Download as a PDF

Water Qual. Res. J. Canada, 2005
•
Volume 40, No. 3, 233–250
Copyright © 2005, CAWQ
Three Decades of Swedish Experience Demonstrates
the Need for Integrated Long-Term Monitoring
of Fish in Marine Coastal Areas
Olof Sandström,1* Åke Larsson,2 Jan Andersson,3
Magnus Appelberg,3 Anders Bignert,4 Helene Ek,2 Lars Förlin5 and Mats Olsson4
1
2
Skärgårdsutveckling SKUTAB AB, Öregrund, Sweden
Göteborg University, Department of Applied Environmental Science, Göteborg, Sweden
3
National Board of Fisheries, Institute of Coastal Research, Öregrund, Sweden
4
Swedish Museum of National History, Stockholm, Sweden
5
Göteborg University, Department of Zoophysiology, Göteborg, Sweden
The first attempts to monitor coastal fish in Sweden were made in the 1960s and 1970s. Ecological, physiological, biochemical and environmental chemistry data were collected in separate projects. When the National Marine Monitoring Programme was revised in 1992, a new strategy was introduced for assessments of long-term trends in coastal fish communities.
Annual integrated monitoring of contaminants, biomarkers and population and community indicators of ecosystem health
was started in selected areas using common sentinel species. Data from one monitoring area at the coast of the Baltic proper
are analyzed in this paper. The results have shown a shift in fish community structure indicating changes in ecosystem productivity. Trends have been detected in growth rate (positive) and relative gonad size (negative) in perch (Perca fluviatilis),
suggesting a metabolic disturbance according to the predictive response model developed for interpretations. One factor
which may have contributed to the reduced GSI was a decrease in mean age of sampled fish during the period of study.
Chemical exposure was indicated by a 3-fold increase of EROD activity during the monitored 15-year period. However, concentrations of most measured contaminants in perch have decreased during the same period. The experience of the integrated approach has shown that a tentative analysis of cause and environmental significance could be made, improving the
assessment, but there still remain unsolved questions to be answered in follow-up studies. The analysis has also shown the
importance of long-term monitoring at several levels of biological organization to distinguish between natural variation and
low-level effects on ecosystems.
Key words: integrated monitoring, the Baltic, perch, contaminants, biomarkers, physiological functions, population characteristics, community structure
Introduction
ized long-term programs. The weaknesses of non-integrated effect-related monitoring could be summarized as:
The use of fish in environmental monitoring has a comparatively long history in Sweden. The monitoring activities started in the 1960s and 1970s as separate series of
measurements directed to: (a) chemical contamination,
(b) biochemistry, physiology and pathology, and (c) populations and communities. There was no comprehensive
strategy and only minor collaboration during sampling
and data interpretations. However, awareness was growing that single chemical, biochemical or ecological data
sets alone could not be used to assess ecosystem health
and to analyze the cause of observed changes. It was felt
that a higher degree of coordination was needed, and
that the possibility to join activities within a common
framework should be explored. To distinguish between
natural variation and low-level pollution effects, monitoring should also be developed towards more standard-
• contaminant concentrations indicate exposure
but can not alone indicate biological effects on
the individual or population levels;
• biomarkers might indicate exposure to chemicals
and give early signs of biochemical/physiological/pathological effects, but they can not alone
identify responsible contaminants;
• there is a limited capacity to extrapolate a certain
change at a low level of biological organization
to an effect on growth, reproduction and survival;
• changes at the level of the individual organism
might indicate risks for population effects, but
the causes are often unknown and correlations
between a specific change and an impact on
recruitment, mortality and abundance are weak
or poorly known;
* Corresponding author; [email protected]
233
234
Sandström et al.
• changes at the level of populations may be ecologically relevant, but in many cases the changes
occur later than biomarker responses and the
causes are often unknown; and
• natural variations may obscure interpretations,
but the understanding of the extent and significance of natural variation is usually lacking.
The first example of a more integrated approach
appeared during the Swedish pulp and paper mill effluent research in the 1980s. It was then realized that the
analysis of observed deviations and their biological significance could benefit from a coordination among studies of ecology, physiology, biochemistry and environmental chemistry (reviewed by Larsson et al. 2003).
An opportunity to further strengthen project cooperation came when the Swedish National Marine Monitoring Programme (NMMP), run by the Swedish Environment Protection Agency (SEPA), was revised in 1992.
As a result of the review, a new strategy of integrated
monitoring of coastal fish, including contaminants, biomarkers, and population and community indicators of
ecosystem health in a common program, was decided for
national monitoring.
This paper presents and evaluates the strategy of the
integrated fish monitoring included in the Swedish
NMMP, and the interpretation model developed to support assessments of observed deviations. Observed
trends indicating changes in the Baltic coastal environment are presented and discussed against this background.
Objectives of the Integrated Fish Monitoring
The main purpose of the integrated fish monitoring is to
provide a contribution for assessments of ecosystem
health (according to the pragmatic model suggested by
Calow [1995]) by analyzing observations from subcellular to population and community levels. Any comprehensive ecosystem health assessment must require measures at several levels of biological organization,
including the suborganismal (Sherry 2003). Integrated
effects of all stressors should ideally be possible to detect
(Munkittrick 1992) and low-level, basin-wide perturbations should be possible to distinguish from natural variation. Although it is not the responsibility of regular
monitoring to provide a full understanding of all observations, the integrated monitoring should allow a primary analysis of observed changes to verify whether or
not changes are relevant and related to environmental
stressors. Monitoring should also provide opportunities
to test whether annual variations of demographics of
studied populations obscure changes that we wish to
detect (Gagnon et al. 1995; Bussieres et al. 1998). When
changes are detected but their relevance is unclear, this
should lead to follow-up studies addressing the cause of
the change and its further ecological significance. The
objectives of the integrated monitoring of coastal fish
were set to:
• document long-term development in biological
variables on different levels of organisation, i.e.,
from subcellular to population and community
levels;
• document long-term development in contaminant
concentrations;
• provide data for comprehensive/integrated interpretations and assessments of the cause and significance of observed changes;
• provide data on natural variations for variables
under study;
• estimate the impacts of reducing discharges of
contaminants and nutrients;
• act as a watchdog to detect renewed usage of
banned contaminants and effects of new contaminants;
• provide time series for contaminants of relevance
for human and wildlife risk assessments; and
• provide reference data for local monitoring.
Monitoring Strategy
An effects-based approach, rather than a strictly stressorbased approach, was chosen for the design of the monitoring strategy, although many potential stressors were
known already when the program started. In the past,
many monitoring activities studied known pollutants and
their specific biological impact. Since the environment
receives a mixture of numerous contaminants, it is hardly
possible to monitor concentrations of them all. The picture may be even more complicated as chemical contamination is often accompanied by high plant nutrient concentrations and other environmental stressors. An
effects-based assessment serves as an indicator of effects
of unforeseen, unknown or known but ignored stressors.
Documentation of stressor identities is not required, and
an initial analysis can be made without knowing the
identity of stressors (Dubé and Munkittrick 2001).
The program is optimized to: (a) indicate trends in
concentrations of certain identified chemical contaminants, (b) indicate response in individuals and populations
of sentinel species caused by chemical exposure, and (c)
indicate fish community response to eutrophication or
other stressors that affect ecological conditions in coastal
habitats. Identification of chemical impacts depends on
the ability of the monitoring to distinguish changes in survival and energy allocation from changes associated with
alterations in habitat and natural variability (Munkittrick
1992; Bussieres et al. 1998). Several natural ambient factors may influence fish, e.g., season, temperature, salinity,
feeding conditions and predation. The sentinel species
monitoring was designed for assessing the influence of
Integrated Coastal Fish Monitoring in Sweden
two of these factors: temperature and feeding conditions,
as they are known to have a major influence on Baltic
perch populations. Salinity may vary in the Baltic due to
irregular inflows of marine water from the North Sea.
Although the effects are small in coastal areas, salinity is
followed in other monitoring and the data may also be
used for interpreting fish monitoring results. Gonad development, growth and many other physiological and biochemical variables often follow distinct seasonal rhythms
influenced by day length. Sampling routines prescribe
short and defined periods for collections to minimize the
effects this may have on trend monitoring.
The biological variables in sentinel species monitoring were selected to indicate population impacts through
either recruitment or adult mortality (Fig. 1). Once
changes are observed, a primary analysis should be made
to evaluate whether impacts are related to toxicity, productivity or temperature. Supporting data from other
Swedish monitoring programs may be used in the assessment. Steps for further follow-up studies should be taken
when the cause of the change is still unclear or when critical stages have to be identified for a deeper understanding of the impact. For example, if a negative trend in relative gonad size (gonadosomatic index, GSI) is observed,
this may indicate a disturbed reproduction. To understand the cause of the change, mechanistic follow-up
studies may be directed to biochemical variables indicating toxic/endocrine disturbance, as well as to measurements of contaminants not included in the program
although judged to be of potential importance. The relevance of the change in GSI at the population level and the
identification of critical life stages can be addressed in
follow-up studies directed to spawning success, embryo
development and larval survival (Karås et al. 1991).
Laboratory follow-up studies are also possible,
allowing verifications of the field observations and closer
analyses of responsible substances once effects have been
documented. Support will also be available from, e.g.,
human and veterinary toxicology, where identical variables are used for indicating toxicity and health impairment in higher vertebrates.
The present monitoring program aims to document
temporal changes in the coastal zone on a basin-wide
scale. With this purpose, historical data are the only possible references, and a statistically well-designed trend
monitoring, covering all relevant levels of biological
organization, is the only option available. There is little
possibility of making retrospective assessments on scattered or insufficient biological data from the past once
changes are detected.
A banking of tissues for future chemical analyses of
persistent compounds can provide an opportunity for
retrospective assessments, and a specimen bank is a
practice included in the contaminants monitoring program (Olsson and Bignert 1997). Similar preservation of
samples is, however, difficult for nonpersistent contami-
235
nants as well as for many biochemical and physiological
purposes, although some materials are kept for future
use, e.g., tissue samples for histological preparations and
blood plasma for hormone analyses.
Criteria for Selecting Monitoring Areas
and Sentinel Species
The selection of adequate sampling areas and sentinel
species is a critical step in the assessment of ecosystem
health (Olsson 1983, 1994; Munkittrick 1992; Bignert et
al. 1994). Much effort was devoted to preparatory studies in potential monitoring areas (Ådjers et al. 1995,
1996) and to increase the knowledge about the biology
of potential sentinel species (Neuman 1976; Karås 1986;
Jacobsson et al. 1986; Sandström et al. 1995; Vetemaa
1998). The Swedish national coastal monitoring has to
provide reference data for local programs at industrial
sites, besides following large-scale effects on the coastal
ecosystem. Thus, selected sites should be as free as possible from local environmental impacts.
Fig. 1. A conceptual model explaining the strategy of effectsdriven fish monitoring.
236
Sandström et al.
The following criteria were set up for the selection
of sampling sites and the delimitation of coastal monitoring areas in the NMMP:
• the area must represent a typical Swedish coastal
environment;
• there should be no local environmental impact;
• the probability of future local impact must be
low;
• the area must be large enough to assure that the
probability of irregular impact from the surroundings is low;
• strong populations of non-migratory fish must
occur, allowing long-term sampling of sentinel
species for monitoring; and
• habitats suitable for all life stages of sentinel
species must be available within the selected area.
It is also important that reference areas provide habitats suitable for other organisms than fish for expanded
integration purposes, e.g., benthic flora and fauna for
eutrophication studies, and bird and mammalian top
predators, i.e., species with a comparatively advanced
metabolic capacity, for monitoring effects of contaminants.
Scanning the coasts, it was found that pulp and
paper mills, metal and petrochemical industries, nuclear
power plants, large cities, shipping and agriculture influenced large parts of the Swedish coast. However, representative areas with little or no local impact could be
identified, and basic studies were made to see if they
could meet the criteria set up for national monitoring.
The first monitoring of fish community and population
variables started in 1962 in Kvädöfjärden, a bay at the
southwest coast of the Baltic proper. Monitoring of contaminants started in the area in 1984, and biomarker
monitoring was included in 1988. The monitoring in
Kvädöfjärden has been expanded to also cover other
ecosystem compartments besides fish.
Besides Kvädöfjärden, three additional coastal reference areas have been accepted so far for the NMMP
(Fig. 2): one in the Northern Quark separating the Bothnian Bay from the Bothnian Sea (Holmöarna), a second
at the coast of the Baltic proper (Torhamn/Gåsöfjärden),
and a third (Väderöarna/Fjällbacka) at the Skagerrak
coast of the North Sea. Additional reference data are
available from regional programs and from other countries bordering the Baltic (Fig. 2).
Research on candidate fish species for sentinel purposes was mainly concentrated to migratory behaviours
and reproduction biology. Herring (Clupea harengus)
and cod (Gadus morhua) were included in the open sea
contaminants program as they represent regional conditions within the Baltic basins. Sentinel species in coastal
monitoring need to be less migratory and stay for long
periods, preferably during all life stages, within the delimited monitoring area. The studies were concentrated on
perch (Perca fluviatilis) and viviparous blenny (Zoarces
viviparus) and they could be recommended for regular
monitoring. Perch belongs to the dominating species in
Baltic archipelagos, while viviparous blenny is common
at the open coasts of the Baltic up to the Northern Quark
and in the Kattegatt and Skagerrak coastal waters.
Variables Selected for Integrated Fish Monitoring
Criteria for Variable Selection
The monitoring variables included in the program (Table 1)
were selected to document:
• trends in contaminant concentrations in fish tissues indicating changes in background levels;
• changes on cellular or subcellular levels following
exposure to toxic/endocrine-disrupting substances;
• changes in vital physiological functions indicating
impacts on recruitment or adult mortality patterns;
• population responses to changes in recruitment
and mortality; and
• changes in fish community structure relative to
eutrophication or climate change.
The contaminant variables were also selected to
serve in human and wildlife risk assessments and to
assess the results of regulatory measures.
Pre-design Studies
Considerable statistical analyses were made to improve
assessments in contaminants monitoring, and the results
showed the importance of location of sampling area,
sample size and sampling frequency in studies of temporal and spatial variation in contaminant exposure (Bignert et al. 1993, 1994; Olsson 1994; Bignert 2003).
Within the financial limits given, sampling was optimized and stratified with regard to species, season, sex,
age and size of the fish. As a consequence of these and
other pre-design studies, monitoring of all variables
included in the integrated program is performed annually according to SEPA guidelines.
Support from Other Programs
Data from several coastal subprograms within the
NMMP and from different monitoring areas can be
compared for comprehensive interpretations. Coastal
data can moreover be compared with open sea monitoring, which can provide more accurate information about
the general development within actual basins, i.e., contaminant trends and nutrient levels. Support from other
programs is also often needed to assess the importance
of changes in natural ambient conditions and effects of
fishing and predation.
Integrated Coastal Fish Monitoring in Sweden
237
Fig. 2. Map showing the sites used for the integrated fish monitoring and for other monitoring activities producing supporting
data on fish populations and contaminants.
238
Sandström et al.
TABLE 1. Environmental contaminants, physiological functions, population and community variables studied in sentinel fish
monitoringa
Response group
Contaminants
Growth, energy
storage and
metabolism
Liver function
and exposure
Immune defence
Hematology and
ion regulation
Pathology
Reproduction
Abundance and
demography
Community
structure
Variables
Cd, Cu, Cr, Ni, Zn, Pb in liver
Hg, PCBs, DDTs, HCHs
and HCB in muscle
Specific growth rate, condition
factor (CF), relative liver size
(LSI), fat content in tissue,
blood glucose, blood
plasma lactate
Liver morphology (necrosis,
degenerated cells), relative
liver size (LSI), EROD,
glutation reductase (GR)
and catalase activities,
metallothionein (MT), DNA
adducts, plasma vitellogenin
Lymphocytes, neutrophilic
granulocytes, thrombocytes
White blood cell counts
Hematocrit (Ht) and hemoglobin
(Hb) (fin erosions, skin)
Plasma Cl-, Na+, K+, Ca2+
Interpretation/significance
Positive trends in concentrations indicate higher exposure and risk
for biological effects, while negative trends reflect the efficiency of
regulatory measures.
Changes in growth, Cf, LSI and fat content may indicate metabolic
disturbance. Temperature, feeding conditions, sex and sex stage
must be considered in assessments. Altered glucose or lactate levels
may indicate sampling stress (increased values), but also metabolic
disturbances due to ambient factors (including toxic exposure).
Structural changes indicate toxic damage or parasite infections.
Change in LSI may reflect nutritional imbalance or high metabolic
activity. EROD and MT reflect detoxification and exposure to
certain organic chemicals and metals. GR and catalase may reflect
oxidative stress and exposure to oxygen radicals. DNA adducts
indicate exposure to genotoxic contaminants. Vitellogenin:
see below.
Low white blood cell counts may indicate suppression of immune
defence. High counts suggest activation due to cell/tissue damage,
inflammations or bacterial infections.
Ht and Hb reflect the capacity of oxygen transport. Together with
the plasma ions Na+ and Cl-, these variables may indicate disturbed
osmoregulation due to gill damage or kidney failure. Increased
plasma K+ may indicate cell or tissue damages, and lowered plasma
K+ reflects impaired gill or intestinal uptake, or reduced kidney
retention. Low plasma Ca2+ levels may indicate kidney damages
and high levels may be due to disturbed Ca regulation.
Pathological changes may indicate toxic exposure, although natural
factors also have to be considered in assessments.
Internal and external pathological
changes (fin erosions, skin
damages, wounds, malformations)
Relative gonad size (GSI)
Lowered GSI indicates reproductive disturbance through impacts on
Plasma vitellogenin
fecundity or oocyte growth rate. Growth and energy storage must
be considered in assessments. Vitellogenin in male plasma indicates
exposure to estrogenic substances. In female fish a low level
indicates hormonal imbalance and a risk for disturbed
reproduction.
Catch-per-unit-of-effort (CPUE)
Changes in CPUE, age and size distributions indicate changes in
Age and size distribution
mortality or recruitment. Fisheries and predation should be
considered in assessments.
Species distribution
Shifts toward cyprinids indicates eutrophication. Changes in total
Total biomass
fish biomass reflect feeding conditions and recruitment. Selective
fisheries and predation should be considered in assessments.
a
Interpretation guidance and significance of change is indicated (modified from Larsson et al. 2000; Sandström et al. 2003).
Information on, e.g., contaminants in different
matrices, nutrient concentrations, pelagic primary production, phyto-benthos and benthic macrofauna, is
accessible from other subprograms in the NMMP. Fishery statistics, although poor in Swedish coastal waters,
are provided by the National Board of Fisheries, and
results from bird censuses may occasionally be available
from regional programs.
Variables in Sentinel Species Monitoring
In this paper we present the variables selected for monitoring of perch. The primary purpose is to produce data
for assessments of toxic/endocrine-disrupting effects,
although responses to temperature variations, predation
and changes in feeding conditions may also be evaluated. The variables analyzed were chosen to reflect population characteristics and to illustrate vital physiological functions: reproduction, growth, energy storage and
metabolism, liver function, immune defence, hematology and ion regulation. Some biomarkers also indicate
chemical exposure, and concentrations of selected contaminants are analyzed. Furthermore, annually collected
tissue samples are stored in a specimen bank (Olsson
and Bignert 1997) to allow future retrospective studies
of contaminants.
Integrated Coastal Fish Monitoring in Sweden
The interpretation and significance of changes in
population characteristics and in single morphometric,
biochemical/physiological or chemical variables is based
on common knowledge about their respective diagnostic
capacity (Larsson et al. 1985, 2000). Table 1 interprets
changes in measured variables and possible ecological
consequences of a direction of change. This primary
interpretation of observed changes has to be followed by
a careful scrutinizing of alternative explanations.
The importance to populations of physiological and
biochemical changes at the level of the organism has
long been a matter of debate. Changes in single biochemical variables may indicate chemical exposure and
provide valuable support in cause-effect analyses. They
may also merit further attention as early signals of toxic
effects, but alone they seldom indicate population
effects, as extrapolations are generally difficult. Our
solution is to analyze functional groups of variables and
decide whether changes are of a magnitude indicating
that the specific physiological function is impacted
(Table 1). Reproductive effects may act on the population through a reduced recruitment, while impairments
of other physiological functions may influence the population through an increase of mortality (Fig. 1).
Variables in Fish Community Monitoring
The coastal fish community monitoring is directed
towards the resident fish assemblage inhabiting Baltic
archipelagos. The primary purpose of this part of the
program is to produce data for assessments of fish community reactions to general ecosystem change caused by,
e.g., eutrophication. The significance of observed changes
is presented briefly in Table 1. Secchi disk depth and
water temperature are measured in connection with test
fishing. Temperature is also recorded throughout the
growth season to provide data for recruitment analyses
and growth comparisons.
Influence from fisheries and predation on the monitored populations may obscure assessments of other
environmental impacts. However, data collected by the
program can be used for evaluating exploitation-related
changes in size distributions and adult fish mortality.
Statistical Power and Trend Analysis
The selection of monitoring variables was influenced by
their respective diagnostic properties. However, when
evaluating long-term monitoring variables, it is important
to determine the capacity to detect temporal changes (i.e.,
the chance to reveal true trends with the chosen sampling
design). Is a negative result of a trend test indicating a
stable situation or is the monitoring program too poor to
detect even a serious change? One approach to this problem is to estimate the statistical power of the time series
based on the observed random among-year variation.
239
The objectives of quantitative long-term monitoring
may be expressed as: (a) the lowest detectable change in
the current time series with an acceptable power, i.e., the
power to detect a log-linear trend (Nicholson and Fryer
1991), or (b) the estimated number of years required to
detect a relevant annual change (e.g., 5–10%) with an
acceptable power. As an example, the objectives of contaminants trend monitoring were set to detect an annual
change of 10% within a 10-year time period with a
power of 80% at a significance level of 5%.
When the statistical power analysis was performed
on the obtained data series, the “random” between-year
variation was calculated using the residual distance from
a log-linear regression line. In many cases the log-linear
line, fitted to the current observations, seemed to be an
acceptable representation of the true development of the
time series. In cases where a significant non-linear trend
was detected, the regression line may not serve this purpose and the estimated power may be biased. Such
results were excluded from estimations of performance.
Another problem is that single outliers can strongly
influence estimations of among-year variation, reducing
the possibilities to detect trends. However, suspected
outliers were not excluded in the current statistical
analysis. If these values are true outliers which do not
belong to the studied population and could be excluded,
the power and sensitivity would in some cases be considerably improved.
Model for Integrated Assessment
of Fish Response
General Aspects
Theoretically, fish populations respond in a predictable
manner to different stressors (Colby 1984). Since fish
represent high trophic levels, populations reflect the flow
of energy through the ecosystem. Eutrophication
changes the energy flows, generally increasing fish production, but also other habitat change can affect fish
populations. Coastal fish are often sensitive to overgrowth of algae on the spawning grounds, changes in
turbidity, and other physical changes in recruitment
habitats (Sandström and Karås 2002). Ecosystem
changes related to eutrophication are known to influence
the relation among species in the Baltic coastal fish community (Neuman and Sandström 1996; Sandström and
Karås 2002). Abundance and biomass of common
cyprinids like roach (Rutilus rutilus) and silver-bream
(Blicca bjoerkna) generally increase during moderate
eutrophication while perch and many other species do
not respond or react negatively. The net result will be a
shift in community structure, which can develop towards
an unwanted reduction of biodiversity.
The sentinel species monitoring is based upon general life-history theory and results from more directed
240
Sandström et al.
research on the selected species. It has many similarities
with the model originally proposed by Colby (1984) and
further elaborated by Munkittrick and Dixon (1989)
and Gibbons and Munkittrick (1994). The conceptual
framework is drawn from the theoretical and empirical
literature on trade-offs and fitness and the patterns of
energy allocation and reproductive response variables.
There is a considerable literature on, e.g., how sexual
development in fish is influenced by growth rate and
energy storage (Policansky 1983; Stearns and Crandall
1984; Roffe 1984; Rowe et al. 1991; Thorpe 1994;
Gagnon et al. 1995; Sandström et al. 1995; Svedäng et
al. 1996). In order to evaluate the trade-offs between
growth, survival and reproduction, mechanistic bioenergetics frameworks have been conceptualized as a base
for developing individual-based population models (Van
Winkle et al. 1997). The models track daily somatic condition, consumption, energetic cost, growth, mortality
risk and reproductive status of each fish from first feeding to death. Potential applications include evaluation of
the effects on populations by different stressors, e.g.,
industrial effluents, food deficits, climate change, chemical exposure and fishing pressure.
More specific data on the life history of perch and
reactions to different stressors were obtained during
research on populations exposed to cooling water and
pulp mill effluents. Interactions between growth, storage
and reproduction were studied in a perch population
exposed to cooling water from a nuclear power plant
(Sandström et al. 1995). The high temperature allowed
very fast juvenile growth, which enabled an early sexual
maturation at a small size. However, spent fish had
depleted their energy stores to such low levels that mortality increased and repeated spawning was inhibited in
surviving fish. The reactions of this perch population
could be explained in terms of life-history strategy
(Stearns and Crandall 1984).
When a perch population exposed to effluents from a
kraft pulp mill was studied, the results were conflicting.
Growth was faster and the condition factor higher in
exposed fish. Reproduction was, however, inhibited
(Sandström et al. 1988) as sexual maturation was delayed
and gonad size reduced. Recruitment was impaired, and
adult fish appeared in low abundance in the effluent area
(Neuman and Karås 1988; Karås et al. 1991). Very similar results were obtained in parallel Canadian pulp mill
research (McMaster et al. 1991; Munkittrick et al. 1994;
Gagnon et al. 1994, 1995; Lowell et al. 2003). The
response pattern indicated a metabolic disturbance stimulating growth and inhibiting reproduction, which differed
from the response seen in cooling-water exposed fish. The
importance of growth as a natural regulator of reproductive investment was evident from the studies by Gagnon et
al. (1995) and Bussieres et al. (1998) on the reproduction
of white sucker along productivity gradients in rivers contaminated with pulp mill effluents and in reference rivers.
Simple upstream-downstream comparisons would not
have been successful unless the relations between productivity and fish growth in both streams had been included
in the interpretations.
Exposure to chemical substances was indicated by
biomarker response in many of the studies. Elevated
EROD activities indicated the presence of potentially
toxic Ah-receptor ligands, and changes in vital physiological functions indicated severe health impairment
and increased mortality risk (Andersson et al. 1988).
Higher mortality in exposed fish could also be verified
when age distributions were analyzed (Sandström and
Thoresson 1988). It was concluded that the response
pattern indicated serious metabolic disturbance caused
by toxic substances in the effluent with effects on all
levels of organization in the exposed fish population
(Larsson et al. 2003).
Response Model for Baltic Perch
At present, there is no individual-based model developed
for Baltic perch populations, but there is enough information to summarize the predicted directions of
responses to different stressors into a conceptual framework. The response model for Baltic perch is based on
the model proposed by Colby (1984) and Gibbons and
Munkittrick (1994). Data on native perch populations
under different kinds of stress were used to elaborate the
model, and additional support was provided by, e.g., the
relations between growth response and reproduction in
white sucker demonstrated by Gagnon et al. (1995) and
Bussieres et al. (1998). The model aims to support analyses of observed patterns of energy allocation and reproduction, and it also includes predicted responses on the
subcellular and cellular levels due to contaminant exposures. The model still is general in many respects, and it
should be used primarily for a preliminary assessment
and to direct further analytical steps towards critical
aspects and key areas. It can be especially difficult to
analyze cause and effect relationships when several stressors simultaneously act on the population.
The predicted response patterns in Baltic coastal
perch populations can be summarized as follows:
• Faster growth, higher condition, larger livers, earlier maturation and increased gonad size indicate
nutrient enrichment (when food is limiting) as
well as increased temperatures (within reasonable
limits in relation to the optima of the fish). Biomarkers of exposure to toxic/endocrine-disrupting
substances do not react. Concentrations of monitored contaminants do not change. This response
is in agreement with basic life-history theory.
• Faster growth, higher condition, larger livers,
later maturation and smaller gonad size indicate
exposure to toxic or endocrine-disrupting sub-
Integrated Coastal Fish Monitoring in Sweden
stances. Biomarkers of exposure may react,
depending upon active substances, and impairments of physiological functions can be expected.
Adult mortality may increase. Concentrations of
monitored contaminants may be elevated. This
response is not in agreement with basic life-history theory, as increased growth should allow
earlier maturation and larger gonads. Higher
energy use for growth and storage and lower
commitment to reproduction should be interpreted as a metabolic disturbance.
• Slower growth, lower condition, smaller livers,
later maturity and smaller gonad size, if feeding
conditions and temperature stay constant, indicate a severe toxic situation. Biomarkers of exposure may react and impairments of physiological
functions are expected. Concentrations of monitored contaminants may be elevated. Adult mortality will increase. This is a strong signal that the
ecosystem really is at risk.
• Increased adult mortality, lower adult abundances and changes in size distributions indicate
exploitation by fisheries or predation by birds
and mammals. Biomarkers of exposure or effect
do not react and there is no change in contaminant concentrations. If there are food limitations,
and if the reduced abundances will lead to lower
competition for resources, growth rate will
increase as well as condition and liver weight.
Maturation will be earlier and gonad sizes larger.
In perch populations not regulated by densitydependent factors, exploitation or predation will mainly
influence abundances and size distributions with no
effects on individual growth, maturation and GSI. This
can be expected in Baltic archipelagos where populations
are primarily regulated by recruitment.
Trends Observed in the Integrated Monitoring
Data from the Kvädöfjärden bay long-term monitoring
were used to analyze the statistical power and to test the
interpretation tools and the integrative properties of the
system, partly because we have the longest time series
from this area, and partly because trends have been documented which need a comprehensive analysis. In the
following presentation we have selected some of these
series for an integrated analysis of observed changes.
The complete results and time trends recovered from
the Kvädöfjärden monitoring area are reported elsewhere, together with more detailed descriptions of sampling methods, materials and analytical procedures
(Ådjers et al. [2001] for fish community and perch population monitoring; Hansson et al. [Submitted for publication] for perch biochemistry/physiology; Bignert and
Asplund [2003] for contaminants).
241
Statistical Power
Assessment of statistical power was primarily made for
variables monitored for more than 8 years. In most cases
the power was high or acceptable for the respective
objectives (Table 2). Long-term perturbations in this
aquatic ecosystem are not expected to be swift or stepwise, and slopes of 5 to 10% per year should be realistic
maximum levels of change. If a 5% annual change can
be detected within 15 years, this can be considered as
acceptable. A 5% annual change corresponds to a total
change of 100% in 14 years.
Reliable estimations based on shorter time series
cannot be expected, and some variables recently
included in the program need to be tested for additional
years before they can finally be approved for long-term
monitoring. However, the calculated statistical power
was generally high also for these variables.
Fish Community Structure,
Perch Abundance and Growth
Test fishing for community studies and collection of
samples for perch age analyses were made in August,
using multi-mesh-size gill nets set in shallow water
(3–5 m). Captured fish were determined to species and
examined for external injuries, malformations, diseases
and parasitic infections. About 300 female perch, distributed over the size range from 12 cm to ca. 30 cm,
were sampled annually from test-fishing catches for age
and growth analysis. Back-calculated growth was estimated for different ages (Thoresson 1996).
The test-fishing data were further elaborated to show
possible effects of eutrophication or climate change on
fish community structure. In the southern parts of the
Baltic, where the influence of brown-coloured river water
is small, the Secchi disk depth is a good indicator of phytoplankton densities (Dennison 1987). The relation
between cyprinids (nine species, dominated by roach, silver bream and rudd [Scardinius erythrophthalmus]) and
perch in catches was calculated and related to the Secchi
disk depth (Fig. 3). There have been rather small variations in Secchi disk depth among years with no significant trend. Cyprinids dominated over perch at the beginning of the study period, but there was a change towards
more equal shares with a ratio close to 1 in 2001 and
2002. The shift in species dominance was a result of both
decreased catches of cyprinids and increased catches of
perch. Total biomass, however, did not change significantly over time (Fig. 4). The catches of 15- to 20-cm
perch, which is the dominating size-class, increased significantly during the study period (Fig. 5).
Growth as indicated by individual length increase
had increased significantly for the first, second and third
growth seasons (Fig. 6). As growth rate may be influenced by temperature, the relation between summer
242
Sandström et al.
TABLE 2. Results of analyses to estimate the power to detect log-linear trends in the time seriesa
Response group
Variables
n of years
available
n of years required
to detect a slope
of 5% at a
power of 80%
Abundance
Reproduction
Growth, energy
storage and
metabolism
CPUE
Gonad size (GSI)
Juvenile (1+) growth
Adult (3+) growth
Condition factor (Cf)
Blood glucose
Plasma lactate
EROD
GR
GST
Catalase
Liver size (LSI)
Lymphocytes
Neutrophilic granulocytes
Thrombocytes
White blood cell count
Hemoglobin
Hematocrit
Plasma chloride
Hg
Pb
Cd
Ni
Cr
Cu
Zn
ΣPCB
CB-153
α-HCH
γ-HCH
HCB
16
15
16
16
15
8
14
15
9
9
8
15
4
4
4
4
8
15
14
19
6
6
6
6
6
6
19
11
13
9
13
17
9
6
10
4
10
18
12
16
14
15
6
9
7
12
8
9
6
4
15
9
13
20
7
14
11
20
20
14
19
20
Liver function
and exposure
Immune defence
Hematology and
ion regulation
Contaminants
Minimum slope
possible to detect in
a period of 10 years
at a power of 80%
12
3.7
1.5
4.7
0.8
4.7
12
6.1
10
8.0
9.3
2.0
3.5
2.5
6.2
3.3
3.5
1.9
0.7
9
3.7
7.5
15
2.3
7.9
5.3
15
15
8.8
14
15
a
Statistical power has only been calculated for variables tested in monitoring for more than three years.
water temperatures (May to September) recorded manually every week in the monitoring area and the annual
length increase was analyzed. The mean annual length
increase over the period 1985 to 2000 was calculated for
each age group each year of catch. A normalized growth
value was calculated as the ratio between the annual
mean and the grand mean for all years. These normalized growth estimates were used to analyze the correlation between temperature and growth. There was a statistically significant dependence between annual length
growth and temperature variations (Fig. 7; r2 = 0.78).
Physiological and Biochemical Analyses
Biochemical, physiological and histological analyses
were made on samples collected during autumn (September), after the start of the gonad growth period. Gill
nets were used to capture 25 mature female perch
within the size range 20 to 30 cm (an additional sample
of 10 males was collected for vitellogenin analyses). The
analytical procedures for the variables presented in
Table 1 are described by Hansson et al. (Submitted for
publication). The age of the sampled fish was estimated
during 1993 to 1995 and 1997 to 2001 according to
Thoresson (1996).
Gonadosomatic index (GSI), vitellogenin in plasma,
condition factor (Cf) and liver-somatic index (LSI)
reflect reproductive capacity and the energy storage of
the fish. Cf and LSI (measured since 1988) differed little
among years with no significant trends (Hansson et al.
Submitted for publication), but a significant reduction of
GSI was found (Fig. 8). The change over the monitored
15-year period was about 40%. Vitellogenin in male
plasma, indicating exposure to estrogenic substances,
was included in the program in 1998. It is thus too early
to expect a trend in plasma vitellogenin concentrations.
The concentrations have, however, been at such low levels that estrogenic exposure seems unlikely.
Integrated Coastal Fish Monitoring in Sweden
243
Fig. 3. Secchi disk depths and the ratio of cyprinids/perch in
test-fishing catches.
Fig. 4. Total biomass (kg) of perch, cyprinids and other
species in test-fishing catches.
Fig. 5. Abundance (CPUE) of perch (15–20 cm) in test
fishings.
Fig. 6. Annual length increment in perch of ages 1 to 6 in
different calendar years during the period 1985 to 2000.
Fig. 7. The relation between normalized growth and summer
temperatures.
Fig. 8. Relative gonad size (GSI) in female perch.
244
Sandström et al.
EROD activities increased significantly about 3fold during the study period (Fig. 9). As a relationship
between gonad size and EROD activity in female fish
may be expected through, e.g., a stimulative influence
of estradiol on gonad growth and a possibly suppressive effect on EROD activities (Andersson and Förlin
1992), the correlation between EROD activities and
GSI was analyzed. There was a statistically significant
correlation between the EROD increase and the GSI
decrease (Hansson et al. Submitted for publication). By
statistical means the EROD activities were adjusted for
the effect of lower GSI. The adjusted values gave a
slightly lower increase rate, but the adjusted EROD
trend still was significant.
Apart from the change in GSI and EROD activities,
a significantly increasing trend of about 1% per year in
plasma chloride concentrations was observed (Hansson
et al. Submitted for publication).
Contaminant Concentrations
Monitoring of contaminants in perch from Kvädöfjärden
started in 1980 as part of the NMMP, using methods
described in Bignert et al. (1998). Ten female perch
within the size-range 15 to 20 cm were collected annually in August using gill nets.
Concentrations of most contaminants analyzed (Cu,
Hg, Pb, Zn, PCBs, DDTs, HCHs, HCB) have decreased
since the samplings started (Bignert and Asplund 2003).
The PCB time series was selected to illustrate this general
development. A significant decrease at a rate of 6% per
year was observed since 1989 (Fig. 10A). The concentration of cadmium in perch liver increased rapidly until
2001, when this trend was broken (Fig. 10B). Cadmium
contamination in the Baltic has been a matter of concern
as concentrations have increased in herring samples from
the Baltic proper (Bignert and Asplund 2003).
Information about tissue fat concentration is made
available when samples for contaminants monitoring are
analyzed. Data exist for two periods, 1980 to 1985 and
1995 to 2003. There was a significant (p < 0.001)
decrease in mean fat content in perch muscle tissue from
0.76 to 0.69% when the periods were compared.
Interpretation of the Monitoring Results
Fish Community Recovery
Community structure, total fish biomass and Secchi disk
depths in the Kvädöfjärden monitoring area did not
change in directions reflecting progressing eutrophication. The shift from a dominance of cyprinids to a more
balanced community with shares close to 50% on the
contrary indicates a commenced recovery towards normal conditions for Baltic archipelagos (Neuman and
Sandström 1996). Supporting information from other
monitoring confirms that an increasing trend in concentrations of nutrients was broken at the end of the 1980s
and that there has been a tendency for decreasing concentrations during later years in surface water (Fig. 11).
The community reaction can thus be seen as an indication of reduced eutrophication.
a
b
Fig. 9. EROD activities in female perch.
Fig. 10. (a) PCB-153 lipid concentrations in perch muscle.
The trend is presented by a regression line (plotted if p < 0.1;
two-sided regression analysis). (b) Cadmium concentrations
in perch liver.
Integrated Coastal Fish Monitoring in Sweden
Decreased GSI and Increased EROD Activity
During the last years several biochemical and physiological exposure and/or effect indicators have been included
in the program, but so far, significant trends are only
seen in GSI (Fig. 8), EROD activities (Fig. 9) and plasma
chloride concentrations. Other variables monitored for
longer periods without showing any trends should indicate stable situations, not that the monitoring program
was too poor to detect relevant change, as the statistical
power generally was acceptable (Table 2). Both the
decreased gonad size and the EROD response may indicate an increased exposure to toxic and/or endocrineactive substances. An elevated EROD activity is a common reaction in fish exposed to complex effluent from
pulp mills, oil refineries, petrochemical plants or other
industrial or municipal activities (Andersson et al. 1988;
Goksøyr and Förlin 1992; Förlin et al. 1994; Vetemaa et
al. 1997; Stephensen et al. 2000; Lindesjöö et al. 2002).
In a review, Sandström (1996) found an indication of
metabolic disturbance (faster growth and inhibited
reproduction) in seven cases where fish had been
exposed to pulp and paper effluent. In all these cases
there was also a significant increase in EROD activity.
Generally, EROD induction in fish is interpreted as
a reaction to exposure to Ah-receptor ligands including
planar compounds, e.g., halogenated dioxins and co-planar PCBs, as well as certain PAHs (Stegeman et al. 1992;
Goksøyr and Förlin 1992; Andersson and Förlin 1992).
However, PCB concentrations have decreased in the
studied perch population during the monitored period
(Fig. 10A). The pattern of generally lower concentrations of PCBs in this part of the Baltic is confirmed by
the monitoring of cod and herring (Bignert et al. 1998;
Bignert and Asplund 2003). As there are no data on
dioxin concentrations in coastal biota; the dioxin expo-
245
sure for the studied perch population is unclear. Monitoring of Guillemot eggs from the central Baltic proper
has shown a significant decrease in dioxin concentrations from the 1970s to the beginning of the 1980s.
Since then, this trend is broken and there is no longer
any significant change (Olsson et al. 2003a,b). Data on
herring collected in the southern Baltic proper and the
Bothnian Bay show an unchanged dioxin level from the
end of the 1980s up to 2000. However, herring collected
in the southwest Bothnian Sea indicate an increase since
the end of the 1970s (Olsson et al. In press).
Among other contaminants not included in the
perch monitoring, the PAHs are known to induce
EROD. Relevant data about PAHs from other programs
are also lacking.
The EROD response may be related to decreasing
GSI suggesting an altered sex hormone metabolism in
mature female perch. Sex hormones such as estradiol
were never measured but it is known that EROD activities are generally higher in juvenile female perch than in
adult females with maturing gonads and high plasma
levels of estradiol. Although there may be other possible
or unknown interactions between EROD activities and
the physiological status of the fish, the trend in EROD
activity in Kvädöfjärden perch after the first interpretation step should be seen as an indication of increased
toxic exposure.
Increased Plasma Chloride
The weak but statistically significant increase of the plasma
chloride level suggests an altered ion regulation. The reason for such an alteration is, at present, difficult to explain,
although a reaction to the increase in cadmium concentrations during 1989 to 1999 might not be excluded.
Increased Growth and Decreased GSI—
Unexpected and Abnormal Response Pattern
Fig. 11. Trends in nutrient (nitrate and phosphate, January
to February) concentrations at station BY 32, northwest
Baltic proper. Data fom the National Marine Monitoring
Programme.
The increased growth rate and reduced gonad size noted
in Kvädöfjärden perch is not an anticipated normal
response pattern. The positive impact on growth should
result in higher commitment to reproduction. A lack of
this coupling between growth and reproduction, should
after a first evaluation step, be considered to indicate
metabolic disturbance or a possible toxic effect on the
reproduction process.
However, predictive response models are general in
character, and results should be tested for natural variability and confounding factors. The integrated fish monitoring allows a second analytical step incorporating
some possible factors. Growth is considered to be a driving force in fish reproduction, and the understanding of
why growth rate has changed is important. In pulp mill
research, which has produced much of the empirical data
behind the theoretic development of response patterns
246
Sandström et al.
(Sandström et al. 1988; McMaster et al. 1991; Gibbons
and Munkittrick 1994; Gagnon et al. 1994; Lowell et al.
2003), growth stimulation in concert with inhibited
reproduction is seen as a change in energy allocation. In
this respect, growth stimulation is not a natural response.
Many different stressors can influence perch growth,
among them temperature, which may be the most forceful environmental factor in Baltic coastal waters (Neuman 1976; Karås and Neuman 1981). The analysis of
temperature versus annual length increase did reveal that
the growth response in Kvädöfjärden perch could be
explained to a considerable extent by higher temperatures. Enhanced growth consequently should be interpreted as a response to a climate change, not a sign of
metabolic disturbance. Cf and LSI—indicating energy
storage—did not react significantly during the study
period. However, a change in fat metabolism in juvenile
fish sampled for contaminants monitoring was indicated
by the significant negative decrease in muscular fat content. The relevance of this change is unclear, and an integrated interpretation is difficult as the samples consisted
of juvenile fish while adults were selected for biochemical/physiological monitoring and as the samples for contaminants were collected earlier in summer.
Although the faster growth could be explained as a
natural response to higher temperatures, this should still
have resulted in a positive effect on reproduction, not a
negative response, and certainly not a reduction of the
magnitude seen during the 15-year monitoring period.
One possible contribution to the trend in GSI may be the
lower age of the sampled fish in later years (Hansson et al.
Submitted for publication). The effect of size on GSI and
other variables is minimized by the sampling strategy, but
an effect of age is possible. When the pooled material of
aged fish was used for a correlation analysis, there was a
significant positive dependence of GSI on age (p < 0.01).
However, the negative GSI trend remained when the effect
of age was eliminated (ANCOVA, p < 0.02, tao = 0.033).
In normal conditions perch begin maturing during
their second and third summers, although slow-growing
individuals may mature later (Sandström et al. 1995). To
minimize the possibility that first-time spawners could
influence the results, a regression analysis was made for
4- and 5-year-old fish over the study period. GSI of these
age classes decreased significantly over time (p < 0.02).
Consequently, the overall GSI trend could not be
explained as only the result of the faster growth, and
hence lower age of sampled fish. A follow-up study of
the relations between size, age, growth, maturation and
GSI may be necessary for a final assessment.
Conclusions
There were only few significant trends detected in the integrated fish monitoring in Kvädöfjärden. As statistical
power was high or acceptable for most variables, the
absence of trends was not an effect of high variability,
masking impacts. It was also evident that the study area
was well selected for assessing long-term low-level effects
of anthropogenic stress, and that more irregular impacts of
a local nature have been small. Most variations observed
were likely of a natural character, which supports the use
of monitoring data from this site in comparisons with fish
responses in contaminated receiving waters.
Apart from the changes in GSI, EROD activity,
muscular fat content and plasma chloride there was no
apparent disturbance of vital physiological functions
pointing to an immediate risk for increased adult mortality, which was also verified by the significant increase in
perch abundance. As perch recruitment is strongly influenced by temperature, the generally warmer springs and
summers during later years is a likely explanation for the
increase in catches due to increased growth, and hence
survival, during the first year of life (Karås 1986). The
trend in abundance also indicates that the possible
impact on reproductive capacity due to the GSI change
was not serious enough to counteract the positive effect
of temperature on recruitment.
After this tentative analysis of the data we can conclude that there are statistically significant time trends
indicating a toxic exposure and/or a metabolic disturbance, which still remain to be explained. It cannot be
excluded that these changes are related to natural events
such as faster growth and lower mean age of sampled
fish, and therefore follow-up studies are needed for a full
understanding of the results. However, at present the
reduced or delayed gonad development, the induced
EROD activity, and the affected chloride regulation
must be regarded as early signals of toxic response eventually leading to disturbed reproduction and/or impaired
osmotic and ionic regulation.
Experiences of the Integrated Approach
Integration starts with the practical arrangements during
sampling and data handling. Our experience has shown
that much was gained from using a common infrastructure during fishing and sample treatment. This has been
an important contribution to quality assurance. Data are
stored in common databases with free access for everyone, and it has been possible to produce integrated
assessments according to common routines, often in single comprehensive analyses.
The next step to consider is the possibility to use all
or most variables in integrated interpretations. The basis
for this is the interpretation model. The analysis presented in this paper has shown that when trends are
observed, the system has an integrative capacity to provide an initial, tentative analysis. Lacking information
could also be identified and at least some of this can be
available after a further treatment of stored materials
and data and through directed follow-up studies.
Integrated Coastal Fish Monitoring in Sweden
The need for long-term commitment was evident
when reviewing the monitoring results. Although much
can be gained by an adequate statistical design, it may
take years of monitoring to distinguish between natural
variation and the basin-wide low-level effects of anthropogenic stress we should be able to detect. We have to
accept that ecosystem health variables tend to fall within a
wide range and follow temporal directions that can only
be revealed by prolonged observations (Sherry 2003).
A well-designed and focused monitoring program
must be able to identify and analyze impacts on all
levels in a cost-effective way. There are, however, several demands set up for the fish monitoring which are
difficult to fit into an integrative framework. Contaminants monitoring must show changes relevant for
human and wildlife risk assessments, even if levels are
too low to harm the fish. Due to biomagnification,
contaminant concentrations not harmful for fish might
increase to harmful concentrations in the fish predator. Similarly, the original contaminants in fish might
be metabolized to a more toxic compound in the metabolic system of birds and mammals.
Other parts of the program fulfil the need of references for local pollution studies. Basic research can also
be justified even if the results do not provide support for
comprehensive interpretations. For example, new biomarkers must be developed and tested somewhere, and
why not in an integrated program where practical
arrangements can be facilitated and where comparisons
with other data will be possible.
As a conclusion, strong as well as weak properties
of the integrated monitoring program could be identified. The positive aspects are:
• it was possible to cooperate in the same areas, on
the same species and sometimes on the same
specimens;
• the statistical design and sampling routine had
resulted in adequate statistical power for most
variables monitored;
• monitoring variables, known to provide reliable
responses in local pollution studies, could also be
introduced in long-term integrated monitoring;
• the long-term approach to monitoring made
assessments of low-level effects possible;
• it was possible to show convincing trends in contaminant concentrations;
• it was possible to show shifts in community
structure which could be interpreted in terms of
ecosystem productivity;
• it was possible to show trends in population
abundance and individual growth rate;
• the trends in abundance and growth could be
explained by a temperature increase;
• it was possible to show trends in GSI, EROD
activity and plasma chloride concentrations;
247
• a first, tentative analysis of these data was possible to perform;
• the use of a predictive response model, based on
life-history theory, was a constructive approach
to interpretations; and
• the concept of physiological functions in monitoring was a constructive approach to interpretations.
Negative aspects or weaknesses of the present monitoring program, and hence challenges for future
improvements, can be summarized as:
• we have a lack of data on certain suggested risk
substances such as the PAHs, dioxins and other
“modern” contaminants such as brominated
compounds;
• we have a lack of data about the identity of suspected endocrine disrupters;
• biomarkers of exposure are included, but we
know that they may not react to all relevant risk
substances;
• the possibility to integrate contaminants and biochemical/physiological monitoring and assessment has not yet been fully explored;
• there is still a lack of good markers for reproduction, e.g., serum steroid hormones, in the program, so further development is needed;
• we lack sufficient data on sexual maturation—a
critical variable in reproduction studies;
• we lack sufficient information for fully understanding the life-history of monitored species, so
further research is needed for development of
response models; and
• we lack sufficient knowledge of how Baltic fish
communities react to different stressors from a
biodiversity perspective.
Acknowledgements
This evaluation of the strategy for integrated fish monitoring was financed by the Swedish Environment Protection Agency (Contract No. 212 0326) and the Faculty of
Science, Göteborg University. Basic data for the evaluation was available from the National Marine Monitoring
Programme, integrated fish monitoring project, which
has been financed by the Swedish Environment Protection Agency on different contracts since 1992.
References
Ådjers K, Andersson J, Böhling P, Mölder M, Neuman E,
Sandström O. 1996. Monitoring in Baltic coastal reference areas. TemaNord 1996:627.
Ådjers K, Appelberg M, Eschbaum R, Lappalainen A, Lozys
L. 2001. Coastal fish monitoring in Baltic reference
areas 2000. Kala-ja Riistaraportteja nro 229. 15 p.
248
Sandström et al.
Ådjers K, Böhling P, Järvik A, Lehtonen H, Mölder M,
Neuman E, Raija T, Storå C. 1995. Coastal fish monitoring in the northern Baltic proper - establishment of
reference areas. TemaNord 1995:596.
Andersson T, Förlin L. 1992. Regulation of the cytochrome
P450 enzyme system in fish. Aquat. Toxicol. 24:1–20.
Andersson T, Förlin L, Härdig J, Larsson Å. 1988. Physiological disturbances in fish living in coastal water polluted with bleached kraft pulp mill effluents. Can.
J. Fish. Aquat. Sci. 45:1525–1536.
Bignert A. 2003. Biological aspects and statistical methods to improve assessments in environmental monitoring. Ph.D. thesis, Department of Zoology, Stockholm University.
Bignert A, Asplund L. 2003. Comments concerning the
National Swedish Contaminant Monitoring Programme in Marine Biota, 2003. Report from the Contaminant Research Group at the Swedish Museum of
Natural History and the Institute of Applied Environmental Research, University of Stockholm.
Bignert A, Göthberg A, Jensen S, Litzén K, Odsjö T, Olsson
M, Reutergårdh L. 1993. The need for adequate biological sampling in ecotoxicological investigations: a
retrospective study of twenty years pollution monitoring. Sci. Total Env. 128:121–139.
Bignert A, Olsson M, de Wit C, Litzén K, Rappe C,
Reutergårdh L. 1994. Biological variation – an important factor to consider in ecotoxicological studies based
on environmental samples. Fresenius J. Anal. Chem.
348:76–85.
Bignert A, Olsson M, Persson W, Jensen S, Zakrisson S,
Litzén K, Eriksson U, Häggberg L, Alsberg T. 1998.
Temporal trends of organochlorines in Northern
Europe, 1967–1995. Relation to global fractionation,
leakage from sediments and international measures.
Environ. Poll. 99:177–198.
Bussieres D, Gagnon MM, Dodson JJ, Hodson PV. 1998.
Does annual variation in growth and sexual maturation
of white sucker (Catostomus commersoni) confound
comparisons between pulp mill contaminated and reference rivers? Can. J. Fish Aquat. Sci. 55:1068–1077.
Calow P. 1995. Ecosystem health – a critical analysis of concepts, p. 33–41. In Rapport DJ, Gaudet CL, Calow P
(ed.), Evaluating and monitoring the health of large-scale
ecosystems. Springer-Verlag, Heidelberg, Germany.
Colby PJ. 1984. Appraising the status of fisheries: rehabilitation techniques. In Cairns VW, Hodson PV, Nriagu
JO (ed.), Contaminants effects on fisheries. Adv. Environ. Sci. Technol. 16:233–257.
Dennison WC. 1987. Effects of light on seagrass photosynthesis, growth and depth distribution. Aquat. Bot.
27:15–26.
Dubé M, Munkittrick K. 2001. Integration of effects-based
and stressor-based approaches into a holistic framework for cumulative effects assessments in aquatic
ecosystems. Human Ecol. Risk Assess. 7(2):247–258.
Förlin L, Goksøyr A, Husøy AM. 1994. Cytochrome P450
monooxygenase as indicator of PCB/dioxin like compounds in fish, p. 135–150. In Kramer KJM (ed.), Biomonitoring of coastal waters and estuaries. CRC Press
Inc., Boca Raton, Fla.
Gagnon MM, Bussieres D, Dodson JJ, Hodson PV. 1995.
White sucker (Catostomus sommersoni) growth and
sexual maturation in pulp mill-contaminated and reference rivers. Environ. Toxicol. Chem. 14:317–327.
Gagnon MM, Dodson JJ, Hodson PV, Van Der Kraak G,
Carey JH. 1994. Seasonal effects of bleached kraft mill
effluent on reproductive parameters of white sucker
(Catostomus commersoni) populations of the St. Maurice River, Quebec, Canada. Can. J. Fish. Aquat. Sci.
51:337–347.
Gibbons W, Munkittrick K. 1994. A sentinel monitoring
framework for identifying fish population responses
to industrial discharges. J. Aquat. Ecosyst. Health
3:227–237.
Goksøyr A, Förlin L. 1992. The cytochrome P450 system in
fish, aquatic toxicology and environmental monitoring.
Aquat. Toxicol. 22:287–312.
Hansson T, Lindesjöö E, Förlin L, Balk L, Bignert A, Larsson Å. Submitted for publication.
Jacobsson A, Neuman E, Olsson M. 1986. The viviparous
blenny as an indicator of environmental effects of
harmful substances. Ambio 15:236–238.
Karås P. 1986. Basic abiotic conditions for production of
perch (Perca fluviatilis L.) young-of-the-year in the
Gulf of Bothnia. Ann. Zool. Fenn. 33:371–381.
Karås P, Neuman E. 1981. First year growth of perch
(Perca fluviatilis L.) and roach (Rutilus rutilus (L.)) in a
heated Baltic bay. Rep. Inst. Freshw. Res. Drottningholm 59:48–63.
Karås P, Neuman E, Sandström O. 1991. Effects of a pulp
mill effluent on the population dynamics of perch,
Perca fluviatilis. Can. J. Fish. Aquat. Sci. 48(1):28–34.
Larsson Å, Förlin L, Grahn O, Landner L, Lindesjöö E, Sandström O. 2000. Guidelines for interpretation and biological evaluation of biochemical, physiological and pathological alterations in fish exposed to industrial effluents.
SSVL Miljö 2000, Rapport nr. 5, Supplement 2.
Larsson Å, Förlin L, Lindesjöö E, Sandström O. 2003.
Monitoring of individual organism responses in fish
populations exposed to pulp mill effluents, p. 216–226.
In Environmental impacts of pulp and paper waste
streams. Proceedings of the 3rd Int. Conf. Environmental Fate and Effects of Bleached Pulp Mill Effluents.
SETAC Press, Rotorua, New Zealand.
Larsson Å, Haux C, Sjöbeck ML. 1985. Fish physiology
and metal pollution: results and experiences from
laboratory and field studies. Ecotox. Environ. Safe.
9:250–281.
Lindesjöö E, Adolfsson-Erici M, Ericson G, Förlin L. 2002.
Biomarker responses and recin acids in fish chronically
exposed to effluents from a total chlorine-free pulp mill
Integrated Coastal Fish Monitoring in Sweden
during regular production. Ecotox. Environ. Safe.
53:238–247.
Lowell R, Ribey S, Khouzam Ellis I, Grapentine L, McMaster ME, Munkittrick KR, Scroggins R. 2003. National
assessment of the pulp and paper environmental effects
monitoring data. Environment Canada, National
Water Research Institute, Burlington/Saskatoon, NWRI
Contribution No. 03-521.
McMaster ME, Van Der Kraak GJ, Portt CB, Munkittrick
KR, Sibley PK, Smith IR, Dixon DG. 1991. Changes in
hepatic mixed function oxygenase (MFO) activity,
plasma steroid levels and age at maturity of a white
sucker (Catostomus commersoni) population exposed
to bleached kraft pulp mill effluent. Aquat. Toxicol.
21:199–218.
Munkittrick K. 1992. A review and evaluation of study
design considerations for site-specifically assessing the
health of fish populations. J. Aquat. Ecosyst. Health
1:283–293.
Munkittrick KR, Dixon DG. 1989. Use of white sucker
(Catostomus commersoni) populations to assess the
health of aquatic ecosystem exposed to low-level contaminant stress. Can. J. Fish. Aquat. Sci. 46:1455–1462.
Munkittrick KR, Van Der Kraak GJ, McMaster ME, Portt
CB, van den Heuvel MR, Servos MR. 1994. Survey of
receiving-water environmental impacts associated with
discharges from pulp mills. 2. Gonad size, liver size,
hepatic EROD activity and plasma sex steroid levels in
white sucker. Environ. Toxicol. Chem. 13(7):1089–1101.
Neuman E. 1976. The growth and year-class strength of
perch (Perca fluviatilis L.) in some Baltic archipelagos,
with special reference to temperature. Rep. Inst.
Freshw. Res. Drottningholm 55:51–70.
Neuman E, Karås P. 1988. Effects of pulp mill effluent on a
Baltic coastal fish community. Water Sci. Technol.
20(2):95–106.
Neuman E, Sandström O. 1996. Fish monitoring as a tool
for assessing the health of Baltic coastal ecosystems.
Builetyn Morsk. Inst. Rybackiego 3(139):3–11.
Nicholson M, Fryer R. 1991. The power of the ICES Cooperative Monitoring Programme to detect linear trends and
incidents. In Anon report of the working group on statistical aspects of trend monitoring. ICES Doc CM 1991.
Olsson M. 1983. Miljöprovbanken vid Naturhistoriska
riksmuseet - mål och metoder (The Specimen Bank at
the Museum of Natural History – objectives and methods). Memoranda Soc. Fauna Flora Fennica 59:93–100
(In Swedish).
Olsson M. 1994. Additional sampling recommendations for
biological samples. In Round table discussions outcome and recommendations. Appendix I. Fres. J. Anal.
Chem. 348:177–178.
Olsson M, Bignert A. 1997. Specimen banking – a planning
in advance. Chemosphere 34(9/10):1961–1974.
Olsson M, Bignert A, Aune M, Haarich M, Harms U,
Korhonen M, Poutanen E, Roots O, Sapota G. 2003a.
249
Organic contaminants, 6.3.1, p. 133–140. In Fourth
periodic assessment of the state of the marine environment of the Baltic Sea area, 1994–1998. Baltic Marine
Environment Protection Commission 2002.
Olsson M, Bignert A, DeWitt C, Haglund P. 2003b. Dioxiner I Östersjöns fisk – ett hot mot svenskt fiske. (Dioxins in Baltic fish – a threat to Swedish fisheries. In
Swedish with English summary). In Miljötillståndet i
Egentliga Östersjön. Rapport 2003. Stockholms
Marina Forskningscentrum, Stockholm.
Olsson M, Bignert A, Haglund P, DeWitt C. Miljögifter i
biota. (Contaminants in biota. In Swedish with English
summary). In Bottniska viken 2003 – årsrapport från
den marina miljöövervakningen, p. 18. Umeå Marina
Forskningscentrum, In press.
Policansky D. 1983. Size, age and demography of metamorphosis and sexual maturation in fishes. Am. Zool.
23:57–63.
Roffe DA. 1984. The evolution of life history parameters in
teleosts. Can. J. Fish. Aquat. Sci. 41:1395–1404.
Rowe DK, Thorpe JE, Shanks AM. 1991. Role of fat stores
in the maturation of male Atlantic salmon (Salmo
salar). Can. J. Fish. Aquat. Sci. 48:405–413.
Sandström A, Karås P. 2002. Effects of eutrophication on
young-of-the-year freshwater fish communities in
coastal areas of the Baltic. Env. Biol. Fish. 63:89–101.
Sandström O. 1996. In situ assessments of the impact of
pulp mill effluent on life-history variables in fish, p.
449–457. In Servos MR, Munkittrick KR, Carey JH,
Van Der Kraak GJ (ed.), Environmental fate and effects
of pulp and paper mill effluents. St. Lucie Press, Delray
Beach, Fla.
Sandström O, Förlin L, Grahn O, Landner L, Larsson Å,
Lindesjöö E. 2003. Assessments of the environmental
impact of Swedish pulp and paper mill effluents at the
beginning of the next century. In Environmental
impacts of pulp and paper waste streams. Proceedings
of the 3rd International conference on environmental
fate and effects of bleached pulp mill effluents,
Rotorua, New Zealand. SETAC Press 2003.
Sandström O, Karås P, Neuman E. 1988. Effects of a
bleached pulp mill effluent on growth and gonad function in Baltic coastal fish. Water Sci. Technol.
20:107–118.
Sandström O, Neuman E, Thoresson G. 1995. Effects of
temperature on life history variables in perch, Perca
fluviatilis. J. Fish. Biol. 47:625–670.
Sandström O, Thoresson G. 1988. Mortality in perch populations in a Baltic pulp mill effluent area. Mar. Poll.
Bull. 19(11):564–567.
Sherry JP. 2003. The role of biomarkers in the health
assessment of aquatic ecosystems. Aquat. Ecosyst.
Health Manage. 6(4):423–440.
Stearns SC, Crandall RE. 1984. Plasticity for age and size at
sexual maturity: a life-history response to unavoidable
stress, p. 13–33. In Potts GW, Wootton RJ (ed.), Fish
250
Sandström et al.
reproduction: strategies and tactics. Academic Press,
London.
Stegeman JJ, Brouwer M, Di Guilio RT, Förlin L, Fowler BA,
Sanders BM, Van Veld PA. 1992. Molecular responses to
environmental contamination: enzyme and protein systems as indicators of chemical exposure and effects, p.
235–335. In Hugget RJ, Kimerle RA, Mehrle PM,
Bergman HL (ed.), Biomarkers: biochemical, physiological, and histological markers of anthropogenic stress.
SETAC Special Publications Series, Lewis Publishers.
Stephensen E, Svavarsson J, Sturve J, Ericsson G, AdolfssonErici M, Förlin L. 2000. Biochemical indicators of pollution exposure in shorthorn sculpin (Myoxocephalus
scorpius), caught in four harbors on the south-west
coast of Iceland. Aquat. Toxicol. 48:431–442.
Svedäng H, Neuman E, Wickström H. 1996. Maturation
patterns in female European eel: age and size at the silver eel stage. J. Fish. Biol. 48:342–351.
Thoresson G. 1996. Guidelines for coastal fish monitoring.
Swedish National Board of Fisheries, Institute of
Coastal Research, Kustrapport 1996:2.
Thorpe JE. 1994. Reproductive strategies in Atlantic salmon,
Salmo salar L. Aquacult. Fish. Manage. 25:77–87.
Van Winkle W, Shuter BJ, Holcomb BD, Jager HJ, Tyler
JA, Whitaker SY. 1997. Regulation of energy acquisition and allocation to respiration, growth and reproduction: simulation model and example using rainbow
trout, p. 103–137. In Chambers RC, Trippel EA (ed.),
Early life history and recruitment in fish populations.
Chapman and Hall, London.
Vetemaa M. 1998. Reproduction parameters of fish as indicators in environmental monitoring. Ph.D. thesis, Institute of Zoology and Hydrobiology, University of
Tartu, Estonia. Dissertationes Biologicae Universitatis
Tartuensis 44.
Vetemaa M, Förlin L, Sandström O. 1997. Chemical industry
effluent impacts on reproduction and biochemistry in a
North Sea population of viviparous blenny (Zoarces
viviparus). J. Aquat. Ecosyst. Stress Recovery 6:33–41.
Received: February 9, 2005; accepted: July 27, 2005.