carbon (DOC and POC, respectively) may be combined in a
single measurement of total organic carbon (TOC) on unltered water. For more than 100 rivers,
mainly located in lowland areas, TOC was found to be generally between 2 and 30 mg/l with a
median value of 10 mg/l (Meybeck, 1982). Relationships between COD and TOC have also been
established in specic rivers. For example, Pettine et al. (1985) found the following linear equation
(SD = 0.9 mg/l) for the Po River:
TOC (mg C/l) = 0.51 + 0.28 COD (mg O
2
/l)
Degradation of organic matter by heterotrophic bacteria is one of the major processes controlling
the oxygen level of aquatic ecosystems and thus their quality. However, parameters that have
been traditionally used in models for describing organic matter degradation, such as BOD, COD
and TOC, are often insufcient to predict correctly the impact of waste water release on aquatic
ecosystems. Ecological models able to describe riverine ecological processes and useful for the
development of rehabilitation scenarios would require a detailed characterization of the quality
of the organic matter (biodegradable and refractory; Servais et al., 1999) and a better
understanding of the complex relationship between the bacterial community (structure and
functioning) and organic matter availability (Fazi et al., 2005).
Biological indicators of organic impact
The enrichment with organic matter can affect the competitiveness of species both through the
change in trophic resources available and because of deoxygenation caused by increased
heterotrophic activities. Severe deoxygenation may lead to a decrease in community diversity with
only a few dominant species able to live under low oxygen conditions. However, river biota can be
affected by organic pollution even if the level of dissolved oxygen is unaffected. Further, the
riverbed can be covered with a bacterial lm, and the oxygen concentration in the sediment can
be deteriorated even without signicant changes in DO in surface waters. Ammonium discharged
or released by decomposition processes may be converted, at high pH and temperature, to
ammonia that in turn is extremely toxic to biota.
In Europe, there is a long tradition in the use of biotic indices to characterize river quality, but only
a few studies focus on the quantitative relations between biological indicators and pollutants
concentrations. The WFD requires the use of ve biological elements to assess the impacts of
pressures on river quality:
Benthic invertebrates
Macrophytes
Benthic algae
Fish
Phytoplankton.
Among these ve key biological elements, benthic algae and benthic invertebrates are widely
used as indicators for the assessment of impacts of organic pressures on river quality. There are
currently two favoured approaches to stream assessment (Milner and Oswood, 2000). One
approach considers multimetric indices, whose goal is to incorporate in a single value different
components of the biological community that are sensitive to a broad range of human activities
(e.g., Barbour et al., 1999; Buffagni et al., 2004; Dahl et al., 2004; Hering et al., 2004). Each
index is limited to cover only organisms within one group of key biological elements (phytoplankton,
periphyton, macrophytes, invertebrates, sh). The other approach relies on multivariate statistical
methods to uncover patterns in taxonomic composition. Examples of this second method include,
multivariate systems like BMWP (Armitage et al., 1983), BEAST (Reynoldson et al., 2000) and
PERLA (Zahrdkov et al., 2000).
4.3 The Saprobic System
The Saprobic System, which is widely used in Central Europe, focuses on organic pollution and
the associated decrease in dissolved oxygen. It
was rst developed by Kolkwitz and Marsson
(1902) and subsequently changed and improved
(Rolauffs et al., 2003). According to the
Saprobic System, water bodies are characterised
by the degree of heterotrophy (ratio between
heterotrophic and autotrophic production), the
resulting trophic scale is reported in Table 4.4.
Saprobic index related to chemical parameters
A determination of the saprobic value is based on the sampling and the identication of species
of fauna and ora and a comparison with the Saprobic characteristics established for each species
(see e.g. Liebmann, 1951 and Sldec
^
ek, 1973). In Table 4.5 average values for BOD
5
, DO and NH
4
+
for the upper limits of the saprobic classes are given, derived from data in Slovakian rivers. The NH
4
+
values represent 90
th
percentiles. Figure 4.3 shows the dependence of the saprobic index on
degradable organic matter concentration in Danish rivers under different hydrological regimes.
Trophic group Denition
(I) oligosaprobic Practically unpolluted
(II) -mesosaprobic Rather slightly polluted
(III) -mesosaprobic Rather heavily polluted
(IV) Polysaprobic Heavily polluted
Table 4.4 Saprobic system trophic groups.
Saprobic classes Saprobic index BOD
5
(mg/l) DO (mg/l) NH
4
+
(mg/l)
Xenosaprobity <0,5 <1 >8 <0.5
Oligosaprobity <1,5 <2,5 >6 <0.9
-mesosaprobity <2,5 <5 >4 <1.7
-mesosaprobity <3,5 <10 >2 <3.3
Polysaprobity <4,5 <50 >0,1 <5.9
Table 4.5 Saprobic classes related to BOD
5
(mg/l), DO (mg/l) and NH
4
+
(mg/l) values.
Figure 4.3 Relationships between the degree of pollution (saprobic index) and the concentration of
degradable organic matter (BOD5) in Danish streams (Andersen, 1994).
BOD
5
(mg/l)
D
e
g
r
e
e

o
f

p
o
l
l
u
t
i
o
n
IV
III
II
I
0 2 4 6 8 10 12 14 16 18 20 22
Slow ow
Moderate ow
Good ow
4.4 Phytobenthos
Attached lamentous green algae such as Cladophora are a frequent nuisance in shallow rivers
polluted by wastewaters. Usually their occurrence is mainly linked to the level of inorganic
nutrients in the river water (Whitton, 1970), but often enhanced growth is not seen in rivers with
elevated nitrate and phosphate concentrations if otherwise unpolluted. Other factors such as
organic compounds can be important.
Benthic diatoms are widely used as water quality indicators in rivers. In particular, diatom species
composition is used to evaluate water quality in terms of organic pollution (Fabri and Leclercq,
1984; Sldec
^
ek, 1986; Descy and Coste, 1991). Most of the diatom indices, e.g., the Trophic
Diatom Indices of Hofmann (1994) and Kelly (1998), the saprobic index of Sldec
^
ek (1973), the
Pollution Index of Descy (1979), and the Generic Diatom Index (Coste and Aypassorho, 1991),
are in essence numerical models that use species indicator values to predict water quality
parameters (Potapova et al., 2004).
Occurrence, species composition and biomass are usually related to the level of inorganic nutrients.
Organic pollution also affects the occurrence of benthic diatoms, and for many species saprobic
values can be found in the literature (Liebmann, 1951). Eichenberger and Wuhrmann (1966)
studied the effects of addition of sewage to experimental channels in Switzerland and found
signicant impacts on autotrophic and heterotrophic microbenthos with organic enrichments
corresponding to a few mg BOD/l.
4.5 Sewage fungus
One of the most well known indicators of pollution in streams with easily degradable organic
matter is probably sewage fungus. This consists of macroscopic colonies of heterotrophic micro-
organisms living upon the organic matter discharged into the stream and covering the stream
bottom and other substrates. The sheath bacterium Sphaerotilus natans is an important
constituent, but many other bacteria, fungi and animals contribute to the formation of sewage
fungus (Hynes, 1960). With readily available organic matter (dissolved sugars, fatty acids and
amino acids), and especially in cold water sewage fungus can develop even at low concentrations
of organic matter.
4.6 Macroinvertebrates
Currently, macroinvertebrates are the most commonly used element for biological classication of
rivers in Europe. Although the details in methodologies might vary from country to country, the
use of macroinvertebrates to assess the effects of organic pollution of rivers has a long history
throughout Europe. Since the key products of organic matter decomposition are nutrients,
eutrophication is very closely linked to organic pollution. Effects of organic pollution and
eutrophication on stream benthic fauna are linked to each other by the processes of organic
matter and nutrient transformation.
The rst biotic index based on macroinvertebrates was the Trent Index developed by the Trent
River Authority (Metcalfe-Smith, 1996). In common with the Saprobic System, the Trent Index
mainly targeted organic pollution. The Trent Index has been the forerunner to all biotic indices
currently in use in Europe. They differ, however, due to the regional differences in macroinvertebrate
species composition in Europe and due to the introduction of various modications related to the
taxonomic level used, the method of index calculation etc.
Currently used macroinvertebrates methods and metrics include:
ASPT, Average Score per Taxon used in the UK (Armitage et al., 1983)
BMWP, Biological Monitoring Working Party Score used in the UK (Armitage et al., 1983) and
Spain (BMWP, Alba-Tercedor and Sanchez-Ortega, 1988)
BBI Belgian Biotic Index (De Pauw and Van Hooren, 1983)
DSFI, Danish Stream Fauna Index (Skriver et al., 2000)
EPT index, number of Ephemeroptera, Plecoptera and Trichoptera taxa usually employed to
derive additional information, not to directly classify river sites
IBE (Indice Biotico Esteso), Italian Extended Biotic Index (Ghetti, 1997; APAT/IRSA-CNR,
2004), is a modied version of the Extended Biotic Index (Woodiwiss, 1964)
IBGN, Indice Biologique Global Normalis (AFNOR, 1992).
The recent EU project The Development and Testing of an Integrated Assessment System for the
Ecological Quality of Streams and Rivers throughout Europe using Benthic Macroinvertebrates
(AQEM) developed alternative approaches to assess the impact of organic pollution by dening
type-specic multimetric indices (Hering et al., 2003, 2004). Many different metrics were tested
and multimetric indices and/or scoring systems were developed for individual stream types and
countries. The majority of these scoring systems show a good relationship with the magnitude of
organic pollution within each stream type (e.g. Buffagni et al., 2004; Pinto et al., 2004; Skoulikidis
et al., 2004). These results highlight the need for an intercalibration of the biological response to
organic pollution across Europe (Sandin and Hering, 2004).
The current river monitoring procedures and assessment methods should be improved to cover
different habitat types in rivers. The results of the AQEM project concerning differences between
depositional and transport reaches of rivers (i.e. pools versus rifes), indicate that habitat type
should be considered (Brabec et al., 2004; Buffagni et al., 2004; Morais et al., 2004).
Examples of the relationship between single biological metrics, multi-metric indices and abiotic
indicators of organic pollution are shown in Figure 4.4. The relationships are based on the data
presented by Buffagni et al. (2004), which refer to small-sized rivers in Southern Italy. Rivers in
this geographic area suffer from a multi-pressure situation but are mainly affected by organic
pollution. A multivariate axis (PCA) based on BOD
5
, NH
3
, NO
2
, PO
4
, Escherichia coli, Cl
-
and O
2

concentration values, which are all tracers of organic pollution, was derived for samples from three
sampling seasons. The obtained PCA axis was related to the number of Trichoptera families (Figure
4.4: top, left), one of the single metrics that performed best in demonstrating organic pollution
(Pearson R
2
=0.587).
Figure 4.4 Examples of biological response to organic pollution in small-sized Southern European rivers
(data from Buffagni et al., 2004). From top to bottom and from left to right, the relationship between
Number of Trichoptera families, ICMi and a Multi Metric Index to a multi-variate axis (PCA) of organic
pollution is shown. The fourth diagram illustrates the relation between the ICM index and BOD
5
values.
M
M
I

i
n
d
e
x
I
C
M

i
n
d
e
x
I
C
M

i
n
d
e
x
#
T
r
i
c
o
p
t
e
r
a

f
a
m
i
l
i
e
s
Organic pollution score (PCA)
BOD
5
(mgl
-1
)
R
2
(Pearson) = 0.587
R
2
(Pearson) = 0.650
R
2
(Pearson) = 0.510
R
2
(Pearson) = 0.307
12
10
8
6
4
2
0
1.0
0.8
0.6
0.4
0.2
1.0
0.8
0.6
0.4
0.2
1.0
0.8
0.6
0.4
0.2
-2.0 -1.0 0.0 1.0 2.0 2.0 1.0 0.0 1.0 2.0
2.0 1.0 0.0 1.0 2.0
0.0 2.0 4.0 6.0 8.0 10.0 12.0
The combination of single metrics into multimetric indices usually enhances the general
performance of the index. In the example, the Intercalibration Common Metric index (ICMi:
Buffagni et al., 2005) developed for the European Intercalibration process and based on a family
level identication, is related to the multivariate axis of organic pollution. Along the x axis, high
values reect conditions of strong organic pollution.
Although the R
2
value slightly decreases, the overall response of the index is more suitable for
describing organic pollution (with the exclusion of heavily polluted sites). The performance of the
Multi-Metric Index dedicated to this stream type (left, bottom) is the best among all the metrics
tested (Pearson R
2
= 0.650). Nevertheless, a certain degree of dispersion of samples around the
regression line can be seen. This is not surprising, because the index is designed to assess general
degradation of river sites (i.e. not only organic pollution; Buffagni et al., 2004). This underlines
the need for considering organic pollution indicators together with e.g. hydro-morphological and
toxic parameters, so that a multi-pressure impact can be properly interpreted. The last graph
(right, bottom), shows the relationship between ICMi and a single organic pollution descriptor
(BOD
5
), which indicates that attempts to nd simple direct relationships betweem biological and
chemical metrics can fail or yield poor results when analysing eld data.
4.7 Secondary indicators (macrophytes, phytoplankton, sh)
Little information is available on the impacts of organic pollution on river macrophytes, although
they are clearly likely to be affected by high levels of organic matter. Two factors are likely to be
important: (i) deposits of organic matter and low redox potentials in the sediments that prevent
the occurrence of many macrophytes; (ii) increased turbidity and increased occurrence of attached
micro-organisms that reduce light availability for macrophyte growth.
It is very unlikely that organic pollution can impact river phytoplankton without leading to a more
signicant impact on heterotrophic organisms or benthic diatoms. Phytoplankton is therefore not
considered a valuable indicator in the assessment of ecological impacts from discharges of organic
pollutants into rivers. The protection of sh populations in rivers from impacts of organic pollution
has traditionally been a key issue in river pollution management. In the Fresh Water Fish directive
(78/659/EEC), focus is placed on criteria for dissolved oxygen, BOD and ammonia. The guidance
and mandatory criteria from the directive are reported in Table 4.6.
These criteria, that take into account the tolerances of sh to low oxygen concentration and
ammonia, have been the basis for the development of river models describing the impacts of
organic matter (from Streeter and Phelps, 1925 to Shanahan et al., 2001). This approach, based
on sh tolerances to oxygen and ammonia concentrations, has been successful regarding the
direct effects of organic pollution on sh, but it does not take into consideration the indirect
effects caused by the other changes induced by organic enrichment in the river ecosystem.
Dissolved oxygen
mg/l
BOD
5
mg/l
Total ammonium
mg/l
Free ammonia
mg/l
Mandatory Guidance Mandatory Guidance Mandatory Guidance Mandatory Guidance
Salmonid waters 7 (100%) 9 (100%)
3 1 0.04 0.025 0.005

Cyprinid waters 7 (50%) 8 (50%)
5(100%)
6 1 0.2 0.025 0.005

Table 4.6 Guidance and mandatory criteria from the Fresh Water Fish directive (78/659/EEC): water
quality criteria related to organic pollution.
4.8 Summary
River ecosystems experience many different types of organic inputs. While, in Europe, wastewater
discharges from urban areas and industry have substantially declined during recent decades,
other sources have become more important. These include urban storm-water, discharges from
agricultural husbandry, sh-farming, wastewater from scattered dwellings, and organic inputs
from eutrophied lakes.
There is substantial evidence in the scientic literature that benthic macroinvertebrates, benthic
algae and sh are effective biological elements for the assessment of the impact of organic matter
on river ecosystems. In the past, the relationships between BOD levels and saprobic indices have
been well described, especially for rivers polluted by untreated wastewaters. However, organic
pollution impacts are not necessary only related to oxygen decits. Changes in biota also occur at
low increases in organic matter (OM) concentration and without a substantial decit of oxygen in
the water column. Likewise, the quality of the organic matter discharged seems to be another
important characteristic in determining the impact on river biota. The chemical structure of
organic compounds inuences the behaviour of OM in aquatic ecosystems affecting its biological
persistence, microbial communities, interaction with suspended minerals, properties in the
production of colloids and gels and its complexing ability towards heavy metals, etc. All these
effects should be further investigated.
Mathematical models are available for the evaluation of chemical water quality. Concentrations
of DO, BOD, ammonium and other chemical constituents are well described in rivers as a function
of discharged organic pollutants. Nevertheless these models usually require rates and kinetic
coefcients not always available at local scale for different river types. Moreover, modes and
resilience time for river ecosystems recovering from organic pollution are not yet fully understood.
Despite the overwhelming scientic knowledge on the biological impacts of organic pollution in
river ecosystems, there is still a lack of operational tools to estimate the amount of OM to be
reduced in order to meet biological quality criteria for different river types in Europe.
4.9 References
AFNOR - Association Franaise de Normalisation. 1992. Dtermination de la qualit biologique des eaux
courantes: Indice Biologique Global Normalis. Norme NFT pp. 90-350
Alba-Tercedor, J.and A. Sanchez-Ortega. 1988. Un mtodo rpido y simple para evaluar la calidad biolgica
de las aguas corrientes basado en el de Hellawell, 1978. Limnetica 4:51-56
Andersen, J.M. 1994. Water quality management in the river Gudenaa, a Danish lake-stream-esturay system.
Hydrobiologia 275/276:499-507
APAT/IRSA-CNR. 2004. Indice Biotico Esteso (I.B.E). In: APAT, Manuali e linee guida 29/2003. APAT-IRSA/
CNR, Metodi analitici per il controllo della qualit delle acque 3:1115-1136
APHA, AWWA, WEF. 1998. Standard Methods, 20
th
ed., APHA Washington, DC
Armitage, P.D., D. Moss, J.F. Wright, and M.T. Furse. 1983. The performance of a new biological water quality
scores system based on macroinvertebrates over a wide range of unpolluted running-water sites. Water
Research 17:333347
Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid bioassessment protocols for use in
streams and wadeable rivers: periphyton, benthic macroinvertebrates and sh. Second Edition, EPA 841-B-
99-002. U.S. Environmental Protection Agency, Ofce of Water, Washington D.C.
Brabec, K., S. Zahradkova, D. Nemejcova, P. Paril, J. Koke, and J. Jarkovsky. 2004. Assessment of organic
pollution effect considering differences between lotic and lentic stream habitats. In: D. Hering, P.F.M.
Verdonschot, O. Moog, and L. Sandin (Eds.) Integrated Assessment of Running Waters in Europe, Hydrobiologia
516:331346
Buffagni A., S. Erba, M. Cazzola, and J.L. Kemp. 2004. The AQEM multimetric system for the southern Italian
Apennines: assessing the impact of water quality and habitat degradation on pool macroinvertebrates in
Mediterranean rivers. In: D. Hering, P.F.M. Verdonschot, O. Moog, and L. Sandin (Eds.), Integrated Assessment
of Running Waters in Europe, Hydrobiologia 516:313-329
Buffagni A., S. Erba, S. Birk, M. Cazzola,

C. Feld, T. Ofenbck, J. Murray-Bligh, M.T. Furse, R. Clarke, D. Hering,
H. Soszka, and W. van de Bund. 2005. Towards European Inter-calibration for the Water Framework Directive:
Procedures and examples for different river types from the E.C. project STAR. 11
th
STAR deliverable, STAR
Contract No: EVK1-CT 2001-00089, Quaderni Istituto Ricerca Acque 123, Rome, IRSA, pp. 468
Chapra, S., and G.J. Pelletier. 2003. QUAL2K: A modelling framework for simulating river and stream water
quality: Documentation and users manual. Civil and Environmental Engineering Dept., Tufts University,
Medford
Coste, M., and H. Ayphassorho. 1991. Etude de la qualit des eaux du bassin Artois-Picardie a laide des
communauts de diatomes benthiques. Application des indices diatomiques au rseau. Cemagref Bordeaux,
Agence de lEau Artois, Picardie, 227 pp. Rapport Convention dtude n. 90 X 3300 du 19 Juin 1990
Dahl, J., R.K. Johnson, and L. Sandin. 2004. Detection of organic pollution of streams in southern Sweden
using benthic macroinvertebrates. In: D. Hering, P.F.M. Verdonschot, O. Moog, and L. Sandin (Eds.), Integrated
Assessment of Running Waters in Europe. Hydrobiologia 516:161-172
De Pauw, N., and G. Van Hooren. 1983. Method for biological quality assessment of watercourses in Belgium.
Descy, J.P. 1979. A new approach to water quality estimation using diatoms. Nova Hedwigia 64:305-323
Descy, J.P., and M. Coste. 1991. A test of methods for assessing water quality based on diatoms. Verhandlungen
der Internationalen Vereinigung fr Theoretische und Angewandte Limnologie 24:21122116
EEA. 1991. Europes Environment. The Dobris Report. European Environment Agency, pp. 676
EEA. 1994. European Rivers and lakes. Assessment of their environmental state. European Environment
Agency, EEA Environmental Monographs 1, pp. 122
EEA. 2003a. Europes environment: The third assessment. Environmental assessment report no. 10. European
Environment Agency
EEA. 2003b. Europes water: An indicator-based assessment. European Environment Agency
Eichenberger, E., and K. Wuhrmann. 1966. ber Jahreszeitliche Vernderungen der Besiedlungsdichte in
Modelliessgewsser mit verschiedene Abwasserbelastung. Verhlungen of Internationale Vereinigung fur
Limnologie 16:888-896
Fabri, R., and L. Leclercq. 1984. Etude cologique des rivires du nord du massif Ardennais (Belgique): ore
et vgtation de diatomes et physico-chimie des eaux. 1. Contexte msologique, mthodes, analyses
physico-chimiques, synthse taxonomique, cologique et oristique, iconographie, bibliographie. Universit
de Lige, Station Scientique des Hautes-Fagnes, Robertville
Fazi, S., S. Amaltano, J. Pernthaler, and A. Puddu. 2005. Bacterial communities associated with benthic
organic matter in headwater stream microhabitats. Environmental Microbiology 7:1633-1640
Ghetti, P.F. 1997. Indice Biotico Esteso (I.B.E.). I macroinvertebrati nel controllo della qualit degli ambienti
di acque correnti. Provincia Autonoma di Trento, pp. 222
Gray, N.F. 1989. Biology of Wastewater Treatment. Oxford University Press, New York, USA
Henze, M., P. Harremoes, J.C. Jansen, and E. Arvin. 1995. Wastewater treatment biological and chemical
processes. Springler-Verlag Berlin Heidelberg
Hering, D., A. Buffagni, O. Moog, L. Sandin, M. Sommerhuser, I. Stubauer, C. Feld, R.K. Johnson, P. Pinto, N.
Skoulikidis, P.F.M. Verdonschot, and S. Zahrdkov. 2003. The development of a system to assess the
ecological quality of streams based on macroinvertebrates design of the sampling programme within the
AQEM project. International Review of Hydrobiology 88:345361
Hering D., O. Moog, L. Sandin, and P.F.M. Verdonschot. 2004. Overview and application of the AQEM
assessment system. In: D. Hering, P.F.M. Verdonschot, O. Moog, and L. Sandin (Eds.), Integrated assessment
of running waters in Europe. Hydrobiologia 516:120
Hofmann, G. 1994. Aufwuchs-Diatomeen in Seen und ihre Eignung als Indikatoren der Trophie. Bibliotheca
Diatomologica 30:1-241
Hynes, H.B.N. 1960. The Biology of Polluted Waters. Liverpool University Press, Liverpool, England
IRSA. 1980. La rimozione biologica dellazoto dalle acque di scarico. Quaderno Istituto di Ricerca sulle Acque
n. 53
Ittekkot V., and R.W.P.M. Laane. 1991. Fate of riverine particulate organic matter. In: E.T. Degens, S. Kempe,
and J.E. Richey (Eds.), Biogeochemistry of Major World Rivers. John Wiley and Sons, New York
Kelly, M.G. 1998. Use of the trophic diatom index to monitor eutrophication in rivers. Water Research 36:
236-242
Kolkwitz, R., and M. Marsson. 1902. Grundstze fr die biologische Beurteilung des Wassers nach seine Flora
und Fauna. Mitteilungen Prfungsanst. Wasserversorg. Abwasserreinig 1:1-64
Liebmann, H. 1951. Handbuch der Frischwasser und Abwasserbiologie. Mnchen
Metcalfe-Smith, J.L. 1996. Biological water-quality assessment of rivers: Use of macroinvertebrate communities.
In: G. Petts, and P. Calow (Eds.), River restoration. Blackwell Science, Oxford, UK, pp. 17-43
Meybeck, M. 1982. Carbon, nitrogen, and phosphorus transport by world rivers. American Journal of sciences
282:401-450
Milner, A.M., and M.W. Oswood. 2000. Urbanization gradients in streams of Anchorage, Alaska: a comparison
of multivariate and multimetric approaches to classication. Hydrobiologia 422/423:209-223
Morais, M., P. Pinto, P. Guilherme, J. Rosado, and I. Antunes. 2004. Assessment of temporary streams: the
robustness of metric and multimetric indices under different hydrological conditions. In: D. Hering, P.F.M.
Verdonschot, O. Moog, and L. Sandin (Eds.), Integrated Assessment of Running Waters in Europe.
Hydrobiologia, 516:229-249
Pettine, M., T. la Noce, R. Pagnotta, and A. Puddu. 1985. Organic and Trophic load of major italian rivers. In:
E.T. Degens, S. Kempe, and R.Herrera (Eds.), Transport of carbon and minerals in major world rivers Part 3.
Mitt.Geol.-Palaont. Inst. Univ. Hamburg 58:417-429
Pinto, P., J. Rosado, M. Morais, and I. Antunes. 2004. Assessment methodology for southern siliceous basins
in Portugal. In: D. Hering, P.F.M. Verdonschot, O. Moog, and L. Sandin (Eds.), Integrated assessment of
running waters in Europe. Hydrobiologia 516:191-214
Potapova, M.G., D.F. Charles, K.C. Ponader, and D.M. Winter. 2004. Quantifying species indicator values for
trophic diatom indices: a comparison of approaches. Hydrobiologia 517:25-41
Reichert, L., D. Borchardt, M. Henze, W. Rauch, P. Shanahan, L. Somlydy, and P. Vanrolleghem. 2001. River
Water Quality Modelling no. 1 (RWQM1): II. Biochemical process equations. Water Science and Technology
43:11-30
Reynoldson, T.B., K.E. Day, and T. Pascoe. 2000. The development of BEAST: a predictive approach for
assessing sediment quality in the North American Great Lakes. In: J.F Wright, D.W. Sutcliffe, and M.T. Furse
(Eds.), Assessing the biological quality of fresh waters RIVPACS and other tecniques. FBA:165180
Rolauffs, P., D. Hering, M. Sommerhuser, S. Rdiger, and S. Jhnig. 2003. Entwicklung eines leitbildorientierten
Saprobienindexes fr die biologische Fliegewsserbewertung. Umweltbundesamt Texte 11/03:1137
Sandin L., and D. Hering. 2004. Comparing macroinvertebrate indices to detect organic pollution across
Europe: a contribution to the EC Water Framework Directive intercalibration. In: D. Hering, P.F.M. Verdonschot,
O. Moog, and L. Sandin (Eds.) Integrated Assessment of Running Waters in Europe. Hydrobiologia 516:55-68
Servais, P., J. Garnier, N. Demarteau, N. Brion, and G. Billeni. 1999. Supply of organic matter and bacteria to
aquatic ecosystems through waste water efuents. Water Research 33:3521-3531
Shanahan, P., D. Borchardt, M. Henze, W. Rauch, P. Reichert, L. Somlydy, and P. Vanrolleghem. 2001. River
Water Quality Modelling no. 1 (RWQM1): I. modelling approach. Water Science and Technology 43:1-9
Skoulikidis, N.Th., K.C. Gritzalis, T. Kouvarda, and A. Buffagni. 2004. The development of an ecological
quality assessment and classication system for Greek running waters based on benthic macroinvertebrates.
In: D. Hering, P.F.M. Verdonschot, O. Moog, and L. Sandin (Eds.), Integrated assessment of running waters in
Europe. Hydrobiologia 516:149-160
Skriver, J., N. Friberg, and J. Kirkegaard. 2000. Biological assessment of running waters in Denmark:
introduction of the Danish Stream Fauna Index (DSFI). Verhandlungen der Internationalen Vereinigung fr
Theoretische und Angewandte Limnologie 27:1822-1830
Sldec
^
ek, V. 1973. System of water qualication from the biological point of view. Archiv fr Hydrobiologie,
Beiheft Ergebnisse der Limnologie 7:1-218
Sldec
^
ek, V. 1986. Diatoms as indicators of organic pollution. Acta Hydrochimica et Hydrobiologica 14:555566
Streeter, W.H., and E.B. Phelps. 1925. A study of the pollution and natural purication of the Ohio River.
Public Health Bull. 146. U.S. Public Health Service, Washington D.C.
Tchobanoglous G., F.T. Burton, and H.D. Stensel. 2002. Wastewater Engineering, Treatment, Disposal and
Reuse, Metcalf and Eddy, McGraw Hill, New York
Vanrolleghem, P., D. Borchardt, M. Henze, W. Rauch, P. Reichert, P. Shanahan, and L. Somlydy. 2001. River
Water Quality Modelling no. 1 (RWQM1): III. Biochemical submodel selection. Water Science and Technology
43:31-40
Whitton, B.A. 1970. Biology of Cladophora in freshwaters. Water Research 4:457-476.
Woodiwiss, F.S. 1964. The biological system of stream classication used by the Trent River Board. Chemistry
and Industry 14:443-477
Zahrdkov, S., J. Koke, J. Hodovsk, D. Vojtkov, D. Scheibov, Y. Pozkov, J. Schenkov, and J. Heleic.
2000. Predikn systm PERLA Prediction system PERLA. In: M. Rulk (Ed.) Limnologie na pelomu tiscilet.
Sbornk refert XII limnologick conference, Olomouc, Univerzita Palackho, pp. 260264
Acidication
1 - Niels Jepsen. Salmon caught by electroshing in river Varde, Denmark
2 - Finn Sivebak. Electroshing a lowland river from boat (River Gudenaa, Denmark)
3 - Laurence Carvalho. North Esk Reservoir (Scotland)
4 - Seppo Rekolainen. Nimphea in a Finnish lake
5 - Gary Free. Upper lake Kerry (Ireland)
1 2 3 4 5
5 Acidication of rivers and lakes
Torleif Bkken, Frode Kroglund, Eli-Anne Lindstrm, and Laurence Carvalho
5.1 Introduction
Acidication of water implies a pH reduction. The pH reduction can be due to anthropogenic
activities (e.g. burning of fossil fuels) but also to natural processes (e.g. presence of organic acids).
It is important to separate these two causes of low pH as the ecological impact of elevated H
+
will
vary. This difference is to a large extent due to the interaction between acid rain and the
mobilisation of aluminum (Al). Al interacts with metal sensitive tissues such as the gills of sh and
invertebrates and will impair physiological processes in both plants and animals affecting growth
and reproduction (Gensemer and Playle, 1999). While low pH will aggravate the effects of Al,
increased concentrations of e.g. calcium (Ca), increased ionic strength and the presence of various
organic and inorganic ligands will act to reduce toxicity (see reviews in: Rosseland and Staurnes,
1994; Gensemer and Playle, 1999). The ecological effect depends on exposure intensity, the
species and life stages exposed, exposure duration and the possibility for post-exposure recovery
of a sub-lethal exposure. As toxicity is related to both H
+
and Al, acidication is a combined
stressor. The two toxicants must be interpreted in combination, but also separately. While some
species are mainly H
+
-sensitive, others are mainly Al-sensitive (Gensemer and Playle, 1999).
The ecological effect depends as such on the relative contribution of the prime toxicants and how
these act in concert. Due to the combined action of H
+
and Al, the pH-limits restricting organisms
in naturally acidic water-bodies are normally lower than the limits restricting the same species in
water having elevated concentrations of Al. pH limits restricting organisms are further inuenced
by all confounding factors giving rise to a large variability in the response surfaces.
The most obvious biological impacts of acidication are related to changes in the composition
and abundance of invertebrate and macrophyte periphyton communities and the reduction or
extinction of sh populations (Raddum and Fjellheim, 1984; Farmer, 1990; Rosseland and
Staurnes, 1994; Gensemer and Playle, 1999). The community structure changes from one
supporting acid sensitive organisms to one supporting acid tolerant organisms. Only few of these
changes are acidication specic, as similar changes can be related to a variety of pressures.
Annex V of the WFD outlines both the physico-chemical and biological elements to be included. The
chemical elements are pH, alkalinity and acid neutralising capacity. All three are related to the true
toxic components: hydrogen activity and Al. The annex does not detail how naturally acidic sites
should be separated from sites being acidic due to anthropogenic activity. The biological elements:
macrophytes and phytobenthos, phytoplankton, benthic invertebrates and sh are largely assessed
using species composition and abundance with the addition of age structure for sh.
5.2 Sources of acidication
Acidication can be brought about by acid rain, but also by activities such as draining of
wetlands and mining activities (Hyne and Wilson, 1997). These latter sources of acidication have
ACIDIFICATION IN RIVERS AND LAKES | 117
only local effects, while acid rain, being trans-boundary, has regional and global effects. Acid rain
mainly results from the burning of fossil fuels and the subsequent emission of sulphur dioxide and
nitrogen oxides into the atmosphere. In addition, ammonia from agricultural activities contributes
locally to acid loading through the nitrication process. In air and in contact with water the oxides
are transformed into sulphuric and nitric acids. These acids can be transported over large distances,
before being deposited as acid rain. In Europe, acidication has a negative impact in water
bodies with a low alkalinity, i.e. in non-calcareous catchments.
The cause and effect chains linking acid rain to acid water were rst established in the 1950s.
Acidication became an international issue around 1970 (Almer, 1974). Since then, focus has
changed from pH being regarded as the main factor for environmental change prior to 1980 to
include the toxicity of Al mobilised at low pH values. Focus has also changed from proving effects
and establishing causal relationships in the 1980s to the present day focus on recovery or the lack
of recovery following deposition reductions (Skjelkvle et al., 2003). Although water quality has
improved, episodes can still be detrimental and hinder the reestablishment of acid sensitive species.
In addition, recolonisation rates are low in regions heavily impacted by acid rain.
Owing to the reduced emission of sulphuric compounds in Europe there has been a substantial
reduction in the atmospheric deposition of acidity since the 1970s. However, there has been no
signicant decrease in nitrogen deposition (Alewell et al., 2001). These trends are visible in the
changes in wet deposition of sulphur and nitrogen over southern Norway (Figure 5.1). Future
effects relating to the deposition of N represent an uncertainty. The reduction in acid loading will
initiate a recovery process, the rate of which is related to changes in the deposition of acids, but
also to local variations in soil properties e.g. weathering rates (Skjelkvle and Wright, 1991). Future
changes in water quality will also depend on climate change, where sea-salt episodes, elevated
temperature and the mobilisation of organic matter will all contribute to changes in acidication.
Figure 5.1 Trends in the wet deposition of sulphur and nitrogen compounds shown as averages from
7 representative sites in southern Norway.
Wet deposition mg/m
2
1500
1000
500
0
1974 1982 1990 2002
SO4-S
NO3-N & NH4-N
5.2 Impacts on water quality
Key chemical water quality elements
The chemical impact acid rain has on surface water quality is well known. The immediate effect
of acidication is a reduction in pH and alkalinity and an increase in Al. In acid water, Al is
present in a number of forms and toxicity is related to the labile/cationic or inorganic form of Al
(LAl). The concentration of LAl depends on many water chemistry properties, including pH and
the presence of various organic (e.g. TOC) and inorganic (e.g. Si, F) ligands. The impact of acid
precipitation on pH and metal concentration in water bodies depends on the watershed geology
and vegetation, primarily the content of limestone, but also on the content of humic matter in the
surface waters.
The changes in water quality can be summarised into changes in alkalinity or acid neutralizing
capacity (ANC
cb
). ANC
cb
is calculated as the base cations - anions (Reuss and Johnson, 1986)
and can be modeled (see later). Based on the empirical relationship between ANC
cb
and the
biological status established for various organism groups (Raddum and Fjellheim, 1984; Bulger et
al. 1993; Lien et al., 1996; Raddum and Skjelkvle, 1995; Juggins et al., 1995; Kroglund et al.,
2002), ANC
cb
is commonly used to hind- and forecast changes in water quality and associated
changes in biological status. The use of ANC
cb
as a surrogate for the true toxic components is
possible due to the close relationship between ANC
cb
and pH and LAl. (Figure 5.2 and 5.3). An
ANC
cb
threshold of 20 eq l
-1
is currently considered as an indicator of water quality limits
identifying good ecological status mandated by the WFD. However, the relationship between
ANC
cb
, water quality and biological health is not straightforward. The relationship between ANC
cb
,
pH and LAl is affected by TOC (Figure 5.2 and 5.3). In high TOC waters, pH is lower and LAl is
higher for a given ANC
cb
value than the relationship observed in low TOC waters. To protect the
biological community, the ANC
limit
must therefore be set higher in water containing organic matter
than in clearwater systems (Lydersen et al., 2004). The current ANC
limit
is to a large extent based
on clearwater lakes. The higher ANC
limit
needed to protect organisms in high-TOC waters can be
considered contradictory to the generally accepted view that organic matter reduces Al-related
toxicity. Organic matter reduces the toxicity of total-Al, but the increased H
+
-activity permits more
Al to remain present in toxic form. Although the relative concentration of LAl to total-Al decreases
with increased TOC, this does not say that toxicity is negligible. Current ANC
cb
limits will not
protect sh in more organic rich sites.
An index based on ANC
cb
is preferable to those based on H
+
and LAl. While ANC can be modelled
using knowledge regarding process-based relationships, H
+
and LAl are more difcult to model.
Al is fractionated and analysed using a large variety of protocols where the reported concentrations
are not comparable. Furthermore, correct determination of the free metal concentration is difcult
due to its sensitivity to changes in pH and temperature. Downstream of the conuence of acid
and more neutral waters LAl will polymerize into lesser or non-toxic forms of Al. As unstable forms
can persist for hours, extensive areas of the river can be under the inuence of this ongoing
polymerizing process. Dose-response relationships that are based on in situ fractionation of Al
(within seconds after sampling) yield better and different relationships from those based on
traditional stored water samples (Kroglund et al., 2001).
Figure 5.2 Relationships between pH (a) and labile aluminium (b) and ANC (base cations - anions)
for Norwegian lakes having low (<1.5 mg C/l) and high (>5 mg C/l) organic content (Data from the
Norwegian 1000-lakes survey in 1986).
350
300
250
200
150
100
50
0
7.5
7
6.5
6
5.5
5
4.5
4
100 5 0 0 50 100 150 100 5 0 0 50 100 150
ANC-1 (CB) ANC-1 (CB)
p
H
L
a
b
i
l
e

A
l
,

g

L
-
1
TOC; 5-10 mg L
-1
TOC; 0-1.5 mg L
-1
TOC; 5-10 mg L
-1
TOC; 0-1.5 mg L
-1
Figure 5.3 Response curves showing the relationship between ANC and the probability of occurrence
of (A) the diatom Achnanthes minutissima, (B) the mayy Baetis rhodani and (C) non-impoverished
macroinvertebrate assemblage and (D) brown trout. Figures redrawn from Juggins et al. (1995) and Lien
et al. (1996).
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
100 50 0 50 100 150 200
100 50 0 50 100 150 200
80 70 60 50 40 30 20 10 0 10 20 30 40
1.0
0.8
0.5
0.3
0.0
ANC (eq/l) ANC (eq/l)
ANC (eq/l)
A
B
C
D
extinct
reduced
not affected
p
r
o
b
a
b
i
l
i
t
y
p
r
o
b
a
b
i
l
i
t
y

o
f

o
c
c
u
r
r
e
n
c
e
Due to the combined action from H
+
and LAl, it is important not to confuse natural acidic
conditions with anthropogenic acidication. While both are associated with a low pH, elevated
levels of Al in its toxic form are mainly found as a result of anthropogenic acidication. Water
bodies acidied by natural processes can have a low pH due to a high content of organic (humic)
matter and following oxidation of sulphide rich soils. Naturally acidic water bodies can as such
have a high ecological status despite having a low pH restricting the diversity of the biological
community.
5.3 Water quality and biological modelling
There are numerous models based on chemical criteria. Most models rely on the critical load
concept. Critical load is dened as the maximum amount of a pollutant that can be deposited on
an ecosystem without having adverse effects (Nilsson and Grennfelt, 1988). Application of the
critical load approach involves the identication of a key organism (or organisms) to be protected
and the critical limit with respect to a water quality indicator. With respect to acidication, the
model should be able to relate changes in deposition to the indicator status. Figure 5.3 shows
examples of relationships between a water chemistry indicator (ANCcb) and the probability of
occurrence of different types of species and species groups.
Several chemical parameters or indices are being used as water quality indicators; e.g. pH, Al,
alkalinity, Ca/H+ ratio and ANCcb which is the preferred indicator. Changes in water quality that
have no biological implication are to be interpreted as acceptable. As the chemical and biological
sensitivity to acidication varies according to bedrock and soil type the critical load varies
accordingly (Figure 5.4).
Critical loads, and if those are exceeded, are calculated using various static models, such as the
steady-sate water chemistry model (SSWC) and the rst-order acid balance model (FAB; Posch
et al., 1997). Steady-state models indicate that something will happen in the future if, for example,
pollution is reduced but does not indicate the timescale of recovery.
Figure 5.4 Critical loads for acid deposition.
Red indicates high sensitivity, blue low.
(Source: The Swedish NGO Secretariat on Acid Rain
1
)
1
http://www.acidrain.org/acidication.htm
<200
200-400
400-700
700-1000
1000-1500
>1500
eq/ha/yr
Geochemistry and soils play a central role in the response of forests, lakes and rivers to acid rain.
Dynamic, process-based models integrate and interpret theoretical knowledge from soil science
and hydrochemistry with results from experiments and monitoring. Response times are related to
factors such as mineral and organic matter content, weathering rates etc; all properties that are
more or less unique for an individual site. Because of the complex nature of interactions, dynamic
models are used to predict the rate of change over time. Experimental data from ongoing
experiments (long time-series) are used to validate these models. The main dynamic model used
in acidication studies is the MAGIC model.
At present there are no dynamic models for biological responses to acidication except for a
model developed for Atlantic salmon in Canada (Korman et al., 1994). There are several non-
dynamic models based on the relationship between species composition and abundance and
water chemistry (see chapter on individual organism groups).
5.4 Macrophytes and attached algae
Aquatic plants are not only impacted because of reduced pH and elevated Al, but also by the
change in inorganic carbon speciation. The importance of inorganic carbon as a key element in the
build up of organic material is often overlooked in acidied ecosystems. Carbon is the main structural
element and is needed in much larger quantities than nitrogen and phosphorus during primary
production (Redeld, 1958). The speciation of inorganic carbon changes with acidication; from
bicarbonate (HCO
3
-
) to free carbon dioxide (CO
2
) in acid water (Wetzel, 1983). This change in the
relative proportions of inorganic carbon species causes extensive changes in the species composition
of freshwater primary producers, because some species use bicarbonate to produce organic material,
while others use carbon dioxide. This is increasingly being acknowledged as the main cause for
changes in species composition and diversity when a water body becomes acidied (Figure 5.5).
Figure 5.5 Relation between pH and the relative proportions of inorganic carbon species of CO
2
(+H
2
CO
3
),
HCO
3
-
and CO
3
- -
. Slightly modied from Wetzel (1983).
pH
4 5 6 7 8 9 10 11 12
100
50
0
%

t
o
t
a
l

C
O
2
Free CO
2
HCO
3
-
CO
3
- -
Some of the structural changes caused by acidication are well documented:
Species composition change, partly because of a change in carbon source
Species diversity decreases
Vegetation biomass may increase.
In addition to the effect of a lower pH and change in carbon source, the increased atmospheric
deposition of nitrogen compounds may increase the availability of this nutrient and increase the
N/P ratio. This may lead to P-limitation of plant growth in the water and cause a shift towards
species requiring a higher N availability. The phosphorus content in acidied Nordic rivers is
usually low, generally less than 5 g/l P, and the N/P ratio was around 20-40 prior to the
increase in N deposition. Today, N/P ratios around 100-200 are common in these rivers. Some
attached algae, mosses and macrophytes are reported to be particularly common in waters with
such elevated N/P ratios (Lindstrm et al., 2004).
Many species of benthic algae, bryophytes and macrophytes have well dened tolerance limits of
acidity (Lindstrm et al., 2004). Surprisingly, many taxa were found to tolerate a pH <5.0 and
were classied as acid-tolerant (Table 5.1). Some even seemed to prot from acidication and
increased signicantly in frequency and biomass in markedly acidied rivers (pH <5.0). In contrast,
a large number of taxa were never found below pH 6.5 and some were even limited to rivers with
a pH permanently above 7.0. Different requirements concerning carbon source is probably the
main reason for these differences.
During acidication aquatic vegetation undergoes a change in species composition and a decline
in diversity. The study by Lindstrm et al. (2004) demonstrated that the impact on regional
diversity was more pronounced than at a given locality. The distribution of species according to
their tolerance of low pH values is shown in Figure 5.6. Acidity-tolerant algae increased in
frequency and occurred in most of the acidied rivers.
River acidication may be accompanied by an increase in plant biomass and occasionally one or
more species occurs in mass abundance. The most common and striking phenomenon is the
proliferation of lamentous green algae covering most of the riverbed. These mats/clouds usually
consist of Zygnemales, mainly one or a few species within the genera Mougeotia and Zygogonium.
However, plant biomass is not a useful indicator of acidication.
Vegetation
type
Total number of
taxa assessed
Acid tolerant
pH < 5
Weakly acid sensitive
pH > 5
Moderately sensitive
pH > 5.5-6.0
Strongly sensitive
pH > 6.5
Benthic algae 127 34 8 36 (divided in 2 groups) 49
Macrophytes 70 12 9 28 29
Bryophytes 11 5 3 3 Not assessed
Total 208 51 20 67 78
Table 5.1 Tolerance limits for acidity for 205 taxa of freshwater vegetation based on large datasets
collected in Norway (Lindstrm et al., 2004).
Critical limits
Lindstrm et al. (2004) showed the relationship between an index based on the presence of acid-
sensitive attached algae (Index of Acid Sensitivity) and the pH values of Norwegian rivers (Figure 5.7).
They concluded that the index showed that a pH > 6.3 was needed to ensure that the algal
community was not impacted by acidication. They also concluded that the algal community had
a substantial probability of being impacted if the pH was less than 5.6. In reference lakes that
have a naturally low pH, the community structure resembles that of an acidied lake.
The variability observed within the pH range 5.6 to 6.3 could have been caused by episodes, but
could also have been due to toxic properties of Al being present in anthropogenically acidied
water, but absent in waters having a naturally low pH (Gensemer and Playle, 1999).
Combined impacts of acidication and eutrophication on macrophytes
Aquatic macrophytes are intrinsically linked to lake water and sediments through their roots and leaves
and individual species are sensitive to physical and chemical changes in these media. Several studies
document signicant alterations of macrophyte assemblages in recent decades as a result of gross
reductions in pH in low alkalinity lakes. This has been due to high inputs of acidifying substances from
the atmosphere (mostly sulphur and nitrogen; Hornung, 1997) or from the drainage basin (Farmer,
1990; Riis and Sand-Jensen, 1998; as cited in Vestergaard and Sand-Jensen, 2000). Eutrophication is
caused by enhanced inputs of nutrients and is generally more of an issue in lowland environments with
higher population densities and more intensive agriculture. In upland regions eutrophication can
potentially be associated with atmospheric deposition of nutrients, primarily nitrogen compounds.
The pressures of eutrophication and acidication combined are, therefore, most likely to be signicant
in predominantly rain-fed, oligotrophic, low alkalinity lakes. These lakes are poor in (calcium)
bicarbonate and nutrients (Roelofs et al., 2002). Acidication and eutrophication cause changes in
Figure 5.6 Changes in the frequency of ve
acid-sensitivity classes of benthic algae along a
pH-gradient. Data based on benthic cyanobacteria
and chlorophyta found in 768 algal samples
collected in Norwegian rivers (Lindstrm et al., 2004).
Cyanobacateria
Chlorophyta
f
r
e
q
u
e
n
c
y

-

w
e
i
g
h
t
e
d

a
v
e
r
a
g
e
f
r
e
q
u
e
n
c
y

-

w
e
i
g
h
t
e
d

a
v
e
r
a
g
e
Tolerant
Slightly sensitive
Moderatly sensitive
Sensitive
Very sensitive
<5 5-5.5 5.5-6 6-6.5 6.5-7 7-7.5 >7.5
<5 5-5.5 5.5-6 6-6.5 6.5-7 7-7.5 >7.5
N, P and C budgets, with particularly apparent impacts on macrophyte community composition.
Nutrient-poor, low alkalinity lakes can have highly diverse plant communities, in which isoetid species
dominate (Lobelia dortmanna, Littorella uniora and Isoetes spp.). During the 20
th
century, however,
isoetid vegetation in lowland lakes has declined greatly and become threatened in many European
countries and North America (Smolders et al., 2002), with particularly devastating declines in the
Netherlands, northern Germany and Belgium (Arts, 2002). This provides important warnings for
other countries/regions, not least in the context of the EU Habitats Directive, and highlights the
need to focus on macrophyte responses to combined eutrophication-acidication pressures.
Alkalinity and transparency appear to be the most inuential drivers of macrophyte species
composition (Vestergaard and Sand-Jensen, 2000). The macrophyte composition of low alkalinity
lakes is very different from that of acidic or alkaline waters (Brouwer et al., 2002). Isoetid species
are most common in low alkalinity, oligotrophic lakes, as a consequence of the restricted availability
of inorganic carbon within the water layer (Sand-Jensen and Sondergaard, 1979). The change in
ammonium availability and inorganic carbon budgets as a consequence of acidication causes
changes from isoetid macrophyte species to more acid-tolerant species such as Juncus bulbosus
and Sphagnum spp in western European lakes (Arts, 2002). Enrichment with phosphate can then
lead to a luxurious growth of more loosely-attached or free-oating macrophytes in shallow waters,
and to plankton blooms and lavish growth of epiphytes in larger deeper waters (Roelofs, 1983).
The Northern Ireland Lake Survey quantied macrophyte species environmental relationships from
over 500 lakes using Generalised Additive Models (GAM) and Canonical Correspondence Analysis
(CCA; Heegaard et al., 2001). Nutrient concentrations appeared inuential in explaining species
distribution, but were highly correlated with alkalinity/pH and altitude. A DCA ordination diagram
with regressions of different physico-chemical environmental variables was used to classify vegetation
types from an extensive set of 600 surface waters in The Netherlands (Smolders et al., 2002).
The relationship between macrophytes and acidity in lakes has received relatively little attention
in the scientic literature despite macrophyte assemblages showing clear changes over a broad
gradient. These changes can often be related to acidity and other parameters, including nutrient
availability. However, there have been few attempts to develop relevant classications for these
two pressures acting together. For a better understanding of the successional differences within
low alkalinity macrophyte assemblages further research and cooperation is needed, in which
vegetation, physico-chemical and atmospheric deposition data are integrated (Arts, 2002).
Figure 5.7 Index of Acid Sensitivity (IAS) as
function of pH-gradient. Data based on attached
algae collected from rivers in Norway (Lindstrm
et al., 2004). pH
4 4.5 5 5.5 6 6.5 7 7.5
n = 365
I
A
S
14
12
10
8
6
4
2
0
5.5 Benthic invertebrates
Indices and metrics
Ecological assessment indices based on macroinvertebrates have been developed during the last
2-3 decades in several countries where acidication of rivers and lakes has been recognised as a
signicant problem. Acidication indices in Europe are primarily based on the presence or absence
of indicator species. However, many of these indicator species are tolerant species commonly
observed both in acidied and not acidied water.
Within the EU project AQEM
2
the acidication assessment tools used by the different countries
participating in the project were:
France: A French biotic index of acidication (B.I.A) has been proposed to assess the acidication
in streams of Les Vosges Mountains. The method uses a multimetric approach and the index has
10 classes that vary from 0 (unaffected) to 9 (strongly acidied; Guerold et al. 1995; 1999).
Germany: German acidication index (Braukmann, 1992).
Sweden: Medins acidication index, a multimetric index with 5 classes based on
macroinvertebrate data from humic rich rivers in Sweden (Henrikson and Medin, 1986).
Norway: Raddums index, based on data from clear, low conductivity rivers in Norway.
Each taxon is given a score corresponding to a pH interval and to an acidication class in the
index system (Raddum and Fjellheim, 1984; Lien et al., 1996). The Raddum index has also
been used on a European scale (Raddum and Skjelkvle, 1995; Raddum, 1999) and is being
further developed (Bkken and Kjellberg, 2004).
The aforementioned indices should be harmonised to derive an index suitable for use at a
European scale.
Impacts of acid water
The concepts of critical load and critical limit of acidifying compounds have been crucial to the
study of biological effects (Nilsson, 1986). The critical limit is usually dened in relation to the
effect acidication has on the most sensitive species. This implies that the limit may vary across
Europe depending on the species composition and their local adaptations. The critical limit for
ANC
cb
is often placed within the range 20 to 50 eq/l. The French B.I.A index observed
impoverished macro-invertebrate communities at ANC values between 50 and 100 eq/l.
However, severe loss of taxa was only observed below mean values of 50 eq/l (Guerold et al.,
1999). Some of this variability in ANC limits is related to the presence of organic matter.
Turnbull et al. (1995) in N-E Scotland concluded that even though critical loads provide a good
predictor of biotic status, it is not as sensitive a parameter as pH or alkalinity. Many acidication
indices are related to pH. For the Nordic indices, the lower limit for the class of intolerant species
is pH 5.5. In the French IBA index, unaffected streams were found only where mean pH values
were above 6.5. The most signicant change in the IBA index values occurred between pH 6.5
2
http://www.aqem.de.
and 5.5, a pH interval that has not received much attention in the Nordic indices (Guerold et al.,
1999). While these various indices give good predictions on a local scale, they are inadequate to
predict subtle, but ecologically important changes on a regional scale.
Humic substances
In general, data from the Nordic countries indicate a higher tolerance to low pH at high levels of
TOC (total organic carbon). It is generally assumed that humic substances detoxify the water
making metals and in particular Al less bio-available. It has been shown experimentally that humic
substances may reduce the toxic effect of low pH on the mayy Baetis rhodani and the crustacean
Gammarus lacustris (Bkken and Aanes, 1990). Using pH as the indicator avoids some of the
difculties in interpretation as discussed for the relationship between ANC
cb
, pH and LAl.
Natural acid water
Parts of Europe, in particular the northern countries, have regions with naturally acid water due to
organic acids. These may be wrongly regarded as temporarily or permanently acidied by
anthropogenic sources and be incorrectly proposed for remediation by liming (Warfvinge et al.,
1995; Kortelainen and Suakkonen, 1995). Dangles et al. (2004) found that the taxonomic richness
of macroinvertebrates as well as the breakdown rate of leaf litter in naturally acid streams was not
signicantly different from circum-neutral streams. Basing an index on pH will not separate
anthropogenically impoverished sites from sites having low species richness due to natural causes.
General toxicity of aluminium and heavy metals
It is difcult to isolate the effects of low pH from associated chemical changes. Labile Al has been
shown to be toxic to a number of macroinvertebrates and is certainly the most important toxic
metal in acidied water. However, it has proven difcult to give general guidelines for Al
(Herrmann, 2001). The concentration of Al varies substantially over a year and biological effects
depend on the timing, duration and intensity of an episode rather than on average concentrations.
Biological effects also depend on the differences in sensitivity between species, strain and life
stage. Water chemistry sampled at the time of a survey may not be representative of the water
quality episode that had the biological effect. Furthermore, some of the variability will be related
to biotic interactions, where the presence of sh and/or invertebrate competitors and predators
can have a gross impact on community assemblage. Although many studies have identied
impoverished communities when LAl is in the range of 0.1 to 0.3 mg/l, many exceptions exist.
There are numerous experimental studies investigating the effects Al and other metals have on
macroinvertebrate health (reviewed by Mance, 1987; Gensemer and Playle, 1999). In general
these studies show responses related to metal concentrations, but tolerance will vary between
species and life stages (Figure 5.8). As Al has an effect experimentally, it is most likely that Al also
has an effect within individual water bodies, despite this being difcult to prove. As there is a
close relationship between pH and LAl, LAl cannot be disregarded as the cause for population
responses in anthropogenically acidied lakes.
5.6 Fish
Fish are generally more sensitive to Al than to the pH reduction per se. For example, while Atlantic
salmon are unaffected by H
+
concentration down to pH values lower than 5.4 (Lacroix and
Townsend, 1987; Fivelstad et al., 2004), in the presence of bio-available Al the species is extinct
in rivers having average pH values lower than 5.5 while its abundance is reduced when the pH is
lower than 5.9. Many of the currently acidied rivers have a background (historic) pH level that
today would indicate toxic conditions. Based on such observations, it is concluded that pH alone
does not give a satisfactory prediction of sh population status (Kroglund et al., 2002). Therefore,
understanding the interaction between pH and Al, but also between Al and sh and Al and other
water quality constituents is fundamental to the understanding of the mechanisms controlling
population changes.
Sensitivity to acidication varies between species, life stages and strains. Where Atlantic salmon
(Salmo salar), brown trout (Salmo trutta), roach (Rutilus rutilus) and Arctic char (Salvelinus
alpinus) are more sensitive than perch (Perca uviatilis). While acid-sensitive sh species have
declined the most, even populations of more acid-tolerant species are affected in some regions of
Europe (Rask and Tuunainen, 1990; Rask et al., 1995). Within Europe, the number of populations
affected runs into thousands. In addition, an even higher number of populations have suffered
reduced viability, i.e. show a non-healthy status. Compared to the number of lakes that have lost
their sh populations, sh kills are rare. In most lakes and rivers, the effects of acidication are
Figure 5.8 The range of observed adverse effect
concentrations for freshwater invertebrates (main
taxa) from laboratory tests. A: annelids, B: bivalves,
C: crustaceans, G: gastropods, I: insects R: rotifers;
(after Mance, 1987). Concentration mg/L
As
Cd
Cu
Pb
Hg
Ni
Zn
0.001 0.01 0.1 1 10 100
only observed as changes in sh species composition and/or abundance (Rask and Tuunainen,
1990; Hesthagen et al., 1999; Tammi et al., 2003). The fact that sh kills are rare events, despite
populations being lost, suggests that density reductions occurs at life stages where kills are not
easily observed (e.g. egg and fry) and that density reduction is a gradual event (e.g. more related
to sh health and susceptibility to diseases, predation etc) rather than an acute event. Fish kills
have been observed under extreme conditions and examples of total sh loss have been recorded
in Norway, Scotland and Wales (Hindar et al., 1994). Fish kills (Salmo salar) were observed in
several Norwegian rivers in the years between 1910 and 1920, kills that were most likely due to
acidication and sea salt episodes (Hindar et al., 1994; 2004). Despite massive sh kills, it took
more than 40 years for the various affected populations to go extinct. This illustrates that linking
dose to response is not necessarily a straightforward process.
The alteration or loss of sh populations is not caused by food shortage. Bottom-up relationships
are therefore only of minor importance in acidied systems (Rosseland and Staurnes, 1994;
Gensemer and Playle, 1999). Fish population health is mainly related to Al, where the activity of
H
+
will enhance toxicity. Since Al is not vital to any aquatic biota, the species have not evolved
any defense mechanism specic for this metal. Al exercises its toxic properties by being accumulated
onto the sh gill affecting both tissue function and properties (Rosseland et al., 1994; Gensemer
and Playle, 1999). The accumulation is directly related to the concentration of in situ fractionated
Al (Gensemer and Playle, 1999). The relationships are poor when based on water samples where
the dynamic nature of Al has been stabilised through ageing (Kroglund et al., 2001). Al is
accumulated onto the mucus layer surrounding the gill, onto cell walls and is over time transported
inside the cell. The toxic role and effect depends on the concentration, exposure duration and the
location where the Al is accumulated. As the speciation and hence toxicity of Al is dynamic, Al
does not t directly into biotic ligand models as these assume that the metals are present in a
free, and stable form (Chapman et al., 2003). Al characteristics are more dynamic in rivers than
in lakes, and more at pH levels around 5.6 to 6.2 than at pH levels lower than 5.2. The probability
for errors is as such related to pH and to water retention time.
The most exhaustive studies have been performed on salmonids. A water quality that has no
appreciable effect on Atlantic salmon fry and smolt production can still have a large effect on the
seawater survival of the postsmolt. Concentrations of LAl that are below the analytical certainty
can impair enzyme activities essential for hypoosmoregulation (Kroglund and Finstad, 2003). Due
to this analytical uncertainty, salmon health is best related to Al accumulated onto the gills. The
present day reduction in Atlantic salmon catches across Europe and the North Atlantic could
therefore be due to acidication related causes in addition to other pressures.
5.7 Summary
Acidication is caused by atmospheric deposition of nitrogen and sulphur compounds which
reduces the pH in water bodies with low alkalinity. Reducing the pH has an impact on the chemical
water quality through increasing concentrations of toxic Al and heavy metal species as well as
changing the CO
2
-bicarbonate equilibrium. A high nitrogen deposition can alter the nutrient
balance of freshwaters (e.g. N:P ratio).
The biological impacts have been demonstrated for sh, invertebrates and plants in rivers and
lakes. The impact increases when pH is lowered below 6.5. The specic impacts also depend on
the concentration of aluminium and total organic carbon (TOC or humus). Naturally acidic water
bodies generally have a more diverse ora and fauna compared to anthropogenically acidied
water bodies with the same pH level.
There are still several major knowledge gaps that need to be lled for successful implementation of
the WFD. Further work is needed on the relationship between atmospheric loading and the resulting
chemical water quality in different types of catchments. It is necessary to develop quantitative metrics
for biotic elements in water bodies susceptible to acidication, and to describe ecosystem effects of
reductions in acid precipitation. Chemical and biological indices must be harmonised. Models currently
in use are awed with respect to their ability to identify dose-response relationships.
5.8 References
Almer, B., W. Dickson, C. Ekstrm, E. Hrnstrm and U. Miller. 1974. Effects of acidication of Swedish lakes.
Ambio 1:30-36
Alewell, C., 2001. Predicting reversibility of acidication: The European sulphur story. Water, Air and Soil
Pollution 130:1271-1276
Arts, G.H.P. 2002. Deterioration of atlantic soft water macrophyte communities by acidication, eutrophication
and alkalinisation. Aquatic Botany 73:373-393
Braukmann, U. 1992. Biological indication of stream acidity in Baden-Wrttemberg by the use of
macroinvertebrates. In: Bhmer, J. and H. Rahmann (Eds.), Bioindikationsverfahren zur Gewsserversauerung.
Karlsruhe (Landesanstalt fr Umweltschutz): 58-71
Brouwer, E., R. Bobbink, J.G.M. Roelofs. 2002. Restoration of aquatic macrophyte vegetation in acidied and
eutrophied softwater lakes: an overview. Aquatic Botany 73: 405-431
Bulger, A.J., L. Lien, B.J. Cosby, and A. Henriksen. 1993. Brown trout (Salmo trutta) status and chemistry from
the Norwegian thousand lake survey: statistical analysis. Canadian Journal of Fisheries and Aquatic Sciences
50:575-585
Bkken,T. and K.J. Aanes. 1990. Bruk av vassdragets bunnfauna i vannkvalitetsklassiseringen. Nr.2.
Forsuring. Report 2491, Norwegian Institute for Water Research, Oslo
Bkken, T. and G. Kjellberg. 2004. Klassisering av surhetsgrad og vurdering av forsuring i rennende vann
basert p forekomst av makrobunndyr. Klassiseringssystem tilpasset humusrike elver og bekker i
stlandsomrdet Report 4923-2004, Norwegian Institute for Water Research, Oslo
Chapman, P.M., F.Y. Wang, C.R. Janssen, R.R. Goulet and C.N. Kamunde. 2003. Conducting ecological risk
assessments of inorganic metals and metalloids: current status. Human and Ecological Risk Assessment
9:641-697
Dangles, O., B. Malmquist and H. Laudon. 2004. Naturally acid freshwater ecosystems are diverse and
functional: evidence from boreal streams. Oikos 104:149-155
Farmer, A.M. 1990. The effects of lake acidiaction on aquatic macrophytes- a review. Environmental
Pollution 65:219-240
Fivelstad, S., A.B. Olsen, S. Stefansson, S. Handeland, R. Waagb, F. Kroglund and J. Colt. 2004. Lack of long-
term sublethal effects of reduced freshwater pH alone on Atlantic salmon (Salmo salar) smolts subsequently
transferred to seawater. Canadian Journal of Fisheries and Aquatic Sciences 61:511518
Gensemer, R.W. and R.C. Playle. 1999. The bioavailability and toxicity of aluminum in aquatic environments.
Critical Reviews in Environmental Science and Technology 29:315450
Guerold, F., L. Giamberini, J.-L. Tourmann, J.-C. Pihan and R. Kaufmann. 1995. Occurrence of aluminium in
chloride cells of Perla marginata (Plecoptera) after exposure to low pH and elevated aluminum concentration.
Bullettin of Environmental Contaminants and Toxicology 54:620-625
Guerold, F.A., N.F. Masson and O.J. Dangles. 1999. Acidication in the Vosges Mountains (North-eastern
France): assessment of biological status by the use of a multimetric approach. In: Raddum G.G., B.O.
Rosseland, and J. Bowman (Eds.), Workshop on biological assessment and monitoring; evaluation and
models. Report 4091-99, Norwegian Institute for Water Research, Oslo
Heegaard, E., H.H. Birks, C.E. Gibson, S.J. Smith and S.A. Wolfe-Murphy. 2001. Species environmental
relationships of aquatic macrophytes in Northern Ireland. Aquatic Botany 70:175-223
Henrikson, L. and M. Medin. 1986. Biologisk bedmning av frsuringspverkan p Lengens tilden och
grundomrden 1986 Aquaekologerna, Rapport til lnsstyrelsen i lvborgs ln
Herrmann, J., E. Degerman, A. Gerhardt, C. Johansson, P-E. Lingdell and I.P. Muniz. 1993. Acid-stress effects
on stream biology. Ambio 22:298-307
Herrmann, J. 2001. Aluminium is harmful to benthic invertebrates in acidied waters, but at what threshold(s)?
Water, Air and Soil Pollution. 130:837-842
Hesthagen, T., I.H. Sevaldrud and H.M. Berger. 1999. Assessment of damage to sh population in Norwegian
lakes due to acidication. Ambio 28:112-117
Hindar, A., A. Henriksen, K. Torseth and A. Semb. 1994. Acid water and sh death. Nature 372:327-328
Hindar, A., K. Torseth, A. Henriksen and Y. Orsolini. 2004. The signicance of the North Atlantic Oscillation
(NAO) for sea-salt episodes and acidication-related effects in Norwegian rivers. Environmental Science and
Technology 38:26-33
Hornung, M. 1997. The role of nitrates in the eutrophication and acidication of surface waters. In: Wilson
W.S., A.S. Ball and R.H. Hinton (Eds.), Managing risks of nitrates to humans and the environment. Cambridge
Royal Society of Chemistry, Colchester, p. 157-174
Hyne, R.V. and S.P. Wilson. 1997. Toxicity of acid-sulphate soil leachate and aluminium to the embryos and
larvae of Australian bass (Macquaria novemaculeata) in estuarine water. Environmental Pollution 97:221-227
Juggins, S., Ormerod, S. J., and Harriman, R. 1995. Relating critical loads to aquatic biota., p. 9-13, In:
Battarbee, R. W., Critical loads of deposition for UK freshwaters., Institute of Terrestrial Ecology, Edinburgh
Korman, J., D.R. Marmorek, G.L. Lacroix, P.G. Amiro, J.A. Ritter, W.D. Watt, R.E. Cutting and D.C.E. Robinson.
1994. Development and evaluation of a biological model to assess regional-scale effects of acidication on
Atlantic salmon (Salmo salar). Canadian Journal of Fisheries and Aquatic Sciences 51:662-680
Kortelainen, P. and S. Saukkonen. 1995. Organic vs minerogenic acidity in headwater streams in Finland.
Water Air and Soil Pollution 85:559-564
Kroglund, F., H.C. Teien, B.O. Rosseland and B. Salbu. 2001. Time and pH-dependent detoxication of
aluminum in mixing zones between acid and non-acid river. Water, Air and Soil Pollution 130:905910
Kroglund, F., R. Wright and C. Burchart. 2002. Acidication and Atlantic salmon; critical limits for Norwegian
rivers. Report 4501, Norwegian Institute for Water Research, Oslo. pp. 61
Kroglund, F. and B. Finstad. 2003. Low concentrations of inorganic monomeric aluminium impair physiological
status and marine survival of Atlantic salmon. Aquaculture 222:119-133
Lacroix, G.L., and D.R. Townsend. 1987. Responses of juvenile Atlantic salmon (Salmo salar) to episodic
increases in acidity of Nova Scotia rivers. Canadian Journal of Fisheries and Aquatic Sciences 44:1475-1484
Lien, L., G.G. Raddum, A. Fjellheim and A. Henriksen. 1996. A critical limit for acid neutralizing capacity in
Norwegian surface waters, based on new analyses of sh and invertebrate responses. The Science of the
Total Environment 177:173-193
Lindstrm, E.-A., Brettum, P., Mjelde, M and Johansen, S.W. 2004. Freshwater vegetation in Norway. Tolerance
limits to acidication and effects of liming. Report 4821-2004, Norwegian Institute for Water Research,
Oslo. pp. 125
Lydersen, E., T. Larssen and E. Fjeld. 2004. The inuence of total organic carbon (TOC) on the relationship
between acid neutralizing capacity (ANC) and sh status in Norwegian lakes. The Science of the Total
Environment 362:63-69
Mance, G. 1987. Pollution threat of heavy metals in aquatic environments. Pollution Monitoring Series.
Elsevier Applied Science, London and New York. pp. 372
Nilsson, J. (ed.) 1986. Critical loads for Nitrogen and Sulphur. Nordic Council of Ministers Report 1986: II,
Nordic Council of Ministers Copenhagen Denmark
Nilsson, J. and Grennfelt, P. 1988. Critical loads for Sulphur and Nitrogen. Nordic Council of Ministers Report
1988:15, Nordic Council of Ministers, Copenhagen, Denmark, pp. 418
Posch, M., J. Kamari, M. Forsius, A. Henriksen and A. Wilander. 1997. Exceedance of critical loads for lakes in
Finland, Norway, and Sweden: Reduction requirements for acidifying nitrogen and sulfur deposition.
Environmental Management 21:291-304
Raddum, G.G. 1999. Large scale monitoring of invertebrates: Aims, possibilities and acidication indexes. In:
Raddum G.G., B.O. Rosseland and J. Bowman (Eds.): Workshop on biological assessment and monitoring;
evaluation and models. Report 4091-99. Norwegian Institute for Water Research, Oslo
Raddum, G.G. and A. Fjellheim. 1984. Acidication and early warning organisms in freshwater in western
Norway. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte Limnologie
22:1973-1980
Raddum, G.G. and B.L. Skjelkvle. 1995. Critical limits of acidication to invertebrates in different regions of
Europe. Water Air and Soil Pollution 85:475-480
Rask, M. and P. Tuunainen 1990. Acid-induced changes in sh populations of small Finnish lakes. In: Kauppi
P., P. Anttila and K. Kenttmies (Eds.), Acidication in Finland, Springer-Verlag, Berlin, pp. 911-927
Rask, M., Mannio, J., Forsius, M., Posch, M. and Vuorinen, P. 1995. How many sh populations are affected
by acid precipitation? Environmental Biology of Fishes 42:51-63
Redeld, A.C., 1958. The biological control of chemical factors in the environment. American Scientist
46:205-222
Reuss, J.O. and D.W. Johnson. 1986. Acid deposition and the acidication of soils and waters. Ecological
Studies. Springer Verlag, New York pp. 119
Riis, T. and K. Sand-Jensen. 1998. Development of vegetation and environmental conditions in an oligotrophic
Danish lake over 40 years. Freshwater Biology 40:123-134
Roelofs, J.G.M. 1983. Impact of acidication and eutrophication on macrophyte communities in soft waters
in the Netherlands.1: eld observations. Aquatic Botany 17:139-155
Roelofs, J.G.M., E. Brouwer, R. Bobbink. 2002. Restoration of aquatic macrophyte vegetation in acidied and
eutrophicated shallow soft water wetlands in the Netherlands. Hydrobiologia 478:171-180
Rosseland, B.O. and M. Staurnes. 1994. Physiological mechanisms for toxic effects and resistance to acidic
water: an ecophysiological and ecotoxicological approach. In: Steinberg, E.W. and R.F. Wright (Eds.),
Acidication of freshwater ecosystems: implications for the future. John Wiley & Sons. pp. 228-246
Sand-Jensen, K., M. Sondergaard. 1979. Distribution and quantitative development of aquatic macrophytes
in relation to sediment characteristics in oligotrophic Lake Kalgaard, Denmark. Freshwater Biology 9:1-11
Skjelkvle, B.L. and R.F. Wright. 1991. Overview of areas sensitive to acidication in Europe. Oslo:Norwegian
Institute for Water Research. 24/1990, 1991
Skjelkvle, B.L., C. Evans, T. Larsen, A. Hindar and G.G. Raddum. 2003. Recovery from acidication in
European surface waters: a view to the future. Ambio 32:170-175
Smolders, A.J.P., E. Lucassen, J.G.M. Roelofs. 2002. The isoetid environment: biogeochemistry and threats.
Aquatic Botany 73:325350
Tammi, J., M. Appelberg, U. Beier, T. Hesthagen, A. Lappalainen and M. Rask. 2003. Fish status survey of
Nordic lakes: effects of acidication, eutrophication and stocking activity on present sh species composition.
Ambio 32:98-105
Turnbull, D., C. Soulsby, S. Langan, R. Owen, and D. Hirst. 1995. Macroinvertebrate status in relation to
critical loads for freshwater: a case study from NE Scotland. Water, Air and Soil Pollution 85: 2461-2466
Vestergaard, O. and K. Sand-Jensen. 2000. Aquatic macrophyte richness in Danish lakes in relation to
alkalinity, transparency and lake area. Canadian Journal of Fisheries and Aquatic Science 57:2022-2031
Warfvinge, P., S. Lofgren and U. Lundstrom.1995. Implications of natural acidication for mitigation strategies
in northern Sweden. Water, Air and Soil Pollution 85: 499-504
Wetzel, R.G. 1983. Limnology. Second edition. Saunders College Publishing, New York. pp. 767
Hydromorphological
pressures
1 - Niels Jepsen. Typical regulated lowland stream (Holme, Denmark)
2 - Anna-Stiina Heiskanen. Urban waterways (Canal Grande, Venezia, Italy)
3 - Gary Free. SubAlpine lake in Trentino (lago di Tenno, Northern Italy)
4 - Anna-Stiina Heiskanen. Nymphea in an articial pond, (Isola Bella, Lago Maggiore, Italy)
5 - Niels Jepsen. Fish sampling from an in-river trap (Bygholm River, Denmark)
1 2 3 4 5
6 Hydromorphological pressures in lakes
Seppo Hellsten, Bernard Dudley
6.1 Introduction
Hydromorphological pressures in lakes are related to the human need to control the water levels
of lakes for various reasons such as production of hydropower, ood prevention, recreation,
navigation, and supply of water for agricultural or human consumption. Regulation practices vary
and depend on the objectives of regulation. In a typical hydropower regulation project in the
northern hemisphere, water level during the summer period is increased, relative to the unregulated
regime, and during the winter period, when the demand for and the price of electricity is highest,
the water level is strongly lowered. Flood prevention regulation follows a similar pattern during
the winter, but in summer some storage capacity is left to allow for ash ood events. When the
major objective of the regulation is recreation or navigation, then regulated water levels are often
more stable than natural ones. If the water level is regulated for water supply use, the water level
uctuation is more irregular and depends on the specic use of raw water.
Although hydromorphological pressures are specic to particular water bodies and their regulation
regime, some generalizations can be made. Hydropower effects are typical in northern and high
altitude lakes, which usually are not impacted by other pressures, whereas pressures due to
regulation for navigation and recreation purposes often affect lowland lakes situated in or near
densely populated areas. Morphological alterations such as dams and weirs affect mostly
continuity of rivers situated downstream from the lake. On the other hand, ood protection
constructions and drainage of ood plains have created embankments, which can signicantly
change the morphology of lowland lakes. Generally, large scale morphological alterations are
more common in smaller lakes that are surrounded by agricultural and urban areas.
Hydromorphological alterations mostly affect the uppermost littoral zone, which is characterized
by the presence of aquatic macrophytes, although changes in retention time can also indirectly
change the trophic status of whole lake and pelagial communities too. This review focuses on the
relationship between macrophytes and hydromorphological pressures, which is the most commonly
known and investigated.
The Water Framework Directive (WFD) refers to the hydromorphological quality elements of lakes
as hydrological regime and morphological conditions. Hydrological regime is dened as the
quantity and dynamics of ow, level, residence time, and the resultant connection to groundwater.
Morphological conditions are dened as lake depth variation, quantity and structure of the
substrate, and both the structure and condition of the lake shore zone. Hydromorphological
quality elements are only used specically to dene high status in lakes, where these elements
should indicate totally, or nearly totally, undisturbed conditions. For good and moderate status
levels, hydromorphological elements must only be consistent with the achievement of the values
specied for the biological quality elements. Divisions between hydrological regime and
morphological conditions are difcult to make, as these are often interdependent. For example,
level (dened as hydrological) and depth variation (dened as morphological) are highly related.
Therefore hydrological and morphological elements are treated as one in this review.
HYDROMOROPHOLOGICAL PRESSURES IN LAKES | 135
6.2 Macrophytes
Macrophytes are one of the key indicators of hydromorphological changes in lakes. Because
macrophytes grow in the littoral zone they are sensitive for changes in water level uctuation
regime. A general zonation of macrophytes can be based on life-form distribution; helophytes
grow in the uppermost zone and isoetids, elodeids and charids occupy deeper areas of lakes. Even
small changes in the dynamics of water level uctuation can affect the distribution and the
elevation of zones. Morphological changes of the littoral zone, caused for example by dredging or
embankments, signicantly disturb the development of vegetation.
Annex V of the WFD outlines three macrophyte-related quality elements that need consideration
in assessing the ecological status of lakes. These are (i) taxonomic composition, (ii) average
macrophyte abundance, and (iii) an ecological assessment which is specic to lake type. This third
requirement, for which reference communities (communities found in non impacted conditions)
need to be identied, also allows for the use of different indices and pressure sensitive species in
different lake types.
Taxonomic composition
In general, taxonomic composition of macrophytes is not a sensitive indicator to describe the
changes in hydromorphology of lakes. However, in many studies it has been shown that macrophyte
diversity is lower in lakes with uctuating water level in North Europe (Rrslett, 1991; Hellsten,
2001) and in Canada (Hill et al., 1998). However, the relationship between water level uctuation
and macrophyte diversity is not simple. An extensive literature survey of Scandinavian lakes
showed that general macrophyte richness correlated mainly with draw-down of water level, but a
regulation amplitude of between 1 and 3 meters supported the highest biological diversity
(Rrslett, 1991). A slight increase in disturbance could even create suitable habitats for European
aquatic macrophytes as noted by Murphy et al. (1990). A similar phenomenon was found in lakes
regulated for hydropower in New Zealand, where an increase in the range of monthly water level
uctuation appears to have increased biodiversity (Riis and Hawes, 2002).
Depth variations are usually related to an articial increase or decrease of water level. Water levels
are increased to extend storage capacity of reservoirs or regulated lakes. A sudden increase of
water level will initiate erosion processes, which lower biodiversity (Nilsson, 1981; Hellsten et al.,
1996). It should be noted that taxonomic composition is a poor indicator of water level increase,
because most of the species are still present after water level increase, although abundance may
differ signicantly (Nilsson and Keddy, 1988; Hellsten et al., 1996). Effects of raised water level
also depend on ageing; after inundation shock of Swedish reservoirs species diversity was highest
30-40 years subsequent to the initiation of the regulation (Nilsson et al., 1997). In most cases
diversity is slightly increased after inundation due to stabilization of the shoreline.
In general, lowering of water level will lead to increased diversity, as found in several studies
(Lohammar, 1949; Toivonen and Nybom, 1989; Rrslett, 1991). The main reason for increased
diversity is that a newly exposed littoral zone or increased shallowness allows the sublittoral zone to
cover the entire water body. Extensive lowering, which affects resuspension in the water body by
allowing waves and currents to reach bottom sediments, can drastically change lake ora, as shown
in several studies in the Netherlands (Scheffer, 1998). Several shallow lake studies have demonstrated
a sensitive balance between different species groups (Best, 1987; Van den Berg et al., 1998).
Abundance
Several studies indicate that abundance is a much more sensitive indicator for hydrological
change than species composition (Nilsson and Keddy, 1988; Coops and van der Velde, 1999;
Hellsten et al., 1996; Hellsten, 2001). Generally, water level uctuation affects zonation patterns,
which are a function of the relative abundances of different species with different degrees of
adaptation to stress caused by depth and drying (Pearsall, 1920). Therefore, changes in the
amount of water level uctuation are reected by changes in the distribution of species (Toivonen
and Lappalainen, 1980).
In addition to the range of water level uctuation, the dynamics of the uctuation signicantly
affects the abundance of macrophytes. For example, the timing and range of the spring ood
clearly affects the zonation of sedge species in northern areas (Walker and Wehrhahn, 1970;
Sjberg and Danell, 1983; Hellsten, 2001). The observed increase of common reed (Phragmites
australis) abundance in Scandinavia may be related to lowered early spring water level (Rintanen,
1996; Partanen and Hellsten, 2005). Reeds also benet from stabilized water levels and growth
periods (Coops and van der Velde, 1995; 1999). A general decline of reed beds in Mid-Europe
seems, however, to be related to changes in sediment due to eutrophication (Weisner, 1991;
Clevering, 1999).
Lowering of the water level while a lake is ice-covered will have signicant effects, especially on
large sized isoetids such as Isoetes lacustris and Lobelia dortmanna. Areas where their decline has
been reported include northern Scandinavia (Quennerstedt, 1958; Rrslett, 1984; Rintanen,
1996; Hellsten, 2001) and Scotland (Smith et al., 1987; Murphy et al., 1990). Additional to the
effect of freezing, changes in sediment quality will also signicantly affect their distribution
(Murphy, 2002).
Very recently, as a direct response to the WFD, and because of the prior lack of suitable classication
schemes, a survey methodology called the Lake Habitat Survey (LHS) was developed in the UK
that directly assesses the hydromorphological elements specied in the WFD. The method is based
on the EMAP guideline and its eld survey method (FOML) developed in the USA. In the LHS
scheme, information on physical habitat (including substrate and riparian vegetation type, but
not species composition) is recorded at sampling plots around the shore, and pressures such as
angling, erosion and grazing are recorded over the entire lake. Data on hydrological regime are
also obtained whenever possible. Apart from the nal report (Rowan et al., 2004), there are no
published reviews of this methodology, however this scheme has the support of the UK authorities
responsible for implementing the WFD and it is intended that it will be adopted throughout the
UK (J. Rowan, pers. comm.). It may also prove to be of use across the rest of Europe, although the
relationship between the LHS score and biological status of lake is not very clear.
Apart from this new development, there are few classication schemes related to relationships
between hydro-morphological factors and macrophytes. The direct response of Isoetes lacustris to
ice penetration enables its distribution to be used for classication purposes (Rrslett, 1989;
Rrslett and Johansen, 1996; Hellsten, 2001). The deepest growing areas of I. lacustris are also
sharply limited by lack of light and therefore their growing niche is easy to predict (Rrslett,
1988). The distribution of other large isoetids such as Isoetes echinospora, Lobelia dortmanna and
Littorella uniora can also be used for classication purposes, because they are all responsive to
ice erosion and changes in sediment structure (Rrslett, 1989; Murphy, 2002).
A classication based on strategy analysis and the division of species into stress-tolerating, ruderal
and competitive categories has been effective in classifying regulated lakes in Norway (Rrslett,
1989) and in other areas (Murphy et al., 1990). Ruderal species, with high resistance to disturbance,
were typical in shallow water communities of regulated lakes, whereas stress-tolerating species
prevailed in deeper areas.
The effects of depth changes have been generally estimated using simple calculation procedures
to describe the available growth area for macrophytes. Known relationships between the deepest
growth limits of bottom-rooted helophytes have produced a large number of different applications
for Finnish lakes (Hellsten, 2001). Hudon (1997) developed similar relationships between average
water level scenarios and areas dominated by different vegetation types in oodplain lakes of the
St Lawrence River. The vegetated area of the littoral zone has also been used as an index to
describe the trophic status of lakes (Leka et al., 2003).
Models
The relationship between hydromorphological factors and species is usually modelled through
case-dependent correlation models. For example, the effects of water level uctuation depend
essentially on the use of the lake and climate conditions. Rrslett (1989) showed in his review an
analysis of 17 Norwegian regulated lakes in which the species richness (S) followed the equation:
S = 16.4 1.34 W 0.013 H + 0.085 A,
where W = mean annual range of water level (m), H = lake altitude (m a.s.l.) and A = lake area
(km
2
).
Further in his analysis of 641 lakes from Norway, Sweden, Denmark and Finland, he found that
lake area was the best predictor of species diversity (Rrslett, 1991). A stepwise prediction model
also included hydromorphology related variables such as water level range and lake lowering with
water conductivity and lake elevation values. A predictive model based on discriminant analysis
and vegetation types with environmental background information was developed by Hellsten
(2001). Variables in the stepwise selection were water level duration of summer, duration of
frozen and non-frozen ice pressure zone, bottom substrate and relative erosion rate. The model
produced good results in regulated Lake Ontojrvi, where eight types out of twelve were classied
correctly. In non-regulated Lake Lentua the results were even better, where ten out of fourteen
vegetation types were classied correctly. In this model, variables were more complicated and
more clearly linked to increasing depth.
Heegaard et al. (2001) also found relatively promising results by using generalised additive model
(GAM) techniques in lakes in Northern Ireland. Although the models were used in evaluating the
effects of water chemistry on aquatic macrophytes, these predictive models could also be used in
lakes where the water level is uctuating.
6.3 Summary
The littoral zone is a highly varying environment and therefore the relationships between
hydromorphological factors and aquatic macrophytes are often site-specic. However, the
relationship between lowering of ice levels and the destruction of large isoetids is relatively clear
and easy to predict in regulated hydroelectric lakes. Water level uctuation range during the summer
denes the extension of the helophyte zone which can be used as a major determinant of change.
Some of the WFD variables, such as species richness, depend on both water quality and sediment
properties, therefore in nutrient-rich waters it is very difcult to nd relationships between
macrophytes and hydromorphological factors. Nutrient enrichment can also compensate for
degradation caused by uctuating water level. In general, the relationship between
hydromorphological change and aquatic macrophytes is clearer in oligotrophic than eutrophic
lowland lakes.
6.4 References
Best, E.P.H. 1987. The submerged macrophytes in lake Maarsseveen 1: Changes in species composition and
biomass over a six year period. Hydrobiological Bulletin 21:55-60
Clevering, O. 1999. The effects of litter on growth and plasticity of Phragmites australis clones originating
from infertile, fertile or eutrophicated habitats. Aquatic botany 64:35-50
Coops, H., and G. van der Velde. 1995. Seed dispersal, germination, and seedling growth of six helophyte
species in relation to water-level zonation. Freshwater Biology 34:13-20
Coops, H.G.N., and G. van der Velde. 1999. Helophyte zonation in two regulated estuarine areas in the
Netherlands: vegetation analysis and relationships with hydrodynamic factors. Estuaries 22:657-668
Heegaard, E., H. Birks, C. Gibson, S. Smith, and S. Wolfe-Murphy. 2001. Species-environmental relationships
of aquatic macrophytes in Northern Ireland. Aquatic Botany 175-223
Hellsten, S., Marttunen, M., Palomki, R., Riihimki, J., and E. Alasaarela. 1996. Towards an ecologically-based
regulation practice in Finnish hydroelectric lakes - Regulated Rivers: Research & Management 12:535-545
Hellsten, S. 2001. Effects of lake water level regulation on aquatic macrophyte stands in northern Finland
and options to predict these impacts under varying conditions. Acta Botanica Fennica 171:1-47
Hill, N.M., P.A. Keddy, and I.C. Wisheu. 1998. A Hydrological Model for Predicting the Effects of Dams on the
Shoreline Vegetation of Lakes and Reservoirs. Environmental Management 22:723-736
Hudon, C. 1997. Impact of water level uctuations on St. Lawrence River aquatic vegetation. Canadian
Journal of Fisheries and Aquatic Sciences 54:2853-2865
Leka, J., K. Valta-Hulkkonen, A. Kanninen, S. Partanen, S. Hellsten, A. Ustinov, R. Ilvonen, and O. Airaksinen.
2003. Use of aquatic macrophytes for assessing and monitoring the ecological status of lakes Evaluation
of eld survey and aerial photography methods used in the Life Vuoksi Project. Regional Environmental
Publications - Alueelliset ympristjulkaisut (In Finnish) 312
Lohammar, G. 1949. Uber die Vrenderungen der Naturverhltnisse gesenkter Seen. Verhandlungen der
Internationalen Vereinigung fr Theoretische und Angewandte Limnologie 10:266-274
Murphy, K.J., B. Rorslett, and I. Springuel. 1990. Strategy analysis of submerged lake macrophyte communities:
an International example. Aquatic Botany 36:303-323
Murphy, K.J. 2002. Plant communities and plant diversity in softwater lakes of northern Europe. Aquatic
Botany 73:287-324
Nilsson, C. 1981. Dynamics of the shore vegetation of a North Swedish hydro-electric reservoir during a 5-year
period. Acta Phytogeographica Suecica 69
Nilsson, C., and P.A. Keddy. 1988. Predictibility of change in shoreline vegetation in a hydroelectric reservoir,
northern Sweden. Canadian Journal of Fisheries and Aquatic Sciences 45:1896-1904
Nilsson, C., R. Jansson, and U. Zinko. 1997. Long-Term responses of river-margin vegetation to water-level
regulation. Science 798-800
Partanen, S., and S. Hellsten. 2005. Changes of emergent aquatic macrophyte cover in seven large boreal
lakes in Finland with special reference to water level regulation. Fennia 183:57-79
Pearsall, W.H. 1920. The aquatic vegetation of the English lakes. Journal of Ecology 8:164-201
Quennerstedt, N. 1958. Effect of water level uctuation on lake vegetation. (Verhandlungen der
Internationalen Vereinigung fr Theoretische und Angewandte Limnologie 13:901-906
Riis, T., and I. Hawes. 2002. Relationships between water level uctuations and vegetation diversity in
shallow water of New Zealand lakes. Aquatic Botany 74:133-148
Rintanen, T. 1996. Changes in the ora and vegetation of 113 Finnish lakes during 40 years. Annales
Botanici Fennici 33:101-122
Rrslett, B. 1984. Environmental factors and aquatic macrophyte response in regulated lakes - a statistical
approach. Aquatic botany 19:199-220
Rrslett, B. 1988. Niche extension of aquatic macrophytes in hydrolakes: Predictive assessments of
environmental impacts. Internationale Revue der Gesamten Hydrobiologie 73:129-143
Rrslett, B. 1989. An integrated approach to hydropower impact assesment. II. Submerged macrophytes in
some Norwegian hydro-electric lakes. Hydrobiologia 175:65-82
Rrslett, B. 1991. Principal determinants of aquatic macrophyte richness in northern European lakes. Aquatic
Botany 39:173-193
Rrslett, B., and S.W. Johansen. 1996. Remedial measures connected with aquatic macrophytes in Norwegian
regulated rivers and reservoirs. Regulated rivers: Research and Management 12:509-522
Rowan, J.S., R.W. Duck, J. Carwardine, O.M. Bragg, A.R. Black, and M.E.J. Cutler. 2004. Development of a
technique for Lake Habitat Survey (LHS): Phase I. Final report of SNIFFER (Scotland and Northern Ireland
Forum for Environmental Research) project WFD40. Available at www.sniffer.org.uk
Scheffer, M. 1998. Ecology of Shallow Lakes. Chapman & Hall
Sjberg, K., and K. Danell. 1983. Effects of permanent ooding on Carex-Equisetum wetlands in northern
Sweden. Aquatic Botany 15:275-286
Smith, B.D., P.S. Maitland, and S.M. Pennock. 1987. A comparative study of water level regimes and littoral
benthic communities in Scottish lakes. Biological Conservation 39:291-316
Toivonen, H., and T. Lappalainen. 1980. Ecology and production of aquatic macrophytes in the oligotrophic
lake Suomunjrvi, eastern Finland. Annales Botanici Fennici 17:69-85
Toivonen, H., and C. Nybom. 1989. Aquatic vegetation and its recent succession in the waterfowl wetland
Koijrvi, South Finland. Annales Botanici Fennici 26:1 - 14
Walker, B.H., and C.F. Wehrhahn. 1970. Relationships between derived vegetation gradients and measured
environmental variables in Saskatchewan wetlands. Ecology 52:85-95
Van den Berg, M.S., H. Coops, J. Simons, and A. de Keizer. 1998. Competition between Chara aspera and
Potamogeton pectinatus as a function of temperature and light. Aquatic Botany 60:241-250
Weisner, S. 1991. Within-lake patterns in depth penetration of emergent vegetation. Freshwater Biology
26:133-142
7 Hydromorphological pressures in rivers
Bente Clausen, Michael Dunbar, Nikolai Friberg and Annette Baattrup-Pedersen
7.1 Introduction
With the word hydromorphological pressures we understand all changes caused by human
inuences to either the ow regime (hydrology) or the morphology of the stream that affect the
biota. The most important hydromorphological pressures are:
building of dams or weirs for hydropower, water supply or other purposes
canalization and/or dredging of rivers or streams to improve drainage or for navigation
weed cutting to improve drainage
abstraction of water directly from the stream or from groundwater for water supply or
irrigation, or diversion (e.g. for hydropower or irrigation).
Other inuences that are not described in detail here include urbanisation, afforestation/
deforestation, draining of wetlands (tiling), transport and supply of water from outside the river
basin to increase river discharge at dry periods, and high discharges of water treatment plants in
small river basins.
7.2 Dams and weirs
The term impoundment is often used to describe any structure that alters river water levels.
Larger height structures (often termed dams) are built to provide a hydraulic head to generate
electricity (hydropower) and to store water for irrigation and water supply. Smaller height structures
or weirs are built to maintain water levels in low ow periods for agriculture, shipping and
recreation and to divert water for supply, hydropower or irrigation. Depending on the height of the
impoundment, the upstream physical environment will change, increasing depths and retention
times and decreasing velocities, with major effects on the ecosystem. Impoundments can also
disrupt the connectivity of the river, alter the ow regime and lower the sediment transport
downstream. Flushing ows may be used to clean out ne sediment from the reservoir and
downstream river bed, and can also re-structure the bed generally.
Usually river sections downstream of dams have lower and less frequent oods than the natural
ow regime (Growns and Growns, 2001), although more extreme oods are often unaffected. The
resulting less variable ow regime and lower sediment transport often leads to higher water
temperatures and clarity. Large dams generally lower the temperature downstream, when cooler
water is drawn from the hypolimnion of the impounded reservoir. Another typical feature of rivers
downstream of hydropower stations are daily and weekly ow uctuations depending on
hydropower pricing. The physical changes also lead to changes in water chemistry.
Plants and animals are affected by the physical and chemical changes. The higher clarity and
reduced variability of ow usually leads to a higher abundance of periphyton and higher plants
(Dessaix and Fruget, 1995; Biggs et al., 1998). In nutrient-rich rivers the occurrence of phytoplankton
HYDROMOROPHOLOGICAL PRESSURES IN RIVERS | 141
is usually increased downstream of the dam as a result of phytoplankton production in the
reservoir. Benthos communities below dams often show a reduction in species richness, while
some species increase in abundance (Fruget, 1991). For example, invertebrate lter feeders
(e.g. Hydropsychidae, Simulidae) often increase in numbers downstream of reservoirs. Migratory
sh (e.g., salmon, trout and eel) are especially affected by the breaking of the connectivity of the
river. The spreading of other animal groups and plants can also be hampered because of the
reservoir (Allan, 1995).
Methods for evaluating the ecological consequences of damming
The ecological consequences of a damming of a river are generally large because the river
ecosystem is changed into a lake ecosystem where the reservoir is established. Descriptions of the
ecological impacts therefore include both a comparison of the reservoir ecosystem with the
original river ecosystem and a description of the ecological consequences of the created river
discontinuity. The ecological consequences of weirs are usually smaller and can often be assessed
by the same methods as used for assessment of other changes in river morphology.
7.3 Changes in river morphology
Canalization and dredging
Rivers and streams are canalized for purposes such as navigation, ood protection, drainage of
agricultural land and urbanisation. For example in Finland, some rivers have been straightened
and narrowed to facilitate transport of logs downstream. The straightening of meandering streams
or rivers is often combined with a recurrent dredging of deposited material to maintain the desired
river prole. Canalization changes a naturally meandering stream with hydromorphological variety
into a uniform channel with homogeneous bed substrate and relatively uniform water velocity
across the stream. The channel is usually constructed to be wider than the natural prole to allow
for a larger conveyance, and the water depth and velocity will decrease. Often the ow variability
and the light conditions at the streambed will be affected as well.
These physical changes cause a reduced variety of habitat and consequently a lower species
richness of the stream ora and fauna compared with undisturbed streams (Hortle and Lake,
1983; Allan, 1995; Dessaix and Fruget, 1995; Giller and Malmqvist, 1999). Many stream
invertebrates are adapted to a life on coarse substrates (stones and gravel) with high current
velocities. Also, the availability of sh spawning habitats is dramatically reduced in rivers that are
canalized and dredged because of the removal of stones and gravel.
Weed cutting
Weed cutting is undertaken in many small and medium sized streams to increase the discharge
capacity and prevent the surrounding areas from inundation. River vegetation is rarely found
deeper than 1.5-2 metres and weed cutting is thus a phenomenon related to smaller streams.
When weeds are cut, the water level drops and the water velocity increases, which again may
increase the sediment transport and temperature. Bank stability may also be reduced. However,
the physical changes and the impacts on the river ecosystem are strongly dependent on the weed
cutting method.
The immediate effect of weed cutting is a direct loss of plants, which serve as a habitat for
invertebrates and a refuge for sh. The lower water level also reduces the available space and the
habitat diversity. Long-term use of weed cutting changes the composition and structural complexity
of the macrophyte community, which becomes poorer in species and spatially more homogeneous.
Also, substantial changes in composition patterns can develop with an enhanced abundance of
fast-growing species with a high dispersal capacity (Baattrup-Pedersen et al., 2003). A reduced
diversity and structural complexity of macrophyte communities can affect invertebrate and sh
communities negatively. This probably relates to a lower spatial and temporal physical
heterogeneity i.e. less varied substrate composition and more narrow range of ow velocities with
decreasing structural diversity of the macrophyte community (Garner and Bass, 1996). Therefore
loss of macrophyte species and homogenisation of communities as a result of weed cutting may
have cascading effects on the whole stream biota.
Methods for evaluating the effects of changes in river morphometry
Morphological classication systems. It is useful for the implementation of the WFD to
characterise streams according to their morphology. This morphological characterisation should
be targeted to be a causal link between the anthropogenic pressure on the river morphology and
the resulting impact on the river biota.
Different methods and indices are used in different countries.
The Austrian Habitat Survey (Werth, 1987; Muhar et al., 1996; Muhar et al., 1998) in Austria
The Danish Stream Habitat Index (Pedersen and Baattrup-Pedersen, 2003) in Denmark
The SEQ Physique (Agences de lEau and Ministre de lEnvironment, 1998) in France
The Ecomorphological Survey for Large Rivers (Fleischhacker and Kern, 2002) in Germany
The River Habitat Survey (RHS) (Raven et al., 1998) in the UK.
The methods use a number of parameters (channel, bank, oodplain, ow-related) and a scoring
system to evaluate the hydromorphological status of streams. Most of these methods are based
on a pseudo reference condition, which is identied on the basis of a top percentage of sites
according to their habitat quality scores. This causes a problem for the type of rivers of which
there are only few, or few with no impacts (e.g., large rivers).
Fish indicators. The occurrence, density and reproduction of shes in streams are very much
inuenced by the stream hydromorphology. The FAME project
1
developed sh-based metrics
sensitive to various pressures, including hydromorphology.
1
http://fame.boku.ac.at.
7.4 Abstraction and diversion of water
Water is abstracted from streams and rivers for water supply or irrigation, rivers are diverted for
various reasons, and sometimes streams are augmented by water from other catchments to
increase the stream ow for agricultural purposes. In the case of abstraction, water is taken either
directly from the stream, in which case the ow is reduced immediately, or it is taken from
groundwater and will affect the stream ow, depending on the distance of the borehole from the
stream and the hydraulic properties of the sediments between the borehole and the stream. Often
the highest demand occurs in summer, when the ow is naturally low. Thus, the most important
effect of abstractions is on the low ow regime, while diversions are more likely to affect the
stream at all times. Reducing the ow will lead to lower depth and velocity and an increase in
temperature (especially in summer), and the dilution of pollutants will reduce. In some cases
when water is diverted or abstracted, the stream may even dry up completely.
A reduction in ow caused by abstraction, or diversion, affects the biota (e.g. Collier, 2002).
Usually the growth of lamentous algae is favoured (mainly due to lower velocities) and the
invertebrate community will change from one that grazes on thin periphyton lms to one that
lives amongst thick periphyton mats. Amphibious plants may invade more central parts of the
stream from which they are normally kept away by high velocity and depth, and the distribution
of plant species will change. Reduced water level and velocity leads to a loss of river habitat,
which may affect biota on all levels.
Methods for the evaluation of changes in ow regimes
Flow criteria. The most well known classication system based on natural, frequently historical,
ows is the Tennant method (Tennant, 1976), sometimes also called the Montana method, which
species that 10% of the average ow is the lower limit for aquatic life and 30% of the average
ow provides a satisfactory stream environment. The Tennant method was based on hydraulic
data from eleven U.S. streams (including streams in Montana) and considerations of what values
of velocity, depth and width were needed for sustaining aquatic life, with a focus on sh.
Historical ows can also be used to dene an ecologically acceptable ow regime, although one
should be careful to distinguish between pre-development historical ows and impacted historical
ows, which may over a period of time have altered the composition of biota. Arthington et al.
(1992) describe a holistic method that considers not only the magnitude of low ows, but also
the timing, duration and frequency of high ows. Such a ow regime would not only sustain biota
during extreme droughts, but it would also provide the high ows and ow variability needed to
maintain the diversity of the ecosystem.
Habitat models. Habitat models describe relationships between the water discharge and the
biota of a river reach. Most habitat models use preference indices, which determine how suitable
a given quality element (velocity, depth, substrate) is for certain species and their individual
developmental/ life history stages. These models then combine the results from hydraulic models
with the preference indices to produce values of river area weighted by habitat quality (weighted
usable area) as a function of ow for a given species and life stage.
Current software includes:
PHABSIM (Physical Habitat Simulation; Bovee, 1982; Milhous et al., 1989), and RHABSIM
(River Habitat Simulation) used in the United States
RHYHABSIM (River Hydraulics and Habitat Simulation; Jowett, 1989) used in New Zealand
EVHA (Evaluation of Habitat; Pouilly et al., 1995) used in France
CASIMIR used in Germany (Jorde, 1997)
RSS (River Simulation System; Killingtviet and Harby, 1994) used in Norway
HABITAT used in the Netherlands (Duel et al., 2003).
Habitat Suitability Indices (HSIs). Physical HSIs allow relation of habitat preferences for a given
species and life stage to hydraulic parameters such as velocity, depth and substrate. An example
is given in Figure 7.1. Most HSIs have been dened as univariate response functions but more
complex multivariate relationships also exist (Parasiewcz and Dunbar, 2001). The relationships
established by HSIs are useful for decisions on the protection of relevant habitats or on the
restoration of habitats.
LIFE Index. The LIFE index (Lotic Invertebrate index for Flow Evaluation, Extence et al., 1999) was
formulated to test whether it is possible to link changes in benthic invertebrate community
structure with indices of historical river ow at a gauge close to the sample site. The LIFE index
can be calculated from species or family-level bio-monitoring data. Each taxon is assigned a
velocity preference from I to VI (based on literature data), and ve abundance categories are used.
Implicit in the velocity preference is preference or avoidance of silty substrates. A matrix is then
used to give a combined score for each taxon in the sample of between 1 and 12. The scores for
all taxa are added together, and the average score is the LIFE index.
It is important to note that the index is expected to be sensitive to natural and articial ow
changes; it thus allows an extrinsic hypothesis to be tested. Extence et al., 1999, demonstrated that
correlations exist between the LIFE score and moving averages of historical ows (e.g. Figure 7.2).
LIFE is currently being used in England and Wales as part of the implementation of Catchment
Abstraction Management Strategies and the Water Framework Directive (Soley et al., 2002;
Dunbar et al., 2004). The LIFE index is also being examined in the STAR project and its sensitivity
to ow tested (Dunbar and Clarke, 2004).
Figure 7.1 Habitat suitability indices for two developmental stages of brown trout (from Bird et al. 1995).
S
u
i
t
a
b
i
l
i
t
y
Velocity (m/s)
1
0.8
0.6
0.4
0.2
0
Adult
Fry/Juvenile
0 0.2 0.4 0.6 0.8 1
S
u
i
t
a
b
i
l
i
t
y
Depth (m)
0 0.5 1 1.5
1
0.8
0.6
0.4
0.2
0
Adult
Fry/Juvenile
Mean ow rank index. A similarly constructed index (MFR mean ow rank) has been developed
by the Environment Agency of England and Wales to relate ows to macrophyte communities
(Soley et al., 2002).
7.5 Summary
Damming, water abstraction, channelisation, dredging and weed-cutting cause major
hydromorphological pressures on river ecosystems. The ecological consequences of damming are
generally well known: A river reach is changed into a reservoir and the river continuity is broken.
The ecological consequences of water abstraction and of changes in river morphometry
(channelisation, dredging, and weed cutting) are known to some extent. However, this knowledge
is not sufcient to establish operational relationships between the degree of pressure (the extent
of anthropogenic impact) and the impacts on river biota needed for the implementation of the
Water Framework Directive (WFD).
7.6 References
Agences de lEau and Ministre de lEnvironnement. 1998. SEQ-Physique: a System for the Evaluation of the
Physical Quality of Watercourses, pp. 15
Allan J.D. 1995. Stream Ecology: Structure and function of running waters. Chapman Hall, London, pp. 388
Arthington, A.H., J.M. King, J.H. OKeeffe, S.E. Bunn, J.A. Day, B.J. Pusey, B.R. Bluhdorn, and R. Tharme. 1992.
Development of an holistic approach for assessing environmental ow requirements of riverine ecosystems.
In: J.J. Pilgram, and B.P. Hooper (Eds.) Water allocation for the environment. The Centre for Water Policy
Research, University of New England, Armidale, pp. 69-76
Baattrup-Pedersen, A., S.E. Larsen, and T. Riis. 2003. Composition and richness of macrophyte communities
in small Danish streams - inuence of environmental factors and weed cutting. Hydrobiologia 495:171-179
Biggs, B.J.F., R.J. Stevenson, and R.L. Lowe. 1998. A habitat matrix conceptual model for stream periphyton.
Archiv fur Hydrobiologie 143:21-56
Bird, D.J., G.W. Lightfoot, and A.P. Strevens. 1995. Microhabitat use by young salmon and trout in a UK chalk
stream. Proc IFM 25
th
Annual Study Course, Lancaster, 99-114
Figure 7.2 Flow and the LIFE score in the Waithe Beck in the UK (data from Extence et al., 1999).
86 88 90 92 94 96 98 00
Year
L
I
F
E

S
c
o
r
e
D
i
s
c
h
a
r
g
e

(
m
3
/
s
)
10.00
1.00
0.10
0.01
0.00
8.0
7.5
7.0
6.5
6.0
Fl ow
Li f e Score
Bovee, K.D. 1982. A guide to stream habitat analysis using the instream ow incremental methodology.
Instream ow information paper 12, U.S. Fish and Wildlife Service Biological Services Program FWS/OBS-
82/26, pp. 248
Collier, K.J. 2002. Effects of ow regulation and sediment ushing on instream habitat and benthic
invertebrates in a New Zealand river inuenced by a volcanic eruption. River Research and Applications
18:213-226
Dessaix, J., and J.F. Fruget. 1995. Changes of the macroinvertebrate communities in the dammed and
by-passed sections of the French Upper Rhone after regulation. Regulated Rivers: Research and Management
10:265-279
Dunbar, M.J., A. Ibbotson, I. Gowing, N. McDonnell, M. Acreman, and A. Pinder. 2001. Further validation of
PHABSIM for the habitat requirements of salmonid sh. Environment Agency R&D Report W6-036, Centre
for Ecology and Hydrology, Wallingford, UK, pp. 218
Dunbar, M.J., M.C. Acreman, and S. Kirk. 2004. Environmental ow setting in England and Wales Strategies
for managing abstraction in catchments. CIWEM Journal 18:5-10
Dunbar, M.J., and R.T. Clarke. 2004. Generalised LIFE response curves. Centre for Ecology and Hydrology
(CEH) Report to Environment Agency
Duel, H., G.E.M. van der Lee, W.E. Penning, and M.J. Baptist. 2003. Habitat modelling of rivers and lakes in
the Netherlands: an ecosystem approach. Canadian Water Resources Journal 28:231-248
Extence, C.A., D.M. Balbi, and R.P. Chadd. 1999. River ow indexing using British benthic macroinvertebrates:
a framework for setting hydroecological objectives. Regulated Rivers: Research and Management 15:543574
Fleischhacker, T., and K. Kern. 2002. Ecomorphological Survey of Large Rivers. German Federal Institute of
Hydrology, Postfach 200 253, D-56002 Koblenz, pp. 41
Fruget, J.F. 1991. The impact of river regulation on the lotic macroinvertebrate communities of the lower
Rhone, France. Regulated Rivers: Research and Management 6:241-255
Garner, P., and J.A.B. Bass. 1996. The effects of weed cutting upon the biota of a large regulated river.
Aquatic Conservation: Marine and Freshwater Ecosystems 6:21-29
Giller P.S., and B. Malmqvist. 1999. The Biology of Streams and Rivers. Oxford University Press, pp. 368
Growns, I.O., and J.E. Growns. 2001. Ecological effects of ow regulation on macroinvertebrate and periphytic
diatom assemblages in the Hawkesbury-Nepean River, Australia. Regulated Rivers: Research and Management
17:275-293
Hortle, K.G., and P.S. Lake. 1983. Fish of channelized and unchannelized sections of the Bunyip River, Victoria.
Australian Journal of Marine and Freshwater Research 34:441-450
Jorde, K. 1997. kologisch begrndete, dynamische Mindestwasserregelungen bei Ausleitungskraftwerken
(Ecology-based dynamic minimum water requirement regulations for discharge from power plants, in
German). Mitteilungen des Instituts fr Wasserbau, Heft 90, Universitt Stuttgart, pp. 155
Jowett, I.G. 1989. River hydraulic and habitat simulation, RHYHABSIM computer manual. New Zealand
Fisheries Miscellaneous Report 49, Ministry of Agriculture and Fisheries, Christchurch, New Zealand, pp. 39
Killingtviet, ., and A. Harby. 1994. Multi-purpose planning with the river system simulator a decision
support system for water resources planning and operation. In: Proceedings of the rst international
symposium on habitat hydraulics, Norwegian Institute of Technology, Trondheim
Milhous, R.T., M.A. Updike, and D.M. Schneider. 1989. Physical habitat simulation system reference manual -
version II. Instream ow information paper 26. United States Fish and Wildlife Service, Fort Collins, Colorado
Muhar, S., M. Kainz, M. Kaufmann, and M. Schwarz. 1996. Ausweisungusstypspezisch erhaltener
Fliessgewsserabschnitte in sterreich - sterreichische Bundesgewsser (In German). BMLF, Wasserwirtscha
ftskataster. Wien, pp. 176
Muhar, S., M. Kainz, and M. Schwarz. 1998. Ausweisungusstypspezisch erhaltener Fliessgewsserabschnitte
in sterreich Fliessgewsser mit einem Einzugsgebiet >500 km
2
ohne Bundessse (In German). BMLF,
BMUJF, Wasserwirtschaftskataster, Wien, pp. 177
Parasiewicz, P., and M.J. Dunbar. 2001. Physical Habitat Modelling for Fish: A developing approach. Large
Rivers 12:2-4
Pedersen, M.L., and A. Baattrup-Pedersen. 2003. kologisk overvgning i vandlb og p vandlbsnre
arealer under NOVANA 2004-2009 (In Danish). Teknisk Anvisning fra DMU nr. 21. National Environmental
Research Institute, pp. 128
Pouilly M., S. Valentin, H. Capra, V. Ginot, and Y. Souchon. 1995. Note technique: Mthode des microhabitats:
principes et protocoles dapplication (Technical note: Microhabitat method: Principles and protocols for
application, in French). Bulletin Franais de la Pche et de la Pisciculture 336:41-54
Raven, P.J., N.T.H Holmes, F.H. Dawson, P.J.A. Fox, M. Everard, I.R. Fozzard, and K.J. Rouen. 1998. River
Habitat Quality - The physical character of rivers and streams in the UK and the Isle of Man. Environment
Agency, Bristol, pp. 86
Soley, R., K. Akande, S. Quinn, N. Holmes, and D. Solomon. 2002. Resource Assessment and Management
Framework Report and User Manual (Version 3). Environment Agency RandD Technical Manual W6-066M.
Environment Agency, Bristol, UK
Tennant, D.L. 1976. Instream Flow Regimens for sh, wildlife, recreation and related environmental resources.
Fisheries 1:6-10
Werth, W. 1987. Okomorphologische Gewasserbewertung (Ecomorphological survey of streams, in German).
Osterreichische Wasserwirtschaft 39:122-128
Toxic pressures
1 - Pirkko Kokkonen. Late summer cyanobacteria from the Baltic Sea
2 - Anders Koed. Trap for catching and monitoring juvenile salmonids (Skjern River, Denmark)
3 - Gary Free. Sieving lake sediment (from Lake Ghirla, Northern Italy)
4 - Gary Free. Sampling lake water (lake Ghirla, Italy)
5 - Laurence Carvalho. Zander (IJsselmeer, Netherlands)
1 2 3 4 5
8 Impacts of selected hazardous and priority
substances (WFD) on freshwater and marine biota
Lena Blom and Eva Brorstrm-Lundn
8.1 Introduction
The main sources of toxic substances to water bodies in Europe may be categorised as agriculture,
sewage treatment plants, urban runoff, industry, lake/river sediment and landlls. Typically, each
of the sources contributes to the aquatic concentrations of particular substances that may have
an effect on the water environment because of their toxicity. Annex VIII of the Water Framework
Directive (WFD) contains an indicative list of the main groups of pollutants relevant to the
directive. Annex X of the directive lists priority substances selected amongst those, which present
a signicant risk to or via the aquatic environment. The list is to be reviewed every four years.
From the priority substances, priority hazardous substances are selected, emissions and losses
within 20 years. The COMMPS procedure (combined monitoring-based and modelling-based
priority setting scheme) was used in determining the list of priority substances. Thus, models were
used in the selection procedure. The WFD requires the identication of pressures to which water
bodies are subject, with particular regard to substances in Annex VIII of the directive. Further, the
member states are required to assess the likelihood that surface water bodies will fail to meet
environmental quality standards. The WFD denes environmental quality standards (EQS) as
concentrations of pollutants in water, sediment or biota that should not be exceeded in order to
protect human health and the environment.
The classication of water bodies according to ecological status following Annex V of the WFD
includes consideration of specic pollutants. For these, concentrations close to zero for synthetic
pollutants and concentrations close to the background level for non-synthetic pollutants are
required for the water body to have high ecological status. For good ecological status,
concentrations below the EQS are required.
The European Commission is preparing the daughter directive of the Water Framework Directive
(WFD) on chemical pollution control. The preparatory work is carried out in close co-operation
with the EU member states through the AMPS (Analysis and Monitoring of Priority Substances)
expert working group. The European Commission proposed to member states the environmental
quality standards applicable to priority substances. For specic pollutants not on the priority
substance list, the member states shall set EQS for water, sediment or biota. Where possible, both
acute and chronic toxicity data should be considered as a basis for the decision. The base set of
organisms to be considered are algae and/or macrophytes, daphnia (or representative organisms
for saline waters) and sh. Depending on the amount and type of toxicity data available different
safety factors are used in setting the EQS as outlined in the Technical Guidance Document on risk
assessment (ECB, 2003). Data on persistence and bioaccumulation are to be taken into
consideration if they are available.
IMPACTS OF HAZARDOUS AND PRIORITY SUBSTANCES ON BIOTA | 151
8.2 Overview of WFD priority substances levels in aquatic matrices
The WFD has identied 33 priority substances (Annex X, decision 2455/2001/EC). These
substances were selected on the basis of their risk to the aquatic environment or to human health
via the aquatic environment (European Commission, 2004). The present knowledge concerning
the environmental occurrence of these substances is scattered, and differs widely between
different substances.
The priority substances (PS) may originate from a variety of sources. Some are emitted from different
point sources and some are unintentionally formed via e.g. incomplete combustion. Many of them
are banned in Europe, but may still be present in the environment due to their persistence. Because
of atmospheric long-range transport, unregistered use or re-emission, these banned PS may still
result in high levels in the water environment in the future. The priority substances are listed in
Table 8.1. Their uses or emission sources are also given to provide information on where they may
derive from (Sternbeck et al., 2003). Eleven of the substances have been identied as priority
hazardous substances, which are of particular concern for inland, transitional, coastal and territorial
waters. These hazardous substances will be subject to a cessation or phasing out of discharges,
emissions and losses within an appropriate timetable that shall not exceed 20 years.
The priority substances have a wide range of physical and chemical properties, which affect their
fate in the aquatic environment. The transport processes for different substances in the environment
are illustrated in Figure 8.1. In water the priority substances may be dissolved or bound to particles
(suspended particle matter). The distribution of the substances depends on octanol water
partitioning coefcients (K
ow
) governing the distribution to either the sediment particles or water.
Their environmental fate is also governed by the adsorption coefcient to soil and sediment (K
oc

which for many substances is linearly correlated to K
ow
). The distribution of the substances between
the different phases may affect transport processes as well as the biological effects of each
substance. Several PS have properties that favour their partitioning to sediment (K
ow
for the PS are
presented in Table 8.1). Sediment represents an important media for accumulation of some of the
substances, even though some of them may undergo chemical changes in the sediment.
8.3 Availability of data
The proposed environmental quality standards (see paragraph 8.1) are for the concentrations of
different substances in the water phase (European Commission, 2004). However, most available
data concerns measurements carried out in the sediment or biota. The data included in this review
Figure 8.1 Substances differ in physical and
chemical properties which affect their transport,
occurrence and fate in the aquatic environment as
well as their biological effect (modied from UNEP,
2002).
Sediment
Air
Water
Vegetation
Soil
Aerosols
Aquatic particles
Name Estimated log Kow
(Syracuse Research
Corporation)
Identied as
priority hazardous
substance
Uses or emission sources
Polycyclic aromatic hydrocarbons, PAH X Incomplete combustion
Anthracene 4.35 Incomplete combustion
Naphthalene 3.17 Incomplete combustion
Trichlorobenzene 3.93 Industrial chemical
Pentachlorobenzene 5.22 X Unknown uses
Hexachlorobenzene, HCB 5.86 X Biocide; unintended formation
Alachlor 3.37 Pesticide
Atrazine 2.82 Pesticide
Chlorfenvinphos 4.15 Pesticide
Diuron 2.67 Pesticide
Endosulfan 3.50 Pesticide
Hexachlorocyclohexanes, HCHs 4.26 X Pesticide
(Lindane, gamma-HCH) 4.26 Pesticide
Isoproturon 2.84 Pesticide
Chlorpyrifos 4.66 Pesticide
Pentachlorophenol, PCP 4.74 Pesticide
Simazine 2.40 Pesticide
Triuralin 5.31 Pesticide
Brominated diphenyl ethers, PBDE X Flame retardant
Nonylphenol,
4-para-nonylphenol
5.99 X Industrial chemical; forms through
degradation of NP-ethoxylates
Octylphenol,
4-tert-octylphenol
5.50 Industrial chemical; forms through
degradation of OP-ethoxylates
Hexachlorobutadiene, HCBD 4.72 X Industrial chemical; by-product from
chlorinated solvent production
C10-13-chloroalkanes, SCCA X Lubricant; cutting uid
Di(2-ethylhexyl)-phthalate, DEHP 8.39 Plasticiser
Benzene 1.99 Incomplete combustion; component i
petroleum products
Trichloromethane (chloroform) 1.52 Solvent
1,2-Dichloroethane 1.83 Solvent
Dichloromethane 1.34 Solvent in medical industry
Tributyltin, TBT 7.35 X Antifoulant; preservative; stabiliser in plastics;
Lead, Pb 0.73 Numerous
Mercury, Hg 0.62 X Numerous
Nickel, Ni -0.57 Numerous
Cadmium, Cd -0.07 X Numerous
Table 8.1 Major uses and selected emissions sources of WFD priority substances (modied from Sternbeck
et al., 2003).
have been taken from scientic literature and national reports. Data from Swedish databases
(Swedish EPA) have been accessible but other nations databases have so far been difcult to
access. The concentrations of PS found in the available literature sources divided in water phase,
sediments and biota are presented in Table 8.2.
The data availability on the water phase is scarce and, in general, the 33 PS have rarely been
found above detectable concentrations in the water phase. For metals in water there are more
data showing detectable concentrations than there are for the organic substances. For example,
concentrations in the water phase of polycyclic aromatic hydrocarbons (PAHs), atrazine, diuron,
lindane (gamma-HCH), isoproturon, pentachlorophenol (PCP), simazine, polybrominated diphenyl
ethers (PBDE ## 47, 99, 100), hexachlorobutadiene (HCBD), benzene and tributyltin (TBT), are
not frequently found in the literature. More measurements have been performed on sediments, as
the 33 PS are more commonly detected in sediments than in the water phase. Data on PS in the
biota are available in the reviewed sources. The following substances were reported to occur in
detectable concentrations: PAHs, naphthalene, trichlorobenzene, pentachlorobenzene,
hexachlorobenzene (HCB), HCHs, gamma-HCH, isoproturon, PCP, PBDEs, HCBD, DEHPs, benzene,
trichloroethane, dichloromethane, TBT and all the priority metals.
Most of the reviewed studies include concentrations of only a few of the 33 PS. One exception was
a Swedish sediment study performed in an urban area and its surrounding lakes and coastal area
that included measurements of all the 33 PS (Sternbeck et al., 2003). The following substances
were detected in sediments: 4-nonylphenol, 4-tert-octylphenol, PAH, di(2-ethylhexyl)phthalate, C
10-13

chloroalkanes, pentachlorophenol, hexachlorocyclohexanes, chlorfenvinphos, polybrominated
diphenyl ethers, tributyltin, As, Cd, Co, Cr, Cu, Hg, Ni, Mn, Pb, Zn, PCBs, DDTs and chlordanes.
Chlorobenzenes were only detected sporadically. Those not detected in any sediment sample
were: alachlor, atrazine, diuron, isoproturon, chlorpyrifos, triuralin, TBBPA, hexachlorobutadiene,
benzene, 1,2-dichloroethane, dichloromethane and chloroform.
To perform measurements of high quality it is essential to have a good strategy. In the Water
Framework Directive a surveillance procedure is described in Annex V for waters. Measurements
of the PS could be carried out in order to determine the concentrations in water, sediment or
biota. The purpose of the measurement determines the sampling strategy such as number of
sampling sites, duration of sampling, type of aquatic media as well as sampling methods.
Measurements could be carried out in order to:
compare water concentrations in relation to EQS
look at the spatial distribution over Europe, in member countries, or in a region
screening
assess time trends and seasonal variations
identify hot spots
assess if the pollution is generated locally or is a result of long range transport
assess whether the pollutants may pose a risk to the ecological system
collect data for model validation.
Name Water (ng l
-1
) Sediment (ng g
-1
dw) Biota (ng g
-1
dw/lipid)
min max Min max Min Max
PAHs 0.13 8500 100 36000 9 250
Anthracene n.d. 2 210 2 10
Naphthalene n.d. 2 150 0.01 79
Trichlorobenzene n.d. n.d. 6 n.d. 16
Pentachlorobenzene n.d. 4 10 0.003 16
Hexachlorobenzene n.d. n.d. 3 0.001 41
Alachlor n.d. <9 <13 n.d. <50
Atrazine n.d. 0.5 <1 <2 n.d. <50
Chlorfenvinphos n.d. <6 236 n.d. <50
Diuron 400 600 n.d. 30000 n.d. <10
Endosulfan n.d. n.d. 10000 n.d. <50
HCHs n.d. n.d. 12.5 5 830
(Lindane, gamma-HCH) n.d. 70 0.06 140 3.5 481
Isoproturon n.d. 500 n.d. 30 n.d. <50
Chlorpyrifos n.d. <1 <2 n.d.
PCP urban* 0.2 14
PCP point* 5 950 2 644 0.19 330
PCP 5 3300 0.02 210 20 1100
Simazine n.d. 200 <1 23 n.d. <50
Triuralin n.d. <3 <6 n.d.
PBDE (## 47, 99, 100) 0.010 0.090 0.01 2.7 0.l4 30
Nonylphenol 180000 80 30 5300 <10 <60
Octylphenol n.d. n.d. 140 n.d. <10
HCBD point * n.d. 950 36
HCBD 0.27 3.2 <1 430000 0.06 <50
C
10-13
-chloroalkanes n.d. 0.12 3300 n.d.
DEHP n.d. 25 35000 350 4500
Benzene 0.05 0.3 <570 <17000 <10 <340
Trichloromethane <1 <28 <50 260
1,2-Dichloroethane <100 <2800 n.d. <50
Dichloromethane n.d. <57000 1100 3400
TBT 0.1 4.7 3.6 5800 0.3 360
Pb 3 15000 30000 430000 2 74800
Hg 0.1 9 10 3300 4 1050
Ni 10 47600 n.d. 63000 4 3180
Cd 1 1070 200 7000 7 66500
Table 8.2 Concentrations of the WFD priority substances in the water, sediment and biological matrixes found
in the data sources available (Muir and Eduljee, 1999; Skjelkvle et al., 1999; Palm et al., 2001; Palm et al.,
2002; Uln et al., 2002; Kaj and Palm, 2003; Tesfalidet, 2003; berg, 2003, Sternbeck et al., 2003; Bignert
and Carln, 2004; Fjeld et al., 2004; Kaj and Dusan, 2004; Sternbeck et al., 2004; Swedish EPA, 2004). Data
sources are background for all except for the ones marked point* and urban* identied specically from point
and urban sources (n.d. corresponds to substances analyzed, but with values below the detection limit).
8.4 Examples of results from studies carried out in Sweden
Measurements of the priority substances have been carried out in several studies in Sweden. The
results from these studies are used here to exemplify concentration levels of the PS in the aquatic
environment. Several of these studies have been carried out as screening studies to assess
concentration levels and the results may be used to identify possible emission sources, estimate
diffuse sources and the need for further monitoring.
As an example we report here the occurrence of PS in sediments in the Stockholm region (Sternbeck
et al., 2003). Main aims of this research were to identify which of the PS pollute an urban region
and its adjoining coastal zone and to assess whether this pollution is generated locally or if it is
the result of large-scale atmospheric transport. Other aims were to assess if the coastal region is
polluted due to the proximity to the urban area and to assess whether certain pollutants pose a
risk to the aquatic biota. Surface sediments were collected from 34 stations in the urban
environment, in small lakes surrounding the urban area and in coastal areas. In addition, the
sludge from two sewage treatment plants was analysed. This study was the rst Swedish study
that included all the 33 PS. In addition, As, Co, Cr, Cu, Mn, Zn, PCB and DDT were also covered.
The results are presented using box-whisker type plots, to display the percentile distribution of
each dataset. Values below the detection limit were included when calculating percentiles. An
example of emission (concentration, PBDE 47) in the Stockholm area is shown in g. 8.2, where
there is a clear enrichment of PBDEs in the central city of Stockholm.
The pattern DEHP, a chemical in use, illustrates the strong inuence of local source in the
catchments (Figure 8.3). Higher concentrations were observed upstream of central Stockholm
with decreasing concentrations in the inner part of central Stockholm and a decline from Stockholm
to coastal areas.
On the other hand, the pattern of HCHs concentrations (Fig. 8.4) does not indicate any strong
local inuence. HCHs are probably an example of substance group with long range transport.
Figure 8.2 Box plot of PBDEs in the three groups.
Lakes central Stockholm Coastal
B
D
E
-
4
7
,

g
/
k
g

d
w
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Figure 8.3 Concentrations of DEHP in the three
groups (note the logarithmic y-axis).
Lakes central Stockholm Coastal
n
g
/
g

d
w
10
5
10
4
10
3
10
2
10
1
To summarise the Stockholm region sediment study:
Most chemicals that are known to be used or unintentionally emitted in Sweden were found.
Several chemicals that are banned in Sweden were found. Of these, mainly those known to be
persistent or to undergo long-range atmospheric transport were found.
Except for a few stations, concentrations of e.g. PAHs, DEHP, PBDEs and PCBs are fairly
homogenous in the coastal region.
Many chemicals are apparently released in the urban and suburban region.
Alkylphenols are of high concern. The concentrations are fairly high and strongly irregular,
indicating a strong local inuence. Concentrations were generally highest in the lakes examined.
The concentrations of several chemicals span wide ranges in the lakes. This is interpreted as
indicating a strong inuence of local sources in the catchments.
Whether benzene and the chlorinated solvents pose a risk to the aquatic environment cannot
be assessed, because these chemicals do not partition strongly enough to sediments.
8.5 Studies carried out in Sweden
Screening involves measuring a substance in many matrices at many sites, but with few samples
per site. In a Swedish screening study sh from background sites (lakes, coastal and sea) were
analysed for a number of PS (Sternbeck et al., 2004). Figure 8.5 shows an example of PBDE
concentrations in sh. In general, only a few PAHs were detected in low concentrations.
Chlorobenzenes, DEHP and PCP were generally not found, while PBDEs were detected in most
samples. Considering that different sh species were studied and that the sites represent different
regions of Sweden, the variation in concentrations must be regarded as moderate.
To provide an example of seasonal variations in PCB and PAH concentrations in water, the
outcome of a study in a contaminated bay in the southern Baltic is presented in Figure 8.6 (Palm
et al., 2001). For PCB, the concentrations at the water surface were found to be generally higher
than at the bottom during the summer period. In contrast, there were no obvious differences
between season or depth for PAHs (Figure 8.6). Examples of EQSs (EQSs) for the PS in water are
Figure 8.4 Percentile distribution of sum-HCHs in
the three groups. Lakes central Stockholm Coastal
s
u
m
m
a
-
H
C
H
,

n
g
/
g

d
w
14
12
10
4
2
0
Figure 8.5 Screening of PBDEs in sh (ng g
-1
lipid) in Sweden (Sternbeck et al., 2004).
Lakes
n
g
/
g

l
i
p
i
d
v
i
k
t
20
15
10
5
0
Coastal
Sea
n
g
/
g

l
i
p
i
d
v
i
k
t
15
10
5
0
n
g
/
g

l
i
p
i
d
v
i
k
t
25
20
15
10
5
0
Bysj n
Hj r t sj n
St ensj n
Vder ar na
Kvdf j rden
Hol mar na
Fl aden
Ut l ngan
Har uf j rden
BDE47 BDE99 BDE100 BDE153 BDE154
Sea-berring
Coastal-Perch/eel-pout
Lakes - Perch
Utlngan
Fladen
Hjrtsjn
Kvdfjrden
Vderarna
Bysjn
Stensjn
Holmarna
Harufjrden
Figure 8.6 PCBs and PAHs concentrations in a contaminated bay in the southern Baltic, time trends and
seasonal variations (Palm et al., 2001).
okt-99 dec-99 feb-00 apr-00 jun-00 aug-00 okt-00
Concentrations in water
C
o
n
c
e
n
t
r
a
t
i
o
n

(
n
g
/
l
)
3
2
1
0
7
6
5
4
3
2
1
0
7
PCB
15
PAH
Surface
Bottom
Figure 8.7 PCBs and PAHs concentrations at different sediment depth in a contaminated bay in the
southern Baltic against sampling station (Palm et al., 2001).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Concentrations in water
C
o
n
c
e
n
t
r
a
t
i
o
n

(
g
/
g

d
w
)
4
3
2
1
0
25
20
15
10
5
0
7
PCB
24
PAH
0-20 cm
20-40 cm
40-60 cm
60-80 cm
Sampling station
presented in Table 8.3. The EQSs are collected from the preliminary Substance Data Sheet for
each PS (AMPS-group). If the water concentrations of PAHs (as
15
PAH) in gure 8.6 are compared
with the proposed EQS values in Table 8.3, it can be seen that the proposed EQS value for PAH
6-ring are exceeded in most cases, meaning that this is probably a contaminated site.
Palm et al. (2001) also provided an example of the variation in sediment concentration with
distance from a point source in comparison to background concentrations in a contaminated bay
in the southern Baltic. Figure 8.7 shows the results for PCBs and PAHs. Stations 1-6 are situated
nearest to the source area. Both PCB and PAHs concentrations generally decrease with distance
from the source (Figure 8.7).
8.6 Overview of the toxic effects of priority substances on biological
quality elements
Data on toxic effects of the PS on organisms representing different trophic levels of aquatic ecosystems
(biological quality elements as required for the classication in the WFD) have been compiled. The
review was based on sources available through a data/literature search and from the ECOTOX
database of the US EPA (http://www.epa.gov/ecotox/). Furthermore, a number of reports on risk
evaluation of the substances were reviewed (Abou-Waly et al. 1991; Zagorc-Koncan et al. 1996).
Substance Proposed Environmental Quality Standard (g l
-1
),
good chemical status concentration < EQS
Polycyclic aromatic hydrocarbons, PAH 0.0016 (6-ring PAHs)
0.03 (5-ring PAHs)
Anthracene 0.063

Naphthalene 2.4
Trichlorobenzene 0.4
Pentachlorobenzene 0.0032
Hexachlorobenzene, HCB 0.0004
Alachlor 0.25
Atrazine 0.6
Chlorfenvinphos 0.06
Diuron 0.2
Endosulfan 0.005
Hexachlorocyclohexanes, HCHs 0.042
(Lindane, gamma-HCH) 0.02
Isoproturon 0.32
Chlorpyrifos 0.03

Pentachlorophenol, PCP 0.22
Simazine 0.7
Triuralin 0.03
Brominated diphenyl ethers, PBDE 0.5 10
-3

Nonylphenol, 4-para-nonylphenol 0.33

Octylphenol, 4-tert-octylphenol 0.061
Hexachlorobutadiene, HCBD 0.003
C
10-13
-chloroalkanes, SCCA 0.41
Di(2-ethylhexyl)-phthalate, DEHP 1.3
Benzene 1.7
Trichloromethane (chloroform) 12
1,2-Dichloroethane 10
Dichloromethane 20
Tributyltin, TBT 0.0001
Lead, Pb 0.4
1
Mercury, Hg
2
Nickel, Ni 1.7
Cadmium, Cd
3
0.08 (40-<100 mgCaCO
3
l
-1
)
0.15 (100-<200 mgCaCO
3
l
-1
)
0.25 (>200 mgCaCO
3
l
-1
)
1
dissolved Pb;
2
no overall MPA/QS can be given;
3
as depends on hardness (mgCaCO
3
l
-1
)
Table 8.3 Environmental Quality Standards for WFD priority substances for inland water (AMPS group,
European Commission, 2004).
Toxic Substance Phytop Macroph. Phytobts Zoobts
Alachlor
Anthracene
Atrazine
Benzene
Brominated diphenylethers
Cd and its compounds
Chloroalkanes, C10-13-
Chlorfenvinphos
Chlorpyrifos
Dichloroethane, 1,2-
Dichloromethane
Di-(2-ethylhexyl)-phthalate (DEHP)
Diuron
Endosulfan
Endosulfan, alpha-
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclohexane
Lindane, -hexachlorocyclohexane
Isoproturon
Pb and its compounds
Hg and its compounds
Naphthalene
Nickel and its compounds
Nonylphenols
Nonylphenol, 4-
Octylphenols
Octylphenol, 4-tert-
Pentachlorobenzene
Pentachlorophenol
Polycyclic aromatic hydrocarbons
Benzo(a)pyrene
Benzo(b)uoroanthene
Benzo(g,h,i)perylene
Benzo(k)uoroanthene
Fluoroanthene
Indeno(1,2,3-cd)pyrene
Simazine
Tributyltin compounds
Tributyltin cation
Trichlorobenzenes (1,2,3-; 1,2,4- and 1,3,5-)
Trichlorobenzene, 1,2,4-
Trichloromethane, chloroform
Triuralin
Table 8.4 Summary of observed toxic effects of a priority set of substances on biological organisms
belonging to the major trophic groups and biological quality elements as required by the WFD
(cells highlighted indicate observed toxic effects). Phytop.= phytoplankton; Macroph.= macrophytes,
Phytobts= phytopbenthos and periphyton; Zoobts= benthic invertebrates and zooplankton.
This information is not complete but also the data is incomplete due to technical impairments
(literature sources may report only parameters in the upper effect scale, whereas levels at the lower
scale would be more relevant). Table 8.4 shows a summary of the detected impacts of a set of
priority substances on groups of biological organisms representing different trophic levels of aquatic
ecosystems. All the 33 PS are classied within the reversed proposal for a list of priority substances
in the context of the Water Framework Directive (COMMPS procedure; COMMPS, 1999).
8.7 Fate models and the WFD
As mentioned above priority substances differ in chemical and physical properties, which determine
their fate in the environment. Due to high measurement costs, the information concerning chemical
Model name Description/Field of application Reference Time Scale Spatially resolved Input requirements Media
Global models
MPI-MCTM Multi-compartmental chemistry transport model Max Planck Institute for
Meteorology, Hamburg
months/
decades
300 km x 300 km
ChemRange Distribution model for the assessment of long-range transport and persistence of chemicals (Scheringer 1996, 1997)
Globo-POP Multi-compartmental mass balance model for evaluation of fate in the global environment (Wania and Mackay, 1995) decades zonally divided properties, time-dependent emission rates air/water/soil/sediment/veg
Regional models
Level I,II,III models Initial fate assessment of chemicals, e.g. for comparative reasons (Mackay et al., 1992) steady state key properties air/water/soil/sed/sh
EQC model Level I, II, III models in one package (Mackay et al., 1996a,b,c) steady state key properties air/water/soil/sed/sh
CalTox Regional multi-media model (McKone and Enoch, 2002) air/water/soil/sediment/veg
ChemCAN Level III fugacity model containing site-specic parameters for the states of Canada www.trentu.ca/cemc steady state properties, emission rates air/water/soil/sediment/veg
SimpleBox Commonly used tool for risk assessment of chemicals within the European Union (van de Meent, 1993) steady state nested properties, emission rates air/water/soil/sediment/veg.
CoZMo-POP A generic, dynamic coastal zone fugacity model, which can be modied by the user (Wania et al., 2000b) decades No properties, time-dependent emission rates air/water/soil/sediment/veg
TaPL3 Level III model especially aimed at calculating potential for long-range transport (Beyer et al., 2000) steady state No key properties air, water, soil, sediment, sh
ELPOS Adapted version of EUSES-SimpleBox aiming to evaluate long-range transport potential and persistence of POPs (Beyer and Matthies, 2001)
POPCYCLING-Baltic Dynamic 6-compartmental model used for assessment of the fate of POPs in the drainage basin of the Baltic Sea (Wania et al., 2000a) decades yes air/water/soil/sediment/veg
EUROS Eulerian transport model www.rivm.nl
EUTREND Lagrangian atmospheric transport model www.rivm.nl
MSCE-POP multi-compartment
models
Models for the assessment of POP fate within the EMEP region and at hemispheric scale (Shatalov and Malanichev,
2000)
decades Properties, emissions, geophysical
and meteorological information
air/water/soil/veg
BETR North America Segmented multi-compartmental model for North America (MacLeod et al., 2001) steady state
Local models
ChemUK Level IV UK scale multi-media model Sweetman et al. 2002)
MUM Level III Urban 6-compartment model parameterized to Downtown Toronto (Diamond et al., 2001)
CeStoc Level III Urban 5-compartment model parameterized to centre of Stockholm (Palm, 2001) steady state No properties, emissions air/water/soil /sediment/
urban lm
QWASI A water-air-sediment exchange model that can be applied to ande lake-or river system (Mackay et al. 1983a,b)
GREAT-ER Riverine catchment model (GIS based) www.great-er.org
Table 8.5 Examples of existing models, model time-scales, spatial resolution and main eld of application.
fate cannot be assembled from monitoring data alone, but can also be captured in mass balance
models. As a result, environmental fate models are widely used to predict contaminant fate and
behaviour proles and are essential tools in the risk assessment process.
Persistent toxic substances (PTS) are included among the WFD priority substances. It is
very important to know how these substances are distributed, magnied and transported within
the environment to evaluate risks and identify sources to be able to reach good water
status. Modelling activities are useful for evaluation of PTS redistribution between various
environmental compartments, long-term trends of environmental contamination, spatial
distribution of concentrations in different media (atmosphere, soil, seawater, and vegetation) and
trans-boundary transport.
Model name Description/Field of application Reference Time Scale Spatially resolved Input requirements Media
Global models
MPI-MCTM Multi-compartmental chemistry transport model Max Planck Institute for
Meteorology, Hamburg
months/
decades
300 km x 300 km
ChemRange Distribution model for the assessment of long-range transport and persistence of chemicals (Scheringer 1996, 1997)
Globo-POP Multi-compartmental mass balance model for evaluation of fate in the global environment (Wania and Mackay, 1995) decades zonally divided properties, time-dependent emission rates air/water/soil/sediment/veg
Regional models
Level I,II,III models Initial fate assessment of chemicals, e.g. for comparative reasons (Mackay et al., 1992) steady state key properties air/water/soil/sed/sh
EQC model Level I, II, III models in one package (Mackay et al., 1996a,b,c) steady state key properties air/water/soil/sed/sh
CalTox Regional multi-media model (McKone and Enoch, 2002) air/water/soil/sediment/veg
ChemCAN Level III fugacity model containing site-specic parameters for the states of Canada www.trentu.ca/cemc steady state properties, emission rates air/water/soil/sediment/veg
SimpleBox Commonly used tool for risk assessment of chemicals within the European Union (van de Meent, 1993) steady state nested properties, emission rates air/water/soil/sediment/veg.
CoZMo-POP A generic, dynamic coastal zone fugacity model, which can be modied by the user (Wania et al., 2000b) decades No properties, time-dependent emission rates air/water/soil/sediment/veg
TaPL3 Level III model especially aimed at calculating potential for long-range transport (Beyer et al., 2000) steady state No key properties air, water, soil, sediment, sh
ELPOS Adapted version of EUSES-SimpleBox aiming to evaluate long-range transport potential and persistence of POPs (Beyer and Matthies, 2001)
POPCYCLING-Baltic Dynamic 6-compartmental model used for assessment of the fate of POPs in the drainage basin of the Baltic Sea (Wania et al., 2000a) decades yes air/water/soil/sediment/veg
EUROS Eulerian transport model www.rivm.nl
EUTREND Lagrangian atmospheric transport model www.rivm.nl
MSCE-POP multi-compartment
models
Models for the assessment of POP fate within the EMEP region and at hemispheric scale (Shatalov and Malanichev,
2000)
decades Properties, emissions, geophysical
and meteorological information
air/water/soil/veg
BETR North America Segmented multi-compartmental model for North America (MacLeod et al., 2001) steady state
Local models
ChemUK Level IV UK scale multi-media model Sweetman et al. 2002)
MUM Level III Urban 6-compartment model parameterized to Downtown Toronto (Diamond et al., 2001)
CeStoc Level III Urban 5-compartment model parameterized to centre of Stockholm (Palm, 2001) steady state No properties, emissions air/water/soil /sediment/
urban lm
QWASI A water-air-sediment exchange model that can be applied to ande lake-or river system (Mackay et al. 1983a,b)
GREAT-ER Riverine catchment model (GIS based) www.great-er.org
Table 8.5 Examples of existing models, model time-scales, spatial resolution and main eld of application.
The Mackay suite of fugacity based models are good examples of steady state mass balance
models, with the simple models relying on equilibrium partitioning and the more complex
assuming that equilibrium is not achieved between compartments. Such multimedia and food-
chain models can be used to predict the aquatic concentrations of selected PS when source
information is available, and with knowledge on effect concentrations this is useful for linking the
chemical and ecological status of a water body (Mackay et al., 1983a; Mackay et al., 1983b;
Mackay et al., 1992; Mackay et al., 1996a; Mackay et al., 1996b; Mackay et al., 1996c; Mackay
and Hickie, 2000; Mackay et al., 2003).
There are also examples of dynamic (time-resolved) fugacity based models. Fugacity models driven
by primary emissions are for example used in the U.K to predict and evaluate the historical fate,
behaviour and distribution of PCBs (Sweetman et al., 2002). The model attempts to re-create the
temporal release trend of PCBs over the last 40 years and to replicate the observed historical
trends in soils and sediments.
It is now increasingly accepted that successful management of chemicals in the environment
requires quantitative information on major sources, environmental concentrations, transport
pathways and routes of exposure in the ecosystem. Also esssential are assessments of risk as
determined by the proximity of concentrations or body residues to those at which effects are
observed (see QSAR models, chapter 9 of this book).
At present, several steady-state and dynamic multi-compartmental PTS transport models are
available, which, describe the PTS fate in the environment. The models used, especially those in
Europe, have been summarised in Table 8.5 taken from the report regionally based assessment of
persistent toxic substances regional report region III Europe (UNEP, 2002).
8.8 Summary
The priority substances may originate from a variety of sources. Some are emitted from different
point sources and some are unintentionally formed via e.g. incomplete combustion. Many of them
are banned in Europe, but may still be present in the environment due to their persistence. Because
of atmospheric long-range transport, unregistered use or re-emission, these banned PS may result
in high levels in the water environment in the future.
The WFD has identied 33 priority substances. These substances were selected on the basis of
their risk to the aquatic environment or to human health via the aquatic environment. The present
knowledge concerning the environmental occurrence of these substances is scattered, and differs
widely between different substances. The European Commission is preparing the daughter
directive of the WFD on chemical pollution control. The preparatory work is carried out in close
co-operation with the EU member states through the AMPS (Analysis and Monitoring of Priority
Substances) expert working group. The European Commission proposed to member states the
environmental quality standards applicable to priority substances. At the time of writing this
report, a proposal for priority substance EQS had been submitted but not accepted.
To perform measurements of high quality it is essential to have a good strategy. In the WFD a
surveillance procedure is described in Annex V for waters. Measurements of the PS could be carried
out in order to determine the concentrations in water, sediment or biota. The purpose of the
measurement determines the sampling strategy such as number of sampling sites, duration of
sampling, type of aquatic media as well as sampling methods. Due to high measurement costs, the
information concerning chemical fate cannot be assembled from monitoring data alone, but can
also be captured in mass balance models. Environmental fate models are widely used to predict
contaminant fate and behaviour proles and are essential tools in the risk assessment process.
8.9 References
Abou-Waly, H., M.M. Abou-Setta, H.N. Nigg, and L.L. Mallory. 1991. Growth response of freshwater algae
Anabaena os-aquae and Selenastrum capricornutum to atrazine and hexazinone herbicides. Bullettin of
Environmental Contamination and Toxicology 46:223-229
Beyer, A., D. Mackay, M. Matthies, F. Wania, and E. Webster. 2000. Assessing long-range transport potential
of persistent organic pollutants. Environmental Science and Technology 34:699-703
Beyer, A., and M. Matthies. 2001. Criteria for Atmospheric Long-range transport Potential and Persistense of
Pesticides and Industrial Chemicals R &D Project FKZ 299 65 402 E. Schmidt Verlag, Berlin, Germany
Bignert, A., and I. Carln. 2004. Screening of pestisides in sh. Swedish museum of Natural History
COMMPS. 1999. Reviesed proposal for a list of prioroty substances in the context of the water framework
directive (COMMPS Procedure)
Diamond, M.L., D.A. Priemer, and N.L. Law. 2001. Developing a multimedia model of chemical dynamics in
an urban area. Chemospere 44:1655-1667
ECB, 2003. European Chemical Bureau. Technical Guidance Document on Risk Assessment. EUR 20418 EN/1
European Commission. (2004). EAF(7)-06/01 AMPS draft nal report-of the expert group on analysis and
monitoring of priority substances.
Fjeld, E., M. Schlabach, J.A. Berge, T. Eggen, P. Snilsberg, G. Kllberg, S. Rognerud, E.K. Enge, A. Borgen, and
H. Gundersen. 2004. Screening of selected new organic pollutants-brominated ame retardants, chlorinated
parranes, bisphenol A and triclosan. Report 4809, NIVA, pp. 895
Kaj, L., and A. Palm. 2003. Screening of hexachlorobutediene (HCBD) in the environment. IVL report B1542
Kaj, L., and B. Dunsan. 2004. Screening of organic pollutants in sh- HCBD and chlorobenzene. IVL report B1557
Mackay, D., M. Joy, and S. Paterson. 1983a. A quantitative water, air, sediment interaction (QWASI) fugacity
model for describing the fate of chemicals in lakes. Chemosphere 12:981-997
Mackay, D., S. Paterson, and M. Joy. 1983b. A quantitative water, air, sediment interaction (QWASI) fugacity
model for describing the fate of chemicals in rivers. Chemosphere 12:1193-1208
Mackay, D., S. Paterson, and W.Y. Shiu. 1992. Generic Models for Evaluating the Regional Fate of Chemicals.
Chemospere 24:695-717
Mackay, D., A. Di Guardo, S. Paterson, G. Kicsi, and C.E. Cowan. 1996a. Assessing the fate of new and
existing chemicals: a ve-stage process. Environmental Toxicology and Chemistry 15:1618-1626
Mackay, D., A. Di Guardo, S. Paterson, and C.E. Cowan. 1996b. Evaluating the environmental fate of a variety
of types of chemicals using the EQC model. Environmental Toxicology and Chemistry 15:1627-1637
Mackay, D., A. Di Guardo, S. Paterson, G. Kicsi, C.E. Cowan, and D.M. Kane. 1996c. Assessment of chemical
fate in the environment using evaluative, regional and local-scale models: illustrative application to
chlorobenzenes and linear alkylbenzene sulfonates. Environmental Toxicology and Chemistry 15:1638-1648
Mackay, D., and B. Hickie. 2000. Mass balance model of source apportionment, transport and fate of PAHs
in Lac Saint Louis, Quebec. Chemosphere 41:681-692
Mackay, D., J. Hubbarde, and E. Webster. 2003. The role of QSARs and fate models in chemical hazard and
risk assessment. Paper prepared for Quantitative Structure-Activity Relationships (QSAR) Proceedings of the
QSAR 2002 Conference, Ottawa May 2002. Qsar & Combinatorial Science 22:106-112
MacLeod, M., D.G. Woodne, D. Mackay, T. McKone, D. Bennett, and R. Maddalena. 2001. BETR North
America: A regionally segmented multimedia contaminant fate model for North America. Environmental
Science and Pollution Research 8:156-163
McKone T.E., and K.G. Enoch. 2002. CalTOX, A Multimedia Total Exposure Model Spreadsheet Users Guide
Version 4.0, Lawrence Berkeley National Laboratory report LBNL 47399
Muir, J., and G. Eduljee. 1999. PCP in the freshwater and marine environment of the European Union. The
Science of the Total Environment 236:41-56
Palm, A., I. Wngberg, and E. Brorstrm-Lundn. 2001. Mercury and organic environmental poisons in
rserumsviken. IVL report B1433
Palm, A., J. Sternbeck, M. Remberger, L. Kaj, and E. Brorstrm-Lundn. 2002. Screening of pentachlorphenol
(PCP) in the environment. IVL report B1474
Scheringer, M. 1996. Persistence and spatial range as end-points of an exposure based assessment of organic
chemicals. Environmental Science and Technology 30:1652-1659
Scheringer, M. 1997. Characterization of the environmental distribution and behaviour of organic chemicals
by means of persistence and spatial range. Environmental Science and Technology 31:2891-2897
Shatalov, V., and A. Malanichev. 2000. Investigation and assessment of POP transboundary transport and
accumulation in different media. Part II. MSCE-E EMEP Report 4/2000
Skjelkvle, B.L., J. Mannio, A. Wilander, K. Johansson, J.P. Jensen, T. Moiseenko, E. Fjeld, T. Andersen, J. Vuorenmaa,
and O. Ryseth. 1999. Heavy Metal Surveys in Nordic Lakes; Harmonised Data for Regional Assessment of
Critical Limits. NIVA-report SNO 4039-99, Norwegian Institute for Water Research, Oslo. pp. 71
Sternbeck, J., E. Brorstrm-Lundn, M. Remberger, L. Kaj, A. Palm, E. Junedahl, and I. Cato. 2003. WFD priority
substances in sedimentts from Stockholm and the svealand coastal region. IVLreport B1538
Sternbeck, J., L. Kaj, M. Remberger, A. Palm, E. Junedahl, A. Bignert, P. Haglund, K. Lindkvist, M. Adolfsson-
Erici, K. Nylund, and L. Asplund. 2004. Organic environmental pollutant in sh from Swedish background
sites, IVL report B 1576
Swedish EPA. (2004). Biota database.
Sweetman, A. J., I.T. Cousins, R. Seth, K.C. Jones, and D. Mackay. 2002. A dynamic Level IV multimedia
environmental model: Application to the fate of PCBs in the United Kingdom over a 40-year period.
Environmental Toxicology and Chemistry 21:930-940
Tesfalidet, S., 2003. Screening of organotin compounds in the Swedish marine environment, (rst draft).
Ume University 2004
Uln, B., J. Kreuger, and P. Sundin. 2002. Survey of pestisides in water from agriculture and society year
2001. SLU report 2002:4
UNEP. 2002. Regionally Based Assessment of Persistent Toxic Substances. United Nations environmental
programme, chemicals. Europe regional report, UNEP Chemicals, Chtelaine, Switzerland.
Van de Meent, D. 1993. SimpleBox: A generic multimedia fate evaluation model. National Institute of Public
Health and the Environment (RIVM). Report no. 672720 001, Bilthoven, The Netherlands
Wania, F., and D. Mackay. 1995. A global distribution model for persistent organic chemicals. The Science of
the Total Environment 160/161:211-232
Wania, F., J. Persson, A. Di Guardo, and M.S. McLachlan. 2000a. The POPCYCLING-Baltic Model. A Non-steady State
Multicompartmental Mass Balance Model of the Fate of Persistent Organic Pollutants in the Baltic Sea Environment.
Norwegian Institute for Air Research (NILU 0R 10/2000). Available at www.scar.utoronto.ca/
~
wania
Wania, F., J. Persson, A. Di Guardo, and M.S. McLachlan. 2000b. CoZMo-POP - A Fugacity-Based Multi-
Compartmental Mass Balance Model of the Fate of Persistent Organic Pollutants in the Coastal Zone, WECC
Report 1/2000. Available at www.scar.utoronto.ca/
~
wania/
Zagorc-Koncan, J. 1996. Effects of atrazine and alachlor on self-purication processes in receiving streams.
Water Science and Technology 33:181-187
9 Applicability of QSAR models for assessing
impacts of toxic chemicals
Erik Furusj
9.1 Introduction
The WFD requires the identication of pressures to which water bodies are subject, in particular in
relation to substances listed in Annex VIII of the directive. Further, the member states are required
to assess the likelihood that surface water bodies will fail to meet environmental quality standards.
The WFD denes environmental quality standards as concentrations of pollutants in water, sediment
or biota that should not be exceeded in order to protect human health and the environment.
Annex VIII contains an indicative list of the main groups of pollutants relevant to the directive.
Annex X of the directive lists priority substances selected amongst those, which present a
signicant risk to or via the aquatic environment. From the priority substances, a list of priority
hazardous substances are selected, which shall be subject to a phase-out of discharges, emissions
and losses within 20 years (EC, 2001). The COMMPS procedure (combined monitoring-based and
modelling-based priority setting scheme) was used in determining the list of priority substances.
Thus, models were used in the selection procedure.
In addition to the specically listed substances, the WFD also requires that account is taken of
other toxic substances that may have ecological impact, including river-basin specic pollutants.
For such less common and less investigated pollutants, experimental data about toxicity or
physico-chemical properties is often scarce. Quantitative Structure Activity Relationships (QSARs)
are a potential means to ll data gaps in such circumstances.
9.2 Scope
Potential use of modeling methods in relation to toxic substances and the WFD includes fugacity
modeling and Quantitative Structure Activity Relationships (QSAR):
QSAR models can provide estimates of acute and/or chronic toxicity of substances to
different trophic levels in cases where experimental data are missing
QSAR models can provide data on partitioning, persistence and bioaccumulation in cases
where experimental data are missing.
Fugacity models can provide a link between pressures and concentrations of specic pollutants in
the environment. Thus, they have potential applicability in assessing the likelihood that surface
water bodies will fail to meet environmental quality standards and can be used to estimate
critical target loads. QSAR models are frequently used to estimate physical parameters needed for
fugacity modeling, i.e. partition coefcients, persistence and other parameters, since experimental
data of this type is often scarce.
It is important to note that although there is a strong focus in the WFD on the priority substances;
this review does not only consider these but also other potentially toxic pollutants. Other
QSAR MODELS FOR ASSESSING IMPACT OF CHEMICAL COMPOUNDS | 167
substances besides the priority substances may cause a deterioration in ecological status. QSARs
are potentially more useful in those cases, since experimental data on properties and effects of
the priority substances are, in general, better known.
The major questions that this review tries to answer are outlined below.
What QSAR models are currently available that have a potential use for implementation
of the WFD? Only models with a reasonable degree of successful validation and testing
are considered.
What are possible developments within the near future for QSAR models with potential
use for implementation of the WFD?
What gaps exist with respect to environmental compartments and different classes
of substances?
QSARs are not considered for inorganic compounds such as metals that are not in metallo-organic
form and small inorganic molecules, such as cyanide. The reason for this is that the amount of
structural information for these substances is very limited. Thus there is very limited possibility to
develop QSARs, which is also reected in the lack of available QSARs for these substances (Comber
et al., 2003). The Environmental Protection Agency of the United States of America (US EPA) uses
toxicity estimates based on water quality standards in such cases. Further, the fate and toxicity of
many metals is well understood and chemical speciation models and biological ligand models
may be used to predict their fate and toxicity in many cases. There is a vast literature on QSAR
model interpretation for understanding the mechanism behind toxic action of different substances.
This is not of primary concern for the WFD and thus not considered in this review.
9.3 Introduction to QSARs
Quantitative structure-activity relationships, collectively referred to as QSARs, are theoretical
models that can be used to predict the physicochemical, environmental and biological properties
of molecules. The notion that there is a relation between chemical structure and biological activity
is not new and QSARs have been developed for over a hundred years (Schultz et al., 2003). The
last 20 years have seen an enormous growth due to the rapid increase in computing power and
new methodology for molecular modeling and statistical analysis.
QSAR modeling can be based on so-called molecular descriptors that describe the structure and/
or properties of the molecule in numerical form. A statistical or mechanistic relation is then built
between the descriptors and the end-point of interest. The descriptors should contain relevant
information for the purpose of the modeling, i.e. the aspect of the behavior modeled. Molecular
descriptors can be classied by origin into measured and calculated descriptors. The major
difference from an application point-of-view is that the chemical substance in question is required
in order to obtain a measured descriptor while the calculated descriptors can be obtained for
substances that cannot be isolated or have not yet been synthesized. Undoubtedly, the single
most important descriptor used in QSAR is hydrophobicity, which is usually measured as the
logarithm of the octanol/water partition coefcient, log K
OW
. Recently, more advanced calculated
descriptors that account for the 3D structure of a molecule are being increasingly used especially
for receptor mediated effects such as endocrine disruption (Perkins et al., 2003).
It is generally accepted that it is important to account for the mode of action (narcosis, electrophilic,
receptor-mediated etc.) when developing toxicity QSARs (Escher and Hermens, 2002; Dearden,
2002; Schultz et al., 2003; Bradbury et al., 2003) since the descriptors and the model must cover
the properties important for that mode of action. This means that QSARs are often, but not
always (Ren, 2003), suitable only for a group of substances having common structural features
and a common mode of action. But QSARs can also be used to infer a likely mode of action from
the chemical structure (Verhaar et al., 1996; Russom et al., 1997).
Models for substance properties other than toxicity are often not called QSARs, since it is not
activity that is modeled. Frequently used names and abbreviations are QSPR (Quantitative
Structure Property Relationships) and QSBR (Quantitative Structure Biodegradation Relationships).
However, the naming is not consistent. In this report we use the QSAR abbreviation for all models
independent of the end-point modeled.
Model selection criteria
The ECVAM (European Centre for the Validation of Alternative Methods
1
, Joint Research Centre,
JRC) is co-ordinating the European activity on QSARs, which has the overall objectives of promoting
the development, validation and implementation of QSARs. QSAR developers are encouraged to
submit their models to the JRC for independent validation, which will be carried out within the
framework of the ECVAM validation and peer review process. Submitted QSARs should be
developed and documented in accordance with a minimal set of principles, referred to as QSAR
development and validation principles of which the most important in the current context are:
1. The QSAR should have a clearly dened scientic and regulatory purpose. In other words, it
should make predictions for a well-dened physical effect or biological endpoint that is of
regulatory relevance.
2. The QSAR should be dened in an explicit manner. For QSARs, the full equation should be
given, with an explanation of the descriptors used.
3. The domain of applicability of the QSAR should be explicitly dened. The QSAR should be
associated with a description of the chemical classes for which it is applicable (inclusion
rules) and/or inapplicable (exclusion rules).
4. A QSAR should be associated with basic statistics for its goodness-of-t to the training set
(e.g. r and R
2
values in the case of regression models).
5. Any information regarding the cross-validation or external validation of the QSAR should be
provided.
In the BMW project (Benchmarking Models for the Water Framework Directive), benchmark criteria
for selecting appropriate models in the eld of water management has been assembled (Saloranta
et al., 2003). Although QSAR models reviewed in this report are not strictly used for water
management, some of the criteria are relevant:
1. How well has the model been tested? (related to 5 above)
2. What is the balance between the models input data requirements and data availability?
1
http://ecvam.jrc.it.
3. How easily are the models results understood and interpreted? (related to 1 above)
4. Has the model received much peer acceptance and what is the models consistency with
scientic theory?
5. How are the models user manual and tutorial?
6. How is the models technical documentation?
7. How is the models interactivity, user-friendliness and suitability for end-user participation?
8. How is the models exibility for adaptation and improvement?
These principles have been applied in the evaluation and analysis of QSAR models in this report.
Validation and model performance measures
Measuring a models performance or accuracy is of great importance, since it is the accuracy that
determines if it useful for the intended purpose or not. In order to obtain a trustworthy estimate
of QSAR model performance, one must test the model with new substances that were not a part
of the training data. The importance of independent validation data deserves to be emphasised;
far too many of the QSAR models found in the literature have not been properly validated.
The two main categories of performance measures are:
Measures of relative explained variance. The most common are R
2
and Q
2
that give the
fraction (between 0 and 1) of explained variance in the response for training and validation
data respectively.
Prediction error measures. The prediction error estimate from training (calibration) data is
called root mean square error of calibration (RMSEC), which is sometimes denoted root mean
square error of estimation (RMSEE). The prediction error from independent validation data is
called root mean square error of prediction (RMSEP). RMSEC and RMSEP have the same unit
as the model response and can be interpreted approximately as the standard deviation of the
prediction, which makes them very useful to determine model performance.
9.4 Current regulatory use of QSARs
Most of the QSARs found in the literature have been developed for risk assessment of new and
existing chemicals. A recent review describes the use of QSARs in screening for environmental risk
assessment in the EU, USA and Canada (Russom et al., 2003). Cronin et al. (2003) reviewed
regulatory use of QSARs in USA, Canada and Europe. QSARs are used for the estimation of
properties such as persistence, bioaccumulation, partitioning in the environment and toxic effects
for a number of different species. The use of QSARs is most established in the USA. In Europe,
QSARs are most used in Denmark, Germany and the Netherlands.
The new EU REACH system emphasises the use of QSARs to decrease animal testing. The current
Technical Guidance Documents (TGD) on risk assessment of chemical substances (ECB, 2003) contains
a chapter with guidance on the use of QSARs. The TGD identify four specic uses of QSARs:
1. Data evaluation. Evaluation of adequacy of experimental data.
2. Decision for further testing. When the PEC/PNEC ratios derived by QSARs are >1 this is
considered an indication that more testing is needed (PEC: predicted environmental
concentration; PNEC: predicted no effect concentration).
3. Establishing specic parameters. Acceptable QSARs can be used to estimate parameters
needed as input for the PEC calculation.
4. Effects of potential concern. Acceptable QSARs can be used for preliminary assessment
of end-points that are not part of the base set of data.
The US EPA and other US authorities routinely use QSARs in chemical risk assessment (Walker,
2003). The Toxic Substances Control Act (TSCA) requires that the TSCA interagency testing
committee use QSARs to predict toxicity in the absence of data. Further, the US EPA uses QSARs
in combination with expert opinion for risk assessment of new chemicals (Comber et al., 2003).
The OECD Manual for Investigation of HPV (high production volumes) chemicals (OECD, 2003)
states that when no information is available for a given data element, calculation or estimates
derived from Quantitative Structure Activity Relationships (QSARs) can sometimes be provided.
The recommended QSARs for ecotoxicity are those implemented in the ECOSAR software package
but other models can be used as long as an indication of the quality of the methods used is given.
Also, QSAR predictions must be accompanied by experimental data for a close analogue. QSARs can
also be used for vapour pressure, K
OW
and water solubility if tests are impractical for some reason.
9.5 QSARs for toxicity
The development of reliable QSARs for toxicity requires a large amount of data from toxicity tests
that have been performed on different substances using a standardised protocol. This prerequisite
cannot be expected to be fullled for tests performed on a large number of species. Indeed, the
primary reason for using toxicity QSARs is the lack of available toxicity data! In chemical risk
assessment, one or a few species from each different trophic level are used to represent the toxicity
to all species in that level. Most regulatory protocols include acute lethality for sh and Daphnia
and chronic or reproduction data for algae, sh and Daphnia. The EU TGD on risk assessment of
chemicals (ECB, 2003) states that the minimum requirement is acute toxicity data for sh, daphnia
and algae. Thus most QSARs are developed for these end-points as discussed below.
Typical species for which QSARs have been developed include: Pseudokirschneriella subcapitata
(a green alga also known as Selenastrum capricornutum or Raphidocelis subcapitata), Daphnia
magna (small crustacean), Tetrahymena pyriformis (ciliated freshwater protozoan) as well as a
number of sh species: Poecilia reticulata (guppy), Lepomis macrochirus (bluegill sh), Danio rerio
(zebra danio), Leuciscus idus (Ide) and Pimephales promelas (fathead minnow). The full OSPAR
marine toxicity data set comprises (OSPAR, 2000): Skeletonema costatum, Acartia tonsa and
Scophthalmus maximus juveniles but, as discussed below, very few QSARs have been developed
for marine organisms.
Over the past 20 years QSARs for toxicity have become increasingly available and reliable and
today they are routinely used by some regulatory agencies, e.g. US EPA. Acute toxicity end-points
have most reference data and, hence, most QSARs have been developed for acute toxicity and
much fewer for chronic toxicity end-points.
The uncertainty of reference data is important to consider when developing QSARs. ECETOC (1989)
reports results from inter-laboratory testing indicating coefcients of variation (relative standard
deviations) of 5-80% with a typical value of around 50% for a range of biological toxicity tests.
Acute aquatic toxicity
The EU TGD on risk assessment (ECB, 2003) recommends four QSARs for acute aquatic toxicity,
two each for Pimephales promelas and Daphnia magna, which are valid for non-polar and polar
narcosis respectively. The prediction errors claimed are 0.33-0.37 log units, for details see Verhaar
et al. (1995). Note that the models were not validated using an external validation set, which
means that no reliable estimate of model performance exists.
ASTER (Assessment tools for the evaluation of risk)
2
was developed by the US EPA to assist
regulators in performing ecological risk assessments. ASTER is an integration of the AQUIRE
(AQUatic toxicity Information REtrieval) toxic effects database and a QSAR based expert system
that selects models according to mode of action. All QSARs in ASTER are linear and use log K
OW

as the only descriptor. ASTER contains models for LC
50
for Pimephales promelas and EC
50
for
Daphnia magna. A validation study for ASTER has been published (Russom et al., 1997) indicating
that it has a reasonable but limited performance (Q
2
=0.61 for a 97 substances test set). ASTER
was also tested in the comparative study by Moore (2003) discussed below but had a poorer
performance than most other software packages evaluated. ASTER is currently not publicly
available due to USEPA restrictions.
ECOSAR
3
was developed by Syracuse Research corporation for the US EPA. The software package
contains an expert system that classies a substance into any of a number of classes for which
QSARs are available. Most ECOSAR QSARs are Hansch type, i.e. based on log K
OW
only. If the
experimental K
OW
value is not available, a QSAR estimate is used. End-points covered by the
models include acute and chronic toxicity to sh, daphnid and algae respectively. Note that not
all end-points can be predicted for all classes, see the technical documentation (Clements et al.,
1996), which can be downloaded from US EPA web site. In a evaluation study performed jointly
by the US EPA and the European Commission in an OECD initiative (OECD, 1994) the QSARs in
the ECOSAR package used by the US EPA were evaluated against measured European MPD data
on acute toxicity to daphnid and sh for a large number of organic substances. The agreement
(dened as EC
50
within a factor 10 of the experimental value) was 79% and 87% respectively
after removal of questionable data from the MPD set. Note, however, that there have been serious
objections to the program and the validation study (Kaiser et al., 1999), claiming that the results
produced by the program are very uncertain. These objections are well founded in our opinion and
are also supported by the Procter & Gamble evaluation described in ECETOC (2003).
TOPKAT
4
can predict LC
50
for Pimephales promelas and EC
50
for Daphnia magna. The Pimephales
promelas predictions are made by eight class-specic QSARs. Cross-validation indicates that
2
http://www.epa.gov/medatwrk/databases/aster.htm.
3
http://www.epa.gov/opptintr/newchems/21ecosar.htm.
4
http://www.accelrys.com
predictions are normally within a factor of 2-3 from the experimental value. Daphnia magna
models exist for four classes of compounds and cross-validation indicates an error of 0.5-1.5 log
units. TOPKAT determines whether the submitted structure belongs to the so-called Optimum
Prediction Space (OPS) of the model in order to evaluate the reliability of the QSAR-based
assessment. The independent comparative study by Moore et al. (2003), discussed further below,
indicates that TOPKAT performs very well within its OPS but that this space is rather limited, i.e.
that the models are only valid for a limited set of substances.
Multicase
5
has a module for Pimephales promelas LC
50
predictions. The model is built from 479
training substances. Klopman et al. (2000) used Multicase to develop a model for the same end-
point based on 685 substances. The model incorporates the concept of a baseline activity as one
of the parameters for the correlation as well as others parameters such as the presence of
biophores, hardness-softness parameters, and other characteristics determined from quantum
mechanical calculations. This model is used by the Danish EPA for prediction of sh toxicity. The
performance was not veried by independent test data to our knowledge.
Kaiser and co-workers have developed probabilistic neural network (PNN) models for large diverse
sets of chemicals for Daphnia Magna (Kaiser and Niculescu, 2001), Tetrahymena pyriformis
(Kaiser et al., 2002) and Pimephales promelas (Kaiser and Niculescu, 1999). They used a large
number of substances and structure fragment based descriptors to build models that are valid for
a variety of toxic modes of action. Model validation shows excellent performance and the large
number of substances in the training set means that the applicability domain of the model is
larger than for many other models. The QSAR for Pimephales promelas (fathead minnow) has
been commercialised by TerraBase Inc
6
.
Furusj et al. (2003; 2006) have developed models for acute toxicity to green alga 96 h
(Selenastrum capricornutum growth rate inhibition EC
50
), Daphnia magna (48 h immobilisation
EC
50
) and sh (Lepomis macrochirus and Leuciscus idus 96 h toxicity LC
50
). They use an outlier
detection methodology to ensure reliable predictions and thus manage to produce a prediction
standard error of 0.5 log units on test data for substances within the valid chemical space of the
model. For substances outside the prediction space, the error is larger but the model automatically
provides diagnostics that shows if the prediction is accurate or not.
A review by Moore et al. (2003) has evaluated the performance of six QSAR packages for acute
aquatic toxicity: ECOSAR, TOPKAT, a Probabilistic Neural Network (PNN; Kaiser and Niculescu,
1999), a Computational Neural Network (Eldred et al., 1999), ASTER and OASIS
7
. Some of these
require input in the form of log K
OW
, which was calculated by the CLOGP
8
software discussed
elsewhere in this report. PNN and ECOSAR perform best in this study and are also the only
packages that are able to produce predictions for all 130 test substances. Still the coefcient of
variance is relatively low (r
2
~
0.3). TOPKAT and OASIS have a large advantage compared to other
packages, they produce prediction outlier diagnostics, i.e. indicates if the model is valid for the
5
http://www.multicase.com.
6
http://www.terrabase-inc.com.
7
http://omega.btu.bg/.
8
http://www.biobyte.com.
substance in question (this is called Optimum Prediction Space, OPS, in TOPKAT). For TOPKAT
there is a huge difference in performance between substances within the OPS (47 out of 127
substances) and those outside the OPS. The performance of TOPKAT for the 47 OPS substances is
far better than for any other package evaluated (r
2
~
0.6). For OASIS, 105 substances do not
generate warnings but the difference in performance between those and the 15 substances with
warnings is much smaller than for TOPKAT, which indicates that the outlier detection in OASIS
does not perform so well. The test set used in the evaluation by Moore et al. (2003) is heavily
weighted with low molecular weight substances that act by non-polar narcosis. A larger and more
diverse test set is needed to evaluate performance for substances with other modes of action.
The results of both evaluation studies discussed above (Moore et al., 2003; OECD, 1994) are
based on the assumptions that the reference toxicity data (LC
50
and EC
50
) are accurate. However,
this is denitely not the case. In both studies sh LC
50
data are, for several species, under non-
standardised test conditions. This should not affect the ranking but means that the performance
of the methods is likely to be better than indicated in the studies, since the reported prediction
errors are actually a sum of the model and experimental errors.
Chronic aquatic toxicity
The EU TGD on risk assessment (ECB, 2003) recommends two QSARs for chronic toxicity to the
sh Brachydanio rerio (28-32 day NOEC) and the daphnia Daphina magna (16 day NOEC for
growth and reproduction). Both models are Hansch type and only valid for substances acting by
non-polar narcosis (Verhaar et al., 1995). The TGD gives standard errors of 0.33 and 0.39 log units
for the two models respectively. The TGD (ECB, 2003) also recommend a semi-chronic algae QSAR
(72-96 h growth inhibition for Selenastrum capricornutum; Van Leeuwen et al., 1992) that is
similarly only valid for non-polar narcosis. Note that the models were not validated using an
external validation set, which means that no reliable estimate of model performance exists.
The US EPA ECOSAR predicts a number of chronic ecotoxicological end-points, e.g. chronic toxicity
to sh, invertebrates, algae and submerged vegetation. The software package contains an expert
system that classies a substance into any of a number of classes for which QSARs are available.
All QSARs are Hansch type, i.e. based on log K
OW
only. Note that not all end-points can be predicted
for all classes, see the technical documentation (Clements et al., 1996), which can be downloaded
from US EPA web site. Only acute toxicity end-points were evaluated in the OECD validation study
of ECOSAR (OECD, 1994) discussed above because the MPD data lacked other toxicity measures.
We have found no independent validation study of chronic end-points in ECOSAR. The recent
ECETOC study on QSARs (ECETOC, 2003) lists ECOSAR as the only available model for chronic
toxicity. Note the serious objections to the program (Kaiser et al., 1999), claiming that the results
produced by the program are very uncertain.
When considering using poorly validated and, sometimes, poorly performing chronic toxicity
QSARs it can be an alternative to use acute toxicity QSAR models and an estimated acute/
chronic toxicity ratio to estimate chronic toxicity. The judgement must be based on a case specic
consideration: is the uncertainty caused by less accurate chronic toxicity QSARs greater than from
generalisation of acute/chronic ratio.
Aquatic no-effect concentrations
Due to the lack of QSARs for no-effect concentrations (NOEC), extrapolation and safety factors are
used to estimate predicted no-effect concentrations (PNECs) from acute and chronic toxicity
estimates (ECB, 2003). Accurate estimation of PNEC requires test data for a large number of
organisms and, hence, is expensive, time-consuming and unwanted from an animal testing point-
of-view. Thus, there is considerable incentive to develop QSARs for aquatic NOEC and/or PNEC.
The ECETOC report on QSARs (ECETOC, 2003) discusses QSARs for PNEC extensively. de Zwart
(2001) showed that the mode of action has a large inuence on the species sensitivity variation.
Roex et al. (2000) showed a similar relation for the acute/chronic toxicity ratio (ACR). Van Leeuwen
et al. (1992) have used QSAR estimates of toxicity of narcotic chemicals to predict no-effect levels
(NELs) at the ecosystem level by means of recently developed extrapolation methods. The approach
was shown to yield estimates in good agreement with established methods for deriving water
quality objectives. A paper describing statistical extrapolation of QSAR toxicity data for phtalate
esters using this approach was recently published (Parkerton and Konkel, 2000).
Sediment toxicity
No models covering more than a very limited group of substances that predicts toxicity to
sediment-living organisms have been found. Approaches to convert PNEC
water
predicted by QSARs
to PNEC
sediment
based on pore water concentrations calculated from equilibrium theory have been
used. An example of this approach in combination with QSARs is the one by van Leeuwen et al.
(1992) at the University of Utrecht. The authors have used QSAR estimates of toxicity of narcotic
chemicals to estimate PNECs from means of extrapolation methods. Equilibrium partitioning
theory was used to derive PNECs for aquatic sediments and internal toxicant concentrations for
aquatic organisms. Calculations were carried out for 102 narcotic compounds. The 1999 revised
proposal for priority substances (Klein et al., 1999) refers to the TGD on risk assessment (ECB,
2003) for calculation of PNEC
sediment
from PNEC
water
for organic substances. The proposed procedure
is not applicable for metals.
Endocrine disruption
Endocrine disrupting chemicals (EDCs) are an area of growing concern in chemical risk assessment.
Most of the end-points related to endocrine disruption used in QSAR modelling are in vitro tests
related to receptor binding or biological effects of receptor binding, e.g. gene activation. QSARs
are considered important and potentially useful due to the complexity and cost of in vitro and in
vivo tests for endocrine disruption.
The current status of QSARs for endocrine disruption (Schmieder et al., 2003) are that most
models are developed for specic substance groups. Some models have been developed based on
diverse sets of chemicals with the purpose of screening new chemicals for EDCs. Focus is on the
human estrogen receptor (hER) with few applications to other human receptors and ERs from
rodents and calves. No QSARs for endocrine disruption in aquatic organisms have been found.
9.6 QSARs for fate of substances
QSARs have an important role to play in providing input data to environmental fate models, e.g.
fugacity models. Mackay and Webster (2003) discuss this in a recent paper and list areas where
QSARs are of primary importance for fate of substance modelling and prediction, namely
partitioning, reactivity and transportation. Due to the very limited amount of QSARs for transport
properties, these are not discussed in this review. Properties in the other two areas are discussed
below together with bioaccumulation which is dependent on both of the other groups.
EUSES
9
, which is used for fate modelling, needs all or some of the following properties related to
the aquatic environment. In addition, a number of properties related to the air and soil
compartments are needed.
melting point (MP), boiling point (BP)
water solubility (S
W
)
octanol-water partitioning coefcients (K
OW
) and organic carbon normalised partitioning
coefcients (K
OC
)
Henrys law constant (HLC) and vapour pressure (VP)
bioconcentration factor (BCF) for aquatic biota
characterisation of biodegradability
degradation rate in sewage treatment plants
rate constants for hydrolysis, photolysis and biodegradation in surface water
rate constants for degradation in aerated and bulk sediment.
Persistency, degradation and reactivity
Biodegradation. Biodegradation is one of the main factors determining the length of time a
chemical persists in the environment. Biodegradation is usually bacterial, so testing is dependent
on the bacterial ora and results are comparable only if determined under the same protocol but
even then there are difculties in reproducing results, which leads to a considerable uncertainty
in biodegradation reference data. Furthermore, the multitude of degradation mechanism can be
expected to make QSAR modelling even more difcult, which is conrmed by the lack of accurate
quantitative models especially for diverse data sets. An often used approach is to develop
classication models that classify substances as readily degradable or not.
The BioWin program
10
, which is a part of the EPI suite software, estimates the probability for the
rapid aerobic biodegradation of an organic chemical in the presence of mixed populations of
environmental micro-organisms. Different types of estimates are obtained:
1. Estimates for biodegradability classication (fast, not fast) based upon fragment constants
that were developed using multiple linear and non-linear regression (Howard et al., 1992).
The methods classify about 90% of the training substances correctly. The EU TGD
9
http://ecb.jrc.it/existing-chemicals.
10
http://www.epa.gov/oppt/exposure/docs/episuitedl.htm.
recommends these models for biodegradation but states that many substances are falsely
predicted to degrade fast, so that only predictions indicating slow biodegradation should be
considered (ECB, 2003).
2. Estimates the time required to achieve primary and ultimate biodegradation on the scale of
hours, days, week, months or longer according to a method described by Boethling et al. (1994).
3. Predictive models for assessing a chemicals biodegradability (readily degradable, not readily
degradable) in the Japanese Ministry of International Trade and Industry (MITI)
biodegradation test (Tunkel et al., 2000). These models were developed using larger data sets
than the above and were validated with external data giving approximately 80% correct
classication.
4. The predictions are combined in a battery approach to obtain a summary prediction of
whether or not the substance is readily biodegradable.
The ECETOC report on available QSARs (ECETOC, 2003) states that BioWin is a simple and very
well working model. It should be noted that BioWin contains six different models that can give
non-uniform results, which hinders easy usage. A recent study shows that although individual
models can misclassify substances the risk of false positives, i.e. to predict a substance as readily
degradable although it is not, can be signicantly reduced by using several models from Biowin
in a battery (Boethling et al., 2004).
The TOPKAT package has an aerobic biodegradability module, consisting of four structurally based
sub-models. A single study which reported the biodegradability of 894 compounds, as assessed by
the MITI test protocol number I, was used to develop these models. The discriminant models
estimate the probability of aerobic biodegradation, where probability values between 0.3 and 0.7
refer to an indeterminate region. TOPKAT uses prediction outlier diagnostics to determine whether
the submitted structure belongs to the Optimum Prediction Space (OPS) of the model, which
indicates a reliable prediction. No external validation has been reported to our knowledge.
Expert systems can be used to predict degradation pathways for organic substances and assess
persistence. Two commercially available expert systems are CATABOL from the Laboratory for
Mathematical Chemistry in Bulgaria and META from Multicase Inc. (see notes 5 and 7). The
CATABOL software uses a mechanistic modelling approach for quantitative assessment of
biodegradability (Jaworska et al., 2002), which predicts biodegradation pathways of chemicals
and classies them as readily degradable or not. CATABOL is unique in that it considers the full
biodegradation pathway and not only the parent compound. META can be used to predict
biodegradation pathways for both anaerobic and aerobic environments. The methodology is an
expert system based on fragment data and transformation rules (Klopman et al., 1995).
OECD (OECD, 1993b) published an early biodegradability QSAR report in 1993. The authors
concluded that only a few of the 75 models evaluated (most of them class specic) can be
recommended for predictive purposes. In most cases the poor performance is due to bad training
data: inhomogeneous end-point determinations or restricted data sets.
ECETOC (ECETOC, 2003) has compiled validation data for BioWin (MITI end-point), META and
CATABOL. The results showed similar performance for the three programs with approximately
80% correct classications into readily or not readily degradable.
A different and potentially more accurate approach is to use a multimedia/fugacity type model
to estimate the total half-life of the substance in the system studied. Different approaches are
possible as discussed by Mackay et al. (2003) but the advantage is that explicit consideration is
given to the partitioning of the substance in the environment. The multimedia model would then
need QSAR estimates of partitioning coefcients as well as half-lives or degradation constants in
different media including biodegradation.
Photolysis. Photolysis in water can be a signicant degradation pathway, especially for substances
that are not biodegradable. The EU TGD currently does not recommend any QSAR models for this
end-point, since the existing models do not meet the requirements (ECB, 2003). Mill (2000)
further discusses QSAR development for aqueous photolytic reactions.
Hydrolysis. The EU TGD (ECB, 2003) recommends ve separate class specic QSARs for
brominated alkanes, esters, carbamates and benzonitriles, which are based on electronic and
steric constants. No validation of the models has been found in the literature. OECD (OECD,
1993a) recommends some class-specic methods based on Hammet and Taft constants but note
that the models do not cover more than a very small amount of chemicals. HydroWin, which is a
part of the EpiSuite program (see footnote 10), is applicable for certain chemical classes. The
program estimates base and acid catalysed hydrolysis rate constants. No validation of the models
has been found in the literature.
Chemical oxidation and reduction. Canonica and Tratnyek (2003) have reviewed QSARs for
oxidation rate constants in aquatic environments. They conclude that many of the reactions
treated in the literature are not relevant for oxidation in the natural environment. Environmentally
relevant oxidants include excited triplet states of dissolved organic substances and photo-
generated singlet oxygen. For the latter, a number of Hammet constant based QSARs exist for
individual groups of chemicals.
QSARs for reduction reactions relevant for the aquatic environment describe dechlorination and
nitro group reduction (Tratnyek et al., 2003). A problem with models yielding observed rate
constants is that the systems for which the experiment is carried out are usually not very well
dened, which makes it difcult to use these QSARs in practice. Models of well-dened systems
also exist, but unfortunately these are not straightforward to use in practice since they are usually
unrepresentative of the complexity of systems occurring in the environment. Thus, the underlying
problem, which is not a QSAR issue, is that the chemical agents responsible for reduction and
oxidation in the environment vary in concentration and type and are often difcult to quantify.
Bioconcentration and accumulation
Bioaccumulation and bioconcentration measure how (persistent) substances are concentrated in
organisms. Bioaccumulation is the term for uptake via the food chain giving higher concentrations
at higher trophic levels. This property has not been subject to QSAR analysis to our knowledge.
Bioconcentration is the term for uptake of substances in the organism from surrounding media
and shows the potential for bioaccumulation. It is usually measured as the bioconcentration
factor (BCF). QSARs have been developed for BCF in sh and algae. Most of them are based on
log K
OW
as the sole descriptor but it has been noted that the correlation between BCF and log K
OW

breaks down at log K
OW
values over 6-7 (OECD, 1993a).
The EU TGD on risk assessment (ECB, 2003) recommends a method implemented in the EUSES
software (footnote 9). A linear QSAR for calculation of BCF in sh from K
OW
was derived for log
K
OW
<6 and is only valid within this range. Validation with a 267-substance test set gave a standard
prediction error of 0.58 log units. A non-linear QSAR is used for log K
OW
>6 but the accuracy is lower
than for the linear model. The models are not applicable for ionic substances and organometallics.
ECETOC (ECETOC, 2003) raises a number of questions relating to these models and the data they
are based on and recommends careful assessment of the model and the purpose of its use.
BCFWIN is a part of the EpiSuite software and is publicly available (see footnote 10). The BCFWIN
method published by Meylan et al. (1999) classies a compound as either ionic or non-ionic. The
QSARs for non-ionic compounds use log K
OW
and structural correction factors to predict BCF. The
correction factors allow for correction of substance metabolism to some extent (ECETOC, 2003), which
is advantageous. Different QSARs are used depending on the K
OW
range. It is noted that organometallics,
substances with long chain alkyls and aromatic azo compounds require special treatment.
CEFIC, the European Chemical Industry Council, validated the BCFWIN model using a test set of
50 substances (ECETOC, 2003). The results show reasonable predictive ability of BCF and about
70-80% correct classication of bioaccumulation potential. It should be noted that several of the
substances included are expected to be bio-transformed and consequently predictions based on
log K
OW
are expected to be conservative. The RMSEP for BCF was 0.96 log units. The authors note
that the BCFWIN model can be expected to give both false positives and false negatives in
classication but still recommend it over the other models evaluated.
Mackay et al. (2003) indicate that although BCFWIN is a successful model, mechanistic models
are preferable. This is especially true for substances that are metabolised readily. QSARs are likely
to overestimate bioaccumulation for these. It should be noted that the mechanistic models are
heavily dependent on QSAR derived partitioning data, as discussed in section 9.3. The POPs
software (Laboratory of Mathematical Chemistry, Bourgas University, Bulgaria) implements a
method for evaluation of sh bioaccumulation potential by a non-linear BCF model. The model is
dened within a large range of hydrophobicity (Dimitrov et al., 2002).
Partitioning and physical properties
Octanol/water - K
OW
. Because of the importance of log K
OW
in pharmaceutical research and drug
design, there are a large number of software packages that contains QSAR models that are either
commercially or freely available, e.g. KowWin which is a part of USEPA Epi Suite (footnote 10),
CLOGP v3
11
and v4.0
12
, SPARC
13
, ALOGPS
14
and CSLogP
15
. The OECD method for K
OW
determination
11
http://daylight.com.
12
http://www.biobyte.com.
13
http://ibmlc2.chem.uga.edu/sparc.
14
http://146.107.217.178/lab/alogps/index.html.
15
http://www.chemsilico.com.
has a reported uncertainty of 0.3 log units (ECETOC, 1989) and all reported QSAR prediction
uncertainties below must be compared with this value. The approach used in KowWin (Meylan
and Howard, 1995) is based on atom/fragment contributions and has been reported to give an
RMSEP of 0.31 log units for 6055 test substances. An independent study by Eros et al. (2002)
indicates that the RMSEP is at least twice that gure for new data.
A recent comparison between some available software packages and a new method (Eros et al.,
2002) revealed that the classic and freely available software package CLOGP v3 performs well
in most cases. It uses fragment contributions and correction factors to account for fragment
interactions and builds on a database of thousands of compounds. The model is thus applicable
for substances in a large chemical domain. The study indicated a prediction standard deviation of
about 0.6 log units (a factor 4) for substances unknown to the model. This model is also the one
recommended by the OECD in a study in 1993 (OECD, 1993a). Both CLOGP and KOWWIN are
recommended by the EU in the TGD for risk assessment (ECB, 2003).
ECETOC (ECETOC, 2003) refers to a validation study using independent test data where the
prediction errors for SPARC, KowWin and CLOGP were 0.33, 0.34 and 0.29 log units respectively.
In another study (ECB, 2003), the prediction standard deviation is approximately 0.3 log units for
both CLOGP and KowWin for a 1166 substance test set. The accuracy is higher for KowWin for log
K
OW
>5, while CLOGP gives better estimates for log K
OW
<0. It should be noted that the CLOGP
program v3 cannot process structures that contain tertiary amines, thiophosphates, phosphites,
phosphines and some other fragments because they are missing in the fragment database.
Also, salts and substances with formal charges are not handled (ECB, 2003).
Methods based on molecular descriptors, as opposed to atom or fragment-based contribution, at
least partly overcome problems that appear when new fragments, for which no contribution is
available, are introduced (Katritzky et al., 2000). Methods based on solvatochromic descriptors
have good performance but these descriptors have to be determined experimentally, which makes
the practical use of the models very limited.
A good and applicable QSAR model based on molecular descriptors (electrotopological state
indices; Gombar and Enslein, 1996) is implemented in the TOPKAT software. A large advantage
is the prediction diagnostics, called optimum prediction space (OPS). The authors report a
prediction error of 0.27 and 1.35 log units for substances inside and outside the OPS respectively,
which clearly indicates the benets of prediction diagnostics.
ALOGPS (footnote 14) is freely available over the internet and uses atom counts and
electrotopological state indices and an articial neural network model to estimate log K
OW
for
substances based on their structure (Tetko and Tanchuk, 2002). Validation with an independent
test indicates an RMSEP of approximately 0.5 log units. An interesting option is to include a
library of molecules with known K
OW
that are similar to the ones one wants to predict. This improves
the performance substantially but it is of course not certain that such data is available.
Soil sorption - K
OC
. Soil sorption is a fundamental process that has a large impact on the
availability, fate and persistency of substances in the environment. Sorption coefcients
quantitatively describe the distribution of a chemical between an environmental solid and the
aqueous phase that it is in contact with. Sorption can take place by a number of mechanisms and
sorption coefcients depend on a large number of factors such as organic matter quantity and
type, clay mineral content and type, particle size distribution, pH, temperature and dissolved
organic material. Fortunately some generalisations can often be made that facilitate the
calculation and measurement of sorption coefcients. Sorption coefcients are normally assumed
to represent equilibrium but sorption kinetics can play an important role for some slowly absorbing
substances, especially in fast processes such as soil resuspension. Incorporation of kinetic
considerations is an important area for future work (Doucette, 2003). Linear sorption behaviour
is usually assumed, although non-linear sorption can be accounted for using the Freundlich or BET
type models.
For organic substances, which are the topic of this review, and especially neutral and hydrophobic
substances, sorption is highly correlated or directly proportional to organic matter content of the
solid. As a consequence, organic carbon normalised sorption coefcients, K
OC
, are used.
This approach works reasonably well for a large number of organic chemicals and is the most
popular for use in environmental applications. For ionizable substances and for soil/sediment
with a low organic content (<0.1%) other approaches should be used (Doucette, 2003).
When evaluating the performance of QSARs for K
OC,
consideration must be given to the accuracy
of experimental K
OC
determinations for two reasons. The rst is that the QSAR model is an
alternative to more expensive and time consuming experimental determinations and hence their
relative performance is of interest. The second reason is that the prediction errors estimated from
model validation are inuenced by inaccurate reference values since it is the difference between
these and the predictions that are used to estimate the error. Usually unstandardised methods for
experimental determination of K
OC
are used, but even for the ASTM standardised method a
coefcient of variation of around 50% can be expected (Doucette, 2003), which corresponds to
0.2-0.3 log units.
The most widely used QSARs for K
OC
are based on simple correlations with K
OW
or water solubility
(S
W
). Some also require the melting point as an input. General or class specic models are available
for examples see the review by Doucette (2003). Class specic models often have a better
performance if they are available for the substance in question. Generally, the accuracy of models
for log K
OC
is within 0.3-1 log units if accurate values of K
OW
or S
W
are available (Doucette, 2003).
If experimental K
OW
and/or S
W
are not available, calculated values can be used but this lowers the
accuracy of the K
OC
predictions.
The system of class-specic QSARs for calculation of K
OC
from K
OW
developed by Sabljic et al.
(1995) are recommended for K
OC
prediction in the EU TGD on risk assessment (ECB, 2003). The
EUSES software (footnote 9), contains an implementation of these that can be used in the absence
of experimental soil sorption data. Calibration errors for different models range from 0.2-0.5 log
units. To our knowledge, the methods have not been validated with external data.
Methods that are based on molecular structure and do not use K
OW
or S
W
as descriptors do not
suffer from lack of experimental data for these parameters. Methods based on molecular
connectivity indices (MCI) are popular but their domain is often limited to a specic group of
substances, which makes them less generic. Meylan et al. (1992) have used MCIs in combination
with fragment based polar correction factors to overcome this drawback. Their model is
implemented in the PCKOCWIN model of the EPI Suite software (footnote 10).
In addition, methods based only on fragment contributions can be used. Tao (1999) developed
such a model with a reported average test set prediction error of 0.47 log units, which is considered
acceptable compared to other approaches and accuracy of experimental reference values. Other
methods for estimating K
OC
based on the quantum mechanical descriptors, molecular surface
area, characteristic root indices (CRI), linear solvation energy relationships (LSER) and
chromatographic retention times are considered less useful with respect to either performance
requirements or from a practical viewpoint and are not discussed further.
Air-water partitioning. Dearden and Schuurmann (2003) summarised some studies that all point
out the great deciency of experimentally determined Henrys law constants (HLCs) for
environmentally relevant substances, which means that QSAR models are of great relevance. Two
recent reviews cover methods and models for calculating HLCs (Staudinger and Roberts, 1996;
Dearden and Schuurmann, 2003). HLCs can theoretically be approximated by the ratio of vapour
pressure and aqueous solubility but the use of experimental values in this relation can hardly be
considered predictive (Dearden and Schuurmann, 2003), due to uncertainties in the experimental
values and non-ideal behaviour of the substance. Methods for measuring non-ideality (activity
coefcients) experimentally exist but also suffer from poor accuracy for many environmentally
relevant substances (Staudinger and Roberts, 1996). Approaches that do not use experimental
data as descriptors are interesting from a practical point of view, since good quality experimental
values of physical properties are often unavailable.
HenryWin is a part of the EpiSuite software (footnote 10) that can estimate HLCs. Two methods
are used, based on bond (Meylan and Howard, 1991) and group contributions respectively.
Validation of the bond method with a test set of 1650 substances gave an RMSEP of 1.43 log
units but a lower RMSEP of 0.79 log units was reported for substances with higher volatility than
what is considered non-volatile from water, which is much more relevant for the practical use of
the model. The group method does not usually perform as well as the bond method. An independent
comparative study of several methods and models for a data set of 700 substances (Dearden and
Schuurmann, 2003) indicated that HenryWin was the best performing of the 12 tested methods
with a standard error of 1.03 log units. Also Staudinger and Roberts (1996) recommend the two
methods implemented in HenryWin among available general (diverse) QSARs. Poor performance
for large molecules is noted in the EU TGD (ECB, 2003) but HenryWin is still the recommended
QSAR model for air-water partitioning in the TGD.
Water solubility. Water solubility S
W
is relevant for the partitioning of substances in the
environment and required as input to most fate models. Temperature effects on S
W
are usually not
accounted for by QSAR models available today, which is a limitation in their applicability and an
area for further research. There are three major groups of models for prediction of S
W
: (1) based
on measured physicochemical properties (most importantly K
OW
, MP, BP and molar volume), (2)
based on group contributions and (3) based on descriptors calculated from molecular structure.
WsKowWin is a part of the EpiSuite software that can estimate water solubility based on log K
OW
,
molecular weight and (if available) melting point (Meylan et al., 1996). Validation with 817 test
substances gave an RMSEP of 0.62 log units and an absolute mean error of 0.48 log units.
WaterNT was recently added to the EpiSuite software. It uses a fragment contribution method
and correction factors to estimate water solubility. The standard deviation of the predictions for
training data is 0.34 log units but external validation with nearly 4000 substances gave an
RMSEP of 0.87 log units.
The freely available SPARC model (footnote 13) uses linear free energy relationships to predict S
W
.
A recent but limited comparison (Letinski et al., 2002) indicates that SPARC performs better than
WsKowWin. Also ALOGPS (footnote 14) is freely available over the internet and uses
electrotopological state indices and an articial neural network model to estimate S
W
for
substances based on their 2D structure
16
(Tetko et al., 2001). Validation with an independent test
set gave RMSEP of 0.63 log units.
QikProp (Jorgensen and Duffy, 2002) is designed for drug discovery purposes and can predict S
W

with RMSEP of 0.70 log units. QikProp requires structure optimisation (e.g. by molecular mechanics)
to give accurate results. The OECD report from 1993 recommends a method by Isnard and Lambert
(1989) with RMSEC=0.58 but with no external validation. Numerous other methods for prediction
of S
W
exist of which some are summarised in Katritzky et al. (2000). Most of them are not validated
well enough to assess their performance for structurally diverse sets of new substances.
Melting point. Melting points are normally not of interest by themselves in determining the
environmental fate of chemical substances but are used as an input in many fate models due to
their relation to solubilities, boiling point and vapour pressure. Melting point can be expected to
be a difcult property to model due to the complex interactions present in the solid state. On the
other hand, they are usually quite easy to measure accurately experimentally, so the interest for
QSARs with this end-point is probably smaller than for other properties. However, there is a need
for MP models since many environmentally interesting compounds lack experimental MPs
(Dearden, 2003).
Most published QSARs for melting points are only valid for a limited set of substances (Katritzky et al.,
1997; Katritzky et al., 2000; Dearden, 2003). Models for more diverse sets are of greater interest in
the current context but, as noted below for MpBpWin, model performance is often disappointing, e.g.
with RMSECs of 30 K for a model covering mono and disubstituted benzenes (Katritzky et al., 1997).
Dearden (2003) reviewed recent QSAR models and SW packages for MP prediction. The three SW
packages tested (MpBpVp, ChemOfce, Propred) had very similar performances with RMSEP between
25 and 27 K for 100 small organic test substances. MpBpWin is a part of the EpiSuite software and
can estimate melting points with two different methods: the Joback method (Joback, 1984) and the
Gold and Ogle method, which is a simple relation between melting point and boiling point. The
program uses a weighted average on these two with weights depending on the type of structure.
The accuracy of the MpBpWin estimate was tested on a 666 compound data set containing a
diverse mix of compounds. The results were Q
2
=0.73 and RMSEC=59 K. The accuracy was further
tested on a data set of 1379 different compounds including many complex pharmaceutical
chemicals and pesticides. MpBpWin estimates yield Q
2
=0.71 and RMSEP=58 K, which is very
similar to the results for the 666 compound data set. Thus, MpBpWin estimates of melting point
are at best useful for screening; a prediction standard deviation of 60 K cannot be considered a
quantitative estimate.
16
given as a SMILES string.
QSAR estimated melting points can only be recommended for screening purposes (at best), which
is also noted by the OECD (OECD, 1993a).
Boiling point. MpBpVp calculates boiling points using an adapted Stein and Brown method
(Stein and Brown, 1994) based on group contributions with a correction term. The Stein and
Brown method was derived from a training set of 4426 diverse organic compounds with a standard
deviation of 24.6 K. It was then validated on an external test data set of 6584 compounds with
the following statistical accuracy: average absolute error = 20.4 K; standard deviation = 38.1 K;
average error = 4.3%. The extensive and diverse validation set makes these performance estimates
reliable for predictions of new compounds.
Dearden (2003) reviewed and tested 6 software packages that can be used for BP prediction and
are available commercially or free of charge. The test was done using 100 small organic substances.
MpBpVp was one of the packages tested and the resulting average absolute prediction error was
13.8 K, which is lower than the number obtained from the larger and more diverse data set. SPARC
produced an average prediction error of 6.3 K for the same (limited) data set, which indicates
superior performance. No study showing the applicability for more diverse sets have been found.
A number of models for boiling point estimates have been developed by Jurs and co-workers
(Egolf et al., 1994; Wessel and Jurs, 1995). The models have, in general, good performance in
their applicability domain with RMSEC values around 12 K for rather large diverse data sets. The
good performance originates partly from the use of so-called charged partial surface area (CSPA)
descriptors, which are 3D descriptors that require specic software for calculation. The models are
thus not straightforwardly applied today. Several other models based on linear QSAR are reviewed
in Katritzky et al. (2000) and Dearden (2003) but the validation of these models is not satisfactory
or not described in enough detail to estimate their reliability.
Non-linear models, primary neural network based models, have received increasing attention.
These types of models seem promising and have lower RMSECs (Katritzky et al., 2000; Dearden,
2003) but as with the linear models the validation is not sufcient. Further validation studies of
existing models to determine their precision and domain of applicability are needed.
Vapour pressure. Vapour pressure (VP) is required as an input to fugacity calculations and can be
used (together with S
W
) to estimate HLC. VP is difcult to determine accurately experimentally for
low boiling compounds. Katritzky et al. (2000) report some different QSAR models for VP. The
models based on diverse sets of chemicals (Liang and Gallagher, 1998; Katritzky et al., 1998) are
most relevant and have prediction errors of 0.3-0.5 log units. The models use quantum chemical
descriptors which means that quantum mechanical modelling software is needed to use them.
The models are not readily available as software packages.
QSAR methods and software packages for VP estimation were also recently summarised by
Dearden (2003). Most models discussed are valid for diverse substances and show relatively high
accuracy (estimated errors around 0.4 log units, which is considered good given the large span in
VPs involved). Four software packages, available commercially or free of charge, were tested using
100 small organic substances. One of the reviewed software packages is MpBpVp. MpBpVp can
estimate vapour pressures using different methods: Antoine (Lyman et al., 1990) and modied
Grain (Lyman, 1985). For solids, the modied Grain estimate is the suggested VP. For liquids and
gases, the suggested VP is the average of the Antoine and the modied Grain estimates. Boiling
and melting points are needed and QSAR estimates of these are used if experimental values are
not available. MpBpVp showed an average absolute prediction error of 0.285 log units while
SPARC produced an average prediction error of 0.105 log units.
In the original work, the MpBpVp VP predictions were validated on an 805 compound data set.
In the validation, all the boiling and melting points used for the VP estimates were estimated by
MpBpWin to introduce the maximum amount of error that would be encountered during normal
use. The validation gave Q
2
= 0.94 and RMSEP = 0.72 log units, which is higher than the errors
reported by Dearden (2003) for the more limited data set. It is likely that much of the error is
introduced by the known poor estimates of boiling and melting points. The same problem, as
discussed for S
W
and HLC, is that only models giving predictions valid at 25C are available,
applies to most VP QSAR models (Dearden, 2003).
9.7 Conclusion and Summary
Diverse and class-specic models
When applying QSAR models for toxicity end-points there are two basic ways to approach the
problem:
1. To develop QSARs that are valid for a specic group of chemicals or a specic mode of toxic
action. This approach makes modelling easier (e.g. by Hansch type models for non-polar
narcosis) but requires that substances are classied before the correct model can be used,
which is not always straightforward since a substance can belong to several or none of the
classes for which QSARs are available. It has been noted that misclassication of substances
can give very large prediction errors (Devillers, 2001).
2. To develop QSARs for diverse sets of chemicals that accounts for multiple modes of action
and several classes of chemicals. This makes modelling more complex, since several
mechanisms have to be modelled. Descriptors that contain information about properties
important for several modes of action are needed, which means that structural descriptors
rather than surrogate physicochemical parameters are often used. It is also important to note
that since no model can cover the full chemical space prediction outlier diagnostics are of
great importance. Successful models based on this approach have been presented above, e.g.
(Kaiser and Niculescu, 1999; Kaiser and Niculescu, 2001; Kaiser et al., 2002; Klopman et al.,
2000) and a comparative study for Tetrahymena pyriformis has been conducted (Ren, 2003).
It is clear that recent development indicates a change of focus from specic (type 1) models to
general models valid for diverse sets of substances (type 2). This development is likely to continue.
This is true also for QSARs for non-biological end-points (Katritzky et al., 2000; ECETOC, 2003). If
mode of action based toxicity models (type 1) are used, they should be accompanied by an
automatic classication system that automatically chooses the correct model. An expert system
of this type is implemented in ASTER and OASIS. ECOSAR also has an automatic model selection
based on chemical class but this system has been criticised for not indicating if it encounters
unknown substructures (ECETOC, 2003).
Applicability domain of QSARs
It is of increasing concern that models are applied outside their domain, i.e. for substances that
are not within the chemical space of the model (ECETOC, 1989; Cronin, 2002), which leads to
poor predictions of the model end-points. This is not surprising from a statistical perspective and
the EU TGD on risk assessment (ECB, 2003) states that the exact domain of denition of the
model should be stated, which can be straightforward for class-specic QSARs. For QSARs
developed using diverse sets of chemicals with the aim of as broad an applicability as possible, it
is however usually not straightforward to determine the valid model domain. It is clear that tools
to detect when the substance in question is outside the models domain are needed. The usefulness
of prediction diagnostics in this respect are clearly indicated in the studies by Furusj et al. (2003;
2006) and by the evaluation results for TOPKAT by Moore et al. (2003). Such tools also highlight
cases where no valid model exists for a substance. The fact that no prediction can be given for a
substance because no model is available must be accepted if this is the case.
Methodology for prediction diagnostics is dependent on the method used for the QSAR in
question. Implementation of prediction diagnostics in new and existing models is a priority area
for further research. It would increase the reliability of QSAR predictions signicantly and promote
both use and acceptance of QSAR models. The importance of further research in this area has
been pointed out previously for risk assessment (Eriksson et al., 2003; ECETOC, 2003).
Model validation framework
Independent assessment of model quality is another important question to increase usefulness
and acceptance of QSAR methodology. Transparent and harmonised principles are necessary.
Work in this area is being carried out by the OECD and ECVAM (Worth et al., 2004). QSAR
developers are encouraged to submit their models to the JRC for independent validation, which
will be carried out within the framework of the ECVAM validation and peer review process. QSARs
should be developed and documented in accordance with a minimal set of principles, referred to
as QSAR development and validation principles.
Reference data availability
With the evolution in computational methodology and capacity that have made 3D molecular
calculations feasible on desktop computers, the fundamental challenge is to build the required
toxic effect databases that can be used for modelling. It is a common understanding that a main
factor limiting further QSAR development is the lack of data needed to build new models (Cronin,
2002; ECETOC, 2003; Russom et al., 2003; Bradbury et al., 2003). Important points to consider
are uncertainty estimates in reference data and testing according to standardised protocols.
Toxicity
Effect and no-effect concentrations. The availability of QSARs for chronic predicted no-effect
concentrations (NOEC) is very limited. In the absence of measured NOECs, PNECs are frequently
calculated by using an assessment factor and predicted acute effect concentrations. In risk
assessment within the EU, an assessment factor of 1000 is used to reect the large uncertainty in
extrapolating acute effect data to no-effect concentrations. If chronic test data are available,
lower assessment factors can be used (ECB, 2003).
The safety margin inherent in the assessment factors is adequate for risk assessment where conservative
estimates are needed. If likely rather than conservative estimates of PNEC are needed as will be the
case for at least some applications related to the WFD this procedure is not efcient. In such cases,
derivation of PNECs from acute or chronic effect concentrations is not straightforward.
There is considerable incentive to develop QSARs for aquatic NOEC and/or PNEC but the work in
this eld has only started. The ECETOC report on QSARs (ECETOC, 2003) discusses QSARs for
PNEC and lists a number of possible advantages compared with current practice and recommends
further research. QSAR for setting water quality objectives is a related issue discussed by Vighi et
al. (2001). NOEC and/or PNEC model development is a priority eld for further research of QSARs
with applicability in the WFD.
End-point gaps. As noted in this study and by others, e.g. Schultz et al. (2003), there are gaps in
QSAR toxicity model availability, both with respect to chemical substance groups within some
end-points and with respect to availability of models for certain end-points. Few QSARs available
are valid for non-organic chemicals (Comber et al., 2003) and also the validity for metallo-organic
chemicals is limited.
Most of the QSAR work has been devoted to freshwater organisms. There are very few QSARs for
toxicity to estuarine and marine organisms and their quality is generally lower than for freshwater
organisms (ECETOC, 2003, Comber et al., 2003). The same is true for sediment-dwelling organisms.
The need for development of new QSARs here is evident as is for example noted in ECETOC
(1989). QSARs cannot be used for estimation of chronic toxic effects in an aquatic environment
and toxicity to sediment-living organisms today. A signicant obstacle to further development in
these elds is the lack of databases with reference toxicity data.
Approximately 70% of all industrial chemicals are estimated to act acutely via non-polar or polar
narcosis. These mechanisms are also the ones for which most models exist. Non-polar narcosis can
in general be modelled based on log K
OW
only for individual groups of chemicals. Naturally, for
models covering diverse sets of chemicals, more descriptors are necessary. A remaining challenge
is to model acute and chronic effects of chemicals whose toxicity is elicited through other
mechanisms (covalent binding, receptor interaction etc.; Bradbury et al., 2003).
Antagonistic and synergistic effects. All methods discussed in this report focus on discrete
chemicals and to not account for additive, synergistic or antagonistic effects. This is a signicant
problem for practical use of QSARs in a WFD perspective.
Mode of action based QSARs, however, open a possibility in this direction by allowing the mode
of action to be identied. For compounds acting via the narcosis mode of action, concentration-
addition, a response is often exhibited but for reactive chemicals this is only true for chemicals
having the same target site (Escher and Hermens, 2002).
Some initial QSAR work in this direction has been carried out by Vighi et al. (2003). Further
testing of this approach would be an interesting area for further research.
Environmental fate
Fate models and the WFD. Mackay et al. (2003) suggest that back-tracking of models should be
used for risk assessment of chemicals in cases where emissions are uncertain or unknown. The
approach is also potentially useful in the context of the WFD. The load (emission) estimates are
often very uncertain but if a PNEC (or similar) concentration from a QSAR model can be used as
a starting point for reverse calculation by a fugacity model. The calculation (which will probably
be of a trial-and-error type) produces an estimate of the critical load that would give an effect on
the organism for which the PNEC is valid. It is then usually feasible to see if the actual loads are
of this magnitude or much smaller. This enables an investigation of which substances that could
be responsible for an observed ecological effect, even in the absence of analytical data (i.e.
environmental concentrations). It could also indicate how much the loads have to be decreased
to lower the concentration of the substance below the PNEC.
End-point gaps. For vapour pressure and water solubility there exists available and well-validated
models covering diverse chemicals that can be used to predict these properties. As noted above,
QSAR estimated melting points can only be recommended for screening purposes (at best). This
failure derives, in part, from ignoring the effects of symmetry in the molecule (Lyman, 1985).
Boiling points are better predicted by existing models but one should be aware of the magnitude
of the error when using these estimates further.
The water-octanol partition coefcient, log K
OW
, is the most studied of all QSAR end points and a
number of useful models implemented in software exist. Log K
OW
is also the most common descriptor
for QSARs for other end-points, e.g. bioaccumulation, and sediment-water partitioning. Water-air
partitioning coefcients, most commonly Henrys law constants, are predicted with lower relative
accuracy but available models still give acceptable accuracy for many applications.
Comber et al. (2003) point out the need for improvement of existing models for soil-water
partitioning and sediment-water partitioning. The most likely area to further increase the
usefulness of QSARs for soil and sediment sorption is not through improved QSAR models for
organic carbon normalised partition coefcients, K
OC
, but through improved understanding of the
sorption process and inclusion of other sediment properties than organic carbon content. Abiotic
aquatic degradation (photolysis, hydrolysis) is poorly covered by existing models as noted in this
study and by others (Comber et al., 2003; ECETOC, 2003).
The QSARs for biodegradation are primarily of classication type while the required input to fate
models, e.g. EUSES, is often quantitative. Thus there is a need for new QSARs for biodegradation
half-lives as noted in this study and by others (Comber et al., 2003; ECETOC, 2003). It should be
noted however that several factors, e.g. the multitude of degradation mechanisms and inaccurate
reference data, could be expected to make QSAR modelling difcult for this end-point. Bioconcentration
factors can be approximately predicted by QSAR models. The need to account for metabolism is
evident and is a weak point of existing models. There are no models available for bioaccumulation.
9.8 References
Boethling, R.S., P.H. Howard, W.M. Meylan, W.M. Stiteler, J. Beaumann, and N. Tirado. 1994. Group
contribution method for predicting probability and rate of aerobic biodegradation. Environmental Science
and Technology 28:459-465
Boethling, R.S., D.G. Lynch, J.S. Jaworska, J.L. Tunkel, G.C. Thom, and S. Webb. 2004. Using Biowin, Bayes,
and batteries to predict ready biodegradability. Environmental Toxicology and Chemistry 23:911-920
Bradbury, S.P., C.L. Russom, G.T. Ankley, T.W. Schultz, and J.D. Walker. 2003. Overview of data and conceptual
approaches for derivation of quantitative structure-activity relationships for ecotoxicological effects of
organic chemicals. Environmental Toxicology and Chemistry 22:1789-1798
Canonica, S., and P.G. Tratnyek. 2003. Quantitative structure-activity relationships for oxidation reactions of
organic chemicals in water. Environmental Toxicology and Chemistry 22:1743-1754
Clements, R.G., J.V. Nabholz, and M. Zeeman, 1996. Estimating toxicity of industrial chemicals to aquatic
organisms using structure-activity relationships - ECOSAR Technical documentation, U.S. Environmental
Protection Agency, Washington, DC
Comber, M.H.I., J.D. Walker, C. Watts, and J. Hermens. 2003. Quantitative structure-activity relationships for
predicting potential ecological hazard of organic chemicals for use in regulatory risk assessments.
Cronin, M.T. 2002. The current status and future applicability of quantitative structure-activity relationships
(QSARs) in predicting toxicity. Atla-Alternatives to Laboratory Animals 30: 81-84
Cronin, M.T., J.D. Walker, J.S. Jaworska, M.H.I. Comber, C.D. Watts, and A.P. Worth. 2003. Use of QSARs in
international decision-making frameworks to predict ecologic effects and environmental fate of chemical
substances. Environmental Health Perspectives 111:1376-1390
de Zwart, D. 2001. Observed Regularities in SSDs for Aquatic Species. In: L. Posthuma, G.W. Suter, and T.P.
Traas (Eds.), Species sensitivity distributions in ecotoxicology. CRC Press, Boca Raton, FL, pp. 133154
Dearden, J.C. 2002. Prediction of environmental toxicity and fate using quantitative structure-activity
relationships (QSARs). Journal of the Brazilian Chemical Society. 13:754-762
Dearden, J.C. 2003. Quantitative structure-property relationships for prediction of boiling point, vapor
pressure, and melting point. Environmental Toxicology and Chemistry 22:1696-1709
Dearden, J.C., and G. Schuurmann. 2003. Quantitative structure-property relationships for predicting Henrys
law constant from molecular structure. Environmental Toxicology and Chemistry 22:1755-1770
Devillers, J. 2001. QSAR modeling of large heterogeneous sets of molecules. Sar and Qsar in Environmental
Research 12:515-528
Dimitrov, SD., N.C. Dimitrova, J.D. Walker, G.D. Veith, and O.G. Mekenyan. 2002. Predicting bioconcentration
factors of highly hydrophobic chemicals. Effects of molecular size. Pure and Applied Chemistry 74:1823-1830
Doucette, W.J. 2003. Quantitative structure-activity relationships for predicting soil-sediment sorption
coefcients for organic chemicals. Environmental Toxicology and Chemistry 22:1771-1788
EC, European Commission. 2001. Decision of the european parliament and of the council: establishing the
list of priority substances in the eld of water policy, COM(2001) 17 nal, European Commision, Brussels
ECB, European Chemicals Bureau. 2003. Technical Guidance Document (TGD) on risk assessment of chemical
substances following European regulations and directives, 2nd edition ed.; European Chemicals Bureau, Ispra
ECETOC, 1989. QSARs in the Assessment of the Environmental Fate and Effects of Chemicals. Technical
report no 74, European Centre for Ecotoxicology and Toxicology of Chemicals, Brussels
ECETOC, 2003. (Q)SARs: Evaulation of the commercially available software for human health and
environmental endpoints with respect to chemical management applications, Technical report no 89,
European Centre for Ecotoxicology and Toxicology of Chemicals, Brussels
ECVAM, 2004. European Centre for the Validation of Alternative Methods, http://ecvam.jrc.it
Egolf, L.M., M.D. Wessel, and P.C. Jurs. 1994. Prediction of boiling points and critical-temperatures of
industrially important organic-compounds from molecular-structure. Journal of Chemical Information and
Computer Sciences 34:947-956
Eldred, D.V., C.L. Weikel, P.C. Jurs, and K.L. Kaiser. 1999. Prediction of fathead minnow acute toxicity of
organic compounds from molecular structure. Chemical Research in Toxicology, 12:670-678
Eriksson, L., J. Jaworska, A.P. Worth, M.T. Cronin, R.M. McDowell, and P. Gramatica. 2003. Methods for
reliability and uncertainty assessment and for applicability evaluations of classication- and regression-
based QSARs. Environmental Health Perspectives 111:1361-75
Eros, D., I. Kovesdi, L. Or, K. Takacs-Novak, G. Acsady, and G. Keri. 2002. Reliability of logP predictions based
on calculated molecular descriptors: a critical review. Current Medicinal Chemistry 9:1819-1829
Escher, B.I., and J.L.M. Hermens. 2002. Modes of action in ecotoxicology: their role in body burdens, species
sensitivity, QSARs, and mixture effects. Environmental Science and Technology 36:4201-4217
Furusj, E., M. Andersson, M. Rahmberg, and A. Svenson. 2003. Estimating environmentally important
properties of chemicals from the chemical structure, B1517, IVL Swedish Environmental Research Institute,
Stockholm
Furusj, E., M. Andersson, M. Rahmberg, and A. Svenson. 2006. The importance of outlier detection and
training set selection for reliable environmental QSAR predictions. Chemosphere 63:98-108
Gombar, V.K., and K. Enslein. 1996. Assessment of n-octanol/water partition coefcient: When is the
assessment reliable? Journal of Chemical Information and Computer Sciences 36:1127-1134
Howard, P. H., R.S. Boethling, W.M. Stiteler, W.M. Meylan, A.E. Hueber, J.A. Beauman, and M.E. Larosche.
1992. Predictive model for aerobic biodegradability developed from a le of evaluated biodegradation data.
Isnard, P., and S. Lambert. 1989. Aqueous solubility and normal-octanol water partition-coefcient
correlations. Chemosphere 18:1837-1853
Jaworska, J., S. Dimitrov, N. Nikolova, P. Masscheleyn, and O. Mekenyan. 2002. Probabilistic assessment of
biodegradability based on metabolic pathways: Catabol system. SAR and QSAR in Environmental Research
13:307-323
Joback, K.G. 1984. SM Thesis in Chemical Engineering, MIT, Cambridge, MA, 6/1984
Jorgensen, W.L., and E.M. Duffy. 2002. Prediction of drug solubility from structure. Advanced Drug Delivery
Reviews 54:355-366
Kaiser, K.L. and S.P. Niculescu. 1999. Using probabilistic neural networks to model the toxicity of chemicals to
the fathead minnow (Pimephales promelas): a study based on 865 compounds. Chemosphere 38:3237-3245
Kaiser, Klaus L.E., J.C. Dearden, W. Klein, and T.W. Schultz. 1999. A Note of Caution to Users of ECOSAR.
Water Quality Research Journal of Canada 34:179-182.
Kaiser, K.L., and S.P. Niculescu, 2001. Modeling acute toxicity of chemicals to Daphnia magna: A probabilistic
neural network approach. Environmental Toxicology and Chemistry 20:420-431
Kaiser, K.L., S.P. Niculescu, and T.W. Schultz. 2002. Probabilistic neural network modeling of the toxicity of
chemicals to Tetrahymena pyriformis with molecular fragment descriptors. SAR and QSAR in Environmental
Research 13:57-67
Katritzky, A.R., U. Maran, M. Karelson, and V.S. Lobanov. 1997. Prediction of melting points for the substituted
benzenes: A QSPR approach. Journal of Chemical Information and Computer Sciences 37:913-919
Katritzky, A.R., Y. Wang, S. Sild, and T. Tamm. 1998. QSPR studies on vapor pressure, aqueous solubility, and
the prediction of water-air partition coefcients. Journal of Chemical Information and Computer Sciences
38:720-725
Katritzky, A.R., U. Maran, V. S. Lobanov, and M. Karelson. 2000. Structurally diverse quantitative structure-
property relationship correlations of technologically relevant physical properties. Journal of Chemical
Information and Computer Sciences 40:1-18
Klein, W., S. Denzer, M. Herrchen, P. Lepper, M. Mller, R. Sehrt, A. Storm, and J. Volmer. 1999. Revised
proposal for a list of priority substances in the context of the Water Framework Directive (COMMPS procedure),
98/788/3040/DEB/E1, Fraunhofer-Institut, Schmallenberg
Klopman, G., Z. Zhang, D.M. Balthasar, and H.S. Rosenkranz. 1995. Computer-automated predictions of
aerobic biodegradation of chemicals. Environmental Toxicology and Chemistry 14:395-403
Klopman, G., R. Saiakhov, and H.S. Rosenkranz. 2000. Multiple computer-automated structure evaluation
study of aquatic toxicity II. Fathead minnow. Environmental Toxicology and Chemistry 19:441-447
Letinski, D.J., M.J. Connelly, D.R. Peterson, and T.F. Parkerton. 2002. Slow-stir water solubility measurements
of selected alcohols and diesters. Chemosphere 48:257-265
Liang, C., and D.A.Gallagher. 1998. QSPR prediction of vapor pressure from solely theoretically-derived
descriptors. Journal of Chemical Information and Computer Sciences 38:321-324
Lyman W.J. 1985. Environmental exposure from chemicals. In: W.B. Neely, and G.E. Blau (Eds.), Environmental
exposure from chemicals. Vol.1. CRC Press, Boca Raton
Lyman, W. J., Reehl, W.F. and Rosenblatt, D.H. 1990. Handbook of Chemical Property Estimation Methods.
American Chemical Society, Washington, DC, 1990
Mackay, D., J. Hubbarde, and E. Webster. 2003. The role of QSARs and fate models in chemical hazard and
risk assessment. Paper prepared for Quantitative Structure-Activity Relationships (QSAR). Proceedings of the
QSAR 2002 Conference, Ottawa May 2002. Qsar & Combinatorial Science 22:106-112
Mackay, D., and E. Webster. 2003. A perspective on environmental models and QSARs. Sar and Qsar in
Environmental Research 14:7-16
Meylan, W.M. and P.H. Howard, 1991. Bond contribution method for estimating Henrys law constants.
Meylan, W.M., P.H. Howard, and R.S. Boethling. 1992. Molecular topology/fragment contribution method
for predicting soil sorption coefcients. Environmental Science and Technology 26:1560-1567
Meylan, W.M. and P.H. Howard. 1995. Atom/fragment contribution method for estimating octanol-water
partition coefcients. Journal of Pharmaceutical Sciences 84: 83-92
Meylan, W.M., P.H. Howard, and R.S. Boethling. 1996. Improved method for estimating water solubility from
octanol water partition coefcient. Environmental Toxicology and Chemistry 15:100-106
Meylan, W.M., P.H. Howard R.S. Boethling, D. Aronson, H. Printup, and S. Gouchie. 1999. Improved method
for estimating bioconcentration/bioaccumulation factor from octanol/water partition coefcient.
Environmental Toxicology and Chemistry 18:664672
Mill, T. 2000. Photoreactions in Surface Waters. In: Boethling, R. S. and D. Mackay (Eds.), Handbook of
property estimation methods for chemicals. Lewis Publishers, Boca Raton, FL, pp. 355-381
Moore, D.R.J., R.L. Breton, and D.B. MacDonald. 2003. A comparison of model performance for six quantitative
structure-activity relationship packages that predict acute toxicity to sh. Environmental Toxicology and
Chemistry 22:1799-1809
OECD, Organisation of Economic Co-operation and Development. 1993a. Application of structure-activity
relationships to the estimation of properties important in exposure assessment. OCDE/GD(93)125, Paris
OECD, Organisation of Economic Co-operation and Development. 1993b. Structure-activity relationships for
biodegradation. OCDE/GD(93)126, Paris
OECD, Organisation of Economic Co-operation and Development. 1994. U.S. EPA/EC joint project on the
evaluation of (quantitative) structure activity relationships, OECD/GD/(94) 28, Paris
OECD, Organisation of Economic Co-operation and Development. 2003. Manual for Investigation of HPV
Chemicals; OECD, Paris
OSPAR, 2000. OSPAR Recommendation 2000/5 on a Harmonised Offshore Chemical Notication Format
(HOCNF), OSPAR, London
Parkerton, T.F., and W.J. Konkel. 2000. Application of quantitative structure-activity relationships for assessing
the aquatic toxicity of phthalate esters. Ecotoxicology and Environmental Safety 45:61-78
Perkins, R., H. Fang, W.D. Tong, and W.J. Welsh. 2003. Quantitative structure-activity relationship methods:
Perspectives on drug discovery and toxicology. Environmental Toxicology and Chemistry 22:1666-1679
Ren, S. 2003. Ecotoxicity prediction using mechanism- and non-mechanism-based QSARs: a preliminary
study. Chemosphere 53:1053-1065
Roex, E.W.M., C.A.M. van Gestel, A.P. van Wenzel, and N.M.van Straalen. 2000. Ratios between acute aquatic
toxicity and effects on population growth rates in relation to toxicant mode of action. Environmental
Toxicology and Chemistry 19:685-693
Russom, C.L., S. Bradbury, S.J. Broderius, D.E. Hammermeister, and R.A. Drummond. 1997. Predicting modes
of toxic action from chemical structure: Acute toxicity in the fathead minnow (Pimephales promelas)
Russom, C.L., R.L. Breton, J.D. Walker, and S.P. Bradbury. 2003. An overview of the use of quantitative
structure-activity relationships for ranking and prioritizing large chemical inventories for environmental risk
assessments. Environmental Toxicology and Chemistry 22:1810-1821
Sabljic, A., H. Gsten, H. Verhaar, and J. Hermens. 1995. QSAR modeling of soil sorption - improvements and
systematics of log k-oc vs log k-ow correlations. Chemosphere 31:4489-4514
Saloranta, T.M., J. Kamari, S. Rekolainen, and O. Malve. 2003. Benchmark criteria: A tool for selecting
appropriate models in the eld of water management. Environmental Management 32:322-333
Schmieder, P.K., G. Ankley, O. Mekenyan, J.D. Walker, and S. Bradbury. 2003. Quantitative structure-activity
relationship models for prediction of estrogen receptor binding afnity of structurally diverse chemicals.
Schultz, T.W., M.T. Cronin, J.D. Walker, and A.O. Aptula. 2003. Quantitative structure-activity relationships
(QSARs) in toxicology: a historical perspective. Journal of Molecular Structure-Theochem 622:1-22
Stein, S.E., R.L. Brown. 1994. Estimation of normal boiling points from group contributions. Journal of
Chemical Information and Computer Sciences 34:581-587
Tao, S., C. Liu, R. Dawson, J. Cao, and B. Li. 1999. Uptake of particulate lead via the gills of sh (Carassius
auratus). Archives of Environmental Contamination and Toxicology 37:352-357
Tetko, I.V., V.Y. Tanchuk, T.N. Kasheva, and A.E.P. Villa. 2001. Estimation of aqueous solubility of chemical
compounds using E-state indices. Journal of Chemical Information and Computer Sciences 41:1488-1493
Tetko, I.V., and V.Y. Tanchuk, 2002. Application of associative neural networks for prediction of lipophilicity
in ALOGPS 2.1 program. Journal of Chemical Information and Computer Sciences 42:1136-1145
Tratnyek, P.G., E. J. Weber, and R.P. Schwarzenbach. 2003. Quantitative structure-activity relationships for
chemical reductions of organic contaminants. Environmental Toxicology and Chemistry 22:1733-1742
Tunkel, J., P.H. Howard, R.S. Boethling, W. Stiteler, and H. Loonen. 2000. Predicting ready biodegradability in
the Japanese Ministry of International Trade and Industry test. Environmental Toxicology and Chemistry
19:2478-2485
Vanleeuwen, C.J., P.T.J. Vanderzandt, T. Aldenberg, H.J.M. Verhaar, and J.L.M. Hermens.1992. Application of
Qsars, Extrapolation and Equilibrium Partitioning in Aquatic Effects Assessment.1. Narcotic Industrial
Pollutants. Environmental Toxicology and Chemistry 11:267-282
Verhaar, H.J.M., W. Mulder, and J.L.M. Hermens. 1995. QSARs for ecotoxicity. In: Hermens J.L.M. (Ed.),
Overview of structure-activity relationships for environmental endpoints, Part 1: General outline and
procedure. Report prepared within the framework of the project QSAR for prediction of fate and effects of
chemicals in the Environment, an international pro-ject of the Environmental Technologies RTD Programme
(DGXII/D-1) of the European Commission under contract no. EV5V-CT92-0211
Verhaar, H.J.M., Ramos, E.U. and Hermens J.L.M.1996. Classifying environmental pollutants. 2. Separation of
class 1 (baseline toxicity) and class 2 (polar narcosis) type compounds based on chemical descriptors.
Journal of Chemometrics 10:149-162
Vighi, M., P. Gramatica, F. Consolaro, and R. Todeschini. 2001. QSAR and chemometric approaches for setting
water quality objectives for dangerous chemicals. Ecotoxicology and Environmental Safety 49:206-220
Vighi, M., R. Altenburger, A. Arrhenius, T. Backhaus, W. Bodeker, H. Blanck, F. Consolaro, M. Faust, A. Finizio,
K. Froehner, P. Gramatica, L. H. Grimme, F. Gronvall, V. Hamer, M. Scholze, and H. Walter. 2003. Water quality
objectives for mixtures of toxic chemicals: problems and perspectives. Ecotoxicology and Environmental
Safety 54:139-150
Walker, J.D. 2003. Applications of QSARs in toxicology: a US government perspective Journal of Molecular
Structure-Theochem 622:167-184
Wessel, M.D. and P.C. Jurs. 1995. Prediction of normal boiling points for a diverse set of industrially important
organic-compounds from molecular-structure. Journal of Chemical Information and Computer Sciences
35:841-850
Worth, A.P., M.T.D. Cronin, and C.J. van Leeuwen. 2004. A framework for promoting the acceptance and
regulatory use of (quantitative) structure-activity relationships. In: M.T.D. Cronin, and D.J. Livingstone (Eds.),
Predicting chemical toxicity and fate. CRC Press: Boca Raton FL, USA, pp. 427-438
10 Advanced Molecular-Based Techniques for
Environmental Toxicology
Teresa Lettieri
10.1 Introduction
The development of new technologies and particularly the genomic revolution has the potential
to completely change the way we address the actions of chemicals and other environmental
stressors on biological systems. To study the effects of environmental contaminants the quite
general concept of biomarkers has been extensively used in the past. According to the National
Academy of Sciences of the USA a biomarker is an indicator signaling events in biological
systems or samples. From a classication point of view there are three broad types of molecular
biomarkers in the eld of environmental toxicology: biomarkers of exposure, biomarkers of effect,
and biomarkers of susceptibility.
Biomarkers of exposure are methods that directly quantify the internal concentration of a given
chemical or metabolites and they are usually applied early in the exposure-disease pathway.
These kinds of markers are very useful for epidemiologists because they allow an accurate
correlation between external and internal dose of an environmental agent. Biomarkers of effect
detect functional changes in the biological system under analyses. This second class of biomarkers
is particularly useful to predict the ultimate outcome of exposure; in particular early markers of
effect can be used to identify populations at risk. Of all the individuals exposed to certain
contaminants only a subclass will in the end develop a disease. The individual susceptibility
depends, in a very complex way, on genetic and environmental factors. These factors include
single nucleotide polymorhisms, age, diet, occupation, and lifestyle.
An ideal molecular biomarker should have several characteristics: highly accurate and reproducible,
specic and sensitive, should provide a plausible biological mechanism for the mode of action, should
be based on a high-troughoutput method of measurements. However, not many biomarkers meet all
the above criteria. One of the best examples is biomarker of exposure to aatoxin. In both animal
models and human studies, DNA adduct level correlates with known external dose of aatoxin and
thus supports the strong epidemiological link between adduct level and risk of liver cancer. A urinary
metabolite of the aatoxin DNA adducts is also used as a quantitative biomarker of exposure.
This chapter reviews the use of Real Time Quantitative PCR, and related methods, in the discovery
and testing of new molecular biomarkers. The chapter focuses on the recent application of
genome-based technologies that propose to apply both mRNA and protein expression technologies
to study chemical effects in biological systems. A common theme of all these techniques is the
objective to develop specic signatures, both at the gene or protein expression level, of exposure
to chemicals and environmental stressors in general. The pattern of gene expression or protein
expression changes induced upon exposure to a chemical both in vitro (Burczynski et al., 2000;
Waring et al., 2001) or in vivo (Hamadeh et al., 2002) would form a sort of genetic signature.
The development of such gene or protein expression signatures would allow a fast screening of
unknown or suspected toxicants based on their similarity to known toxicants.
ADVANCED MOLECULAR-BASED TECHNIQUES FOR ENVIRONMENTAL TOXICOLOGY | 193
10.2 Overview of Real Time PCR
Real Time PCR (RT-PCR) or Quantitative Real Time PCR (QRT-PCR) was developed to allow
researchers to study differences, at the quantitative level, in mRNA expression. It is based on
detecting and quantifying the uorescence signal that increases in direct proportion to the
amount of PCR product (or amplicon) in a reaction (Heid et al., 1996). By recording the level of
uorescence emission at each amplication cycle it is possible to monitor a reaction during its
exponential phase, where the replicate samples amplify exponentially and the increase in the
amount of PCR product correlates to the initial amount of target template. The two predominant
methods for the quantitative detection of amplicon use generic DNAbinding agents and
uorescent oligonucleotide probes.
The original detection scheme for the Real Time PCR used an intercalating dye that uoresces
more brightly when bound to double-stranded (ds) DNA binding. Early work in this area used
ethidium bromide which is approximately 40 times brighter when bound to dsDNA than when
free in solution in the PCR mix (Higuchi et al., 1993). More recently, the greater dynamic range of
SYBR-green, which shows little uorescence in solution but has a 200 fold increase in uorescence
intensity upon binding to the dsDNA, has made it the dye of choice (Wittwer et al., 1997).
Although the dyes are not expensive and require only a primer set for the gene of interest, there
are a few caveats to their use because they bind to all (ds) DNA e.g. primer dimer generating
many unspecic products.
The second method for quantitative detection is based on uorescence resonance energy transfer
(Fluorescence Resonance Energy Transfer, FRET) probes that are specically designed for the gene
target, thus overcoming the problem of the intercalating dye. There are several different probe
technologies which involve the specic oligonulceotide hybrization to the specic DNA sequence
and the dually labeling with a fuorochrome and quencher in close proximity. When the probe is
free in solution, the energy released by the uorochrome is absorbed by the quencher. When the
probe binds its target, the uorochrome and quencher are physically separated, which allows light
to be emitted by the uorochrome (Lakowicz, 1999). The FRET systems are the TaqMan assay, the
TaqMan Minor Groove Binding (MGB).
The most popular system, TaqMan assay is shown in Figure 10.1. The system is based on the
5-3exonuclease activity of Amplitaq Gold polymerase that cleaves the labeled probe bound to
the specic target (Holland et al., 1991). As shown in the gure the probe is designed between
the primer and forward, with the uorescence dye at the 5end and the quencher at 3end.
The PCR kinetic is designed such that after melting (denaturation) of the DNA duplex, the probe
will hybridize to the specic target at higher temperature than the PCR primers. Extension from
the Taq polymerase subsequently annealing forward primer is then blocked by the probe.
The 5end of the probe is then replaced and the resulting overhang 5-3- is cleaved by the
polymerase. Consequently the uorescence dye will be physically separated by the quencher and
hence an increase in uorescence directly proportional to the DNA concentration.
Figure 10.2 shows the amplication plot of the yeast housekeeping gene Glyceraldehyde
3-phosphate dehydrogenase (TDH3) (E. Bizzotto and T. Lettieri, unpublished data). In the
amplication the TDH3 gene expression was analysed using the TaqMan MGB probe. The TaqMan
MGB probe is a modication of the linear TaqMan probe (Kumar et al., 1998) which is conjugated
with a Minor Groove Binder (MGB) at 3end, thus allowing a tighter binding to the sequence. In
the reaction the probe was labeled with the uorescence dye FAM (6- carboxyuorescein). The
template, the cDNA (the total RNA reverse transcribed to a DNA copy) was diluted up to a few
micrograms. Considering that the mRNA is just 1-2% of the total RNA, the sensitivity of this
technique allows the detection of just a few copies of the gene. Indeed the RT-PCR is extreme
sensitive, it is able to detect less than ve copies of poorly expressed RNA, reveling very small
changes (Bustin, 2000).
Figure 10.1 TaqMan assay. The RT-PCR reaction exploits the 5 nuclease activity of AmpliTaq Gold DNA
polymerase to cleave a TaqMan probe during the PCR. The Taqman probe contains a reporter dye at 5 end
(R) and a quencher (Q) dye at 3. During the rst step, the primers and the probe anneal at specic target,
and the polymerization starts. During the extension the probe is displaced and cleaved by the exonuclease
activity 5-3. The cleavage of the probe separates the reporter dye from the quencher, resulting in
increased uorescence of the reporter. Accumulation of the PCR products is detected directly by monitoring
the increase in uorescence of the reporter. When the probe is intact, the proximity of the reporter dye to
the quencher results in suppression of the uorescence emission primarily by Foerster-type energy transfer.
Polymerization
Strand displacement
Cleavage
Polymerization
completed
5
5
5
5
5
5
5
5
3
3
3
3
3
3
3
3
5
5
3
5
5
3
5
5
3
5
5
3
Probe
Forward
primer
Reverse
primer
R
R
R
R
Q
Q
Q
Q
10.3 Application of Real Time PCR to Ecotoxicology
In the last few years Real Time PCR has become an invaluable tool for scientists in many disciplines.
In fact its sensitivity, together with the genome sequencing of many organisms, has revolutionised
the eld of molecular diagnostics such as microbiology (Belgrader et al., 1999; Jordan and Durso
2000; Yang et al., 2002), clinical oncology in cancer research (Bernard and Wittwer, 2002), and
gene expression analysis (Bustin, 2000).
In ecotoxicology, RT-PCR has only recently been used to estimate the expression of genes of
interest as molecular indicators diagnostic for exposure to various stressors. Nakanishi et al.,
(2002) described an in vitro system to investigate the effects of trialkytil compound on endocrine
activity in the human placenta using the Real Time PCR. Tributyltin (TBT) is used as biocide and
agricultural fungicide (Boyer, 1989). Most of its toxic effects have been documented as sex organ
alterations in mollusks (Matthiessen and Sumpter, 1998) but in humans no studies have examined
disturbances in sexual development and reproductive function including placental function.
Therefore the authors used the choriocarcinoma cell lines as a placental model to monitor the
mRNA expression level of cytochrome CYP19 and human protein hormone CG by Real Time PCR
upon exposure to TBT in a dose-time dependent manner.
More recently, Islinger et al., (2003) studied the effects of estrogenic compounds in male zebrash
analysing the expression of three estrogenresponse genes by semi-quantitative PCR. Semi-
quantitative PCR is similar to Real Time PCR in that only the linear range of the amplication plots
Figure 10.2 Amplication plot. Exponential amplication of TDH3 amplied using TaqMan MGB probe assay.
The reporter dye of the probe is at 5end FAM (6- carboxyuorescin) and at 3end a quencher dye non-orescent
with a Minor Groove Binder (MGB) which increases the binding efciency. The cDNA template was a dilution
series showing different Ct (cycle threshold indicating the rst detection of the TDH3 PCR product) while Rn
represents the uorescence signal. The PCR reaction was performed in triplicate for each cDNA sample.
Cycle
R
n
2500
2000
1500
1000
5000 E-1
0 5 10 15 20 25 30 35 40
are examined comparing the PCR products only in this range. In order to characterise the hormonal
responsiveness of the three genes: vitellogenin (vtg), estrogen receptor (ER ) and ZP2, adult male
zebrash were exposed to synthetic estrogen 17-ethinylestradiol in a time-dose dependent manner
and the mRNA expression level from liver and gonads was measured. Their study showed Real Time
PCR to be a very sensitive method. Especially for vtg and ER , they were able to detect the presence
of the 17-ethinylestradiol at very low concentration and at short exposure time. In fact the LOEC
(Low Observed Effect Concentration) was 2.5 ng/l for both the biomarkers which lies within the
range of ethinylestradiol concentrations measured in wastewater efuent (Larson et al., 1999).
Galay-Burgos et al., (2003) have described well how Real Time PCR can be applied to assess
exposure to contaminants and how it is possible to differentiate between the effects of different
chemicals. They monitored the expression changes of two selected biomarkers in the earthworm
(Lumbricus rubellus) on control or cadmium-copper spiked soils. They quantied the changes in
the expression of gene encoding for the Methallothionein isoform 2 (MT-2) protein and for the
ribosomal subunit (MLRS) against the internal control, the gene -actin (Kammenga et al., 2000).
The reliability of the assay was demonstrated by the fact that there were no failed reactions of any
of the measured transcripts. The assay reproducibility was also good since they could detect very
low variability from the same transcripts among the experiments.
Rees and Li (2004) developed a RT-PCR assay to monitor the expression changes of CYP1A in the
liver of three salmonid species induced upon -naphtoavone (BNF) exposure. CYP1A or
Cytochromes P4501A constitute a protein family associated with the detoxication of organic
compounds. These compounds induce the CYP1A gene in several tissues of many different shes
(Huuskonen et al., 1998; Levine and Oris 1999). Consequently, the CYP1A changes have been
used as biomarkers for contaminant exposure in sh populations (Cousinou et al., 2000). In the
paper they described the set up of the Real Time PCR assay using the TaqMan probe system. They
designed a primer/probe set to amplify only the CYP1 family and not other P450 (as CYP2 or
CYP3). In addition from the alignment of the known CYP1A sequence from the three salmonid
species, they could identify a conserved sequence to get a highly specic amplication and taking
this advantage to carry the PCR in the same conditions for different species. Furthermore, they
showed the high sensitivity of the assay, measuring CYP1A mRNA expression level down to 1000
transcripts according to the standard curve.
Real Time Quantitative PCR has also been applied to conrm the gene expression changes
obtained by microarray as described by Williams et al., (2003; see Microarray section). They could
compare the expression changes observed in the microarray and conrm the differences giving
condence of the high sensitivity of the two systems.
10.4 Introduction to DNA Microarray technology
Gene expression is a sensitive indicator of toxicant exposure, disease states, and cellular
metabolism and thus it represents a unique way of characterising how cells and organisms adapt
to changes in the external environment. The measurements of gene expression levels upon
exposure to a chemical can be used both to provide information about the mechanism of action
of a toxicant and also to form a sort of genetic signature from the pattern of gene expression
changes it elicits both in vitro (Burczynski et al., 2000; Waring et al., 2001) or in vivo (Hamadeh
et al., 2002). The development of such gene expression signatures would allow a fast screening
of unknown or suspected toxicants based on their similarity to known toxicants.
The development of high quality, commercially available, expression arrays has allowed gene
array technology to become a standard tool in molecular toxicology. With this technology cells or
tissues are exposed to toxicants and then gene expression is measured by collecting mRNA,
converting mRNA to labeled cDNA, hybridizing it to the cDNA array, staining it with an appropriate
dye, and visualizing the hybridized genes using a uorometer (Schena et al., 1995; DeRisi et al.,
1996; Lashkari et al., 1997; see Figure 10.3). The raw data are analysed using bioinformatics
software and databases. The aim is to obtain meaningful biological information such as patterns
of relative induction/repression levels of gene expression, participation in biochemical pathways,
and (in the most favorable cases) genetic signatures.
Figure 10.3 Gene expression analyses by microarray. a) One-color expression analysis uses a single
uorescent label and two arrays to generate expression proles for two cell or tissue samples (test and
reference samples). Activated and repressed genes are obtained by superimposing images obtained by
the two arrays. b) Two-color expression analysis uses two different uorescent labels and a single array
to generate expression proles for the test and reference samples. Activated and repressed genes are
obtained by superimposing images generated in different channels on a single array. In both cases, the
monochrome images from the scanner are imported into software in which the images are pseudo-colored
and merged. Data is viewed as a normalized ratio in which signicant deviation from 1 (no change) are
indicative of increased (> 1) or decreased (<1) level of gene expression relative to the reference sample.
Activated gene
Activated gene
Repressed gene
Repressed gene
Superimpose Scan and superimpose
mRNA mRNA
cDNA cDNA
labelling labelling
mix
Hybridize, wash
and scan
Hybridize, wash
and scan
Hybridize and wash
a b
The possibility of analysing the effect of chemicals and environmental stressors on a large number
of genes in a single experiment has led to the development of the eld of toxicogenomics. Proponents
of toxicogenomics propose to apply both mRNA and protein expression technology to study chemical
effects in biological systems (Afshari et al., 1999; Lovett 2000). The availability of the (almost)
complete human genome and several other organisms (see http://www.ncbi.nlm.nih.gov/
Genomes/index.html for a list of nished and ongoing projects) allows the application of microarray
technology to several model organisms (from bacteria, to yeast, to sh) and mammalian cell lines.
To expand the application of microarray technology to toxicology several national and international
initiatives have been started in order to obtain a better standardisation and harmonisation of the
technology. In fact, one of the early concerns has been how to properly compare experiments that
use a wide variety of commercially available and proprietary platforms, protocols, instrumentation,
and analysis methods.
In the United States the National Institute of Environmental Health Sciences (NIEHS) has created
the National Center for Toxicogenomics (NCT) to provide a reference system of genome-wide gene
expression data and to develop a knowledge-base of chemical effects in biological systems (Tennant,
2002). The NCT has conducted some proof-of-principle experiments to establish signature proles
and to link the pattern of altered gene expression to specic parameters of conventional indices of
toxicity (Bartosiewicz et al., 2001; Bulera et al., 2001; Hamadeh et al., 2002). These studies have
shown that it is possible to identify signature of expressed gene patterns (Tennant, 2002).
The Health and Environmental Sciences Institute (HESI) of the International Life Sciences Institute
(ILSI) has coordinated an international study, involving more than thirty pharmaceutical
companies, to evaluate the harmonization of gene expression data and analyses (Pennie et al.,
2004). In the ILSI Genomic Project common pools of RNA were analysed in more than 30 different
laboratories on both similar and different technical platforms. Overviews of the design and
objectives of the experimental program and more technical articles were published in the mini-
monograph Application of Genomics to Mechanism-Based Risk Assessment published in
Environmental Health Perspectives, 2004, vol. 112, number 4. For example, Amin et al., (2004)
identied gene markers of renal toxicity, Thompson et al., (2004) identied markers of cis-platin
nephrotoxocity, while three groups reported an overview on the inter-laboratory collaboration to
evaluate the effects of nephrotoxicants (Kramer et al., 2004), genotoxic chemicals (Newton et al.,
2004), and hepatotoxicants (Ulrich et al., 2004) on gene expression.
The experimental program detailed in Pennie et al., (2004) has shown that: a) patterns of gene
expression relating to biological pathways are robust enough to allow insight into mechanisms of
toxicity; b) gene expression data can provide meaningful information on the physical location of
the toxicity; c) dose-dependent changes can be observed; and d) concerns about oversensitivity of
the technology may be unfounded.
10.5 Overview of gene expression analyses
The eld of microarray has evolved from Ed Southerns key insight (Southern, 1975), twenty ve
years ago, who showed that labeled nucleic acid molecules could be used to interrogate nucleic
acid molecules attached to a solid support. The resulting Southern blot can be viewed as the rst
array (Southern, 2000). It was only a small step to lter-based screening of clone libraries, which
introduced a one to one correspondence between clone and hybridization signals (Grunstein and
Hogness, 1975). The next advance has been the use of gridded libraries, stored in microtiter
plates, and stamped onto lters in xed position. With this system each clone could be uniquely
identied and information about it accumulated. Several groups have explored expression analysis
by hybridizing mRNA to cDNA libraries gridded on nylon lters. The subsequent explosion of array
technologies has been sparked by two key innovations. The rst has been the use of non porous
solid support, such as a glass, which has facilitated miniaturization and uorescence-hybridization
detection (Schena et al., 1995; Lockhart et al., 1996; Schena et al., 1996). The second critical
innovation has been the development of methods for high-density spatial synthesis of
oligonucleotide which allows the analysis of thousands of genes at the same time.
Recently, a signicant technical achievement was obtained by producing arrays with more than
250,000 oligonucleotide probes or 10,000 different cDNAs per square centimeter. DNA microarrays
are fabricated by high speed robotics, generally on glass. Since the DNA cannot bind directly to the
glass, the glass is rst treated with silane to covalently attach reactive amine, aldehyde, or epoxide
groups. Such surfaces allow stable attachment of DNA, proteins, and other molecules.
The nucleic acid microarrays use short oligonucleotide (15-25 nt), long oligonucleotide (50-120
nt) and PCR-amplied cDNAs (100-3,000 bp) as array elements. The short oligonucleotide is
primarily used for detection of single nucleotide polymorphism (SNPs). Because this application
requires the discrimination of only one mismatch, the presence of a short oligonucleotide
maximises the destabilisation caused by mispairing (Lockhart et al., 1996). On the other hand, the
PCR-amplied cDNAs produce strong signals and high specicity (DeRisi et al., 1996). The cDNAs
elements are obtained readily from cDNA libraries, and are typically used for organisms for which
the available genomic sequence is limited. The long nucleotides offer strong hybridization signal,
good specicity, unambiguous sample identication and affordability (Schena et al., 1998;
Hughes et al., 2000; Kane et al., 2000).
10.6 Recent applications of DNA microarray to ecotoxicology
Gene expression prole represents the primary level of integration between environmental factors
and the genome, providing the basis for protein synthesis, which ultimately guides the response
of organisms to external changes. Thus the analyses of gene expression changes is a powerful tool
both to diagnose major stressors in natural populations and to analyse the causes of such stress
response. When using gene expression proles in Ecotoxicology there is a need to properly account
for sources of variation in gene expression in natural populations; in fact differences in gene
expression due to sex, genotype, age, and intrinsic genetic variability can be quite substantial (Jin
et al., 2001; Oleksiak et al., 2002; Ranz et al., 2003; Townsend et al., 2003).
This technology has been extensively applied to the analyses of natural and anthropogenic factors
in yeast where whole genome chips have been available for a few years (Gasch et al., 2000;
Causton et al., 2001; Momose and Iwahashi 2001). Causton et al., (2001) analysed how the
whole genome of yeast is remodeled in response to environmental stressors such as temperature,
pH, oxidation, and nutrients. The stress response was dependent on the level of the stress and
showed an additive effect for multiple stressors. Similar results were found when using different
stressors like temperature shock, amino acid starvation, nitrogen source depletion (Gasch et al.,
2000) and cadmium (Momose and Iwahashi, 2001). The same approach has been used to
characterise the alteration of gene expression in yeast induced by the pesticide thiuram (Kitagawa
et al., 2002). The results obtained for stress response in yeast will be likely to provide a reference
frame for similar experiments that use more complex organisms.
This methodology has been recently applied to a typical model organism: zebrash, to understand
responses to endocrine modulators (Hoyt et al., 2003). Hoyt and coworkers used a microarray
approach to study the response of zebrash embryos exposed in vitro to the environmental
contaminant 4-nonylphenol (4NP). Using a custom microarray with 230 genes they were able to
identify a set of nine genes associated with the function of estrogen response that are indicative
of embryo exposure to 4NP, even at low concentrations. The application of gene expression
proles is not limited to model organisms for which the complete (or almost complete genome) is
available. Several strategies are available to apply a genomic approach in ecotoxicology even
before genomes are completed.
One approach consists in heterologous hydridization. In fact, due to the length of the probes,
cDNA microarrays can be used in heterologous hydridizations across strains and relatively closely
related species as long as sequence divergence is limited for a given gene (Rise et al., 2004a,b).
For example, this approach has been used to study the molecular basis of traits like hibernation
not present in model species (Hittel and Storey, 2001). More recently heterologous hybridization
has been used to study gene expression proling across a wide range of different species of
African cichlid sh (Renn et al., 2004).
Another possible approach consists of identifying stress-induced genes using special techniques
based on polymerase chain reaction (PCR), such as differential display PCR (DD PCR), suppressive
subtractive hybridisation (SSH PCR), and representational difference analyses (RDA). The
application of these techniques to ecotoxicology has been reviewed by Snell et al., (2003). Gracey
et al., (2001) have used cDNA microarrays to identify hypoxia-induced genes in a non model sh
for which sequence data were unavailable. Their analyses revealed that although some changes
in gene expression mirror the changes that take place in mammals, novel genes are differentially
expressed in sh and that there are tissue-specic patterns of gene expression during hypoxia.
Larkin et al., (2003) have described an expression proling model system for endocrine-disrupting
compounds (EDCs) that mimic estrogens. The team created a gene array by cloning 30 genes
from sheepshead minnows. The genes had previously been identied by differential display
reverse transcriptase-polymerase chain reaction, a method that screens thousands of RNA
messages to identify genes that are turned on or off by specic treatments. They treated the sh
with a constant concentration of weak and strong environmental estrogens and then analysed
which genes were differentially expressed in the livers of treated and control sh. They found
genes that were upregulated by all the test compounds while other genes were specic for a
single compound. Exposure to different concentrations of the strong estrogen 17-ethynyl
estradiol revealed that their microarray is dose-sensitive and that exposure thresholds vary for
different genes. A similar approach has been used to identify alterations in gene expression due
to exposure to androgen hormones in largemouth bass sh (Blum et al., 2004).
Williams et al., (2003) used a cDNA microarray-based approach to analyse the expression level
changes of recognised biomarkers in a relevant sh species: European ounder (P. esus). They
arrayed 160 genes, among which 110 were already known in the literature as biomarkers for sh
and mammalian toxic response. Five adult male and ve adult feral female P. esus were caught
from Tyane (polluted) and Alde (unpolluted) estuaries. They observed that eleven genes were
expressed differently between the Tyane and Alde males. Such differences were not statistically
signicant in females due to the inter-individual variations. In fact the vitellogenin levels differed
radically among the female sh suggesting that their reproductive cycles were at different stages.
Very recently, microarray analyses have also been successfully applied to identify molecular
markers of pathogen infection in salmonids (Rise et al., 2004a,b).
10.7 Overview of Proteomics
Proteomics is part of a growing list of genome-wide analyses that have been developed in response
to the recent focus on system biology. Proteomics aim to analyse the complete set of proteins
synthesised by a cell or tissue under a given set of physiological or developmental conditions
(Wilkins et al., 1996). Compared to the genome-wide expression analyses at the level of mRNA
(usually performed with DNA microarray) it has some strengths and limitations. One particular
concern is the correlation between mRNA abundance and protein expression levels. Detailed
analyses in yeast have shown that although in standard conditions there is a good correlation
between mRNA and protein levels (Futcher et al., 1999), in certain conditions such a correlation
is insufcient (Gygi et al., 1999), thus highlighting the fact that quantitative mRNA data cannot
be used to predict protein expression levels. To understand the cellular role of a gene product it is
also necessary to know the post-translational modications (for example glycosilation, removal of
targeting sequences, phosphorilation, ubiquitination). This information is crucial to understand
protein localisation, function and interaction with other biological molecules and can be studied
only at the protein level.
The most common application of proteomics has been, up to now, in the biomedical sector.
Typically cells (or tissue) derived from a disease state and a normal state (for example a cancerous
cell line and normal cell line) is harvested and the proteins solubilised (see Figure 10.4). The
protein mixture is then separated using two dimensional gels, where the proteins are separated
rst on the basis of their isoelectric point and then on the basis of their size. The gel is then xed
and the proteins visualised by staining and the resulting protein spots are recorded and quantied.
Image software is used to compare the two spectra and to detect any difference between the
normal and disease state (upregulation or downregulation of proteins). The spots of interest are
then excised and subjected to mass spectrometric analyses.
The systematic analysis of expressed proteins requires three key steps:
Separation of the proteins
High throughput method of protein identication
Data analyses and storage.
Separation of proteins is usually achieved using two-dimensional polyacrilamide gel electrophoresis
(2D PAGE). The mixture of proteins is resolved by two orthogonal separation steps, rstly by
charge (isoelectric focusing) and then by protein mass (PAGE). In some experiments, before the
gels are run, the proteins are tagged with uorescent markers to allow later detection. In other
cases, the proteins are chemically stained after separation. The end result is a series of spots and
smears that represent a portion of the proteins in the sample. With this technique a thousand
protein spots are readily visible on a single gel (Blomberg et al., 1995) and the separation of up
to 10,000 proteins has been reported (Klose and Kobalz, 1995).
Even though whole cell analyses still relies primarily on the use of 2DE as its primary separation
technique, it is being replaced by multidimensional HPLC. In a typical application, microscale two-
dimensional liquid chromatography (2DLC) is coupled with electrospray ionization (ESI) tandem
mass spectrometry (MS/MS) (Liu et al., 2002; Aebersold and Mann 2003). This combination of
powerful separation and identication techniques has allowed, for example, a very important
breakthrough in the understanding of the Plasmodium falciparum life cycle (Florens et al., 2002).
The identication of the individual proteins present in the system under analysis is done using mass
spectrometry (MS). A mass spectrometer discriminates between chemical compounds by separating
molecular ions according to the mass-to-charge (m/z) ratio, allowing the determination of molecular
mass. The key improvement that has allowed the use of MS for the identication of proteins has been
the development of soft methods of ionization such as matrix-assisted laser desorption ionization
Figure 10.4 Schematic picture of the two-dimensional gel electrophoresis approach to proteomics. Cells (or
tissue) derived from two different conditions: A and B, are collected and the proteins solubilised with an
appropriate buffer. The crude protein mixture is then separated based on the different isoelectric point of
the proteins (rst dimension). Then the proteins are reduced and applied to a second dimension SDS-
PAGE gel where they are separated on the basis of protein size. The gels are then xed and the proteins
visualised by silver staining. After staining, the resulting protein spots are recorded and quantied with
image analyses software. The spots of interest are then excised and analysed with mass spectrometry.
Condition A
Condition A
Condition B
Condition B
Solubilize proteins
from cells
pH
2 4 7 10
200
100
70
45
20
Reduction
and alkylation
Mass spectrometric
identication of spots
SDS-PAGE
(second dimension)
Excision of spots of interest
Image analysis
Isoelectric focusing on
narrow- or broad-range pl strips
(rst dimension)
Staining
1. Silver
2. Coomassie
3. Fluorescent
4. Autoradiography
+
+

(MALDI) and electrospray ionization (ESI). These two methods allow the ionization of large
biomolecules such as proteins with very little or no fragmentation (thus the term soft ionization).
The third key part of the proteomics technique is the availability of a large number of fully
sequenced genome that allows the peptide mass ngerprinting (Sagliocco et al., 1996; Shevchenko
et al., 1996). In this method the protein to identify is subjected to proteolytic digestion with a
preotease (usually trypsin). The masses of the resulting peptides are acquired by mass spectrometry
to give an experimental peptide ngerprint. This experimental ngerprint is then compared with
all the predicted peptide ngerprint of that organism available in the complete proteome
database. A protein that has the most theoretical peptide fragments in accordance with the
experimental data is identied as the best matched protein.
Early applications of proteomics have focused on the differences, at the protein level, between
normal and cancerous cells. A study compared the proteome of normal human luminal and
myoepithelial breast cells using immunopuried cell populations. It detected 170 spots with a
twofold difference in expression levels (Page et al., 1999) and identied 52 of them. Several
groups have probed two-dimensional gels of proteins from allergy-causing organisms using
antibodies derived from allergic patients (Breitenbach et al., 1996; Sander et al., 1998). Very
recently the technique has been used to follow the time course of the phosphotyrosine dependent
signaling network. Blagoev et al., (2004) grew individual cell cultures in medium containing
isotope-labeled arginine variants, and then stimulated each with EGF for different lengths of time.
Tyrosine-phosphorylated proteins were afnity puried and subjected to mass spectrometry.
10.8 Applications of Proteomics to Environmental Toxicology
Proteomics has been used in a series of papers by Witzmann and colleagues to study the exposure
to jet fuel vapor in rats (Witzmann et al., 1999; Drake et al., 2003; Witzmann et al., 2003). They
found a low but signicant effect of exposure on protein expression in the lungs, kidney, and liver.
A similar approach has been used by Hogstrand et al., (2002) in the study of a non-model species.
They exposed rainbow trout to a sub-lethal concentration of waterborne zinc for up to six days.
The response in gill tissue was investigated by differential screening of a heterologous cDNA array
and by protein proling. Seven proteins were found to be present only in zinc-exposed gills, and
four proteins were present only in control sh. They could not identify the proteins due to the lack
of genome and protein information on the rainbow trout. The major limitation in the proteomic
approach is that unless there is a substantial database of protein and/or sequence information,
it is difcult to identify proteins of interest.
Several studies have attempted to identify protein expression signature in response to
environmental stressors. Bradley and coworkers have analysed the response, at the protein level,
of marine invertebrates to salinity and temperature, and to chemicals such as polychlorinated
biphenyls and copper (Shepard and Bradley 2000; Shepard et al., 2000; Bradley et al., 2002).
Shrader et al., (2003) exposed embryonic zebrash to sub-lethal concentrations of 17beta-estradiol
(E2) and the estrogen mimic 4-nonylphenol (4-NP). Protein expression signatures were derived
from embryo homogenates by two-dimensional electrophoresis and digital imaging. While there
are many proteins that respond specically to each chemical, relatively few are common to the
two chemicals suggesting that the response pathways of the two chemicals are distinct.
10.9 Summary
Gene expression and protein expression proling represent a unique way of characterising how
cells and organisms adapt to changes in the external environment. The measurements of such
changes upon exposure to a chemical can be used both to provide information about the
mechanism of action of toxicants and also to form a sort of signature for the identication of
toxic products. Several key technical advances have allowed proteomics to become an established
discipline in system biology and in the biomedical sector. At the same time the development of
high quality, commercially available, gene arrays has allowed this technology to become a
standard tool in molecular toxicology. Several national and international initiatives have provided
the proof-of-principle tests for the application of gene expression and proteomics to the study of
toxicology of new and existing chemical compounds. In the last few years the eld has progressed
from evaluating the potential of the technologies to illustrating the practical use of gene
expression proling and proteomics in toxicology.
The application of the above technologies to ecotoxicology is at an early stage, due mainly to
more complex problems at hand, and to the many more variables involved in analysing the status
of natural populations in a real ecosystem rather than in a laboratory environment. Nevertheless,
signicant studies have been carried out to study responses to environmental stressors in both
model and non-model organisms. It can be easily predicted that the development of stressors-
specic signatures in ecotoxicology will have a similar potential as the toxicant signatures in
toxicology studies.
One of the best ways to advance the eld is probably to focus on more precise objectives that have
been neatly outlined in a recent paper by Snell et al., (2003): identication of conserved genes
that are up-regulated in response to toxicant exposure; how these gene-expression proles can be
used to diagnose stressors; and determination of which genes are most informative to incorporate
into stress gene arrays.
10.10 References
Aebersold, R., and M. Mann. 2003. Mass spectrometry-based proteomics. Nature 422:198-207
Afshari, C.A., E.F. Nuwaysir, and J.C. Barrett. 1999. Application of complementary DNA microarray technology
to carcinogen identication, toxicology, and drug safety evaluation. Cancer Research 59:4759-4760
Amin, R.P., A.E. Vickers, F. Sistare, K.L. Thompson, R.J. Roman, M. Lawton, J. Kramer, H.K. Hamadeh, J. Collins,
S. Grissom, L. Bennett, C.J. Tucker, S. Wild, C. Kind, V. Oreffo, J.W. Davis 2nd, S. Curtiss, J.M. Naciff, M.
Cunningham, R. Tennant, J. Stevens, B. Car, T.A. Bertram, and C.A. Afshari. 2004. Identication of putative
gene based markers of renal toxicity. Environmental Health Perspectives 112:465-79
Bartosiewicz, M.J., D. Jenkins, S. Penn, J. Emery, and A. Buckpitt. 2001. Unique gene expression patterns in
liver and kidney associated with exposure to chemical toxicants. Journal of Pharmacology and Experimental
Therapeutics 297:895-905
Belgrader, P., W. Benett, D. Hadley, J. Richards, P. Stratton, R. Mariella, Jr., and F. Milanovich. 1999. PCR
detection of bacteria in seven minutes. Science 284:449-450
Bernard, P.S., and C.T. Wittwer. 2002. Real-time PCR technology for cancer diagnostics. Clinical Chemistry
48:1178-1185
Blagoev, B., S.E. Ong, I. Kratchmarova, M. Mann. 2004. Temporal analysis of phosphotyrosine-dependent
signaling networks by quantitative proteomics. Nature Biotechnology 22:1139-1145
Blomberg, A., L. Blomberg, J. Norbeck, S.J. Fey, P.M. Larsen, M. Larsen, P. Roepstorff, H. Degand, M. Boutry,
A. Posch, and A. Gorg. 1995. Interlaboratory reproducibility of yeast protein patterns analyzed by immobilized
pH gradient two-dimensional gel electrophoresis. Electrophoresis 16:1935-1945
Blum, J.L., I. Knoebl, P. Larkin, K.J. Kroll, and N.D. Denslow. 2004. Use of suppressive subtractive hybridization
and cDNA arrays to discover patterns of altered gene expression in the liver of dihydrotestosterone and
11-ketotestosterone exposed adult male largemouth bass (Micropterus salmoides). Marine Environmental
Research 58:565-569
Boyer, I.J. 1989. Toxicity of dibutyltin, tributyltin and other organotin compounds to humans and to
experimental animals. Toxicology 55:253-298
Bradley, B.P., E.A. Shrader, D.G. Kimmel, and J.C. Meiller. 2002. Protein expression signatures: an application
of proteomics. Marine Environmental Research 54:373-377
Breitenbach, M., F. Ferreira, A. Jilek, I. Swoboda, C. Ebner, K. Hoffmann-Sommergruber, P. Briza, O. Scheiner,
and D. Kraft. 1996. Biological and immunological importance of Bet v 1 isoforms. Advances in Experimental
Medicine and Biology 409:117-126
Bulera, S.J., S.M. Eddy, E. Ferguson, T.A. Jatkoe, J.F. Reindel, M.R. Bleavins, and F.A. De La Iglesia. 2001. RNA
expression in the early characterization of hepatotoxicants in Wistar rats by high-density DNA microarrays.
Hepatology 33:1239-1258
Burczynski, M.E., M. McMillian, J. Ciervo, L. Li, J.B. Parker, R.T. Dunn 2
nd
, S. Hicken, S. Farr, and M.D. Johnson.
2000. Toxicogenomics-based discrimination of toxic mechanism in HepG2 human hepatoma cells.
Toxicological Sciences 58:399-415
Bustin, S.A. 2000. Absolute quantication of mRNA using real-time reverse transcription polymerase chain
reaction assays. Journal of Molecular Endocrinology 25:169-193
Causton, H.C., B. Ren, S.S. Koh, C.T. Harbison, E. Kanin, E.G. Jennings, T.I. Lee, H.L. True, E.S. Lander, and
R.A. Young. 2001. Remodeling of yeast genome expression in response to environmental changes. Molecular
Biology of the Cell 12:323-337
Cousinou, M., B. Nilsen, J. Lopez-Barea, and G. Dorado. 2000. New methods to use sh cytochrome P4501A
to assess marine organic pollutants. The Science of the Total Environment 247:213-225
DeRisi. J., L. Penland, P.O. Brown, M.L. Bittner, P.S. Meltzer, M. Ray, Y. Chen, Y.A. Su, and J.M. Trent. 1996.
Use of a cDNA microarray to analyse gene expression patterns in human cancer. Nature Genetics 14:457-460
Drake, L.A., F.C. Dobbs, and R.C. Zimmerman. 2003. Effects of epiphyte load on optical properties and
photosynthetic potential of the seagrasses Thalassia testudinum Banks ex Konig and Zostera marina
L. Limnology and Oceanography 48:456-463
Florens, L., M.P. Washburn, J.D. Raine, R.M. Anthony, M. Grainger, J.D. Haynes, J.K. Moch, N. Muster, J.B. Sacci,
D.L. Tabb, A.A. Witney, D. Wolters, Y. Wu, M.J. Gardner, A.A. Holder, R.E. Sinden, J.R. Yates, and D.J. Carucci.
2002. A proteomic view of the Plasmodium falciparum life cycle. Nature 419:520-526
Futcher, B., G.I. Latter, P. Monardo, C.S. McLaughlin, and J.I. Garrels. 1999. A sampling of the yeast proteome.
Molecular and Cellular Biology 19:7357-7368
Galay-Burgos, M., D.J. Spurgeon, J.M. Weeks, S.R. Sturzenbaum, A.J. Morgan, and P. Kille. 2003. Developing
a new method for soil pollution monitoring using molecular genetic biomarkers. Biomarkers 8:229-239
Gasch, A.P., P.T. Spellman, C.M. Kao, O. Carmel-Harel, M.B. Eisen, G. Storz, D. Botstein, and P.O. Brown. 2000.
Genomic expression programs in the response of yeast cells to environmental changes. Molecular Biology of
the Cell 11:4241-4257
Gracey, A.Y., J.V. Troll, and G.N. Somero. 2001. Hypoxia-induced gene expression proling in the euryoxic
sh Gillichthys mirabilis. Proceeding of the National Academy of Sciences of United States of America
98:1993-1998
Grunstein, M., and D.S. Hogness. 1975. Colony hybridization: a method for the isolation of cloned DNAs that
contain a specic gene. Proceeding of the National Academy of Sciences of United States of America
72:3961-3965
Gygi, S. P., Y. Rochon, B. R. Franza, and R. Aebersold. 1999. Correlation between protein and mRNA abundance
in yeast. Molecular and Cellular Biology 19:1720-1730
Hamadeh, H.K., P.R. Bushel, S. Jayadev, K. Martin, O. DiSorbo, S. Sieber, L. Bennett, R. Tennant, R. Stoll, J.C.
Barrett, K. Blanchard, R.S. Paules, and C.A. Afshari. 2002. Gene expression analysis reveals chemical-specic
proles. Toxicological Sciences 67: 219-231
Heid, C.A., J. Stevens, K.J. Livak, and P.M. Williams. 1996. Real time quantitative PCR. Genome Research
6:986-994
Higuchi, R., C. Fockler, G. Dollinger, and R. Watson. 1993. Kinetic PCR analysis: real-time monitoring of DNA
amplication reactions. Biotechnology 11:1026-1030
Hittel, D., and K.B. Storey. 2001. Differential expression of adipose- and heart-type fatty acid binding proteins
in hibernating ground squirrels. Biochimica Biophysica Acta 1522:238-243
Hogstrand, C., S. Balesaria, and C.N. Glover. 2002. Application of genomics and proteomics for study of the
integrated response to zinc exposure in a non-model sh species, the rainbow trout. Comparative Biochemistry
and Physiology B-Biochemistry and Molecular Biology 133: 523-535
Holland, P.M., R.D. Abramson, R. Watson, and D.H. Gelfand. 1991. Detection of specic polymerase chain
reaction product by utilizing the 5----3 exonuclease activity of Thermus aquaticus DNA polymerase.
Proceeding of the National Academy of Sciences of United States of America 88:7276-7280
Hoyt, P.R., M.J. Doktycz, K.L. Beattie, M.S. Greeley Jr. 2003. DNA microarrays detect 4-nonylphenol-induced
alterations in gene expression during zebrash early development. Ecotoxicology 12:469-474
Hughes, T.R., C.J. Roberts, H. Dai, A.R. Jones, M.R. Meyer, D. Slade, J. Burchard, S. Dow, T.R. Ward, M.J. Kidd,
S.H. Friend, and M.J. Marton. 2000. Widespread aneuploidy revealed by DNA microarray expression proling.
Nature Genetics 25:333-337
Huuskonen, S.E., M.E. Hahn, and P. Lindstrom-Seppa. 1998. A sh hepatoma cell line (PLHC-1) as a tool to
study cytotoxicity and CYP1A induction properties of cellulose and wood chip extracts. Chemosphere
36:2921-2932
Islinger, M., D. Willimski, A. Volkl, and T. Braunbeck. 2003. Effects of 17a-ethinylestradiol on the expression
of three estrogen-responsive genes and cellular ultrastructure of liver and testes in male zebrash. Aquatic
Toxicology 62:85-103
Jin, W., R.M. Riley, R.D. Wolnger, K.P. White, G. Passador-Gurgel, and G. Gibson. 2001. The contributions of
sex, genotype and age to transcriptional variance in Drosophila melanogaster. Nature Genetics 29:389-395
Jordan, J.A., and M.B.Durso. 2000. Comparison of 16S rRNA gene PCR and BACTEC 9240 for detection of
neonatal bacteremia. Journal of Clinical Microbiology 38:2574-2578
Kammenga, J.E., R. Dallinger, M.H. Donker, H.R. Kohler, V. Simonsen, R. Triebskorn, and J.M. Weeks. 2000.
Biomarkers in terrestrial invertebrates for ecotoxicological soil risk assessment. Review of Environmental
Contamination and Toxicology 164:93-147
Kane, M.D., T.A. Jatkoe, C.R. Stumpf, J. Lu, J.D. Thomas, and S.J. Madore. 2000. Assessment of the sensitivity
and specicity of oligonucleotide (50mer) microarrays. Nucleic Acids Research 28:4552-4557
Kitagawa, E., J. Takahashi, Y. Momose, and H. Iwahashi. 2002. Effects of the pesticide thiuram: genome-wide
screening of indicator genes by yeast DNA microarray. Environ Science and Technology 36:3908-3915
Klose, J., and U. Kobalz. 1995. Two-dimensional electrophoresis of proteins: an updated protocol and
implications for a functional analysis of the genome. Electrophoresis 16:1034-1059
Kramer, J.A., S.D. Pettit, R.P. Amin, T.A. Bertram, B. Car, M. Cunningham, S.W. Curtiss, J.W. Davis, C. Kind,
M. Lawton, J.M. Naciff, V. Oreffo, R.J. Roman, F.D. Sistare, J. Stevens, K. Thompson, A.E. Vickers, S. Wild, and
C.A. Afshari. 2004. Overview on the application of transcription proling using selected nephrotoxicants for
toxicology assessment. Environmrntal Health Perspectives 112:460-464
Kumar, S., M.W. Reed, H.B. Gamper, Jr., V.V. Gorn, E.A. Lukhtanov, M. Foti, J. West, R.B. Meyer, Jr., and
B.I. Schweitzer. 1998. Solution structure of a highly stable DNA duplex conjugated to a minor groove binder.
Nucleic Acids Research 26:831-838
Lakowicz, J.R. 1999. Principles of uorescence spectroscopy. Kluwer Academic/Plenum, New York
Larkin, P., L.C. Folmar, M.J. Hemmer, A.J. Poston, and N.D. Denslow. 2003. Expression proling of estrogenic
compounds using a sheepshead minnow cDNA macroarray. EHP Toxicogenomics 111:29-36
Larson G.L., G. Lomnicky, R. Hoffman, W.J. Liss, and E. Deimling. 1999. Integrating Physical and Chemical
Characteristics of Lakes into the Glacially Inuenced Landscape of the Northern Cascade Mountains,
Washington State, USA. Environmental Management 24:219-228
Lashkari, D.A., J.L. DeRisi, J.H. McCusker, A.F. Namath, C. Gentile, S.Y. Hwang, P.O. Brown, and R.W. Davis.
1997. Yeast microarrays for genome wide parallel genetic and gene expression analysis. Proceeding of the
National Academy of Sciences of United States of America 94:13057-13062
Levine, S.L., and J.T. Oris. 1999. Noncompetitive mixed-type inhibition of rainbow trout CYP1A catalytic
activity by clotrimazole. Comparative Biochemistry and Physiology C-Pharmacology Toxicology and
Endocrinology 122:205-210
Liu, H., D. Lin, and J.R. Yates. 2002. Multidimensional separations for protein/peptide analysis in the post-
genomic era. Biotechniques 32:898-902
Lockhart, D.J., H. Dong, M.C. Byrne, M.T. Follettie, M.V. Gallo, M.S. Chee, M. Mittmann, C. Wang, M. Kobayashi,
H. Horton, and E.L. Brown. 1996. Expression monitoring by hybridization to high-density oligonucleotide
arrays. Nature Biotechnology 14:1675-1680
Lovett, R.A. 2000. Toxicogenomics. Toxicologists brace for genomics revolution. Science 289:536-537
Matthiessen, P., and J.P. Sumpter. 1998. Effects of oestrogenic substances in the aquatic environment. In:
T. Braunbeck, D.E. Hinton, and D. Streit (Eds.), Fish Ecotoxicology. Birkhauser Verlag, Basel, Switzerland,
pp. 319-335
Momose, Y., and H. Iwahashi. 2001. Bioassay of cadmium using a DNA microarray: genome-wide expression
patterns of Saccharomyces cerevisiae response to cadmium. Environmental Toxicology Chemistry 20:
2353-2360
Nakanishi, T., J. Kohroki, S. Suzuki, J. Ishizaki, Y. Hiromori, S. Takasuga, N. Itoh, Y. Watanabe, N. Utoguchi, and
K. Tanaka. 2002. Trialkyltin compounds enhance human CG secretion and aromatase activity in human
placental choriocarcinoma cells. Journal of Clinical Endocrinology and Metabolism 87:2830-2837
Newton, R.K., M. Aardema, and J. Aubrecht. 2004. The utility of DNA microarrays for characterizing
genotoxicity. Environmetal Health Perspectives 112:420-422
Olden, K., and J. Guthrie. 2001. Genomics: implications for toxicology. Mutation Research-Fundamental and
Molecular Mechanisms of Mutagenesis 473:3-10
Oleksiak, M.F., G.A. Churchill, and D.L. Crawford. 2002. Variation in gene expression within and among
natural populations. Nature Genetics 32:261-266
Page, M.J., B. Amess, R. R. Townsend, R. Parekh, A. Herath, L. Brusten, M.J. Zvelebil, R.C. Stein, M.D. Watereld,
S.C. Davies, and M.J. OHare. 1999. Proteomic denition of normal human luminal and myoepithelial breast
cells puried from reduction mammoplasties. Proceeding of the National Academy of Sciences of United
States of America 96:12589-12594
Pennie, W., S.D. Pettit, and P.G. Lord. 2004. Toxicogenomics in risk assessment: an overview of an HESI
collaborative research program. Environmental Health Perspectives 112:417-419
Ranz, J.M., C.I. Castillo-Davis, C.D. Meiklejohn, and D.L. Hartl. 2003. Sex-dependent gene expression and
evolution of the Drosophila transcriptome. Science 300:1742-1745
Rees, C.B., and W. Li. 2004. Development and application of a real-time quantitative PCR assay for
determining CYP1A transcripts in three genera of salmonids. Aquatic Toxicology 66:357-368
Renn, S.C., N. Aubin-Horth, and H.A. Hofmann. 2004. Biologically meaningful expression proling across
species using heterologous hybridization to a cDNA microarray. BMC Genomics 5:42
Rise, M.L., S.R. Jones, G.D. Brown, K.R. Von Schalburg, W.S. Davidson, and B.F. Koop. 2004a. Microarray
analyses identify molecular biomarkers of Atlantic salmon macrophage and hematopoietic kidney response
to Piscirickettsia salmonis infection. Physiological Genomics 20:21-35
Rise, M.L., K.R. von Schalburg, G.D. Brown, M.A. Mawer, R.H. Devlin, N. Kuipers, M. Busby, M. Beetz-Sargent,
R. Alberto, A.R. Gibbs, P. Hunt, R. Shukin, J.A. Zeznik, C. Nelson, S.R. Jones, D.E. Smailus, S.J. Jones, J.E. Schein,
M.A. Marra, Y.S. Buttereld, J.M. Stott, S.H. Ng, W.S. Davidson, and B.F. Koop. 2004b. Development and
application of a salmonid EST database and cDNA microarray: data mining and interspecic hybridization
characteristics. Genome Research 14:478-490
Sagliocco, F., J.C. Guillemot, C. Monribot, J. Capdevielle, M. Perrot, E. Ferran, P. Ferrara, and H. Boucherie.
1996. Identication of proteins of the yeast protein map using genetically manipulated strains and peptide-
mass ngerprinting. Yeast 12:1519-1533
Sander, I., M. Raulf-Heimsoth, C. Siethoff, C. Lohaus, H.E. Meyer, and X. Baur. 1998. Allergy to Aspergillus-
derived enzymes in the baking industry: identication of beta-xylosidase from Aspergillus niger as a new
allergen (Asp n 14). Journal of Allergy and Clinical Immunology 102:256-264
Schena, M., D. Shalon, R.W. Davis, and P.O. Brown. 1995. Quantitative monitoring of gene expression patterns
with a complementary DNA microarray. Science 270:467-470
Schena, M., D. Shalon, R. Heller, A. Chai, P.O. Brown, and R.W. Davis. 1996. Parallel human genome analysis:
microarray-based expression monitoring of 1000 genes. Proceeding of the National Academy of Sciences of
United States of America 93:10614-10619
Schena, M., R.A. Heller, T.P. Theriault, K. Konrad, E. Lachenmeier, and R.W. Davis. 1998. Microarrays:
biotechnologys discovery platform for functional genomics. Trends in Biotechnology 16:301-306
Shepard, J.L., and B.P. Bradley. 2000. Protein expression signatures and lysosomal stability in Mytilus edulis
exposed to graded copper concentrations. Marine Environmental Research 50:457-463
Shepard, J.L., B. Olsson, M. Tedengren, and B.P. Bradley. 2000. Protein expression signatures identied in
Mytilus edulis exposed to PCBs, copper and salinity stress. Marine Environmental Research 50:337-340
Shevchenko, A., O.N. Jensen, AV. Podtelejnikov, F. Sagliocco, M. Wilm, O. Vorm, P. Mortensen, H. Boucherie,
and M. Mann. 1996. Linking genome and proteome by mass spectrometry: large-scale identication of yeast
proteins from two dimensional gels. Proceeding of the National Academy of Sciences of United States of
America 93:14440-14445
Shrader, E.A., T.R. Henry, M.S. Greeley, Jr., and B.P. Bradley. 2003. Proteomics in zebrash exposed to endocrine
disrupting chemicals. Ecotoxicology 12:485-488
Snell, T.W., S.E. Brogdon, and M.B. Morgan. 2003. Gene expression proling in ecotoxicology. Ecotoxicology
12:475-483
Southern, E.M. 1975. Detection of specic sequences among DNA fragments separated by gel electrophoresis.
Journal of Molecular Biology 98:503-517
Southern, E.M. 2000. Blotting at 25. Trends in Biochemical Sciences 25:585-588
Tennant, R. W. 2002. The National Center for Toxicogenomics: using new technologies to inform mechanistic
toxicology. Environmental Health Perspectives 110:8-10
Thompson, K.L., C.A. Afshari, R.P. Amin, T.A. Bertram, B. Car, M. Cunningham, C. Kind, J.A. Kramer, M. Lawton,
M. Mirsky, J.M. Naciff, V. Oreffo, P.S. Pine, and F.D. Sistare. 2004. Identication of platform-independent gene
expression markers of cisplatin nephrotoxicity. Environ Health Perspectives 112:488-494
Townsend, J.P., D. Cavalieri, and D.L. Hartl. 2003. Population genetic variation in genome-wide gene
expression. Molecular Biology and Evolution 20:955-963
Ulrich, R.G., J.C. Rockett, G.G. Gibson, and S.D. Pettit. 2004. Overview of an interlaboratory collaboration on
evaluating the effects of model hepatotoxicants on hepatic gene expression. Environmental Health
Perspectives 112:423-427
Waring, J.F., R. Ciurlionis, R.A. Jolly, M. Heindel, and R.G. Ulrich. 2001. Microarray analysis of hepatotoxins
in vitro reveals a correlation between gene expression proles and mechanisms of toxicity. Toxicology Letters
120:359-368
Wilkins, M.R., C. Pasquali, R.D. Appel, K. Ou, O. Golaz, J.C. Sanchez, J.X. Yan, A.A. Gooley, G. Hughes,
I. Humphery-Smith, K.L. Williams, and D.F. Hochstrasser. 1996. From proteins to proteomes: large scale
protein identication by two-dimensional electrophoresis and amino acid analysis. Biotechnology 14:61-65
Williams, T.D., K. Gensberg, S.D. Minchin, and J.K. Chipman. 2003. A DNA expression array to detect toxic
stress response in European ounder (Platichthys esus). Aquatic Toxicology 65:141-57
Wittwer, C.T., M.G. Herrmann, A.A. Moss, and R.P. Rasmussen. 1997. Continuous uorescence monitoring of
rapid cycle DNA amplication. Biotechniques 22:130-138
Witzmann, F.A., M.D. Bauer, A.M. Fieno, R.A. Grant, T.W. Keough, S.E. Kornguth, M.P. Lacey, F.L. Siegel, Y. Sun,
L.S. Wright, R.S. Young, and M.L. Witten. 1999. Proteomic analysis of simulated occupational jet fuel exposure
in the lung. Electrophoresis 20:3659-3669
Witzmann, F.A., A. Bobb, G.B. Briggs, H.N. Coppage, R.A. Hess, J. Li, N.M. Pedrick, G.D. Ritchie, J.R. Iii, and
K.R. Still. 2003. Analysis of rat testicular protein expression following 91-day exposure to JP-8 jet fuel vapor.
Proteomics 3:1016-1027
Yang, S., S. Lin, G.D. Kelen, T.C. Quinn, J.D. Dick, C.A. Gaydos, and R.E. Rothman. 2002. Quantitative
multiprobe PCR assay for simultaneous detection and identication to species level of bacterial pathogens.
Journal of Clinical Microbiology 40:3449-3454
Combined pressures
1 - Michela Ghiani. Sampling in the Mediterranean Sea
2 - Antonio Ruggiero. Mountain stream in the Appennines (Torrente Simbrivio, Central Italy)
3 - Antonio Ruggiero. Mountain shallow lake on Gran Sasso (Lake Petranzoni, Central Italy)
4 - Anna-Stiina Heiskanen. Urban waterway (Venezia, Italy)
5 - Anna-Stiina Heiskanen. Rocky shores in the Mediterranean (Cinque Terre, Italy)
1 2 3 4 5
11 Large scale evaluation of pressures
Ana Garcia, Bertrand Villeneuve, and Jean-Gabriel Wasson
11.1 Introduction
The purpose of this chapter is the description of relationships between a combination of pressures
coming from human activities (agriculture, urbanisation, etc.) and ecological status of European
rivers, with a large scale approach. However, across Europe, the natural and human geographical
contexts are very diverse, and thus the relationship between anthropogenic pressures and
ecological status may vary according to the sensitivity of river ecosystems and combinations of
pressures. A regional framework related to the dominant ecological processes is useful to properly
identify the strength of the pressures-impacts relationships.
This chapter is a review of the relevant literature on this kind of models, linking pressures and river
ecological status (through biological indicators) at large scales, with emphasis on the possible
variability related to the geographical context. In a recent, very clear and comprehensive review,
Allan (2004) addressed similar questions and most of our ndings are consistent with his
conclusions.
WFD requirements
The Water Framework Directive (WFD) requires the identication of signicant anthropogenic
pressures and the assessment of their impacts on water bodies (Annex II, WFD). These ecological
impacts are not only determined by clearly identied point sources discharges, but also by series
of complex human inuences including diffuse pollution, alteration of water and sediment
regimes, hydro-morphological changes, connectivity breaks etc. (Borchardt and Richter, 2003).
Ecological quality is integrating all pressures and showing the overall status of the ecosystem.
(EEA, 2003). In most cases various pressures are acting simultaneously, and managers must
dene a hierarchy amongst these to identify priority actions. Moreover, these pressures are not
evenly distributed on the territory, and decisions makers have to decide where to act in priority.
Finally, pressures are generated by different kinds of human activities, and integrated policies to
restore river ecosystems must be tailored towards whole socio-economical structures, such as
agricultural sectors, industry, urban areas, etc. This means that an integrated approach and policy
should include possibilities to have differentiated measures according to the nature of the
different waterbodies and to the socio-economic conditions within the catchment.
A large scale view
In this sense, a large-scale view is necessary to take political decisions. A successful implementation
of the WFD will require a common framework and policies for large territories, which needs
scientic support and tools. However, from a scientic point of view, the problem is complex for
the following two reasons.
LARGE SCALE EVALUATIONS OF PRESSURES | 213
First, the criteria to identify and assess pressures must be based on available data. But many
pressures are difcult to identify and measure, especially when working in large territories. In most
cases, we have not the same level of information upon the various kinds of pressures, such as
chemical and hydro-morphological, point sources or diffuse, at the local or basin scale. For this
reason, it is often necessary to rise up at the level of the driving forces to get spatially homogeneous
information (Figure 11.1). For example, it is very difcult to determine in a large basin the amount
of pesticides and nutrients rejected, the sediment yield due to erosion, the degree of river
channelization etc., but we have access to the percentage of land dedicated to one kind of
agriculture that generates these pressures.
Second, in large spatial contexts, the inuence of scales when studying ecological processes is a
key issue (Allan, 2004). One important problem is the upscaling of ecological relationships: how
a model developed at a local scale can be safely extrapolated to larger areas? Another one is
related to the interactions between natural characteristics that determine the sensitivity of river
ecosystems and anthropogenic alterations: has a given pressure the same impact in different
regions?
To address these questions, it is necessary to study pressure-impact relationships in geographically
homogeneous areas, in order to identify the relative inuence of human and natural drivers of
ecosystem responses (Allan, 2004). The proper way for doing that is to work at the level of hydro-
ecoregions, which are considered homogeneous with respect to the natural factors (geology,
relief, climate) that control the functioning of the aquatic ecosystems (Wasson et al., 2002).
11.2 Combined pressures: driving forces
Following the DPSIR
1
(EEA, 2003) (Fig. 11.1) concepts, the driving forces are human activities
reected by land use (agriculture, urbanization); these activities generate a combination of
pressures (pollution discharges, physical structures etc.) which alter the abiotic components of the
ecosystem (hydro-morphology, physico-chemistry, etc.). These alterations have an impact on
biological communities that determine the ecological status. Agriculture and urbanization are
considered the principal impairment sources in the literature, but the relative inuence of combined
pressures coming from agricultural or urban land use is not well established (Allan, 2004).
Agriculture. Agriculture is addressed by the EEA (2003) as a signicant driving force in terms of
ecological quality by the way of various pressures: increased sediment yield (especially siltation),
nutrients and organic pollution, hazardous substances, as well as indirect hydro-morphological
modications like river channelization, habitat fragmentation, water abstractions, altered hydrological
regime (Cormier et al., 2000), etc. The impact of agriculture on Europes water will have to be
reduced if good surface and groundwater status are to be achieved (EEA, 2003). Especially intensive
agriculture is designated by many authors as a cause of degradation of stream habitats (Roth et al.,
1996; Allan et al., 1997; Allan, 2004). Livestock inuences especially nitrogen levels in rivers.
1
DPSIR (Driving forces-Pressures-State-Impacts-Responses) is an analytical framework to assess environmental issues,
considering all these element and their inter-linkages.
The biological effects of agriculture have been noted by several authors: decrease in the number
of species (taxa richness) and abundance of sh and invertebrates, shift in communities
composition for algae (Bis et al., 2000; Cuffney et al., 2000).
Urbanization. The primary impacts of urbanization are related to direct point sources pollution,
due to raw or treated sewage inputs. Pollution problems are well known by the managers, and
most of the nancial investments are dedicated to reduce sewage pollution. However, many others
pressures are related to urban areas (Allan, 2004; Walsh, 2000). Housing, roads and other
infrastructures constitute impervious surfaces. Increased impervious area may result in altered
hydrological regimes: changes in timing, frequency, duration and magnitude of ood events
(Cormier et al., 2000), with hydro-morphological and biological consequences. Untreated storm
water runoff may cause also acute problems of organic or toxic pollution. And nally, many
channelization works are dedicated to the ood protection of urbanized areas.
Other driving forces. More specic problems are related to large industries, energy production,
navigation, tourism, etc. However these problems are generally well known by the managers, and
depend on local specic measures, or will induce the designation of heavily modied water bodies
with adapted environmental objectives. These pressures (often concerning large rivers) are not
easily integrated in a large scale view, and deserve a more targeted approach.
In most cases, it is impossible to get precise information on these multiple pressures at large
scales. However, analysing in a rst approach the relative impacts of agriculture and urbanization
can provide to the decision makers, at the district or state level, useful information for the choice
of management priorities.
Figure 11.1 Possible indicators that could be used in the development of pressure- impact models.
However, when studying the large scale effects of combined pressures, it is often necessary to use data
from the level of the driving forces, in order to get spatially homogeneous indicator data.
Driving forces Pressures Impact Status
Population
density
urbanization
Agriculture
Land use
Industries
Water uses
energy
transport
Tourism
Abiotic parameters
Bioindicators
Pollution inputs
point
non point
Dams
Abstraction
Derivation
Channelization
Invertebrates
Fish
Diatoms
Pollution
Organic (MO)
Nutrients N,P
Toxics, acidication
Physical alteration
Hydrology
Morphology
11.3 Models, quality element and metrics
We describe here studies linking biological indicators to pressures by rising up at the driving
forces level. In the case of combined pressures, this level is represented by land use. The aim is to
identify 1) the main drivers of ecosystem alteration, 2) their relative importance, 3) the effect of
spatial scale on the relationships, 4) the possible effect of the natural geographical context, and
5) the existence of impact thresholds.
A selection has been made taking into account the number of sites considered in the work, what
gives us an idea of the spatial coverage; local studies were not considered. Nevertheless the cases
with more than 30-40 studied sites are rare. The impact of combined pressures is generally
assessed through the response of sh or macroinvertebrates, and sometimes algae. Physical
habitat state is also often measured as it inuences biota.
Generally, many authors use multimetric indices able to detect a wide range of alterations. For
sh communities, the multimetric index IBI (Index of Biological Integrity; Karr, 1981) is the most
commonly used (Allan et al., 1997; Lammert and Allan 1999; Cormier et al., 2000; Cuffney et al.,
2000; Wang et al., 2001; Stewart et al., 2002). Benthic macroinvertebrate communities are
assessed through classical quality index like BMWP (Bis et al., 2000) and EPT (Stewart et al.,
2001), or also by multimetric indices like the ICI (Cormier et al., 2000). Diversity indices (Shannon,
Hilsenhoff) are sometimes used (Wang et al., 2001; Wang and Kanehl 2003). For diatom
assemblages, Pan et al. (2004) measure several autoecological metrics (based on Lange-Bertalot,
1979 and Van Dam, 1994), as well as the Trophic Diatom Index. Habitat quality is measured by
Habitat Index (Roth et al., 1996), or US-EPA protocols (Dovciak and Perry, 2002). Many of these
metrics and indices are discussed in the other chapters of this report.
Examples of European compilations
At the European level, successive studies of the European Environment Agency attempted to
evaluate the global status of river quality and the intensity of main pressures (EEA 1991; 2003).
In the most recent one is developed a draft core set of indicators for water concerning ve sectors:
agriculture, energy, sheries, tourism and transport. One indicator is length of rivers less than
good quality in national classications, but this status is not related to the pressures, and it is
clearly stated that none of the national classication schemes meets the WFD requirements so
there is not at present information enabling to assess the situation in relation to the objectives of
the directive. Different types of schemes cannot be quantitatively compared (EEA, 2003).
Existing models and relationships
We are not aware, within Europe, of large scale models linking driving forces (land use) to
biological indicators, although works could exist in the grey literature. We review here some
selected examples of studies done abroad, even if their validity for European countries is untested.
The methods used are quite diverse, but most of the literature reviewed relies on regressions and
multivariate analysis, while more sophisticated statistical models are scarce.
In one of the pioneer works, Steedman (1988) evidenced positive relationships between the sh
communities (IBI) and the percentage of forest in the basin, the proportion of channel with
riparian forest, and a strong negative impact of watershed urbanization. He emphasised also the
need of more detailed evaluation of agricultural land use, and streamside vegetation to improve
predictive models. As stated by Allan (2004), most of these issues are still open, and the challenge
for the research is to shift from descriptive relationships to predictive models, usable for the
denition of management policies.
11.4 Spatial scale: basin vs. riparian effect
Land use was found to be a strong predictor of biological and habitat integrity (Allan et al., 1997).
But studies of the scale effect (basin vs. riparian) led to apparently contradictory conclusions (Allan,
2004). In search of effective scale for stream management for a large basin, Dovciak and Perry
(2002) tested the effectiveness of two landscape classication systems (watershed vs. agro-ecoregion)
in explaining habitat and invertebrate features. They observed that local habitat conditions, which
strongly inuence invertebrate assemblages, are determined by regional scale landscape factors.
In a literature review, Gergel et al. (2002) aimed to demonstrate how landscape indicators (amount
and arrangement of land cover, spatial distribution of patches etc. derived from aerial and satellite
images and mapping) complement traditional riverine indicators, resulting in a broader perspective
on rivers monitoring. They reviewed various studies relying landscape and biological indicators at
different spatial levels (catchment, riparian zone, river banks), and concluded that a variety of
investigators have tried to determine the spatial extent, or distance from the water body, over which
landscape patterns inuence water quality or aquatic biota, yet this question remains unresolved.
Sponseller and co-workers (Sponseller et al., 2001) found in nine headwater basins that water
chemistry was generally related to features at the catchment scale, while temperature, substratum
and macroinvertebrates were strongly inuenced by the land-cover patterns of a 200 m riparian
corridor. Stewart and co-workers (Stewart et al., 2001) found strong relations between both watershed
and riparian corridor land cover versus sh and invertebrates assemblages in 38 streams.
For Roth and co-workers (Roth et al., 1996), the regional ratio between agricultural and forested
areas is the primary determinant of stream conditions (IBI, HI), while riparian vegetation is a weak
secondary predictor of stream integrity. On the other hand, Lammert and Allan (1999) found no
relationship between regional land use and biological integrity; only habitat and immediate land
use (100 m buffer) predicted biotic integrity. In this study, sites were located close together and
regional land use did not differ greatly. It is possible that the contrasting results are a consequence
of differing scales of the studies, coupled with the differences in the scale of the various effects
(sediment control, etc.; Allan et al., 1997; Allan, 2004).
Water physico-chemistry and riparian condition may be better indicators of sh community
condition than the land use in basinwide scale, as found by Meador and Goldstein (2003) in their
study using data from 172 sites. In agricultural areas, the sh index was closely related to the
riparian condition index, but this relationship was weaker in urban areas. In their conclusions,
Meador and Goldstein (2003) emphasised the universal importance of riparian zones for the
maintenance and restoration of diverse sh communities in streams.
11.5 Urban vs. agricultural land use
The most often studied driving forces are urban and agricultural land use, and there is contrasting
evidence in the literature about their relative impact. Negative relationships between the
percentage of agricultural area in the watershed and sh communities indices were found by Roth
et al. (1996) and Walser and Bart (1999). Cuffney et al. (2000) evidenced impacts on sh,
invertebrates and algae along an agriculture intensity gradient. Pan et al. (2004) found weak
linkages between agricultural land use and lotic diatoms communities during summer base ow.
Longer temporal scale is integrated by Harding et al. (1998), who related the present-day diversity
of stream invertebrates and sh to the inuence of past land use (1950).
Negative correlations between urbanisation and invertebrate diversity indices is shown by
Stepenuck et al. (2002), who study the invertebrates response through a gradient of imperviousness
caused by urbanisation. A similar negative effect of imperviousness is evidenced by Wang and
Kanehl (2003). Roy et al. (2003a) showed strong negative relationships between urban land
cover and invertebrate indices, and described what habitat variables are most important in
explaining this effect.
Among the few large scale studies trying to identify the relative impact of agricultural and urban
land use, Meador and Goldstein (2003) found a complex relationship. Good condition sh
community could be found even in areas where agricultural land use was relatively high (>50% of
the basin). Other factors like urbanization seemed to be more detrimental to stream shes. Often
high levels of agriculture were associated with low levels of urbanization. They concluded that the
extension of agricultural land use can be used only as a general descriptor, while the importance
of different agricultural practices must be taken into account to explain the impacts biota.
11.6 Biological response thresholds
In the USA, Cuffney et al. (2000) studied the different forms of biological response for each
community (sh, invertebrates and algae) along a gradient of varying agricultural intensity;
25 sites were sampled in 3 different ecoregions. They estimated the pressure from agriculture
using multimetric indices such as pesticide load and NPAI (Non-Pesticide Agriculture Index).
Regardless of the reason for impairment, different responses were detected: linear relations for
sh, and threshold-response for invertebrates and algae. They concluded that the forms of the
biological responses to land use gradients are critically important to the development and
implementation of cost-effective mitigation and monitoring procedures.
Some authors have proposed thresholds values corresponding to a signicant biological alteration
along a gradient of urbanization. Stepenuck et al. (2002) found a sharp decline of invertebrates
communities above a threshold around 8% to 12% of imperviousness. A similar result is obtained
by Wang and Kanehl (2003); beyond 7% of impervious area in the watershed, invertebrate index
values tended to be consistently poor. Roy et al. (2003b) identied a breakpoint of about 15-20%
urban land cover above which the invertebrate indices indicated a clear degradation.
More conceptual approaches have been also proposed. For instance, Addicott et al. (1987) emphasized
the importance of environmental patterns, together with spatial and temporal scales of ecological
processes. Bryce and Clarke (1996) linked landscape-level ecoregions within a basin to bridge the gap
between stream habitat and state-level ecoregions classications. Boulton (1999) reviewed the
denitions for the concept of river health, and the indicators that were chosen for different scales. The
large scale approach is also addressed by Naiman and Turner (2000), who recognised the need for
improved understanding of the processes at the watershed and landscape levels.
11.7 Large scale pressure-impact models: a possible approach
Currently there is no model to link combined pressures and river ecological status (through
biological indicators) at a large or even regional scale that could be applied in Europe. However,
we are currently developing large scale modelling approaches in France. Some of our results are
summarized below.
The models that we are developing aim to relate the ecological status of rivers to indicators of
driving forces and pressures, at the member state level, or for the main hydro-ecoregions. These
models are designed to:
1) extrapolate spatially the possible ecological status of water bodies where there is no monitoring
data; type feature can also be used to evaluate the general pressure level matching the limit of
the good ecological status
2) diagnose: i.e. to set a hierarchy between the sources of impact, and identify the dominant
pathologies of a given region or a river type
3) set restoration priorities: to point out the factors that can be managed, and where to act in
priority.
The rst models have been developed with using the Ecological Quality Ratio (EQR) value of the
French invertebrate index IBGN
2
. The pressure indicators have been estimated using the CORINE
land cover information; a methodological report is available (Wasson et al., 2005). Two examples
are given below.
Spatial extrapolation with decision tree models. The objective is to estimate the current
probability of good ecological status for the main branches of the hydrographic network using
the IBGN index. The method is based on the Decision Tree models (Breiman et al., 1984), which
include four steps (Figure 11.2):
The Land cover (simplied in four categories) is calculated for the catchment of each IBGN
site, and a decision tree is developed relating land cover with the ecological status, simplied
in two categories: good (high and good status) or bad (less than good status).
A decision tree is used to predict the status (good or bad) of the 6100 Hydrological Zones
(sub-catchments) according to the land cover upstream of each subcatchment.
An external validation of this model is made by comparing the status of the sub-catchments
predicted by the model, and the observed status of the sites situated on the main stem of the
corresponding sub-catchment.
2
The IBGN: Indice Biologique Global Normalis, is the standardised invertebrate monitoring method in France; a description
is available at http://starwp3.eu-star.at/detail.php?id=29 and http://www.eu-star.at/pdf/IBGNpresentation.pdf.
From these results, predictive outcomes are used to map the present probability of good ecological
status, based on the IBGN index, for the main branch of each subcatchment (Figure 11.3).
The whole process is repeated using a slightly higher good status boundary to identify
borderline situations, i.e. the sub-catchments that are at the edge of the present good status
denition (Figure 11.3).
Diagnostic of dominant driving forces with Partial Least Square regression models. The
objective here is to run an explicative analysis to identify the categories of land cover that have the
strongest negative or positive correlation with the EQR value of the IBGN index. For each IBGN site,
the land cover is evaluated in the corresponding catchment and in a 3 km long riparian buffer. The
model used is a Partial Least Square regression (PLS; deJong, 1993; Tenenhaus, 1998).
An example of the model results for the whole France a given in Figure 11.4. The results clearly
indicate that the constructed areas in the catchment are impacting most, whereas forested and
natural areas and the low impact agriculture (pastures) in the riparian buffer have the most
positive effect on the IBGN. Further models have been run in the level of the hydro-ecoregions
using more precise land use categories in order to provide more precise diagnostics relevant for
the hydro-ecogerion scale.
Figure 11.2 Pressure-impact models: methodological approach for the spatial extrapolation.
Present probability of good/bad ecological status based on decision tree models at the sub-catchment
scale. G = good status; B = Bad status; W = well predicted; G/B = predicted good/observed bad;
B/G = predicted bad/observed good. See text for details.
Pressures
(land cover)
Biological
stations
Pressures
(land cover)
Subcatchments
Observed proportion
of good sites in
each subcatchment
Pressure/impact models: Principle of spatial extrapolation
Mapping the
probalility of
good/bad status
Application of the
predictive model to
the pressure of the
subcatchments
Creation and validation of a
predictive decision tree from
the le stations-pressures
External
validation
Probable GS
Probable BS
Articialised < X%
Articialised < Y%
Intensive agriculture < Z%
Yes
Yes
Yes
No
No
No
G/B W
5
B/G
1.0
0.8
0.6
0.4
0.2
0.0
G
G B B
11.8 Summary
Despite the strong need, research approaches at the large spatial scale are not well developed.
Due to the traditional focus on one single basin, watershed, or ecoregion, there is a lack of studies
comparing different regions. However, in order to provide tools for large scale decision-making,
especially within the diverse territories in Europe, an regional approach would be very useful to
identify 1) comparable issues even in regions spatially separated and 2) different regional river
Figure 11.3 Map of the present probability of good/bad status and borderline situations at the
subcatchment scale, from pressure/impact models relating invertebrate index (EQR-IBGN) to land cover
(CORINE). Six different models have been elaborated for ve groups of hydro-ecoregions and for the whole
France. On the right, external validation (see gure 11.2 for legend).
Probable good status
Borderline situation
Probable bad status
HER (Hydro-
ecoregions
FRANCE MEDITER.
TYPE
HER 5
HER 1
AND 2
HER 9,
20, 12
HER 3,
4, 21,
17, 19
G/B W B/G G/B W B/G
0.8
0.4
0.0
0.8
0.4
0.0
0.8
0.4
0.0
Figure 11.4 Results of the PLS regression for France. Cross-validated, R2 = 16.3%. Correlation coefcient
between EQR (IBGN) and land cover (from CORINE Land Cover) of the station catchment and riparian
buffer (analysis on 3640 stations).
Station
catchment
Station
riparian
buffer
Built up areas
High impact agriculture
(arable land)
Low impact agriculture
(pastures...)
Forest and natural areas
Built up areas
High impact agriculture
(arable land)
Low impact agriculture
(pastures...)
Forest and natural areas
PLS regression coef. EQR (IBGN)/land cover
0.3 0.2 0.1 0 +0.1 +0.2 +0.3
pathologies that should be managed differently even within a country. This would set a coherent
natural framework for monitoring and management of rivers. So far ecoregional scale has not
been widely used in large-scale studies on pressures-impact relationships.
The proportion and the hierarchy of the different driving forces impacting stream ecosystems are
not clear in the light of the papers reviewed. Although the negative effects of agriculture on the
integrity of the stream ecosystems are generally indicated by several studies, there is not sufcient
scientic basis to quantify the degree of impairment, or discriminate the effects from different kinds
of agricultural practises, or to set the thresholds (of % of agricultural land use, for example).
Differences in the methodologies (selection of bioindicators, methods, or models) may be the
reasons that no clear differentiation has been found in several studies, whether the share of
agricultural vs. urban areas could be most impairing activity. This could be also because the
hierarchical relations between the driving forces vary depending on the level and type of development,
or within the ecoregional context. Ecosystems sensitivity, related to the natural characteristics, may
inuence the intensity of the impairment. Due to the covariation of anthropogenic and natural
landscape features (Allan, 2004), natural constraints may determine both ecosystems sensitivity
and land use practices in different regions, thus leading to different regional responses.
However, the key issue is that the national socio-economic context determine the intensity of the
pressures generated by a given land use, as well as the policies adopted to alleviate the impacts.
For example, the same category of agricultural land cover may have different effects in terms of
pollution, sediment yield etc. on water ecosystems, depending if the practices are intensive or
extensive. In the same way, depending on waste-water treatment practises, two urban areas of
same size may not have the same impacts. In this way the treatment practices may increase or
reduce the intensity of the pressures depending on the local, regional or national socio-economical
context. In conclusion, more detailed treatment practices must be introduced in the system to
better understand the relationships between driving forces and pressures.
Pressures Driving forces
Treatment
practices
The different processes that take place at different spatial scales, like within a riparian corridor vs.
basin scale, need to be understood better (Allan, 2004). It is difcult to conclude which is the
inuence of the local riparian corridor, or the upstream riparian corridor, or the basin or sub-basin
land use conditions on the impacts on the biota at a given site. The review of the results from
previous studies indicate contrasting results; some studies identify landscape-level factors as the
main variables for predicting biological indicators, while others nd no relationships beyond the
local scale measurements. It is not clear whether these discordances are due to the different
anthropogenic factors, natural characteristics, or due to the different methods or indicators used.
Nevertheless, all these factors and scales have to be considered when developing models. The
restoration of river corridors is often seen as the most important action to improve the ecological
status. Still, the actual buffering capacity of riparian areas needs to be properly evaluated.
The same applies with the existence of thresholds in the relations between driving forces and
ecological status. Some managers are wondering if a good ecological status is attainable in
heavily urbanized basins, or in intensively cultivated areas. Some of the models indicate threshold
levels, beyond which streams are systematically impaired. A comparison and validation of such
models should be made within the European territory. This is a complicated issue as the different
biological compartments may not have the same response curve to a given pressure. Another
important related question is the feasibility of spatial extrapolation of ecological status on the
basis of land use in water bodies which are not monitored.
In general, it seems that the present knowledge is not sufcient to give clear advice to decision
makers (Allan, 2004). The large scale approaches are relatively recent, so there is a lack of
conceptual models and experiences. Nevertheless, large scale analysis comparing pressure-impact
relationships across different spatial scales, countries, and ecoregions would still be necessary for
the implementation of the WFD in Europe.
11.9 References
Addicott, J.F., J.M. Aho, M.F. Antolin, D.K. Padilla, J.S. Richardson, and D.A. Soluk. 1987. Ecological
neighbourhoods: scaling environmental patterns. Oikos, 49: 340-346
Allan, J.D. 2004. Landscapes and Riverscapes: the inuence of land use on stream ecosystems. Annual
Review of Ecology Evolution Systematic, 35: 257-284
Allan, J.D., D.L. Erickson, and J. Fay. 1997. The inuence of catchment land use on stream integrity across
multiple spatial scales. Freshwater Biology, 37: 149-161
Bis, B., A. Zdanowicz, and M. Zalewski. 2000. Effects of catchment properties on hydrochemistry, habitat
complexity and invertebrate community structure in a lowland river. Hydrobiologia 422: 369-387
Borchardt, D., and S. Richter. 2003. Identication of signicant pressures and impacts upon receiving waters.
Water Science and Technology, 48: 33-38
Boulton, A.J. 1999. An overview of river health assessment: philosophies, practice, problems and prognosis.
Freshwater Biology, 41: 469-479
Breiman, L., J.H. Friedman, R.A. Olshen and C.J. Stone. 1984. Classication and regression trees. Wadsworth
International Group, Belmont, California, USA
Bryce, S.A., and S.E. Clarke. 1996. Landscape-level ecological regions: linking state-level ecoregion frameworks
with stream habitat classications. Environmental Management 20: 297-311
Cormier, S.M., M. Smith, S. Norton, and T. Neiheisel. 2000. Assessing ecological risk in watersheds: A case
study of problem formulation in the Big Darby Creek watershed, Ohio, USA. Environmental Toxicology and
Chemistry, 19: 1082-1096
Cuffney, T.F., M.R. Meador, S.D. Porter, and M.E. Gurtz. 2000. Responses of physical, chemical, and biological
indicators of water quality to a gradient of agricultural land use in the Yakima River Basin, Washington.
Environmental Monitoring and Assessment, 64:259-270
deJong S. 1993. SIMPLS: An alternative for partial least squares regression. Chemometrics and Intelligent
Laboratory Systems, 18:251-263
Dovciak, A.L., and J.A. Perry. 2002. In search of effective scales for stream management: Does agroecoregion,
watershed, or their intersection best explain the variance in stream macroinvertebrate communities?
Environmental Management, 30: 365-377
EEA (European Environment Agency). 1991. Europes Environment. The Dobris Report. European Environment
Agency, pp. 676
EEA (European Environment Agency). 2003. Europes water: an indicator-based assessment. Luxembourg:
Ofce for Ofcial Publications of the European Communities. pp. 97
Gergel, S.E., M.G. Turner, J.R. Miller, J.M. Melack and E.H. Stanley. 2002. Landscape indicators of human
impacts to riverine systems. Aquatic Sciences, 64: 118-128.
Harding, J.S., E.F. Beneld, P.V. Bolstad, G.S. Helfman, and E.B.D. Jones. 1998. Stream biodiversity: the ghost
of land use past. Proceedings of the National Academy of Science USA, 95: 14843-14847
Karr, J.R. 1981. Assessment of biotic integrity using sh communities. Fisheries, 6: 21-27
Lammert, M., and J.D. Allan. 1999. Assessing biotic integrity of streams: Effects of scale in measuring the
inuence of land use/cover and habitat structure on sh and macroinvertebrates. Environmental
Management, 23:257-270
Lange-Bertalot, H. 1979. Pollution tolerance of diatoms as a criterion for water quality estimation. Nova
Hedwigia 64:285-304
Meador, M.R., and R.M. Goldstein. 2003. Assessing water quality at large geographic scales: Relations
among land use, water physicochemistry, riparian condition, and sh community structure. Environmental
Management, 31:504-517
Naiman, R.J., and M.G. Turner. 2000. A future perspective on North Americas freshwater ecosystems.
Ecological Applications, 10:958-970
Pan, Y.D., A. Herlihy, P. Kaufmann, J. Wigington, J. Van Sickle, and T. Moser. 2004. Linkages among land-use,
water quality, physical habitat conditions and lotic diatom assemblages: A multi-spatial scale assessment.
Roth, N.E., J.D. Allan, and D.L. Erickson. 1996. Landscape inuences on stream biotic integrity assessed at
multiple spatial scales. Landscape Ecology, 11:141-156
Roy, A.H., A.D. Rosemond, D.S. Leigh, M.J. Paul, J.B. Wallace. 2003a. Habitat-specic responses of stream
insects to land cover disturbance: biological consequences and monitoring implications. Journal of the
North American Benthological Society 22:292-307
Roy, A.H., A.D. Rosemond, M.J. Paul, D.S. Leigh, and J.B. Wallace. 2003b. Stream macroinvertebrate response
to catchment urbanisation. Georgia, USA. Freshwater Biology, 48:329-346
Sponseller, R.A., E.F. Beneld, and H.M. Valett. 2001. Relationships between land use, spatial scale and
stream macroinvertebrate communities. Freshwater Biology 46:1409-1424
Steedman, R.J. 1988. Modication and assessment of an index of biotic integrity to quantify stream quality
in Southern Ontario. Canadian Journal of Fisheries and Aquatic Science, 45:492-501
Stepenuck, K.F., R.L. Crunkilton, and L.Z. Wang. 2002. Impacts of urban land use on macroinvertebrate communities
in south-eastern Wisconsin streams. Journal of the American Water Resources Association, 38:1041-1051
Stewart, J.S., L. Wang, J. Lyons, J.A. Horwatich, and R. Bannerman. 2001. Inuences of watershed riparian-
corridor, and reach-scale characteristics on aquatic biota in agricultural watersheds. Journal of the American
Water Resources Association 37:1475-1487
Tenenhaus, M. 1998. La rgression PLS, thorie et pratique. Technip, Paris
Van Dam, H., A. Mertens, and J. Sinkeldam. 1994. A coded checklist and ecological indicator values of
freshwater diatoms from the Netherlands. Netherlands Journal of Aquatic Ecology 28:117-133
Walser, C.A., and H.L. Bart. 1999. Inuence of agriculture on in-stream habitat and sh community structure
in Piedmont watersheds of the Chattahoochee River System. Ecology of Freshwater Fish 8:237-246
Walsh, C.J. 2000. Urban impacts on the ecology of receiving waters: a framework for assessment, conservation
and restoration. Hydrobiologia, 431:107-114
Wang, L., J. Lyons, P. Kanehl, and R. Bannerman. 2001. Impacts of urbanization on stream habitat and sh
across multiple spatial scales. Environmental Management, 28:255-266
Wang, L., and P. Kanehl. 2003. Inuences of watershed urbanization and instream habitat on
macroinvertebrates in cold water. Journal of the American Water Resources Association 39:1181-1196
Wasson, J.G., A. Chandesris, H. Pella, and L. Blanc. 2002. Typology and reference conditions for surface water
bodies in France: the hydro-ecoregion approach. TemaNord, 566:37-41
Wasson, J.G., B. Villeneuve, N. Mengin, H. Pella, and A. Chandesris. 2005. Modles pressions/impacts: Approche
mthodologique, modles dextrapolation spatiale et modles de diagnostic de ltat cologique bass sur les
invertbrs en rivire (IBGN). Rapport CEMAGREF Lyon BEA/LHQ, pp. 61
RESPONSES OF AQUATIC BIOTA TO COMBINED PRESSURES | 225
12 Responses of aquatic biota to combined pressures
from nutrients and priority substances
Constanze OToole and Kenneth Irvine
12.1 Introduction
Eutrophication and pollution are both major pressures on aquatic systems that often co-occur.
Studies (Schaaning et al., 1996 and Gunnarson et al., 1996) show that the eutrophication can
inuence the fate of contaminants in the system and that pollutants can affect the response of
ecosystems to nutrients. The interactions between the pressures are highly complex and may
inuence all levels of the ecosystem. The effects of eutrophication and contamination have been
mostly investigated separately in the past and potential interactions were not considered. However,
the reduction/increase of the concentration of only one pressure may alter the fate of the other.
For the prediction of effects of changes in inputs of nutrients, organic matter and contaminants
on the aquatic system, interactive mechanisms that may alter bioavailability and the fate of
contaminants have to be considered.
Two closely connected research programmes investigating interactions between eutrophication
and contamination in aquatic systems were launched in 1995 for a ve-year period. The Swedish
EUCON (Eutrophication and Contaminants in the aquatic environment) project examined
interactive processes between eutrophication and contaminants (selected organic pollutants and
some metals) in marine and freshwater environments. The connected Norwegian strategic
programme MILE (Interaction between contaminants and eutrophication) investigated interactive
processes relating to the benthos and pelagial in the marine environment focussing on selected
heavy metals, metal organic compounds (e.g. TBT), polycyclic aromatic hydrocarbons (PAHs),
polychlorinated biphenyls (PCBs) and pesticides (e.g. DDT).
Interactive mechanisms included both in EUCON, MILE and other combined effect studies were:
biodilution and growth dilution, sedimentation and sediment sequestering, the importance of
lipids, food quality, organism characteristics and food web structure. In this chapter we summarise
the main interactive mechanisms and introduce case studies investigating the responses of biota
to contaminants in systems of different trophic regimes for the pelagial and benthos. Pioneer
models that are capable of simulating chemical fate/effect and food chain accumulation at
varying trophic levels are reviewed in regards to risk assessment and as decision tool for integrative
ecosystem management.
12.2 Biomass and growth dilution
Increased inputs of nutrients into aquatic systems can lead to eutrophication which includes
increase in productivity and mean standing biomass in the aquatic system. When the system is
impacted by both contaminants and increased nutrients the elevated levels of particulate organic
carbon (POC) may increase the uptake of certain toxic compounds leading to a dilution effect.
Olsson and Jensen (1975) rst formulated the biomass dilution concept assuming that if the
same amount of a bioaccumulating substance is given in equal amounts to eutrophic and
oligotrophic waters, higher concentrations of contaminants will accumulate in the biota of the
oligotrophic system. Especially, hydrophobic organic compounds (HOCs) have the ability to bind
to various POCs matrices due to their limited water solubility (Jonsson et al., 1996).
A process similar to biomass dilution is growth dilution. This analogous reaction occurs if growth
rates of autotrophic organisms e.g. phytoplankton are enhanced. Contaminants bound to
phytoplankton are diluted by growth of the phytoplankton in such way that equilibrium
concentrations may never be attained. Taylor et al. (1996) recommended that the impact of
growth dilution should be related to the relative magnitudes of the uptake constant for
contaminants and their exponential rate of population increase.
12.3 Pollutant uptake mechanisms by organic matter
Mackay and Powers (1987) suggested that the binding process between HOCs and POCs is loose
sorption in which an organic chemical reduces the total organic-water interfacial area. Energy is
released when the pollutant associates loosely with the natural organic surface and water is
displaced from the surface and from part of the chemical. Binding is reversible as a result of
particle collisions. Sorption can also occur in pores or sites in the organic matter. In that case,
HOCs become surrounded with organic matter that has aggregated and the sorption becomes
more non-liable or resistant (Jonsson et al., 1996). The relationship was quantied using a mass
transfer approach. This equilibrium partitioning theory suggests that organic contaminant uptake
is inversely related to organic carbon content of food particles.
In a more recent study, Swackhammer and Skoglund (1993) observed in PCB uptake experiments
using the algae Scenedesmus sp. under growth-promoting and growth-inhibiting conditions that
bioaccumulation of PCB in lipid cells was slower than previously thought. Swackhammer and
Skoglund (1993) and Stange and Swackhammer (1994) describe the uptake of HOCs by
phytoplankton as a two-step process. First, a fast sorption from the water to the cell wall then a
slower transfer of the pollutant to the interior of the cell occurs. The second step increases the
concentration of HOCs in the plankton cell substantially. This process prevents the HOC
concentration from reaching equilibrium and levels become lower in the fast growing biomass of
eutrophic systems in comparison with oligotrophic systems (see section 12.2, growth dilution).
Therefore, models predicting the transport and fate of PCBs in aquatic systems have to consider
uptake kinetics.
Sijm (1996) describes an analogous process to the sorption mechanisms of HOC to phytoplankton, in
which organic carbon is returned to the aqueous phase by algae through the production and excretion
of exudates, which are also capable of sorption of HOCs. The study revealed that the standard
centrifugation technique to separate algae and water to measure the bioconcentration factor (BFC)
does not separate the exudates from the aqueous phase. As a result, the presence of exudates with
bound HOCs might lead to an overestimated concentration of pollutants in the water.
Metal binding processes to organic matter are more complex and probably only indirectly related
to POCs. Certain metals such as lead (Pb) and cadmium (Cd) have the capacity to react directly
with phosphate groups in the lipid bilayer of various cell membranes (Jonsson et al., 1996).
The nature and composition of organic matter is important for the HOC dynamics in water.
The sorption and accumulation of HOCs in phytoplankton are inuenced by a variety of factors:
The binding of lipophilic persistent compounds such as HOCs is related to the amounts
of lipids and their quality in the different algae species (Larsson, 1997 and Skei et al., 2000).
It is believed that the cell wall composition can inuence the initial adsorption of the
pollutant (Skei et al., 2000). The lipid composition of the pelagic community inuences
the uptake of HOCs from water.
Dissolved organic carbon (DOC) matrices inuence the solubility of HOCs (Gjessing and
Berglind, 1981; Hassett and Anderson, 1979 and McCarthy and Jimenez, 1985) and are
crucial for the availability of HOCs in the water to the biota. According to Jonsson et al.
(1996), DOC also has a gluing effect during the aggregation process that facilitates the
formation and/or growth of different particles and colloid types. Taylor et al. (1997) point out
that DOC is a large pool relative to POC in lakes and that changes in DOC caused by changes
in the trophic status can inuence the bioaccumulation of HOCs signicantly.
The importance of DOCs for the distribution and dynamics of many metals (e.g. Pb, Cu, Ni,
Co, Zn, Cd, Fe, Mn and Mg) has been demonstrated in natural sea water. DOC forms
complexes with the metals increasing their solubility and altering the distribution between
the oxidised and reduced form (Jonsson et al., 1996). Frstner and Wittman (1983) mention
that DOCs can inuence the extent to which metals adsorb on particular matter and thereby
affect the stability of colloids that contain metals.
The trophic status of the system impacts on the phytoplankton composition. The biomass of
many eutrophic lakes is dominated by blue-green algae such as Planktothrix, Anabaena and
Apanizomenon (Larsson, 1997). These species contain lower amounts of lipids and posses
different types of cell walls in comparison with diatoms and green algae. Further,
cyanobacteria can be less impacted by zooplankton through food selectively. Therefore, there
may be no signicant biomagnication of HOCs along the pelagic food chain (phytoplankton-
zooplankton- sh) through the blue-green algae link (Larsson, 1997). Oligotrophic lakes
normally display a diverse phytoplankton community that is mostly dominated by diatoms
and green algae (Larsson, 1997). Both of these groups contain high amounts of lipids and
may absorb higher concentrations of lipophilic pollutants (Skei et al., 2000).
Larsson et al. (2000) point out that growth rate of phytoplankton can be high in eutrophic
systems, inducing different persistent organic pollutants (POP) uptake kinetics in comparison with
nutrient-poor systems. The eutrophic system may display lower concentrations of POPs in pelagic
producers. The effect is most obvious during seasons (e.g. spring and summer) or situations in
which plankton have high growth rates.
12.4 Pollutant transfer at air-water interface
Ecological changes such as the transition of aquatic system from oligotrophic to eutrophic can
inuence the exchange of PCBs at the air-water interface. During the eutrophication process the
phytoplankton abundance increases resulting in the partitioning of a higher portion of PCB in the
particulate state (see section 12.2, biomass and growth dilution). The lower concentration of PCBs
in the dissolved state that is subject to volatilisation leads to a decrease in the ow of PCB from
water to the air. The new equilibrium supports the antagonistic process of PCB transport from air
to water. This process is believed to be affected by the abundance of phytoplankton in the water
column (Larsson, 1997).
12.5 Hydrophobic organic compounds in pelagic food chain
and in different trophic regimes
In the pelagic ecosystem there is a classical distinction between the herbivorous (e.g. phytoplankton)
and microbial (e.g. bacteria, protozoa) food webs. Plankton organisms are the rst link for pollutant
transfer in pelagic systems. Mainly phytoplankton are considered the initial step in the POPs
transfer but the importance of bacteria was recently demonstrated in the initial transport of POPs
in the pelagic food chain (Swackhammer and Skoglund, 1993; Stange and Swackhammer, 1994;
Wallberg and Andersson, 1999; and Larsson et al., 2000). The high abundance and small size of
bacteria result in fast turnover times (hours to days) and the effect of growth dilution (see section
12.2, growth dilution) is reduced, because of the shorter time span needed to reach equilibrium.
Bacteria cells also often form the largest particulate biological surface area in natural waters.
According to Larsson et al. (2000) bacteria can potentially take up larger proportions of POPs
from the water column than phytoplankton.
Wallberg and Andersson (1999) developed a simple procedure (ethanol wash) to distinguish
between adsorbed and absorbed PCBs in seawater microorganisms (e.g. bacteria, nanoagellates
and ciliates). Absorption of higher chlorinated PCBs was mainly observed in microplankton
(>10 m, diatoms and dinoagellates), but not for smaller sized bacteria and dinoagellates
(size) that adsorbed HOCs. It was suggested that bacterial membrane physiology, feeding
behaviour of agellates and low incubation time are responsible for the observed differences.
Mesocosm experiments with seawater showed that high proportions of added PCBs initially sorped
to the picoplankton fraction (0.2 2 m) and a transfer of the higher-chlorinated PCBs from the
bacterial size fraction to the predator size fraction (>10 m, e.g. bacterivorous ciliates) occurred.
Environments with dominating heterotrophic microbial food webs (e.g. as might be found in many
nutrient poor systems) had larger food web transfers of PCBs (Larsson et al., 2000). The environmental
fate of the pollutant depends on a number of processes in the pelagic zone such as the partitioning
between water and particles and lipophilic compartments in the water (see section 12.3). Factors
inuencing the available time period for the establishment of the equilibrium are durations of
various biological cycles (e.g. growth and degradation) involved in the partitioning processes (e.g.
adsorption, absorption and desorption) and sedimentation rates (Skei et al., 2000).
The length of the food chain affects increases in POP concentrations in pelagic food webs. The top
consumer lake trout in systems with an additional trophic level displayed higher levels of POPs in
comparison with lake trout in systems with shorter food chains (Larsson et al., 2000).
Larsson et al. (1992) investigated the concentration of organochlorines in pike in contaminated
Swedish lakes. The concentrations of organochlorines were signicantly lower in pike from
eutrophic lakes than in comparable contaminated oligotrophic lakes. Larsson et al. (1992) did not
attribute the differences to dilution effects and rather argued that higher growth rates, higher
turnover times and increased sedimentation of pollutants adsorbed to particles lead to the
decreased contamination levels of biota in eutrophic systems.
Taylor et al. (1996) reported high contamination levels of HOCs in large piscivorous sh (especially
salmonids) from the Great Lakes in southern Ontario. Fish in smaller oligotrophic to mesotrophic
lakes had higher HOC concentrations than eutrophic lakes and eutrophic areas within the Great
Lakes. A number of mechanisms were stated to be responsible for changes in nutrient loading
that may change contaminant levels in biota such as: biomass dilution; changes in atmospheric
and/or sediment exchanges; changes in storage lipid content of organisms and changes in food
web structure. It was also revealed that food web structure and lipid content (especially in sh)
have a strong impact on contaminant levels.
Larsson et al. (1998) investigated an oligotrophic and eutrophic lake in southern Sweden that
both received approximately the same atmospheric deposition of PCB. Higher concentrations of
PCB were recorded in the phytoplankton of the oligotrophic lake in comparison with the eutrophic
lake. This was attributed to the higher lipid content found in the phytoplankton community in the
nutrient-poor system. The biota in the oligotrophic system also contained higher amounts of
hydrophobic substances than in the eutrophic system. The concentration of pollutants in plankton
was lower than that found in the sedimenting material indicating that the degradation of lipid
and organic carbon exceeded the release of persistent pollutants from settling particles. Also, no
biomagnication from phytoplankton to zooplankton was detected. Larsson et al. (1998)
concluded that the higher lipid content and higher amounts of pollutant found in the primary
producers suggests a mechanism that lead to high pollutant concentrations in zooplankton and
sh in oligotrophic systems. The major part of the persistent pollutants cycling in the investigated
Swedish lakes was attributed to internal lake processes.
Jeremiason et al. (1999) examined the stresses of trophic condition on airwater exchange and
settling uxes of polychlorinated biphenyls (PCBs) in a paired lake experiment in the Experimental
Lake Area (ELA) in Canada. In the eutrophic lake lower PCB concentrations were detected on
suspended particles, higher PCB settling uxes and lower volatilisation compared to the
oligotrophic lake.
Hylland (1999a,b) carried out a time-trend analysis (1984 1996) of contaminants in cod (Gadus
morhua) in two areas with different eutrophication regimes in the Norwegian Oslofjord. Inputs of
PCBs and trace metals were considered constant between 1981 and 1997. The inner Oslofjord
was characterised by major changes in trophic status over time, while the outer Oslofjord displayed
smaller changes at the same time and was used as a reference site. Cod quality collected from the
outer Oslofjord was reasonably homogenous over time. Cod samples from the inner Oslofjord were
markedly smaller and had lower lipid content from 1984 to 1987 than cod collected between
1988 and 1996. Liver concentrations of the selected PCB, CB 153, differed at both sites between
the years. A similar pattern was observed for the DDT metabolite, DDE. There was no obvious link
between tissue concentrations of the organic compounds and eutrophication status for the inner
Oslofjord. It was concluded that environmental inuences such as oods and not eutrophication
are likely to be responsible for the observed pollutant levels in cod.
Sderstrm et al. (2000) investigated interactions between eutrophication and organic pollutants:
PCB; DDTs and -HCH in a eutrophic and oligotrophic lake during 1996-1997 in Sweden. The
morphometry, water chemistry and atmospheric exposure of both lakes were similar. The eutrophic
lake displayed higher levels of PCB based on organic carbon and lipid mass in sedimenting matter
in: surface sediment; biota (Perch); higher settling uxes of particulate matter and sediment traps.
No biomass dilution effect was detected. It was suggested that phytoplankton was responsible for
most of the PCB transport to the sediment in the eutrophic lake. Higher DDD/DDE ratios were
related to a high load of p,p-DDT rather than to anaerobic conditions. The oligotrophic lake
showed high DDD/DDE ratio and high sDDT levels in the deep sediment core. The highest ratio
in the sediment traps coincided with the highest runoff of parental DDT from the watershed.
Lithner et al. (2000) studied the direct and indirect effects of eutrophication on turnover of metals
in a paired lake experiment in Sweden. The eutrophic and oligotrophic/mesotrophic lakes were
the same study sites used by Sderstrm et al. (2000; see above). Decreased levels of Pb and Hg
in the eutrophic lake sediments were attributed to signicant dilution effects of pollutants in
seston. The biota (sh) in the nutrient-rich system showed decreased bioavailability of Cd and Tl
but not for Pb and Hg. The observed higher uxes of metals in the eutrophic lake were caused by
the larger biomass and by a different geology and pH. The observed decreased residence time of
Cu, Ni and Tl was attributed to sulphide xation and a reduced concentration of Cd and Mo in
the water column caused by the increased biomass and efcient scavenging of metals.
12.6 Sedimentation and sediment dynamics of hydrophobic
organic compounds
In eutrophic systems the increased phytoplankton biomass after a bloom may lead to increased
settling of cell-bound POPs after a time lag between the collapse of the bloom and the gravitational
settling of organic matter. In stratied and stagnant areas anoxia might also develop as a result
and laminated sediments may indicate long periods of anoxia (Skei et al., 2000). Another effect
of eutrophic systems is an increased sedimentation rate of organic matter that scavenges HOCs
from the water mass. This might lead to an increased fraction of nonmetabolised organic matter,
which can result in increased loads of organics and HOCs to the sediment and their biota (Skei et
al., 2000). The sedimentation of POPs has greater signicance in nutrient-rich systems compared
with nutrient-poor systems in which the transfer of POPs through the pelagic food web is of
greater importance (Larsson et al., 2000).
Sedimentation has an impact on retention time of pollutants in the water mass. The linkage of
the empirical relationship between primary production and sedimentation rates with the trophic
status of aquatic systems is essential to determine retention time. Jonsson (1997) estimated an
average retention time for sPCB in the Baltic Sea water mass of less than one year indicating that
sediments act as efcient traps of PCB.
It was hypothesized that increased sedimentation rates due to eutrophication may have a
signicant impact on the scavenging of contaminants from the water mass to the sediment
(Jonsson, et al., 1996). Jonsson et al. (2000) found that eutrophication-induced scavenging of
PCBs and DDTs are only of minor importance in the Baltic Sea. The main controlling factors for
scavenging were resuspension and erosion. It was concluded that it is most likely that eutrophication-
induced scavenging occurred in the Baltic Sea, but natural processes that caused highly variable
burial rates obscured the evidence.
In contrast to the equilibrium partitioning theory (see section 12.3) recent studies (Axelman et al.,
2000; and Larsson et al., 2000) demonstrated that the uptake of organic contaminants is
proportional to the biomass quality. During the gravitational settling of seston from the photic to
the profundal zone, the organic carbon and lipid contents decreased. Larsson et al. (2000)
attributes this decrease to microbial attacks on settling particles. Through this process energy and
nutrients are released and made available for microorganisms. The POP concentration does not
decrease proportionate to the lipid and organic carbon contents during this process but increases
by approximately 10-fold, instead. The POP increase in settling particles varied with environment,
season and concentration. Axelman et al. (2000) point out that traditional equilibrium partitioning
models might underestimate the prediction of HOC concentrations in settling particles and
sediments from concentrations dissolved in the aqueous phase.
The increase of POP concentrations during vertical particle settlement due to microbial attack has
consequences for POP cycling behaviour: the reduction of lipid and carbon traces of particles at
the sediment surface (Axelman et al., 2000; and Larsson et al., 2000); greater incorporation of
POPs into eutrophic, organic/lipid-rich sediments and, consequently, lower pollutant exposure to
pelagic biota in comparison with oligotrophic systems (Larsson et al., 1998; and Larsson et al.,
2000) and high exposure of POPs to profundal benthos foraging on sedimenting matter (Kucklick
and Baker, 1998; and Larsson et al., 2000). Eckhll et al. (2000) and Persson and Jonsson (2000)
investigated laminated sediments in offshore areas in the Baltic Sea and discovered that gross
deposition rates are governed by frequencies of high winds.
Ignatius (1958) rst estimated a mean accumulation rate of approximately 1 mm year
-1
over
1000 to 7000 years for the Baltic Sea based on varves in old postglacial Baltic sediments. Jonsson
et al. (1990) used the same technique and estimated accumulation rates in the North-West Baltic
proper of 1.5 4.2 mm year
-1
in the unconsolidated upper 5 cm of the sediment during the last
few decades. Jonsson (2000) estimated accumulation rates of 1 - 4.3 mm year
-1
for deep areas in
the Baltic Sea based on several studies (Balzer, 1984; Kunzendorf and Christiansen, 1997;
Neumann et al., 1997; Niemist and Voipio, 1974; Niemist and Voipio, 1981; stlund and
Hallberg, 1991; Suess and Erlenkeuser, 1975) that used radionuclids for sediment dating.
Accumulation rates varied with time and between areas (Jonsson, 2000).
Wassmann (1980) investigated the load of POC during lengthy time intervals to the aphotic zone
and the bottom of coastal areas of the boreal North Atlantic using the power model P
E
= 0.049
P
T
1.41
(n = 16; r
2
= 0.94), where P
E
is the sedimentation of POC out of the euphotic zone and P
T
is
the total primary production. The trend in annual P
T
estimates from the Kattegat was studied for
a time interval of 40 years. Substantial increase of P
T
was recorded in the southern Kattegat and
interpreted as a consequence of eutrophication. P
E
increased by about 140 and 250% during that
time period. It was concluded that the increase in POC ux to the aphotic zone was sufcient to
reduce the oxygen concentration in an 8 m deep layer of bottom water to zero. Widespread
occurrences of hypoxia and anoxia in the sediment were attributed to increases in export of
organic matter to the bottom layer caused by eutrophication.
Sequestering in the sediment
The sequestering factors relate to the control of subsequent capturing of contaminants in the
sediment. According to Jonsson et al. (1996), variables determining sequestering efciency are:
redox conditions; abundance of benthic fauna; bioturbation depths and sediment carbon content.
Jonsson et al. (1996) suggested the use of areas of laminated, anoxic sediments for a rough
estimate of the sequestering capacity for some contaminants in the Baltic. Jonsson and Carman,
(1994) suggested that the sequestering of carbon deriving from primary production and external
sources has increased by a factor of 5-10 from 1900 to the late 1970s in the Baltic proper.
Sequestering of metals in the sediments is inuenced by sulphides. Eutrophication may increase
the ux of metals sorbed to organic matter and, therefore, the probability of anoxia, sulfphide
formation and sequestering of metals in sediment as sulphides (Lithner et al., 2000).
Resuspension of contaminants
PCB carrying particles normally complete a number of resuspension events before being trapped
in laminated depositional areas of the sediments. The time between the rst deposition of particles
and their associated pollutant to its nal burial can be years to decades. During the transport
particles are subjected to oxic and anoxic degradation that can later lead to slow degradation of
the PCBs once the particle is nally trapped in the laminated sediment (Skei et al., 2000). Strong
energy inputs from waves, currents or submarine slides are mainly responsible for resuspension
events of particle bound PCBs in sediments (Jonsson, 2000). In sediment studies (especially,
retrospective studies) substantial delay mechanisms through resuspension in pollutant-associated
particles have to be considered.
Sulphide oxidation and suspended particulate matter (SPM) uxes in aquatic systems are related
closely to metal resuspension. Lithner et al. (2000) investigated turnover of metals in lakes with
different trophic regimes and found that resuspension and water circulation dominated the ux
of particles at times and biased the overall trend of decreasing Pb loads in sediments due to
decreased atmospheric Pb loads. For substantial improvements of contaminant concentrations
after remedial measures a time lag has to be expected, since sediments can still act as pollutant
sources through resuspension events.
Sediment burial of pollutants and laminated sediments
Eutrophication-induced hypoxia or anoxia can change the benthic system profoundly. Macrobenthic
fauna may become extinct due to a lack of oxygen and the sediments are no longer bioturbated.
Under these circumstances, homogenous laminated sediments might form (Jonsson, 1996). In the
Baltic proper spreading oxygen deciency due to eutrophication led to great changes of the
benthic system since the mid-20
th
Century. Recent estimates suggest that one-third of the bottom
of the Baltic proper was laminated by the late 1980s without any benthic faunal activity (Skei et
al., 2000). Eckhll et al. (2000) and Persson and Jonsson (2000) investigated laminated sediments
in offshore and archipelago in the Baltic Sea suggesting that wind speed frequencies govern gross
deposition rates. As a consequence, particles that nally settled in areas with continuous
deposition of ne material show a highly variable age (days to thousands of years) of carbon (Skei
et al., 2000). Jonsson (2000) stated that Baltic Sea sediments are competent traps of PCB based
on the estimated average retention time of sPCB in the water mass of less than one year. High
sPCB burial rates in the Baltic proper also indicate that laminated sediments are more efcient
traps of PCBs than bioturbated sediment. A comparison between annual burial of PCB in
predominantly laminated sediments of the Baltic proper and bioturbated sediments in the Gulf of
Bothnia indicated that the Baltic had 4 times higher annual burial of PCB.
A large-scale investigation of PCBs in the Baltic Sea sediments was carried out in 1993 in
cooperation of the Baltic States. An increase of sPCBs (IUPAC # 52, 101, 118, 105, 153, 138, 180)
from North to South in surcial sediments was observed indicating atmospheric inuences. Dated
sediment cores from the Baltic proper showed increased or constant concentrations of sPCB during
recent decades, while the pelagic biota contained decreasing PCB concentrations. Down-core
congeners patterns displayed no changes. The calculation of the sediment burial of sPCB suggested
that laminated sediments in the Baltic proper act as efcient traps. A simple mass balance model
indicated that sPCB retention time in the water mass was less than one year (Jonsson, 2000).
Sediments can not be considered as nal sinks for metals due to resuspension (see above,
resuspension of contaminants) events that can reactivate metals bound to sulphides and organic
matter when in contact with oxygenated waters.
Bioturbation of sediments and its effect on contaminant cycling
Bioturbation of sediments by benthic fauna can increase signicantly the release of organic
contaminants to the water column, especially in heavily contaminated sediments. Organic
contaminants are redistributed in the top sediment by bioturbation activities of the benthic fauna.
The process of bioturbation involves: mixing of particles; increased oxygenation; increased break-
down and transformation of organic-rich and contaminated particles (Jonsson, 1996). During this
process a great proportion of the organic contaminants may be cycled back to the pelagic system.
Benthic invertebrates also metabolise contaminants changing their chemical structure or
detoxifying them (Skei et al., 2000). Eckhll et al. (2000) observed in sediments that changed
from laminated to oxygenated status a recolonisation by macrobenthic fauna. The animals
bioturbated the sediment down to approximately 3 cm, which may increase the contaminant ux
out of the top sediment. Skei et al. (2000) reported that a turnover from bioturbated to laminated
sediments might improve PCB sediment concentrations in the order of 50%. Many benthic species
also pump water into the sediment during bioturbation affecting chemical benthic processes such
as: nitrication and denitrication.
12.7 Hydrophobic organic compounds in benthic food chain
The eutrophication process initially increases the benthic invertebrate fauna through increased
production of organic matter and, therefore, increased food availability; while at a later stage loss
in diversity may occur due to deteriorated oxygen conditions in the sediment. Species tolerant to
low oxygen conditions may become dominant (Skei et al., 2000). In nutrient-rich systems increased
sedimentation rates of organic matter, additionally, scavenges HOCs from the water column and
organic and HOC loads to the sediment may increase as a consequence (Skei et al., 2000). Hoc
concentrations might further increase during vertical settlement of particles associated with
contaminants (see section 12.6) depending on season and environment. These processes may
lead to higher exposure of benthic fauna to HOCs. Experimental studies by Gunnarson et al.
(1996) indicate that the addition of labile organic matter increases the bioaccumulation of
organic contaminants (PAHs) in benthic organisms. These ndings contradict previous assumptions
that enrichment of organic matter in sediments can counteract the bioavailability of HOCs that
might be more rmly attached to organic matter in comparison with minerogenic matter
(Horzempa and DiToro, 1983).
Evers and Smedes (1996) looked at the seasonal variations of PCB and PAH concentrations in
SPM and blue mussels (Mytilus edulis) in the North Sea. Exposure experiments were carried out in
the North Sea at different times of the year (February, July and October) at 16 locations over a
period of 5 years. Additionally, SPM samples were collected quarterly by a continuous-ow
centrifuge. Short-term changes in HOC concentrations were detected in SPM and phytoplankton
that were synchronised within and between different areas. It was suggested that the synchrony
indicated large-scale forces (e.g. certain weather patterns) controlling short-term patterns of HOC
concentrations in suspended matter and various marine organisms. Spatial and temporal trends
in PCB concentration were detected in benthic mussels. Estuarine and coastal locations that
receive contaminants from uvial sources displayed higher PCB concentrations than offshore
locations. Increases of dissolved and particulate organic carbon during periods of high growth
inuenced contaminant concentrations in offshore-SPM and body burdens in mussels. The used
dynamic model predicted the observed SPM accumulation more accurately in comparison with
traditional steady-state models.
Gilek et al. (1996) examined the bioaccumulation and cycling of HOCs in the blue mussel (Mytilus
edulis) from the Baltic Sea at different trophic regimes in a combined laboratory/eld
experimentation and dynamic modelling approach. The rate of HOC uptake was inuenced by
using up to two orders of magnitude by changes in algal food concentrations, while HOC depuration
remained unaffected. It was suggested that trophic-dependent changes were caused primarily by
the inuence of food ration on the physiological activity of the bivalves. Changes in PCB
bioavailability governed by partitioning between water and organic matter (algal food) inuenced
the HOC uptake rate to a lesser degree. For the establishment of a dynamic model, budgets of the
ow of organic carbons and important HOC groups (PAHs, polychlorinated dibenzo-p-dioxins,
furans and PCBs) through the M. edulis community inhabiting the Swedish coastal zone of the
Baltic proper were constructed. The HOC budgets revealed that the bivalve community signicantly
modied the cycling of HOCs in the Baltic proper coastal zone. The net deposition of HOCs to the
bottoms was increased by 13-19%. This increase resulted in a higher availability of HOCs to other
benthic organisms. High bioaccumulation of the contaminants in the tissue of the mussels also
resulted in a higher contamination of demersal sh and waterfowl that feed on M. edulis.
Maloney (1996) investigated interactive processes between organic enrichment and cadmium
(Cd) using 4 benthic species: Amphiura liformis, Amphiura chiajei, Abra alba and Nereis
diversicolor in microcosm experiments over a 10 week period. The partitioning of Cd between the
sediment and pore water phases and bioaccumulation of Cd in the test species were investigated
under different trophic regimes. Sediments and pore water in systems without organic enrichment
displayed signicant increases in Cd concentrations. As a consequence, the benthic species:
Amphiura liformis; Amphiura chiajei and Abra alba had increased Cd accumulation in these
systems. Nutrient enriched systems (through additions of cellulose bres, sewage and benthic
algae) at low and high doses signicantly altered the Cd partitioning in the sedimentary matrix.
Both algal treatments and high sewage levels signicantly decreased the sediment redox potential,
with a corresponding increase in sulphide activity. Cadmium concentrations in the pore water
were reduced by all organic types and had a highly signicant organic dose response. Different
type or sources of organic matter (bres<sewage<algae) caused signicant variations in the
bioaccumulation of Cd in 3 benthic species. The polychaete (N. diversicolor) was used for growth
response measurements to benthic food quality. The results showed that growth followed the
same organic enrichment sequence (bres<sewage<algae) as observed for the Cd bioaccumulation.
Maloney (1996) concluded that the quality of organic matter or a related parameter form a most
signicant factor in determining Cd bioaccumulation in the marine benthos. Furthermore, some
sources of organic enrichment have the ability to reduce metal bioaccumulation.
Bakke (1999) investigated the effect of food quality and starvation on the bioaccumulation of
PAHs in the marine blue mussel (Mytilus edulis). The experiments were carried out in ow through
systems with an exposure time of 30 days. In total 8 experimental conditions and 3 replicates of
groups of 20 were tested. The TOC:TSM ratio for the high organic treatment was 0.09 0.12 and
for low organic treatment 0.04 - 0.05. The ndings suggested that organic enrichment of food
particles resulted in lower total accumulation of PAHs and lower accumulation of the harmful
PAH compound (KPAH) in the bivalve. The nutritional level did not inuence the PAH proles.
However, mussels both fed and pre-starved in the low TOC:TSM ratio group displayed more
variable PAH proles, while groups with high TOC:TSM ratio had the same prole. In general PAH
proles differed clearly from the PAH source.
Hylland (1999a,b) investigated the effects of organic enrichment, decreased oxygen availability and
moderate contamination on growth of a polychaete (Nereis diversicolor), brittle star (Amphiura
liformis) and bivalve (Abra alba) in microcosm experiments. Possible interactions between the factors
and responses of biomarkers in the sediment dwelling organisms were considered. The study showed
that an increase in sediment organic carbon resulted in increased growth of all test species. Hypoxia
reduced growth of all three species, but interactive effects between hypoxia and organic enrichment
differed among the species. The polychaete and brittle star were able to adjust their metabolism to
low oxygen levels, while the bivalve used the available organic matter inefciently under hypoxia.
Generally, the presence of contaminants had minor effects on growth activities of all test species.
Biomarker responses were affected by tissue contaminant levels in the brittle star and to some extent
in the polychaete. No biomarker responses occurred in the bivalve. Hylland (1999) concluded that the
results of the study show the need for further studies that will focus on the questions of if and how
sediment dwelling organisms are affected by contaminants in sediments or overlying water.
Hylland (1999a,b) studied the relationships between changes in biomass caused by increased food
availability and changes in bioaccumulation of contaminants (Cd, Hg and benzo[a]pyrene) due to
interactive processes between pollutants and organic material. The marine test species: a polychaete
(Nereis diversicolor); brittle star (Amphiura liformis) and bivalve (Abra alba) were exposed in
laboratory experiments to sediment-bound contaminants in a ow-through system for 93 days.
A number of aquaria were treated with additional carbon matter and others were kept under
deoxygenated conditions. The study showed that the benthic invertebrates accumulated a higher
amount of organic and inorganic contaminants under eutrophic conditions in comparison with the
oligotrophic controls. The bioaccumulation rate of contaminants differed among the selected test
species. Highest contaminant accumulations were found for metals in the bivalve, followed by
metals and benzo[a]pyrene in the polychaete and then benzo[a]pyrene in the brittle star.
Gunnarsson et al. (2000) did laboratory experiments on the effects of inputs of organic matter
(microalgae additions) on bioaccumulation of HOCs (PCBs and PAHs) by the blue mussel (Mytilus
edulis), brittle star (Amphiura liformis) and polychaete (Nereis diversicolcor). Bioaccumulation
appeared proportional to the concentrations of algae and organic carbon. Gunnarsson et al.
(2000) suggested that high nutritional quality of algal organic carbon and feeding rather than
equilibrium partitioning governed the bioaccumulation of HOCs by the benthic test species.
Additionally, eld studies were carried out to measure the annual mass uxes of PCBs in mussels
and brittle stars and contaminant transfer to higher trophic levels. The ndings suggested that
eutrophication processes might contribute to increased HOC accumulation in benthic species. It
was also stressed that temporal variations in quality and quantity of organic carbon have to be
taken into account for assessments of contamination in aquatic systems. Macrofaunal feeding
activities were identied as important for the benthic-pelagic coupling of HOCs. Bioturbation by
the benthic fauna improved the released of HOCs from the sediment to overlying water.
12.8 Integrated modelling of the combined pressures
eutrophication and organic toxicants on biota
Koelmans et al. (2001) reviewed integrative models that are capable of simulating chemical
fate/effect and food chain accumulation at varying trophic levels and categorised them as EUTOX
models. The pioneer models EUTOX models are:
AQUATOX, see Box 12.1
CATS-5 (Contaminant in Aquatic and Terrestrial ecosystems), see Box 12.2
GBMBS model framework (Green Bay Mass Balance Study), see Box 12.3
HOC bioaccumulation model, see Box 12.4
IFEM (Integrated Fates and Effects Models), see Box 12.5
QWASI (Quantitative Water, Air, Sediment Interactions), see Box 12.6.
Koelmans et al. (2001) classied AQUATOX as, currently, the most complete of the EUTOX models.
The necessary linkage in EUTOX models of single issue (sub-) models dealing with: eutrophication;
contamination; food web and food chain activities resulted in highly complex model types. The
complexity is necessary for the inclusion of complex interactive processes between eutrophication
and contamination, but carries the risk of over parameterisation and causes difculties in calibration
and validation. Principal limitations in food web modelling are the restricting factor when
superimposing chemical fate model descriptions on the ecological model construct (Koelmans et
al., 2001). Therefore, long term simulations of exposure, food chain bioaccumulation and toxicity
remain uncertain on the ecosystem level. Koelmans et al. (2001) recommended that dominant
processes should be identied, described and estimated rather than incorporation of more detail.
Despite their limitations, EUTOX models are an essential tool for risk assessment of aquatic
systems that are impacted by combined pressures of eutrophication and contamination. The
alternation of only single pressures (e.g. reduction of nutrient load) may have fundamental
consequences to the ecosystem (e.g. increased bioavailability of contaminant in water column).
Integrative models can simulate the interactive scenario of eutrophication and contamination
that can support the environmental decision making process.
Description:
AQUATOX is a freshwater ecosystem simulation model which predicts the fate of various pollutants, such as nutrients and organic
chemicals, and their effects on the ecosystem, including sh, invertebrates, and aquatic plants. The model simulates multiple
environmental stressors (including nutrients, organic loadings and chemicals, and temperature) and their effects on the algal,
macrophyte, invertebrate, and sh communities. Therefore, AQUATOX can help identify and understand the cause and effect
relationships between chemical water quality, the physical environment, and aquatic life. The model can represent a variety of
aquatic ecosystems, including: vertically stratied lakes; reservoirs and ponds and rivers and streams.
Data Requirement:
AQUATOX comes bundled with data libraries that provide default data. This is of particular importance for the biological data,
which are probably the most difcult for a user to obtain.
Required input data are:
loadings
1
to the waterbody
general site characteristics
chemical characteristics of any organic toxicant
biological characteristics of the plants and animals.
Output:
AQUATOX provides output in terms of time varying biomass of the various plants and animal, chemical concentrations in water,
and concentrations of the organic toxicant in water, organic sediments and biota. It has numerous features to assist in display
and analysis of results.
Graphing capability
Easy export to spreadsheet programs
Option to save time varying rates, such as consumption and photosynthesis
Uncertainty and sensitivity analysis capability.
Version and System Requirements:
Minimum: Pentium PC, 133 MHz; Windows 95, 98, 2000 or NT; 64 MB RAM and 30 MB free disk space.
Recommended: Pentium PC, 600MHz or higher; Windows NT, 2000 or XP; 128 MB RAM and 75 MB free disk space.
In comparison to the previous versions of Aquatox distributed by the USEPA in release phase 1 and 1.1. the updated version has
following enhancements: improved stream simulation, by allowing habitat differentiation, simulation of inorganic sediments, and
loss of plant and animal biomass due to stream ow effects; more realistic simulation of sh, by improving representation of
respiration, consumption, spawning and elimination of toxicants; enhanced output for nutrient analysis by computing chlorophyll
a for periphyton, bryophytes, and phytoplankton, and computing the limitation factors for photosynthesis; increased number of
organic toxicants (up to 20) that can be modelled simultaneously and enhanced uncertainty analysis capabilities.
AQUATOX is now an extension to BASINS (EPAs GIS-based watershed modeling system), providing linkages to geographic
information system data, and output from BASINS watershed models (HSPF and SWAT). Following errors were detected in release
1 and 1.1 and corrected in release 2.: a change in the bathymetric computations affecting the areas of the thermocline and littoral
BOX 12.1 AQUATOX, A simulation model for aquatic ecosystems.
1
Environmental loadings can be from multiple sources such as: constant or dynamic; point or nonpoint sources;
upstream contributions and atmospheric deposition.
Continued
zone; removal of an unnecessary conversion from phosphate and nitrate, assuming that all nutrient input is in terms of N and P;
this could affect nutrient limitations; inclusion of an oxygen to organic matter conversion factor (a factor of 1.5) and inclusion of
specic dynamic action in the allometric computation of sh respiration; adding a conversion factor for wind measured at 10 m
height to wind occurring at 10 cm above the water surface in the volatilization computations; for some compounds this could
result in a two-fold reduction in volatilization; nitrication is formulated to occur only at the sediment-water interface and
bioaccumulation, and hence toxicity, are constrained by the life span of an animal.
Assumptions and Limitations:
The aquatic system is assumed well mixed (point model) but stratication can be taken into account.
Model Validation Reports:
The following validation reports are from Release 1 and 1.1. They have not been re-done using the Release 2 code, so the model
results would be somewhat different. But the overall results and conclusions should still be valid.
Download of nutrient analysis on the Onondaga Lake, New York
Download of nutrient analysis of the Coralville Reservoir, Iowa
Download of bioaccumulation of PCBs in Lake Ontario
References:
Koelmans, A.A, A. Van der Heijde, L.M. Knijff and R.H. Aalderink. 2001. Integrated modelling of eutrophication and organic
contaminant fate & effects in aquatic ecosystems. A review. Water Research 15: 3517-3536
Park, R.A., B. Firlie, R. Camacho, K. Sappington, M. Coombs and K.D. Mauriello. 1995. AQUATOX, a general fate and effect
model for aquatic ecosystems. In Proceedings for the Toxic Substances in Water Environments; Assessment and Control, Water
Environment Federation: (3)7-(3)-17
Website:
http://www.epa.gov/waterscience/models/aquatox/
Description:
CATS-5 has been constructed to model the combined effects of micropollutants and nutrients in aquatic systems. An existing
eutrophication module (Janse et al., 1992 and Janse, 1996) was combined with a fate module from SIMPLEBOX, an existing
food-web of micropollutants (Traas et al., 1995; 1996) and an effects module (Traas et al., 1998; 2001).
The aquatic system is assumed well mixed (point model) and average daily conditions are modelled. Equilibrium partitioning of
pesticides among the organic phases is assumed, except for sediment where diffusion-limited exchanges between water and pore
water is taken into account. Mortality of invertebrates, algae and macrophytes is based on the toxicant concentration in the
water. CATS-5 is not applicable to hydrophobic toxic compounds that show signicant association with biotic particles and are
readily taken up in the food chain. Trophic transfer is not included into the model. Additional toxicity by contaminants uptake by
food or bioconcentration is not simulated.
Application History:
Simulations were tested against chlorpyrifos data collected in aquatic microcosm experiments CATS-5 reproduced the observed
effects of nutrients, chlorpyrifos and their interaction reasonably well (Janse and Traas, 1996).
References:
Janse, J.H. 1996. A model of nutrient dynamics in shallow lakes in relation to multiple stable states. Hydrobiologia 342:1-9
Janse, J.H., T. Aldenberg and P.R.G. Kramer. 1992. A mathematical model of the phosphorus cycle in Lake Loosdrecht and
simulation of additional measures. Hydrobiologia 233:119-136
Koelmans, A.A., A. Van der Heijde, L.M. Knijff and L.M. Aalderink. 2001. Integrated modelling of eutrophication and organic
contaminant fate & effects in aquatic ecosystems. A review. Water Research 15:3517-3536
Traas, T.P., T. Aldenberg, J.H. Janse, T.C.M. Brock and C.J. Roghair. 1995. Hazardous concentrations for ecosystems: calculation
with CATS models: Chapter 3. National Institute of Public Health and Environmental Protection (RIVM) Report no. 719102037.
The Netherlands
Traas, T.P., J.A. Stb, P.R.G. Kramer, W.P. Cono, T. Aldenberg. 1996. Modeling and risk assessment of tributyltin accumulation in
the food web of a shallow freshwater lake. Environmental Science and Technology 30:1227-1237
Traas, T.P, J.H. Janse, T. Aldenberg and T.C.M. Brock. 1998. A food web model for fate and direct and indirect effects of Dursban
4E (active ingredient chlorpyrifos) in freshwater microcosms. Aquatic Ecology 32:179-190
Traas, T.P, J.H. Janse, P.J. Van den Brink and T. Aldenberg. 2001. A food web model for fate and effects of toxicants and
nutrients in aquatic microcosms. Model description. Report 60151006, National Institute of Public Health and Environmental
Protection (RIVM), Bilthoven, The Netherlands
BOX 12.2 CATS-5, Contaminant in Aquatic and Terrestrial ecoSystems.
Description:
The Green Bay Mass Balance Study (GBMBS) was conducted in 1989-90 to pilot the technique of mass balance analysis in
understanding the sources and effects of toxic pollutants in the Great Lakes food chain. The overall goal of study was to test the
mass balance approach for tracking toxicants (mainly PCBs at the congeners level) in a system under additional pressure from
nutrients. The integrated modelling framework consisted of a suite of individual models to simulate hydraulics, sorbent dynamics
and toxic chemicals in Green Bay, Lake Michigan.
The study, headed by EPAs Great Lakes National Program Ofce (GLNPO) and the Wisconsin Department of Natural Resources,had
many participants from the Federal, state, interagency, and academic communities. The study focused on four representative
chemicals or chemical classes: PCBs, dieldrin, cadmium, and lead.
Input and Output Data:
The eutrophication model component generates organic carbon-related solids which are input to a long term solid model (GBTS)
and a sorbent dynamics model (Green Bay Organic Carbon Sorbent; GBOCS). The output of GBOCS forms an input to the
contaminant exposure model (GBTOX) the output of which forms the input to the food chain model.
The PCB model framework included interactions between: air water; water sediment and the food chain.
Results of the sub-model component GBTOX were successfully compared with particulate phase and dissolved PCB concentrations
that were collected for the GBMB study. The GBMBS modelling framework was successfully applied to a major follow up project:
the Lake Michigan Mass Balance Study (LMMBS) in the Great Lakes.
Availability:
Contact: Dave DeVault
Telephone: 612-296-7253
Email: dave.devault@pca.state.mn.us
Information Type: Data, Text
The GBMBS data sets are available at: http://www.epa.gov/grtlakes/gbdata/gbay2.html
References:
USEPA. 1989. Green Bay/Fox River Mass Balance Study. EPA-905/8-89/001. US Environmental Protection Agency, Great Lakes
National Program Ofce, Chicago, IL
Website:
http://www.epa.gov/grtlakes/gbdata/
Box 12.3 GBMBS, Green Bay Mass Balance Study.
Description:
Ashley (1988) presented a dynamic environmental fate and population-based food chain accumulation model to determine the
inuence of varying trophic status on hydrophobic organic compound (HOC) transfer and fate in shallow aquatic systems. The
model is based on detailed carbon-based benthic-pelagic food chain and tracks the transfer and accumulation of organic carbon
and HOCs among 16 biotic and abiotic pools over time scales from days to years. The following mechanisms of contaminant
transfer are included: binding to DOC; the bacterial loop; redistribution through bioturbation and re-suspension events.
The model assumes a spatially homogenous system. Temperature, light and nutrient load are the three forcing functions that drive
the production of algal biomass in the modelled system. The three functions are dened over an annual scale. The biomass
supports production in the water column and sediment. The food web consists of ve trophic levels in the water column: bacteria;
protozoa; phytoplankton; zooplankton and piscivores. Three levels are described in the benthic environment: deposit feeder; lter
feeder and benthicvores/forage feeders. Both water column and sediment associated POC and DOC pools are incorporated. The
carbon exchange at the sediment-water interface is dened by: pelagic species supplying deposit feeders with settling organic
matter (e.g. through excretion); lter feeder ingestion of algal and settling POC and carbon recycling to the water column by
piscivores grazing on benthic (forage) sh. HOC transfer is described through: diffusion, air-water exchange, bioconcentration,
utilisation of DOC or POC; trophic transfer; mortality and excretion losses; sedimentation and re-suspension by bioturbation.
HOC concentrations are assumed to be remain below deleterious levels so that the carbon ows are not altered by toxicity.
Therefore, the inuence of toxicity is not included in the model.
The biomass output of the model was veried and validated with respects to measured periodicity and ranges of values, using
data from Chesapeake Bay and the Baltic Sea. However, validation with respect to HOC fate was not yet been done. The model
was used to calculate HOC scenarios in aquatic systems with different trophic states.
References:
Ashely, J.T.F. 1998. Habitat use and trophic status as determinants of hydrophobic organic contaminants bioaccumulation
within shallow systems. Ph.D. Dissertation, University of Maryland at College Park
Koelmans, A.A., A. Van der Heijde, L.M. Knijff and R.H. Aalderink. 2001. Integrated modelling of eutrophication and organic
BOX 12.4 HOC BIOACCUMULATION, model for shallow aquatic systems.
Description:
IFEM is a combination of the toxic effect model SWACOM and the dynamic fate model FOAM. The model is designed for PAHs
and accounts for sorption to suspended and settled solids, dissolved organic matter (DOM), detritus and sediment, volatilisation,
photolysis and uptake and depuration by the model populations.
The foodweb consists of 11 functional groups. The biomass is calculated from daily light conditions, temperature and predator-
prey relationships. Sublethal effects on growth processes are determined using dose-effect and time-varying body burdens,
calculated from uptake, degradation and depuration. The effect is modelled using a population specic toxic effect factor,
adjusting the rate of population growth.
Nutrients are not included in model.
IFEM has been used to study the fate and effects of naphthalene in the water column and shore of a freshwater lake in a
temperate climate zone, but has so far not been validated.
References:
Bartell, S.M., R.H. Gardner and R.V. ONeill. 1988. An integrated fates and effects model for estimation of risk in aquatic systems.
Aquatic Toxicological Hazard Assessment 10: 261-274
Bartell, S.M., R.H. Gardner and R.V. ONeill. 1992. Ecological Risk Estimation. Lewis Publishers, Chelsea
Koelmans, A.A., A. Van der Heijde, L.M. Knijff and R.H. Aalderink. 2001. Integrated modelling of eutrophication and organic
BOX 12.5 IFEM, Integrated Fates and Effects Model.
Description:
QUASI was originally designed to describe the steady state behaviour of an organic chemical in a lake subject to chemical
inputs by direct discharge, inow in rivers, and deposition from the atmosphere. The mass balance equations for the well-mixed
water column and the well-mixed layer of surcial sediment also include sediment-water exchange by diffusion, deposition, and
re-suspension. Wania (1996) presented a modied version of QUASI to study the interaction of nutrients and contaminants. The
model was parameterised for several typical HOCs in a generic lake and input parameters related to the organic carbon
dynamics were adjusted to reect different trophic conditions.
Most important modications from original QUASI model:
Inclusion of the atmosphere above the lake as a compartment, the concentration of which is calculated rather than being a
model input. This allows for the lake to have an inuence on the concentration in the air above the lake, which may be of
importance for large lakes and chemicals with large air-water exchange uxes. The user has to input the background
concentration in the air owing into the lake area.
Description of the particle-associated transport of HOC on the basis of particulate organic carbon (POC) rather than total
solids, and the derivation of the transport parameters for POC settling, re-suspension and burial within the model.
Calculation of some of the transport parameters describing the atmosphere-water exchange from meteorological and
chemical specic data instead of using generic values.
Programming the model in Stella II-format to facilitate the modelling g of non-steady state scenarios with time-variant input
parameters.
The partitioning of HOCs to DOC or colloidal material is not taken into account.
Wania (1996) simulated the interactions between HOCs and nutrients in an eutrophic lake and oligotrophic lake.
Availability and Website:
The original version of QWASI is available at: http://www.trentu.ca/cemc/models/Qwasi.html
References:
Wania, F. 1996. Modelling the Interaction of Eutrophication and hydrophobic organic contaminants behaviour in aquatic
systems. P.R.G. Kramer, D.A. Jonkers and L. van Liere. RIVM report no. 703715001:89-95
BOX 12.6 QUASI Quantitative Water, Air and Sediment Interaction Modied version for EUTOX approach.
12.9 Summary
Possible interactive mechanisms of HOCs in eutrophic systems are: biomass dilution; growth
dilution, increased sedimentation of contaminant bound pollutants and contaminant burial in
sediments. Consequences for the pelagic biota may be lowered exposure to HOCs in the water
mass and lower bioaccumulation of contaminants. In extreme cases, eutrophication alone may
lead to anoxic conditions in the bottom sediment and the benthic fauna becomes extinct.
Increased sedimentation rates and the effect of increased POP concentrations during vertical
settlement of particles due to microbial attacks may increase exposure of benthic fauna foraging
on sedimenting matter. Pollutants may be increasingly bioaccumulated in the benthic organisms
and contamination of predatory animals feeding on benthic invertebrates (e.g. demersal sh and
waterfowl feeding on bivalves) can occur.
In oligotrophic systems smaller biomass in the water column may result in higher concentrations
of HOCs in the dissolved aqueous phase in comparison with nutrient-rich systems. Pelagic biota
may bioaccumulate higher amounts of pollutants. Sedimentation of particle bound HOCs is
decreased compared with eutrophic systems and benthic fauna less exposed to pollutants in
sediments.
The case studies presented in this report conrmed that the pelagic biota accumulated lower
amounts of organic contaminants in eutrophic systems. The important interactive mechanisms
identied were: biodilution, changes in storage lipid content of organisms, and changes in food
web structure. Higher lipid contents and higher amounts of pollutants in primary producers can
lead to mechanisms that result in higher pollutant concentrations in the biota of oligotrophic
systems. However, some episodic events, such as oods can override interactive processes and no
obvious link between pollutant levels in the pelagic biota and eutrophication may be found.
Most case studies relating to the benthos investigated the impact of organic enrichment on
bioaccumulation of pollutants in benthic biota. Benthic invertebrates accumulate higher amounts
of organic and inorganic contaminants under eutrophic conditions. Bioaccumulation was found
to be proportional to concentrations of algae and organic carbon and high nutritional quality of
algal organic carbon and feeding were responsible for bioaccumulation of HOCs. Further, HOC
uptake was dependent on food concentrations, and trophic-dependent changes were caused by
physiological activities of bivalves inuenced by food rations.
Internal factors in aquatic systems can inuence interactive mechanisms between eutrophication
and contamination. The composition of lipids in aquatic organisms, for example, is crucial for the
fate of hydrophobic contaminants. The equilibrium partitioning between contaminants in the
water column and biomass in aquatic systems needs to be quantied for different scenarios since
it is decisive for the biodilution effect.
The single pressures of eutrophication and contamination may both induce changes in the aquatic
system. Effects of eutrophication may be: increased productivity; increased sedimentation after a
bloom; organic enrichment of sediments resulting in hypoxia or anoxia. These and other alterations
(e.g. reduced light conditions) related to eutrophication can inuence the species composition
and diversity structure of phytoplankton, zooplankton, pelagic and benthic communities (Skei et
al., 2000).
Contaminants can cause direct toxic effects when released in certain concentrations into the
aquatic environment. Toxicant induced lethality can decimate sensitive species and/or sublethal
effects may change physiological activities (e.g. reproduction, growth, maturation, longevity) of
the biota. These ecological alterations can lead to a trophic cascade or release from competition
inuencing secondarily tolerant species (Fleeger et al., 2003). The extent of the direct and/or
indirect effects of contaminant exposure on biota is concentration dependent and environmental
quality standards (EQS) regulate the amounts of discharge of selected priority substances.
Interactive effects of chemical mixtures occurring in aquatic systems and their additive, synergistic
and antagonistic effects on biota have not been investigated adequately and relationships have
not yet been quantied.
Additionally to knowledge based on quantied impacts of the single pressures eutrophication/
contamination and risk assessment using ecological models (e.g. ECOPATH), pollutant interaction
has to be considered for environmental decisions and water management. A reduction of nutrients,
for example, improving the trophic status may also lead to an increase of contaminant concentrations
in the system and higher bioaccumulation of toxicants in the pelagic biota. Additionally, improved
oxygen conditions at the bottom may lead to recolonisation of laminated sediments by benthic
fauna. As a result of bioturbation and geochemical processes, contaminants that were previously
trapped in the sediments may be mobilised. The level of contaminants may, vice versa, inuence
the primary production and indirectly the status of eutrophication (Skei et al., 2000).
Interactions between eutrophication and contaminants have to be investigated further to predict
their effects under changing environmental conditions. There has been some work done on
interactions between hydrophobic organic compounds and trophic status (Skei et al., 2000) but
more knowledge is needed on interactive processes between metals and other toxic substances
and eutrophication. Relationships should be quantied to support water management and
remedial measures.
Ecological models from the category of the new EUTOX models (e.g. AQUATOX) are capable of
simulating chemical fate/effect and food chain accumulation at varying trophic levels. The
restricting factor of the new models types are that long term simulations of exposure, food chain
bioaccumulation and toxicity remain uncertain on the ecosystem level due to the superimposition
of chemical fate model descriptions on ecological model constructs (Koelmans et al., 2001).
Despite their limitation, EUTOX models are useful tools for risk assessment of habitats under
combined pressures of nutrients and contaminants.
The identied interactive mechanisms between eutrophication and contaminants raise further
questions about their importance in the aquatic system. The following ten questions reect
knowledge gaps that future research need to answer to enable water policy makers to consider
combined effects of eutrophication and contaminants in water management.
1. What are the consequences for aquatic systems under pressure from both eutrophication
and contamination if only one pressure is altered?
2. To what extent will the concentrations of contaminants in the water mass and pelagic biota
increase if the nutrient input is reduced without simultaneous reduction of contaminant
inputs?
3. Under what circumstances do sediments act as sinks or sources for contaminants in systems
with different trophic regimes?
4. To what extent will decreased eutrophication inuence the bioavailability of contaminants?
5. Can the relationships between the combined pressures of eutrophication and contamination
and the pelagic and benthic biota be quantied?
6. Can key species be identied that indicate the interactive effect of eutrophication and
contamination?
7. Under what circumstances do the effects of biodilution and growth dilution override other
internal factors in eutrophic systems and lead to lowered exposure of the pelagic biota to
contaminants?
8. To what extent will a reduced sedimentation of organic matter inuence the concentrations
of contaminants in the benthic biota?
9. To what extent does the benthic biota reintroduce contaminants into the pelagic biota
through the food chain in eutrophic systems that show lowered contaminant concentrations
in the water mass and higher concentrations of contaminants in the sediments?
10. To what extent does remobilisation of contaminants by bioturbation of re-colonised
laminated sediments after improvement of oxic conditions occur?
12.10 References
Axelman, J., D. Broman, and Carina Nf. 2000. Vertical ux and particulate/water dynamics of polychlorinated
biphenyls (PCBs) in the Open Baltic Sea. Ambio 29:210-216
Bakke, T. 1999. Effects of organic enrichment on the accumulation of PAH in blue mussels, Mytilus edulis.
In: Interaction between contaminants and eutrophication (MILE), report 4129-99, pp. 24-25
Balzer, W. 1984. Organic matter degradation and biogenic element cycling in a nearshore sediment (Kiel
Bight). Limnology and Oceonography 29:1231-1246
Eckhll, J., P. Jonsson, M. Meili, and R. Carman. 2000. Storm inuences on the accumulation and lamination
of sediments in deep areas of the Northwestern Baltic Proper. Ambio 29:238-245
Evers, E.H.G., and F. Smedes. 1996. Seasonal variations of PCB and PAH concentrations in SPM and blue
mussels in the North Sea. In: P.R.G. Kramer, D.A. Jonkers, and L. van Liere (Eds.), Interactions of nutrients and
toxicants in the food chain of aquatic ecosystems., L. RIVM report no. 703715001, pp. 41-45
Fleeger, J.W., K.R. Carman, and R.M. Nisbet. 2003. Indirect effects of contaminants in aquatic ecosystems.
The Science of the Total Environment 317:207-233
Frstner, U. and G.T.W. Wittman. 1983. Metal pollution in the aquatic environment. Springer-Verlag, New
York, New York, pp. 486
Gilek, M., M. Bjrk, N. Kautsky, and U. Kautsky. 1996. Bioaccumulation and cycling of hydrophobic organic
contaminants by Baltic Sea Mytilus edulis: implications of changes in eutrophication status. A summary of
an ongoing research project. In: P.R.G. Kramer, D.A. Jonkers, and L. van Liere (Eds.), Interactions of nutrients
and toxicants in the food chain of aquatic ecosystems., L. RIVM report no. 703715001, pp. 31-40
Gjessing, E.T. and L. Berglind. 1981. Adsorption of PAH to aquatic humus. Archiv fr Hydrobiology 92:24-30
Gunnarsson, J.S. and R. Rosenberg. 1996. Eutrophication increases the association of PCB to dissolved
organic matter in marine microcosms. Marine Pollution Bulletin 33:100-111
Gunnarsson, J.S., M.T. Schaanning, K. Hylland, M. Skold, D.O. Eriksen, J.A. Berge, and J. Skei. 1996. Interactions
between eutrophication and contaminants. III. Mobilization and bioaccumulation of benzo(a)pyrene from
marine sediments. Marine Pollution Bulletin 33:80-89
Gunnarsson, F., M. Bjrk, M. Gilek, M. Granberg and R. Rosenberget. 2000. Effects of eutrophication on
contaminants cycling in marine benthic systems. Ambio 29:252-259
Hasset, J.P., and M.A. Anderson. 1979. Association of hydrophobic organic compounds with dissolved organic
matter in aquatic systems. Environment Science and Technology 13:1526-1529
Horzempa, L.M., and D.M. Di Toro. 1983. The extent of reversibility of polychlorinated biphenyl adsorption.
Water Research 17:851-859
Hylland, K. 1999a. Interactions between eutrophication and contaminants - effects on sediment dwelling
organisms. In: Interaction between contaminants and eutrophication (MILE), report sno: 4129-4199, pp. 18-21
Hylland, K. 1999b. Time-trend analysis of contaminants in cod (Gadus morhua) in two ares with different
eutrophication regimes. In: Interaction between contaminants and eutrophication (MILE), report sno: 4129-
4199, pp. 27-28
Ignatius, H. 1958. On the rate of sedimentation in the Baltic Sea. Bulletin Commission Gologique de
Finlande 180:135-145
Jeremiason, J., S. Eisenreich, M.J. Paterson, K.G. Beaty, R. Hecky, and J.J. Elser. 1999. Biogeochemical cycling of
PCBs in lakes of variable trophic status: A paired lake experiment. Limnology and Oceonography 44:889-902
Jonsson, P., R. Carman, and F. Wulff. 1990. Laminated sediments in the Baltic - A tool for evaluating nutrient
mass balance. Ambio 19:152-158
Jonsson, P., and R. Carman. 1994. Changes in deposition of organic matter in the Baltic Sea during the
twentiest century and notes on sediment dating problems. Marine Pollution Bulletin 28:417-426
Jonsson, P. (Ed.), 1996. EUCON - Interactions between eutrophication and contaminants in the aquatic
environment. Research Programme for the period 1995-1999. Swedish Environmental Protection Agency.
Report 4690, pp. 52
Jonsson, P. 1997. The Swedish EUCON project - priorities and some results in October 1996. In: P.R.G. Kramer,
D.A. Jonkers, and L. van Liere (Eds.), Interactions of nutrients and toxicants in the food chain of aquatic
ecosystems. RIVM report no. 703715001, pp. 9-13
Jonsson, P. 2000. Sediment burial of PCBs in the Offshore Baltic Sea. Ambio 29:260-267
Koelmans, A.A., A. Van der Heijde, L.M. Knijff, and R.H. Aalderink. 2001. Integrated modelling of
eutrophication and organic contaminant fate & effects in aquatic ecosystems. A review. Water Research
15:3517-3536
Kucklick, J.R., and L.E. Baker. 1998. Organochlorines in Lake Superiors Food Web. Environmental Science and
Technology 32:1192-1198
Kunzendorf, H., and C. Christiansen. 1997. Estimation of recent sedimentation histories in the Gotland Basin
using cores from the GOBEX study. In: I. Cato, and F. Klingberg (Eds.), Proceedings of fourth marine geological
conference- the Baltic. Uppsala, 1995, pp. 111-116
Larsson, P., L. Collvin, L. Okla, and G. Meyer. 1992. Lake productivity and water chemistry as governors of the
uptake of persistent pollutants in sh. Environment Science and Technology, 26:346-356
Larsson, P., 1997. Eutrophication and persistent pollutants - effects on plankton and sh in lakes. In: P.R.G.
Kramer, D.A. Jonkers, and L. van Liere (Eds.), Interactions of nutrients and toxicants in the food chain of
aquatic ecosystems. RIVM report no. 703715001, pp. 95-101
Larsson, P., L. Okla, and G. Cronberg. 1998. Turnover of poly chlorinated biphenyls in an oligotrophic and an
eutrophic lake in relation to internal lake processes and atmospheric fallout. Canadian Journal of Fisheries
and Aquatic Science, 55:1926-1937
Larsson, P., A. Andersson, D. Broman, J. Nordbck and E. Lundberget. 2000. Persistent Organic Pollutants
(POPs) in pelagic systems. Ambio 29:202-209
Lithner, G., H.E. Borg, J. Ek, E. Froberg, K. Holm, A.M. Johansson, P. Karrhage, M. Soderstorm, and G. Rosen.
2000. The turnover of metals in a eutrophic and an oligotrophic lake in Sweden. Ambio 29:217-229
Mackay, D., and B. Powers. 1987. Sorption of hydrophobic chemicals from water: A hypothesis for the
mechanism of the particle concentration effect. Chemosphere 16:745-757
Maloney, J. 1996. Inuence of organic enrichment on the partitioning and bioavailability of cadmium in a
microcosm study. Marine Ecology Progress Series 144:147-161
McCarthy, J.F., and B.D. Jimenez. 1985. Interactions between polycyclic aromatic hydrocarbons and dissolved
humic material: binding and dissociation. Environment Science and Technology 19:1072-1076
Neumann, T., C. Christiansen, S. Clasen, K.C. Emeis, and H. Kunzendorf. 1997. Geochemical records of salt-
water inows into the deep basins of the Baltic Sea. Continental Shellsh Research 17:95-115
Niemist, L., and A. Voipio. 1974. Studies on the recent sediments in the Gotland Deep. Merentutkimuslait.
Julk./Havsforskninginst. Skr. 238, pp. 17-32
Niemist, L., and A. Voipio. 1981. Notes on the sediment studies in the Finnish pollution research in the
Baltic Sea. Rapp. P.-v. Run. Cons. int. Explor. Mer 181, pp. 87-92
Olsson, M., and S. Jenssen. 1975. Pike as test organism for mercury, DDT and PCB pollution. A study of the
contamination in the Stockholm archipelago. Institute of fresh water research, Drottingholm 54, pp. 83-106
stlund, P., and R. O. Hallberg. 1991. Plutonium in sediments - apparent half-lives. Environmental Geology
and Water Science 17:195-200
Persson, J., and P. Jonsson. 2000. Historical Development of Laminated Sediments - an Approach to Detect
Soft Sediment Ecosystem Changes in the Baltic Sea. Marine Pollution Bulletin 40:122-134
Schaanning, M. T., K. Hylland, D.O. Eriksen, T.D. Bergan, J.S. Gunnarson, and J. Skei. 1996. Interactions
between eutrophication and contaminants. II. Mobilization and bioaccumulation of Hg and Cd from marine
sediments. Marine Pollution Bulletin 33:71-79
Sijm, D.T.H.M. 1996. The inuence of algae exudates on the bioavailability of organic toxicants in the
aquatic system. In: P.R.G. Kramer, D.A. Jonkers, and L. van Liere (Eds.), Interactions of nutrients and toxicants
in the food chain of aquatic ecosystems. L. RIVM report no. 703715001, pp. 19-25
Skei, J., P. Larsson, R. Rosenberg, P. Jonsson, M. Olsson, and D. Broman. 2000. Eutrophication and
contamination in aquatic systems. Ambio 29:184-194
Sderstrm, M., K. Nylund, U. Jrnberg, G. Lithner, G. Rosn, and H. Kylin. 2000. Seasonal variations of DDT
compounds and PCB in a eutrophic and an oligotrophic lake in relation to algae biomass. Ambio 29:230-237
Stange, K., and D.L. Swackhamer. 1994. Factors affecting phytoplankton species-specic differences in
accumulation of 40 polychlorinated biphenyls (PCBs). Environmental Toxicology and Chemistry 13:1849-1860
Suess, E. and H. Erlenkeuser. 1975. History of metal pollution and carbon input in Baltic sediments.
Contribution Nr. 29 of the Sonderforschungsbereich 95 at the University of Kiel Interaction Sea - Sea
Botoom, Meyniana 27
Swackhamer, D.L., and R.S. Skolund. 1993. Bioaccumulation of PCBs by algae: kinetics versus equilibrium.
Environmental Toxicology and Chemistry, 12:831-838
Taylor, W.D., E. Bentzen, B. DeLuca, B.E. Hickie, D.R.S. Lean, D. Mackay, and G. Patterson. 1996. Empirical
evidence for the importance of lake trophic status and food web in bioaccumulation in Southern Ontario
lakes. In: P.R.G. Kramer, D.A. Jonkers, and L. van Liere (Eds.), Interactions of nutrients and toxicants in the
food chain of aquatic ecosystems., L. RIVM report no. 703715001, pp. 101-104
Wallberg, P., and A. Andersson. 1999. Determination of adsorbed and absorbed polychlorinated biphenyls
(PCBs) in seawater microorganisms. Marine Chemistry 64:287-299
Wassmann, P. 1980. Calculating the load of organic carbon to the aphotic zone in eutrophicated coastal
waters. Marine Pollution Bulletin 21:183-187
European Commission
EUR 22314EN DG Joint Research Centre, Institute for Environment and Sustainability
Editors: Angelo G. Solimini, Ana Cristina Cardoso and Anna-Stiina Heiskanen
Luxembourg: Office for Official Publications of the European Communities
2006 VIII, 252 pp. 17 x 24 cm
Scientific and Technical Research series
ISSN 1018-5593
ISBN 92-79-02646-1
The Water Framework Directive (WFD) defines the ecological status of surface waters as an expres-
sion of the quality of the structure and functioning of aquatic ecosystems. This book provides an
overview of current knowledge on the relationships between anthropogenic pressures and biologi-
cal indicators for rivers, lakes and coastal waters. It is targeted at scientists working in environmen-
tal agencies, managers working on WFD implementation, university lecturers, students of applied
ecology, water management and those interested in limnology and coastal management.
Each section is dedicated to one of the major pressure types: eutrophication and organic pollution,
acidification, hydromorphological, toxic, and combined pressures which are currently most widely
impacting surface waters in Europe. In each chapter we aim to provide a concise and factual over-
view on the known relationships between the pressure type in question and the biological respons-
es, for different surface water categories (rivers, lakes, coastal and transitional waters). We have
compiled information on the applicability of the current existing biological classification tools and
metrics, as well as the empirical and deterministic models which include biological quality elements
as parameters, and could be used for WFD classification and management purposes. Finally, we
also provide information on some new emerging approaches and methodologies that could be ap-
plied in the future.
This book is a deliverable of the EU FP6 REBECCA (Relationships between ecological and chemical
status of surface waters) project. The general objective of the REBECCA project is to provide sci-
entific support for the implementation of the WFD. The two specific aims of the project are, firstly, to
establish links between ecological status of surface waters and physico-chemical quality elements
and pressures from different sources, and, secondly, to develop and validate tools that member
states can use in the process of classification, in the design of their monitoring programs, and in
the design of measures in accordance with the requirements of the WFD.

Indicator and Methods For The Ecological Status Assessment Under The WFD Linkages Between Chemical and Biological Quality of Surface Waters

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

Indicator and Methods For The Ecological Status Assessment Under The WFD Linkages Between Chemical and Biological Quality of Surface Waters

Загружено:

Авторское право:

Доступные форматы

Institute for

Kothes species decit

Odums species per 1000 individuals

NUTRIENTS AND EUTROPHICATION IN COASTAL AND TRANSITIONAL WATERS | 79

where CVS = SCV x STR

3 1 0.04 0.025 0.005

6 1 0.2 0.025 0.005

204 | LINKING ECOLOGICAL STATUS OF SURFACE WATERS WITH ANTHROPOGENIC PRESSURES

Вам также может понравиться