Research Process in Aquatic Biomonitoring

SEGNER H，HAWLICZEK A

(University of Berne，Centre for Fish and Wildlife Health，Institute of Animal Pathology，Berne 3001，Switzerland.)

The decline of wildlife populations，changes in community structure，as well as other adverse ecological effects have long been associated with contaminated water and sediments［1-4］.It is thus of great relevance to characterize the status of ecological systems，to identify contaminated sites and evaluate ecological consequences of site pollution，as well as the cause of their toxicity in order to take concerted remediation actions.A widespread approach to assess the status of the environment is to implement a monitoring strategy.Monitoring represents a descriptive approach aiming to characterize the status of environmental systems and at the same time provides a basis to assess changes of the status.Monitoring of the environment includes observations and measurements of biological，chemical and physical parameters［5］.

The idea of monitoring the environment arose with the industrial revolution and the concomitant pollution of the environment around the beginning of the 20th century，when the impact of anthropogenic activities particularly on aquatic systems was recognised.A good example for the influence of the industrial revolution and subsequent foundation and enlargement of agglomerations on the aquatic environment is the increase of algae blooms in Tolo-Habor，Hong Kong prior to the late 1970's when two new town areas in the north-east of the territory gave rise to massive increases in nutrient loading into the Tolo-Habour，a semi-enclosed embayment［6］.In Europe，at the early beginning of the 20th century local fish kills due to the depletion of oxygen by high organic load stemming from untreated wastewater were observed and subsequently lead to the awareness of the environmental，especially the aquatic environment and the associated possible threat for human safety.

Industrial activities such as pulp mills［7］as well as toxic chemicals such as mutagenic compounds［8］or endocrine disruptors［9］contained in effluents from sewage treatment plants were found to have profound influence on the ecological status of aquatic systems.In addition to site-specific pollution，also problems arising from global contamination by toxic chemicals，in particular by persistent organic pollutants became evident［10-11］.A variety of biotic indices have been developed to assess environmentalpollution［12-13］，many of them relying on benthic macroinvertebrates since they can fairly easily be sampled and determined，and since sediments often contain elevated levels of contaminants［14］.In addition，for many invertebrates the indicator value for organic pollution or specific pollutant groups is known［15］.

To monitor aquatic pollution and its impact on aquatic ecosystems，mainly two approaches are used.(1)Chemical monitoring，this approach relies on chemical-analytical and bio-analytical tools applied to measure the presence，level and nature of the pollutants in the biota，water and sediment. Chemical monitoring has two main limitations. One is the lack of a time-integrated response.If e.g.a chemical is present for only a short time，the analytical sample may miss it，while the biological response might persist over an extended period and thus would indicate the exposure still for some time after the pollution episode.A classical example for such a scenario is the pesticide applications［16-17］.A possibility to overcome this limitation of chemical monitoring is the use of passive samplers［18］.A second limitation of chemical monitoring is that it informs on the presence of a chemical but not-at least not directly-if the presence of the chemical leads to adverse effects.(2) Biological monitoring which utilises biological and ecological parameters to inform on the status of aquatic biota and systems in relation to stressors，be they chemical，physical or biological stressors.Biomonitoring can be performed as either“passive”or“active”biomonitoring:passive biomonitoring refers to the observation of the resident flora and fauna，whereas active biomonitoring refers to the transfer of organisms into the field and subsequent observation of their response to the field condition.This review discusses approaches and methods for the passive and active biomonitoring of chemical pollution in the aquatic environment.A limitation of biological monitoring is that it can be difficult to establish cause-effect relationships for observed impairments of biological and ecological parameters.

In general there are different rationales as why to install biomonitoring programmes on aquatic systems e.g.:

* to record and describe the actual status of the aquatic ecosystem(surveillance monitoring);

* to indicate that a monitored system is contaminated with chemicals and those can be found in the biota(accumulation/exposure monitoring);

* to describe spatial or temporal trends in environmental contamination and effects(trend monitoring);

* to determine the agreement of the resource condition with relevant regulations(compliance monitoring);

* to evaluate associations between natural and/or anthropogenic pressures and the status of the aquatic resources(impact monitoring);

* to identify the causes of an impairment of the aquatic system(investigative monitoring);

* to assess the effectiveness of measures enacted to improve an impaired status of the system (operational monitoring).

1 Design of Biomonitoring Programmes

A clear definition of the objectives and the strategy is essential for the success of the monitoring programme.Misinterpretations of biomonitoring findings usually occur due to problems in selecting appropriate sampling or reference sites，because of natural variation as well as stochastic events，or because of underestimating the influence of multiple stressors on the monitored parameters. In order to design an appropriate monitoring program to survey chemical impact on ecosystems，a few measures should be taken into account.Bio-logical parameters to be measured should be selected on the basis of the objectives and the strategy defined，and thus be relevant to the identified assessment questions.The selected biological parameters to be measured should further be feasible to be implemented practically in field studies and hence be evaluated with respect to their variability e.g.season，migration，gender，age and life stage of organisms，or choice of species(e.g.［19-21］).

Understanding the sources of variability of biological indicators and markers is necessary in order to distinguish extraneous factors from true pollution signals.This variability has a big impact on，and is thus to be taken into account when choosing the right testing frequency，the number of sampling sites and further when choosing an adequate reference site.

Finding a non-polluted or reference site(reference condition respectively)is a difficult task as pristine sites are basically non-existing in anthropologically impacted areas.Thus initially the term reference site or-condition should be defined.A reference site or condition is referred to as the standard against which impacted sites are compared to in order to be able to separate impacts on the average status from the effect of natural variation［22］.Thus a“reference condition”is defined in terms of“naturalness”of the biota(structure，composition，function，diversity)and that naturalness implies the absence of significant human disturbance or alteration.Thus，reference condition refers to a condition of biodiversity，together with the natural physico-chemical boundary conditions as it would be in the absence of human disturbance.As such，reference conditions have to be formulated for different types of water bodies(lakes，small rivers，large rivers etc.).

If it is not possible to obtain reference sites in the region or aquatic system of interest，reference conditions may be derived from similar aquatic systems in similar regions，from minimally or slightly disturbed water bodies，from historical data［23-24］，or from modeling［25］.It is of great importance to understand that reference conditions and hence also reference sites are defined not by single values but by a range of indicator values resulting from natural biological and chemical variation，both in time and space.Once these indicator values have been defined，parameters within these values can be selected as thresholds in order to classify the condition of individual sites and thus assess sites of interest.The selection of these thresholds can be as much a political decision as a scientific one.

Typical strategies on where to sample in respect to reference sites are the BACI(Before and After Control Impact)，the benchmarking and the gradient approach.The BACI approach is a method for measuring the potential impact on an aquatic system by comparing conditionsbefore a planned pollution event or remediation activity and after the event-an approach that is applicable for comparing the affects of anticipated future activities［22］.More often，however，biomonitoring programs have to deal with the retrospective evaluation of chemical impact.In these cases，studies often make use of an unaffected or control site or even a laboratory control and use those data as benchmark to compare affected orimpacted sites［26］.With the so called benchmarking two sites are compared to each other，where one site represents the natural version of the other“actual test”site.Thus in order to find a suitable reference site one has to have an idea of how the ideal natural site would be constituted(see below).A similar approach is the gradient approach［22］.Here，different segments along a river system are compared to each other［27］.The underlying assumption is that a river upstream is less influenced by anthropogenic measures and chemical pollution e.g. runoff.It has to be taken into account though，that upstream the parameters of the river(water temperature and speed，oxygen and nutrient content) are different from the parameters downstream with an effect not only on the chemical behaviour.In cases where the existence of a pollution gradient is not clearly present，multivariate statistics to compare pollution induces and biological indices can be helpful for elucidating the relation between contamination and ecosystem status［28］.

It is not only a fundamental consideration where to sample，but also how to sample.The most frequently used sampling strategy is the random sampling.Random sampling is referred to as taking a number of independent samples from the same probability distribution.This means that the samples should be selected so that all samples of the same size have an equal chance of being selected from the entire population or sampling ground.As this approach is rather unrealistic in execution，a compromise has to be found in selecting the actual sampling size.Sample size is crucial for the success of a monitoring programme and has to consider parameters such as the expected prevalence of the measurement parameter in the monitored population，or the size of the population.An alternative to random sampling is the riskbased sampling strategy.This strategy samples based on pre-knowledge e.g.analytical or toxicity data already obtained and sites are pre-selected. Further epidemiological aspects that need to be considered in designing a monitoring activity include repetition of sampling，be it temporal or spatial，or the genetic constitution of the sampled population.The latter aspect is neglected by most ecotoxicologicalbiomonitoring programmes.Toxic chemicals act as selective factors，what means that microevolution can take place，so that populations of a species at the polluted site may differ in their genotype，and，associated with this，in their phenotypic plasticity or reaction norm from conspecifics of the same species in the population at the reference site.A prominent example is the tomcod population in the Hudson River which-due to historical PCB-contamination has developed resistance to dioxin-like contaminants due to a genotypic change in the aryl hydrocarbon receptor［29-30］. However，even if the effect of genotypic change is more subtle，the reaction norm of individuals from a polluted site may differ from individuals at the reference site what can obscure biomonitoring results on pollution impact［31-32］.

2 Passive vs.Active Monitoring

Passive monitoring refers to the observation of test organisms living naturally within the monitored system whereas active monitoring refers to the active exposition(by actively placing them or exposing them to the aquatic system of interest) of selected test organisms into the monitored systems.Possible techniques to actively expose selected test organisms to the aquatic system of interest include:

*“Caging”and Bypass-Systems

* Transplants/Colonization

*“Enclosures”

* On-line Biotests

The caging technique plants individual species(indigenous or surrogate)or whole communities(e.g.biofilms)within cages into the aquatic system of interest.The species and communities are taken from natural or uncontaminated locations (e.g.from culture).A Bypass-System keeps the species and communities in an aquarium and pumps the aquatic system of interest through the aquarium［33-34］.The reaction of the fish to the assessed aquatic system is then monitored and examined.An advantage of the Bypass-System for an extended monitoring is to be less susceptible of possible destruction or heist.

The transplant/coloniaIization studies test the well being of individuals or communities between different sites.A contamination is indicated when an inhibition of the colonialization or the decline in well being after a transfer of individuals or communities from one site to another is observed. Vice versa if the transplanted organisms recovering and showing growth and well-being it is indicated the former site is contaminated.These approaches have a further advantage of circumventing the problem of migration，e.g.the exposure time is exactly known.

Enclosures refer to the enclosing of e.g.entire trophic communities(except top consumers). With this approach one can analyze the impact of possible contamination on community structures.

On-line biomonitoring is used e.g.to control waste water effluents.Test organisms e.g.bacteria，algae，mussels，Daphnia ssp.or fish are continuously exposed to the aquatic system of interest.If the status or the compilation of the water changes，the organisms respond to it.This response can be e.g.bioluminescence in reporter-bacteria，fluorescence in algae or shell movement in mussels etc. and is detected by a coherent detection system e. g.a luminometer for reporter-bacteria.Appropriate actions can then be implemented.

3 Tools and Test-systems for Biomonitoring of Chemical Pollution Exposure and Effects

In the following section we will introduce biological tools and test-systems for biomonitoring of chemical pollution.Many of these tools are routinely applied in regulatory monitoring programmes，for instance，in the EU Water Framework Directive or Marine Strategy Framework Directive［35］.Importantly，combinations of indices instead of single indices are used［36］，as this improves the ability to detect adverse pollution impact on the environment.It has to be stated though，that biomonitoring of pollution impact should be integrated with chemical monitoring which informs on the chemical status of the monitored system and identifies sites that are likely to be impacted by chemical stressors.

3.1 Ecological Indices

In order to survive，species have certain requirements on the quality of their habitats.If the physical，chemical and nutritional requirements are not met a species will disappear or simply never emerge.If the requirements of a species are defined，the absence or presence of a species indicates the absence or presence of certain physical，chemical and nutritional requirements.Special caution has to be taken into account if a species is absent.This is not necessarily attributable to abiotic or nutritional factors but can be account for by other factors e.g.natural species competition or species turnover etc.However，the presence，absence，or relative abundance of a species(or species assemblage)，either suddenly or gradually，may be used as an indicator of changes in the environment quality or conditions［5］.

Diversity indices investigate the community structure of an aquatic system and primarily depict the biodiversity in an ecosystem e.g.the Shannon index or the Saprobic index.Diversity indices may indicate a polluted aquatic system but are not to be considered reliable.

Biology-based pollution indices are a specific measure of pollution and are based on the reaction of physiologically sensitive species［5］.These species that are most sensitive to pollution are called indicator species.Once the most sensitive species to pollution impact for designated aquatic systems have been identified，an index can be established that classifies the degree of pollution in the aquatic system with respect to the presence or absence of the indicator species，an example is provide by the SPEAR index［37］.The indices may be specific for specific pollutant classes or modes of toxic action，depending on the physiological and ecological factors which drive the species'sensitivity to the pollutants.Thus，if the impact of pollution on an aquatic system is to be assessed，the aquatic system is screened for the absence/presence of the respective indicator species and depending on the outcome in terms of presence or absence of species，or the degree of their presence/absence，conclusions on the nature and intensity of the pollution can be drawn.The more prominent of these diversity-based pollution indices are introduced below.

3.1.1 Shannon Index

The Shannon index(H')has probably been the most widely used index in community ecology.The index is based on information theory.As defined by Meerman 2004:the Shannon index is a measure of the average degree of“uncertainty”in predicting to what species an individual chosen at random from a collection of S species and N individuals will belong.This average uncertainty increases as the number of species increases and as the distribution of individuals among the species becomes even.Thus，H'has two properties that have made it a popular measure of species diversity:(1)“H'=0 if and only if there is one species in the sample，and(2)H'is maximum only when all S species are represented by the same number of individuals，that is，a perfectly even distribution of abundances.When all species in a sample are equally abundant，it seems intuitive that an evenness index should be maximum and decrease toward zero as the relative abundances of the species diverge away from evenness［38］.

The Shannon index is a descriptive index with which the diversity and thus the ecological condition of the aquatic system of interest is assessed.It cannot be differentiated whether the obtained diversity classification is due to pollution and if so，what type of pollution it is accountable to.

3.1.2 Saprobic Index

The Saprobic index is a widely accepted and used system to assess the level of contamination of aquatic systems.The main focus of the Saprobic index is the reduction of oxygen by organic substances.Several indicator species(micro-and macrosaprobia)are jointed in order to assess aquatic systems.Among the species are bacteria，fungi，volvocales，ciliates，insects and further more.The Saprobic index illustrates again the aforementioned important aspect of natural variation.For a reliable assessment of water quality usually not a single indicator species，but biotic indices derived from the investigation of species groups or communities are used.A biotic index takes account of the sensitivity or tolerance of individual species to pollution and assigns them a value，the sum of which gives an index of pollution for a site.The data may be qualitative(presence/absence)or quantitative(relative abundance).The Saprobic index system recognized four stages in the oxidation of organic matter-poylsaprobic，α-mesosaprobic，β-mesosaprobic and oligosaprobic.The presence or absence of indicator species for these four Saprobic stages is recorded.

3.1.3 SPEAR(SPEcies At Risk)Index

The SPEAR index is based on biological features or traits of stream invertebrates［15，39-40］.It is focused on various types of contaminants in fresh waters but was initially developed to test for pesticide contamination in aquatic systems.The e.g. invertebrate species are classified according to their sensitivity to toxicants into sensitive(at risk) and non-sensitive(not at risk)and thus，gives a measure for the loss of species sensitive to toxicants.The SPEAR index not only uses traits that are responsive to the effects of particular toxicants (e.g.physiological sensitivity)but further integrates the associated recovery(e.g.generation time).Thus the system was developed to identify species at risk being affected by pesticides，with reference to life-history and physiological traits［41］.SPEAR bioindicators are developed to complement existing bioassessment methods and indices in order to assess effects of toxicants.A clear advantage is that for large-scale river systems the SPEAR index proved to be relatively stable over altitudinal and longitudinal gradients，thus the structure of invertebrate communities described with SPEAR is independent of the biogeographical region［42-43］.Currently，two SPEAR-indicators exist:SPEARpesticidesand SPEARorganicdesigned to detect and quantify effects of pesticides(insecticide toxicity)and general organic toxicants(e.g.petrochemicals，synthetic surfactants)respectively.

3.2 Biological Indices

Biological indices comprise organismic，physiological and histological parameters.Hence，while the ecological indices target the community，the biological parameters assess the stress/health status of the individual organism.In the following section，frequently used biological indices are shortly introduced.

3.2.1 Histopathological Assessment

Histopathological assessment refers to the examination of histopathological changes in organs from a monitored organism.The organs of interest are usually dissected，chemically fixed，cut into thin sections and stained(e.g.haematoxylin-eosin (HE))for general histological evaluation or for more specific stainings，e.g.，immunohistochemical staining of specific target proteins.To overcome the somewhat subjective nature ofhistopathologicalassessments，standardized，semiquantitative assessment schemes as developed e. g.by Bernet et al.(1999)，may be used.The scheme of Bernet et al.(1999)classifies lesions in organs into five groups

* circulatory，

* regressive，

* progressive，

* inflammatory，

* neoplastic.

With each group having an importance factor between 1 to 3 characterizing the pathological relevance.The degree of the histopathological alteration is scored using a scale e.g.0 to 6.Finally for every organ examined an organ index is calculated by multiplying the importance factor and the degree value.The higher the obtained value the more severe is the histopathological change in the examined organ［44-45］.

Another important histopathological assessment approach is the monitoring of neoplastic changes，for instance，liver tumors in flatfish，as it has been intensively used for pollution monitoring in theNorth Sea and in thePugetSound，USA［46-47］classified hepatic lesions in dab and flounder into four main categories，with each class being further refined in the technical paper.

* Early non-neoplastic toxicopathic lesions

* Foci of cellular alteration(FCA)

* Benign neoplasms

* Malignant neoplasms

The recorded lesions then indicate a range from environmental stress to contaminant-related carcinogenesis.An excellent example of the value of neoplastic changes in pollution monitoring is the correlation between frequency of liver neoplastic changes in flatfish of the Piuget Sound and PAH contamination status of the sediments［48］.

3.2.2 Condition factor and Organ/Somatic Indices

These indices are used to compare the condition of test animals(e.g.fish)or the condition of specific organs of test animals either with test animals from a reference site or within a group［49］. The condition factor is calculated by measuring the weight of the organism as well as the total length of the，the following formula is applied for the example of fish:(K)=(fish wt(g)/total length(cm))×100.The organ/somatic indices apply the same formula and substitute the length by the organ weight e.g.the liver somatic index: (K)=(fish wt(g)/liver wt(g))×100.Differences in condition factor and organ/somatic index are always subject to statistical validation.

3.3 Biomarker Indices

The various ecological and biological indices and parameters introduced above provide good information on the community，population or organism status.Their weakness however，is the difficult causal relation to chemical pollution，as these parameters，due to their interactive nature，are under the influence of many more factors than toxicants. Thus，for the purpose of pollution biomonitoring，it needs indicators which are more directly related to chemical impact.Here，the so-called biomarkers have attracted particular attention.Biomarkers are usually defined as sub-organismic(molecular，biochemical，cellular，physiological)responses that can be related to exposure to or toxic effects of chemicals［50］.The big advantage of biomarker is their use as a specific diagnostic tool.While molecular，cellular and physiological responses are directly involved in the toxic mechanisms induced by the chemical pollution or are at least closely associated with the initial chemico-biological interactions，the causal relationship between the biological change and the toxicant action is getting increasingly confounded at higher levels of organization，due to the action of compensatory processes and of new，level-specific properties［51］.Thus，biomarker help to target detailed chemical and biological analysis of water，sediments and biota.It has to be stated though，that in turn the advantage of biomarker as diagnostic tools on the molecular level has the disadvantage of a limited ecological relevance.Hence，biomarkers should always be integrated into a greater biomonitoring study and not be used as a stand-alone tool.A well-known example of aquatic pollution where biomarker played a major role in detecting and defining the case is endocrine disruption［52］:the fact that in many aquatic environments organisms suffer from exposure to or effects of endocrine active compounds was brought to awareness mainly through biomarker-based monitoring studies，for instance，the surveys on fish populations in English rivers and estuaries based on measuring the egg-yolk protein，vitellogenin，as a biomarker of exposure to estrogen receptor-binding compounds，and analysing the presence of intersex gonads，i.e.gonads containing both male and female germ cells，as a biomarkerof the impactof endocrine disruptors［53-54］.The use of biomarker in aquatic biomonitoring has been reviewed，in more detail，by van der Oost et al(2003)［55-57］.

3.4 Laboratory based in Vitro and in Vivo Assays on Environmental Samples

Sometimes，field sampling of resident biota can be difficult due to sparse animal abundance or because the target species is protected.In this case it is a frequently used strategy to collect environmental samples and applythese to laboratory based in vitro or in vivo assays.This strategy is further regularly applied to test toxic activities of sediments or of tissue extracts.An advantage of this approach is that it provides direct evidence on the potential biological activities of the contaminants contained in the samples.However，it is a clear disadvantage that the results may be difficult to transfer to the field situation，as conditions of pH，oxygen contents，conductivity or nutrients usually differ between the lab assay and the field sites.Apart from standard exposure assays of environmental samples e.g.direct water phase testing by exposing laboratory organisms，in vivo or in vitro laboratory tests are frequently applied.

Possible toxicological endpoints that can be screened for using in vitro test systems include for example

* genotoxicity(e.g.umuC test［58-59］)，

* mutagenicity(e.g.AmesⅡ［60］)，

* estrogenic effects(ER-Calux，yeast estrogen screen(YES)［61-62］)，

* androgen receptor agonism and antagonism(AR-CALUX［63-64］，yeast androgen screen (YAS)［65-66］，

* aryl hydrocarbon receptor(AhR)receptor mediatedeffects(DR-CALUX，EROD induction［67-68］)，

* tumor promotion by inhibition of gapjunctional intercellular communication［69-70］，

* binding to transport protein transthyretin (TTR-binding assay［71-73］)and antibiotic activity［74］.

Possible in vivo laboratory tests include e.g. the sediment contact assay，available in particular for benthic invertebrates as they are most affected by a sediment contamination.Sediment contact assays have partly environmental realism but limited diagnostic power since the response of the test organisms not only rely on organic toxicants but further on factors like grain size，organic matter content，oxygen concentrations，ammonia concentration，hydrogen sulfide concentration，pH and conductivity.Thus it is hard to discriminate toxic effects from confounding factors.Another possibility to test for toxicity of sediment is the extraction and subsequent testing of pore-water，since pore water is believed to represent a major exposurepathway for sediment inhabiting organisms.However，since most pore-waters are anoxic and invertebrates need oxygen to survive the assay has to be carried out under oxic conditions.This leads to the oxidization of constituents such as hydrogen sulfide，the changing of metal bioavailability［75］.

The use of extracts is another frequently used alternative to test for the potential toxicity being present in either environmental compartments like sediment or being accumulated in biota or passive samplers.Extraction methods like(i)Soxhlet，(ii) pressurized liquid extraction as well as(iii)ultrasonic or microwave extraction with strong organic solvents further also eliminates the above stated factors which influence the toxicity in situ，including inorganic contaminants.The disadvantage in this approach is that it neglects the role of bioavailability in toxic effects;related to the extraction method，these methods represent a worst case scenario with a 100% bioavailability，and thus should be interpreted with caution.It needs a combination with other approaches(e.g.bioavailability tests)in order to estimate how much of the potentially observed toxicity can be explained by monitored compounds［76］.

It should be stressed that a direct prediction of adverse effects in the environment from in vitro results is not possible［77］.It may be hypothesized that in future new“omics”techniques including genomics［78-79］，proteomics［80］and metabolomics［81］may provide powerful diagnostic tools.

3.5 Effect Directed Analysis

While the aforementioned parameters or tests may indicate a potential ecological impact of chemical pollution，they provide an overall effect and do not indicate which specific chemicals are the cause of observed adverse effects［82］.In order to identify those chemicals being responsible for a biological response observed in environmental biota or in laboratory tests on environmental samples or extracts，it is essential to identify the active constituent that cause of the toxicity［82］.Many polluted sites，are contaminated by a complex mixture of different chemicals as well as their derivates and metabolites and it is difficult if not impossible to judge which of these many chemicals is responsible for the observed effectv［83］.To address this issue，effect-directed analysis(EDA)was developed.EDA combines chemical analytical methods with fractionation procedures and biotesting，leading to the linkage of effect data to causative hazardous compounds［84-85］.Thus，complex mixtures are reduced in complexity by extraction and fractionation proceduresbased on physiochemical properties.Fractions are tested using in vitro or in vivo bioassays and active fractions are subject for further examination.Eventually，after reduction of the complex mixture to a few individual fractions，active fractions undergo chemical identification and quantification，mostly by using liquid and gas chromatograph mass spectrum(GC-MS)chemical screening.A final step is to confirm the identified toxicants as the cause of toxicity in order to establish a reliable cause-effect relationship［84］.EDA hence focuses on the identification and assessment of toxicants in complex environmental mixtures which are responsible for observed adverse effects.The major goal of EDA is to unravel cause-effect relationships for a reliable risk assessment of environmental contamination.

4 Outlook Biomonitoring in the 21st Century

Biomonitoring approaches will succeed in describing toxic chemical effects on ecosystems and in establishing cause-effect relationships between chemical contamination and ecological impairment when the toxic impact represents a dominant environmental factor.However，in situations with low dose exposures and subtle effects，biomonitoring programmes will encounter problems in assessing the possible impact of chemical contamination，because the confounding influence of other factors is getting more important.For instance，for aquatic systems with a high variation in toxicant load (e.g.river systems in the area of sewage treatment plants)it can be difficult to link effects observed in biota to the exposure concentration in their surrounding environment［86］.As chemical testing usually refers to grab sampling at single time points and subsequent analysis in the laboratory no time-integrated response of the highly variable aquatic system is taken into account as is the case for biota.For example a fish living in a toxic environment may be exposed to a certain concentration or a certain range of concentrations of a toxicant over time and thus bioaccumulate，whereas sampling for chemical testing on the other hand usually shows a snapshot of the concentration at the moment of sampling.Thus with grab sampling a time-integrative change in concentration is not detected und might lead to misinterpretation and non-correlating data when looking at effects in biota in comparison to concentration found in the surrounding environment.Further trace samples of the target analyte might be under the detection limit and thus not be detected at all［87］.Progress has been made in this field in recent years by the development of passive sampler.The so called POCIS(polar organic integrative sampler)for polar toxicants or the SPMD(semi permeable membrane device)for non-polar toxicants are examples of passive sampler.Passive samplers are used as bioaccumulators of environmental toxicants over time.These passive sampler may e.g.be placed in the aquatic system of interest or even possibly co-exposed with caged fish and be exposed to the same concentrations of toxicants as the fish.Thus if later extracted and analysed passive sampler allow a specific analysis of the kind and the amount or concentration of toxicant the fish was exposed to over the time caged in the aquatic system.

It has to be taken into account that the uptake routes of biota in comparison the passive sampler are different and thus may lead to different results，most for substances that are metabolised and/or detoxified and excreted in the biota.In general using a passive sampler is a possibility to decrease the gap between grab sampling with subsequent chemical analysis and effects observed in biota［88］.

Another challenge to biomonitoring the impact of chemical exposure and effects are multiple stressors.Biota and ecosystems are usually not only exposed to one but to several stressors，and they respond to all of them，in particular also to their interactions and the cumulative effects.Major environmentalstressors beyond toxic chemicals which may act on aquatic resources include habitat degradation and fragmentation，changes of hydromorphology，or the invasion of alien species and pathogens.The ecological status of a water body reflects the integrated impact of all stressors being active in the system，and often it is not the influence of a single but the combined impact of chemical，physical and biological stressors that causes ecosystem impairment［89］.Thus，for assessing risks to and causes of ecological impairment，we must move beyond the focus on toxicants towards multiple stressors-what，however，is a complicated issue.Stressor interactions can take place at the exposure side，for instance，toxification of PAHs by UV sunlight，but they can also take place at the effect side，e.g.，the sensitization of organisms towards pathogens by pre-exposure to toxicants.The possible combinations of chemical，physical and biological stressors are endless，and it is thus not possible to study them on a case by case basis. Models that enable qualitative or quantitative predictions on stressor interactions would be needed，unfortunately，however，this type of models is not currently available.Assessing the combination effects is complicated by the different metrics defining the chemical，physical and biological stressors.While dose-response relationships of toxicants are typically sigmoid，they are more variable in the case of chemical and biological stressors.A second problem is that diverse stressors may target different biological processes and receptors，and they may be active at different levels of biological organization，so that the resulting ecological change may arise from indirect rather than direct interactions. These problems complicate not only predictive assessment of stressor combinations，but also the retrospective，“forensic”approach as it is practised in biomonitoring to identify the causes of ecological impairment.It is this multiple stressor field where there exists an urgent need for developing novel approaches and concepts in biomonitoring.

［1］ Anderson J，Birge W，Gentile J，et al.Biological Effects.Bioaccumulation and Ecotoxicology of Sediment Associated Chemicals［C］//Dickson K L，Maki A W，Brungs W A.Fate and Effects of Sediment-bound Chemicals in Aquatic Systems. New York:Pergamon Press，1987:267-296.

［2］ Bailey H C，DiGiorgio C，Kroll K，et al.Development of Procedures for Identifying Pesticide Toxicity in AmbientWaters:Carbofuran，Diazinon，Chlorpyrifos［J］.Environmental Toxicology and Chemistry，1996，15(6):837-845.

［3］ Hartwell S I，Dawson C E，Durell E Q，et al.Correlation of Measures of Ambient Toxicity and Fish Community Diversity in Chesapeake Bay，USA，Tributaries——Urbanizing Watersheds［J］.Environmental Toxicology and Chemistry，1997，16 (12):2556-2567.

［4］ Keiter S，Rastall A，Kosmehl T，et al.Ecotoxicological Assessment of Sediment，Suspended Matter and Water Samples in the Upper Danube River.A Pilot Study in Search for the Causes for the Decline of Fish Catches［J］.Environmental Science and Pollution Research，2006，13(5):308-319.

［5］ Rand G M，Wells P G，McCarty L S.Introduction to Aquatic Toxicology［C］//Rand G M.Fundamentals of Aquatic Toxicology.Florida:Taylor and Francis，North Palm Beach，1995:3-67.

［6］ Morton B.Pollution of the Coastal Waters of Hong Kong［J］.Marine Pollution Bulletin，1989，20(7): 310-318.

［7］ Munkittrick K R，Servos M R，Van Der Kraak G J，et al.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［J］.Environmental Toxicology and Chemistry，1994，13(7):1089-1101.

［8］ White P A，Rasmussen J B，Blaise C.Genotoxic Substances in the St.Lawrence SystemⅠ:Industrial Genotoxins Sorbed to Particulate Matter in the St.Lawrence，St.Maurice，and Saguenay Rivers，Canada［J］.Environmental Toxicology and Chemistry，1998，17(2):286-303.

［9］ Purdom C E，Hardiman P A，Bye V V J，et al.Estrogenic Effects of Effluents from Sewage Treatment Works［J］.Chemistry and Ecology，1994，8 (4):275-285.

［10］ Ueno D，Inoue S，Ikeda K，et al.Specific Accumulation of Polychlorinated Biphenyls and Organochlorine Pesticides in Japanese Common Squid as a Bioindicator［J］.Environmental Pollution，2003，125(2):227-235.

［11］ Scheringer M.Long-range Transport of Organic Chemicals in the Environment［J］.Environmental Toxicology and Chemistry，2009，28(4):677-690.

［12］ Hellawell J M.Biological Indicators of Freshwater Pollution and Environmental Management［J］.Environmental Pollution，1987，45:239-240.

［13］ Mason C F.Biology of Freshwater Pollution［M］. Essex:Prentice Hall，2002.

［14］ Cairns J J，Pratt J R.A History of Biological Monitoring Using Benthic Macroinvertebrate［M］// Rosenberg D M，Resh V H.Freshwater Biomonitoring and Benthic Macroinvertebrates，New York: Chapman＆Hall，1993:10-27.

［15］ Liess M，Von Der Ohe P C.Analyzing Effects of Pesticides on Invertebrate Communities in Streams［J］.Environ Toxicol Chem，2005，24(4):954-965.

［16］ Nauman C H，Santolucito J A，Dary Curtis C C.Biomonitoring for Pesticide Exposure［M］.USA:A-merican Chemical Society，1993:1-19.

［17］ Beketov M A，Liess M.Acute and Delayed Effects of the Neonicotinoid Insecticide Thiacloprid on Seven Freshwater Arthropods［J］.Environmental Toxicology and Chemistry，2008，27(2):461-470.

［18］ El-Shenawy N S，Greenwood R，Abdel-Nabi I M，et al.Comparing the Passive Sampler and Biomonitoring of Organic Pollutants in Water:A Laboratory Study［J］.Ocean Science Journal，2009，44(2): 69-77.

［19］ Metcalfe-Smith J L.Biological Water-quality Assessment of Rivers:Use of Macroinvertebrate Communities［M］.Oxford:Blackwell Scientific Publications，1994.

［20］ Carlisle D M，Clements W H.Sensitivity and Variability of Metrics Used in Biological Assessments of Running Waters［J］.Environmental Toxicology and Chemistry，1999，18(2):285-291.

［21］ Behrens A，Segner H.Cytochrome P4501A Induction in Brown Trout Exposed to Small Streams of an Urbanised Area:Results of a Five-year-study［J］.Environmental Pollution，2005，136(2):231 -242.

［22］ Downes B，Barmuta L，Fairweather P，et al.Monitoring Ecological Impacts:Concepts and Practices in Flowing Waters［M］.Cambridge:Cambridge U-niversity Press，2002.

［23］ Butcher J B，Gauthier T D，Garvey E A.Use of Historical PCB AroclorMeasurements:Hudson River Fish Data［J］.Environmental Toxicology and Chemistry，1997，16(8):1618-1623.

［24］ Hanson N，Förlin L，Larsson A.Spatial and Annual Variation to Define the Normal Range of Biological Endpoints:An Example with Bbiomarkers in Perch［J］.Environmental Toxicology and Chemistry，2010，29(11):2616-2624.

［25］ Reece P F，Richardson J S.Biomonitoring with the Reference Condition Approach for the Detection of Aquatic Ecosystems at Risk［EB/OL］.(1999-03 -15)［2012-04-28］.http://faculty.forestry. ubc. ca/richardson/abstructs RE% 2011% 20Reece.pdf.

［26］ Dubé M G.Cumulative Effect Assessment in Canada:A Regional Framework for Aquatic Ecosystems［J］.Environmental Impact Assessment Review，2003，23(6):723-745.

［27］ Burki R，Vermeirssen E L M，Körner O，et al.Assessment of Estrogenic Exposure in Brown Trout (Salmo trutta)in a Swiss Midland River:Integrated Analysis of Passive Samplers，Wild and Caged Fish，and Vitellogenin mRNA and Protein［J］.Environmental Toxicology and Chemistry，2006，25 (8):2077-2086.

［28］ Smolders R，De Coen W，Blust R.An Ecologically Relevant Exposure Assessment for a Polluted River Using an Integrated Multivariate PLS Approach［J］.Environmental Pollution，2004，132(2):245 -263.

［29］ Roy N K，Courtenay S C，Chambers R C，et al. Characterization of the Aryl Hydrocarbon Receptor repressor and a Comparison of Its Expression in Atlantic Tomcod from Resistant and Sensitive Populations［J］.Environ Toxicol Chem，2006，25(2): 560-571.

［30］ Wirgin I，Roy N K，Loftus M，et al.Mechanistic Basis of Resistance to PCBs in Atlantic Tomcod from the Hudson River［J］.Science Magazine，2011，331(6022):1322-1325.

［31］ Myrand B，Tremblay R，Sévigny J M.Selection A-gainst Blue Mussels(Mytilus edulis L.)Homozygotes Under Various Stressful Conditions［J］.Journal of Heredity，2002，93(4):238-248.

［32］ Coutellec M A，Collinet M，Caquet T.Parental Exposure to Pesticides and Progeny Reaction Norm to a Biotic Stress Gradient in the Freshwater Snail Lymnaea Stagnalis［J］.Ecotoxicology，2011，20 (3):524-534.

［33］ Triebskorn R，Böhmer J，Braunbeck T，et al.The Project VALIMAR(VALIdation of bioMARkers for the Assessment of Small Stream Pollution): Objectives，Experimental Design，Summary of Results，and Recommendations for the Application of Biomarkers in Risk Assessment［J］.Journal of A-quatic Ecosystem Stress and Recovery，2001，8(3/ 4):161-178.

［34］ Triebskorn R，Adam S，Behrens A，et al.Establishing Causality between Pollution and Effects at Different Levels of Biological Organization:The VALIMAR Project［J］.Human and Ecological Risk Assessment:An International Journal，2003，9 (1):171-194.

［35］ Lyons B P，Thain J E，Stentiford G D，et al.Using Biological Effects Tools to Define Good Environmental Status under the European Union Marine Strategy Framework Directive［J］.Mar Pollut Bull，2010，60(10):1647-1651.

［36］ Broeg K，Westernhagen H V，Zander S，et al.The“Bioeffect Assessment Index”(BAI).A Concept for the Quantification of Effects of Marine Pollution by an Integrated Biomarker Approach［J］.Mar Pollut Bull，2005，50(5):495-503.

［37］ Liess M，Schäfer R B，Schriever C A.The Footprint of Pesticide Stress in Communities——Species Traits Reveal Community Effects of Toxicants［J］.Science of the Total Environment，2008，406 (3):484-490.

［38］ Meerman J.Rapid Ecological Assessment In The Columbia River Forest Reserve Past Hurricane Iris. in T.I.f.D［J］.Environment，2004，3(1):1-5.

［39］ von der Ohe P C，Prüss A，Schäfer R B，et al.Water Quality Indices Across Europe——A Comparison of the Good Ecological Status of Vive River Basins［J］.J Environ Monit，2007，9(9):970-978.

［40］ von der Ohe P C，de Deckere E，Prüss A，et al.Towards an Integrated Risk Assessment of the Ecological and Chemical Status of European River Basins［J］.Integr Environ Assess Manag，2009，5 (1):50-61.

［41］ Liess M，Schäfer R B，Schriever C A.The Footprint of Pesticide Stress in Communities——Species Traits Reveal Community Effects of Toxicants［J］.Science of the Total Environment，2008，406 (3):484-490.

［42］ Beketov M A，Liess M.An Indicator for Effects of Organic Toxicants on Lotic Invertebrate Communities:Independence of Confounding Environmental Factors over an Extensive River Continuum［J］. Environ Pollut，2008，156(3):980-987.

［43］ von der Ohe P C，De Deckere E，Prüss A，et al.Toward an Integrated Assessment of the Ecological and Chemical Status of European River Basins［J］.Integr Environ Assess Manag，2009，5(1):50 -61.

［44］ Bernet D，Schmidt-Posthaus H，Wahli T，et al.E-valuation of Two Monitoring Approaches to Assess Effects of Waste Water Disposal on Histological Alterations in Fish［J］.Hydrobiologia，2004，524 (1):53-66.

［45］ Zimmerli S，Bernet D，Burkhardt-Holm P，et al.Assessment of Fish Health Status in Four Swiss Rivers Showing a Decline of Brown Trout Catches［J］.Aquatic Sciences-Research Across Boundaries，2007，69(1):11-25.

［46］ Feist S W，Lang T，Stentiford G D，et al.Biological Effects of Contaminants:Use of Liver Pathology of the European Flatfish Dab(Limanda Limanda L.) and Flounder(Platichthys flesus L.)for Monitoring［J］.ICES Tech Marine Environ Sci，2004 (38):1-42.

［47］ Stentiford G D，Viant M R，Ward D G，et al.Liver Tumors in Wild Flatfish:A Histopathological，Proteomic，and Metabolomic Study［J］.OMICS:A Journal of Integrative Biology，2005，9(3):281-299.

［48］ Horness B H，Lomax D P，Johnson L L，et al.Sediment Quality Thresholds:Estimates from Hockey Stick Regression of Liver Lesion Prevalence in English Sole(Pleuronectes Vetulus)［J］.Environmental Toxicology and Chemistry，1998，17(5): 872-882.

［49］ Zimmerli S，Bernet D，Burkhardt-Holm P，et al.Assessment of Fish Health Status in Four Swiss Rivers Showing a Decline of Brown Trout Catches［J］.Aquatic Sciences-Research Across Boundaries，2007，69(1):11-25.

［50］ Peakall D B.The Role of Biomarkers in Environmental Assessment(1).Introduction［J］.Ecotoxicology，1994，3(3):157-160.

［51］ Segner H.Ecotoxicology——How to Assess the Impact of Toxicants in a Multi-factorial Environment?［J］.NATO Security Through Science Series，2007:39-56.

［52］ Sumpter J P，Johnson A C.10th Anniversary Perspective:Reflections on Endocrine Disruption in the Aquatic Environment:from Known Knowns to Unknown Unknowns(and Many Things in Between)［J］.Journal of Environmental Monitoring，2008，10(12):1476-1485.

［53］ Allen Y，Matthiessen P，Scott A P，et al.The Extent of Oestrogenic Contamination in the UK Estuarine and Marine Environments——Further Surveys of Flounder［J］.Science of The Total Environment，1999，233(1/3):5-20.

［54］ Jobling S，Williams R，Johnson A，et al.Predicted Exposures to Steroid Estrogens in U.K.Rivers Correlate with Widespread Sexual Disruption in Wild Fish Populations［J］.Environ Health Per-spect，2006，114(S-1):32-39.

［55］ van der Oost R，Beyer J，Vermeulen N P.Fish Bioaccumulation and Biomarkers in Environmental Risk Assessment:A Review［J］.Environ Toxicol Pharmacol，2003，13(2):57-149.

［56］ Hagger J A，Jones M B，Leonard D R，et al.Biomarkers and Integrated Environmental Risk Assessment:Are There More Questions Than Answers?［J］.Integr Environ Assess Manag，2006，2(4): 312-329.

［57］ Hanson N.Utility of Biomarkers in Fish for Environmental Monitoring［J］.Integr Environ Assess Manag，2009，5(1):180-181.

［58］ Reifferscheid G，Heil J，Oda Y，et al.A Microplate Version of the SOS/umu-test for Rapid Detection of Genotoxins and Genotoxic Potentials of Environmental Samples［J］.Mutation Research/Environmental Mutagenesis and Related Subjects，1991，253(3):215-222.

［59］ Reifferscheid G，Heil J.Validation of the SOS/umu Test Using Test Results of 486 Chemicals and Comparison with the Ames Test and Carcinogenicity Data［J］.Mutation Research/Genetic Toxicology，1996，369(3/4):129-145.

［60］ Flückiger-Isler S，Baumeister M，Braun K，et al. Assessment of the Performance of the AmesⅡAssay:A Collaborative Study with 19 Coded Compounds［J］.Mutat Res，2004，558(1/2):181-197.

［61］ Hilscherova K，Machala M，Kannan K，et al.Cell BioassaysforDetection ofArylhydrocarbon (AhR)and Estrogen Receptor(ER)Mediated Activity in Environmental Samples［J］.Environ Sci Pollut Res Int，2000，7(3):159-171.

［62］ Houtman C，Swart C P，Lamoree M H，et al.DR and ER CALUX Assays as Tools to Direct Toxicity Identification and Evaluation of Endocrine Disrupting Chemicals［J］.Organohalogen Compounds，2002，58:349-352.

［63］ Sonneveld E，Jansen H J，Riteco J A C，et al.Development of Androgen-and Estrogen-responsive Bioassays，Members of a Panel of Human Cell Line-based Highly Selective Steroid-responsive Bioassays［J］.Toxicol Sci，2005，83(1):136-148.

［64］ Van der Linden S C，Heringa M B，Man H Y，et al.Detection of Multiple Hormonal Activities in Wastewater Effluents and Surface Water，Using a Panel of Steroid Receptor CALUX Bioassays［J］. Environ Sci Technol，2008，42(15):5814-5820.

［65］ Sohoni P，Sumpter J P.Several Environmental Oestrogens are Also Anti-androgens［J］.Journal of Endocrinology，1998，158(3):327-339.

［66］ Thomas K V，Hurst M R，Matthiessen P，et al.An Assessment of in Vitro Androgenic Activity and the Identification of Environmental Androgens in United Kingdom Estuaries［J］.Environ Toxicol Chem，2002，21(7):1456-1461.

［67］ Behnisch P A，Hosoe K，SakaiS.Bioanalytical Screening Methods for Dioxins and Dioxin-like Compoundsa Review ofBioassay/Biomarker Technology［J］.Environ Int，2001，27(5):413-439.

［68］ Legler J，Leonards P，Spenkelink A，et al.In Vitro Biomonitoring in Polar Extracts of Solid Phase Matrices Reveals the Presence of Unknown Compounds with Estrogenic Activity［J］.Ecotoxicology，2003，12(1/4):239-249.

［69］ Bláha L，Kapplová P，Vondrácek J，et al.Inhibition of Gap-junctional Intercellular Communication by Environmentally Occurring PolycyclicAromatic Hydrocarbons［J］.Toxicol Sci，2002，65(11):43 -51.

［70］ Machala M，Bláha L，Vondrácek J，et al.Inhibition of Gap Junctional Intercellular Communication by Noncoplanar Polychlorinated Biphenyls:Inhibitory Potencies and Screening for Potential Mode(s)of Action［J］.Toxicol Sci，2003，76(1):102-111.

［71］ Lans M C，Klasson-Wehler E，Willemsen M，et al. Structure-dependent，Competitive Interaction of Hydroxy-polychlorobiphenyls，-dibenzo-p-dioxins and-dibenzofurans with Human Transthyretin［J］. Chem Biol Interact，1993，88(1):7-21.

［72］ Hamers T，Kamstra J H，Sonneveld E，et al.In Vitro Profiling of the Endocrine-disrupting Potency of Brominated Flame Retardants［J］.Toxicol Sci，2006，92(1):157-173.

［73］ Weiss J M，Andersson P L，Lamoree M H，et al. Competitive Binding of Poly-and Perfluorinated Compounds to the Thyroid Hormone Transport Protein Transthyretin［J］.Toxicol Sci，2009，109 (2):206-216.

［74］ Smith A J，Balaam J L，Ward A.The Development of a Rapid Screening Technique to Measure Antibiotic Activity in Effluents and Surface Water Samples［J］.Mar Pollut Bull，2007，54(12):1940-1946.

［75］ Wang F.Pore Water Toxicity Testing:Does it Make Sense?［J］.SETAC-Europe News，1999， 10:6-7.

［76］ Brack W，Burgess R M.Considerations for Incorporating Bioavailability in Effect-directed Analysis and Toxicity Identification Evaluation［J］.Effectdirected Analysis of Complex Environmental Contamination:The Handbook of Environmental Chemistry，2011，15:41-68.

［77］ Chapman P M.Determining When Contamination is Pollution-weight of Evidence Determinations for Sediments and Effluents［J］.Environment International，2007，33(4):492-501.

［78］ Hamadeh H K，Bushel P R，Jayadev S，et al.Gene Expression Analysis Reveals Chemical-specific Profiles［J］.Toxicological Sciences，2002，67(2): 219-231.

［79］ Roh J Y，Sim S J，Yi J，et al.Ecotoxicity of Silver Nanoparticles on the Soil Nematode Caenorhabditis Elegans Using Functional Ecotoxicogenomics［J］. Environ Sci Technol，2009，43(10):3933-3940.

［80］ Nesatyy V J，Suter M J.Analysis of Environmental Stress Response on the Proteome Level［J］.Mass Spectrometry Reviews，2008，27(6):556-574.

［81］ Kluender C，Sans-Piché F，Riedl J，et al.A Metabolomics Approach to Assessing Phytotoxic Effects on the Green Alga Scenedesmus Vacuolatus［J］. Metabolomics，2009，5(1):59-71.

［82］ USEPA.EPA/600/R-07/080 Sediment Toxicity I-dentification Evaluation(TIE)PhasesⅠ，Ⅱ，andⅢGuidance Document［S］.Washington DC:Office of Research and Development，2007.

［83］ Ankley G T，Mount D R.Retrospective Analysis of the Ecological Risk of Contaminant Mixtures in A-quatic Sediments［J］.Human Ecological Risk Assessment:An International Journal，1996，2(3): 434-440.

［84］ Brack W.Effect-directed Analysis:A Promising Tool for the Identification of Organic Toxicants in Complex Mixtures?［J］.Anal Bioanal Chem，2003，377(3):397-407.

［85］ Schymanski E L，Meinert C，Meringer M，et al.The Use of MS Classifiers and Structure Generation to Assist in the Identification of Unknowns in Effectdirected Analysis［J］.Anal Chim Acta，2008，615 (2):136-147.

［86］ Vermeirssen E L，Körner O，Schönenberger R，et al.Characterization of Environmental Estrogens in River Water Using a Three Pronged Approach:Active and Passive Water Sampling and the Analysis of Accumulated Estrogens in the bile of Caged Fish［J］.Environ Sci Technol，2005，39(21):8191-8198.

［87］ Söderström H，Lindberg R H，Fick J.Strategies for Monitoring the Emerging Polar Organic Contaminants in Water with Emphasis on Integrative Passive Sampling［J］.J Chromatogr A，2009，1216 (3):623-630.

［88］ Addeck A，Croes K，Van Langenhove K，et al. Dioxin Analysis in Water by Using a Passive Sampler and CALUX Bioassay［J］.Talanta，2012，88: 73-78.

［89］ Ormerod S J，Dobson M，Hildrew A G，et al.Multiple Stressors in Freshwater Ecosystems［J］.Freshwater Biology，2010，55(S1):1-4.