aquatic ecotoxicology: from the ecosystem to ... · system. ecosystems are not under steady-state...

Aquatic Ecotoxicology: From the Ecosystemto the Cellular and Molecular LevelsAlain Boudou and Francis RibeyreLaboratoire d'Ecotoxicologie, Universite Bordeaux I/CNRS,Talence, France

This review of aquatic ecotoxicology is presented in three parts. First, we discuss thefundamental concepts and stress the importance of its ecological basis and the complexity anddiversity of the field of investigation, which result from actions and interactions between thephysicochemical characteristics of the biotopes, the structural and functional properties of theliving organisms, and the contamination modalities. Ecotoxicological mechanisms, regardless ofthe level of biological complexity, primarily depend on the bioavailability of the toxic products.Numerous processes control the chemical fate of contaminants in the water column and/orsediment compartments; accessibility to the biological barriers that separate the organisms fromtheir surrounding medium depends directly on bioavailability. Second, we review the principalmethodologies of the field, from in situ studies at the ecosystem/ecocomplex level to bioassaysor single-species tests. Third, we focus on mercury, selected as a reference contaminant, in orderto illustrate the main ecotoxicological concepts, the complementarity between field andlaboratory studies, and the indispensable multidisciplinarity of the approaches. Environ HealthPerspect 105(Suppl 1):21-35 (1997)

Key words: aquatic environment, bioaccumulation, bioamplification, ecotoxicology,methodologies

IntroductionEcotoxicology has been defined in theliterature in many ways over the past twodecades. Depending on their researchobjectives, in particular their fundamentalor applied training, many authors restrictthemselves to a more or less limited area ofinvestigation. For example, according toButler (1), "Ecotoxicology is concernedwith the toxic effects of chemical andphysical agents on living organisms, espe-cially on populations and communitieswithin defined ecosystems; it includes thetransfer pathways of those agents and theirinteractions with the environment."

Manuscript received 9 August 1996; manuscriptaccepted 28 October 1996.We gratefully acknowledge the assistance of R.

Maury-Brachet in the preparation of the figures. Wealso thank H. Koziol from the Department of ForeignLanguages (University of Bordeaux Il) for improvingthe narrative style of the manuscript.

Address correspondence to Dr. A. Boudou,Laboratoire d'Ecotoxicologie, Universit6 Bordeauxl/CNRS, Avenue des Facult6s, 33405 Talence Cedex,France. Telephone: 05-56-84-88-08. Fax: 05-56-84-88-05. E-mail: [email protected]

Abbreviations used: BCF, bioconcentration factor;CVAF, cold vapor atomic fluorescence spectrometry;DLs, detection limits; DOC, dissolved organic carbon;DR, direct route; MFO, mixed-function oxygenase; MTL,Mercury in Temperate Lakes project; NAPAP, NationalAcidic Precipitation Assessment; OSARs, quantitativestructure-activity relationships; TR, trophic route; U.S.EPA, U.S. Environmental Protection Agency.

The main objective of ecotoxicology is tostudy structural and functional disturbancesinduced in the short, medium, and long-term by contamination factors on ecologicalsystems. These factors, including all physical,chemical, and sometimes biological agents,result essentially from the direct and indirecteffects of anthropogenic activities.We present the fundamental concepts

of aquatic ecotoxicology and stress theimportance of the ecological basis and thecomplexity and diversity of the field ofinvestigation. We also discuss the princi-pal study methods and the relationshipbetween reductionism/representativity inthe processes that occur in the natural envi-ronment. Additionally, we focus on mer-cury as a reference contaminant in order toillustrate the main ecotoxicological conceptsand the complementarity between field andlaboratory studies in aquatic ecotoxicology.Fundamental Conceptsin Aquatic EcotoxicologyEcological Basis

Ecotoxicology is based on the fundamentalconcepts of ecology, which aims to under-stand relationships between living organ-isms and their surrounding environmentthrough a systemic analysis using functional

units of varying size and level of complexity.These units range from the ecosphereto terrestrial or aquatic microsystems.Ecological systems function on the basis ofactions and interactions between abioticfactors, which characterize the physi-cochemistry of the biotopes, and bioticfactors, which relate to the biological com-ponent. This component may be examinedat many levels, from the cellular and mole-cular basis to the biocenosis (all the speciesat the ecosystem level), via the intermedi-ary levels of organism and population(all the individuals of a single species).Transition to higher biological levels isbased on the rule of additivity; each higherlevel is the sum of the basic structures thatmake up the level below and also is basedon the emergence of new properties in rela-tion to both structure and function. Undernatural conditions, the abiotic and bioticfactors are extremely diversified and varyconstantly both in time and in space: "Thecomplexity and the individual history ofeach ecosystem give them unique proper-ties which are not duplicated at anotherplace and in many cases not even at thesame place at different times" [Seitz (2)].

Lake systems, considered to be proto-types of ecosystems, correspond to more orless closed units; they are well defined phys-ically and are characterized by a certainfunctional autonomy. However, they areopen systems with permanent exchangeswith the atmosphere and particularly withtheir surrounding terrestrial basins: "Theyoperate as retention zones for inorganic andorganic components transported from thewatershed, the soil/vegetation complex rep-resenting the major compartment of trans-fer between these two systems" [Capblancq(3)]. As in any ecological system, the nor-mal functioning and natural evolution oflake systems rely on the dependence andinterdependence of the various communi-ties of living organisms and their interac-tions with the abiotic environment. Aglobal approach to the metabolic activity ofliving organisms enables us to distinguishtwo main components: First, the photosyn-thetic production of organic material byautotrophic organisms (phytoplankton,macrophytes, periphyton, bacteria); second,the consumption and degradation of thismaterial by heterotrophs (animal species,microbes), with the concomitant recyclingof nutrients, by mineralization (Figure 1).Production of organic matter and biode-gradation are regulated by a complex set of

Environmental Health Perspectives * Vol 105, Supplement 1 * February 1997 2

BOUDOU AND RIBEYRE

AtmosphereUght energy

o Live cellsSeston * Organic detritus

Inorganic solids

Figure 1. Schematic representation of a lake ecosystem showing the main physicochemical (left) and biological(right) processes operating in the production and transformation of dissolved and suspended matter in lake water.Modified from Capblancq (3).

morphodynamic and climatic factors(depth, residence time of water, solar andwind energy, etc.), which impose varyinglevels of constraint on the chemical andphysical properties of the water column(temperature, transparency, gas concentra-tions, dissolved salts, etc.). These propertiescontinuously change in response to seasonalvariations and to feedback effects from themetabolic activity of biota. In contrast toterrestrial systems, primary production inlakes is mainly carried out by microphytes,the phytoplankton being rapidly consumedby herbivorous species, thus inducing a fastrenewal of the biomasses and a rapid recy-cling of the nutrients (4,5). Stratificationcycles in the water column, caused by ther-mic energy exchanges between the atmos-phere and the surface water layers, act incombination with the water temperature/density relationship. This combined action,together with the progressive shift fromoligotrophy to eutrophy, thus makes themedium richer in nutrients-nitrogen andphosphorus-and increases primary produc-tion. These processes lead to a perturbationin the overall system function and to

the natural fate of shallow lakes-theirultimate disappearance via eutrophicationand filling-in processes (5,6). Parallellingthese natural processes, anthropogenicactions contribute to the rate of accelerationat which these systems evolve, which leadsin turn to dysfunctions of varying severity(e.g., dystrophy). Agricultural activity onwatersheds further increases, throughfertilizer spreading (nitrates, phosphates,etc.), the amount of inorganic nutrients orintroduces products toxic for numerousspecies (herbicides, insecticides, fungicides)into the aquatic biotopes, thus increasingdisequilibrium within the systems.

Contemporary limnology has become amuch more experimental and mechanisticscience. Many parameters are quantified,including flux between compartments, andthe processes responsible for the evolutionof the overall systems are investigated (5).Recent systemic studies based on multidis-ciplinary approaches have led to the formu-lation of dynamic and functional models(7); experiments ranging from the labora-tory level to manipulation of the overallecosystem provide a better approach to

fundamental mechanisms (5). Nevertheless,our knowledge of global processes remainsincomplete and often is compartmental-ized. Ecological uncertainties make it verydifficult, if not impossible, to select levelsand analysis criteria sufficiently representa-tive of the ecosystem studied in order toproduce a specific diagnosis on its functionand state of development.

Complexity and Diversityofthe Ecotoxicological FactorsBasic knowledge of the ecological processesis essential in any ecotoxicological study.These ecological processes are the frame ofreference to define the degree of distur-bance to the system, its reversibility, andultimately the qualitative and quantitativeestimates of the medium- and long-termrisks. The variety of ecosystems and thediversity of the natural and anthropogenicperturbations to those systems are the rea-sons for the lack of adequate baseline datafor comparison between disturbed andundisturbed ecosystems, i.e., the classicproblem of distinguishing signal fromnoise (8). The stability and recovery capac-ity of ecosystems depend on the character-istics of the stress (duration, frequency,intensity, novelty) and the history of thesystem. Ecosystems are not under steady-state conditions, even in the absence ofhuman interference.

At the aquatic ecosystem level, the con-tamination of living organisms, whatevertheir level of biological complexity andtheir position in the food webs, closelydepend on the bioavailability of the toxicproducts present in the biotopes (9). Thiskey concept results from the numerousmechanisms controlling the chemical fateof contaminants in the water columnand/or sediment compartments. As shownin Figure 2, metal bioavailability is stronglylinked to the direct and indirect effects ofthe abiotic factors that determine chemicalspeciation reactions in the medium (com-plexation with the different inorganic andorganic ligands in the dissolved and partic-ulate phases) and to the exposure regime(11). Accessibility to the biological barri-ers, which separate the organisms fromtheir surrounding medium, cell mem-branes and epithelial structures (gills, gutwall, integument), and control the ad- andabsorption processes, directly depend onbioavailability. Moreover, abiotic factorsact simultaneously on living organisms;depending on their variation ranges and onthe adaptative capacities of the individuals,they can induce structural and functional

Environmental Health Perspectives * Vol 105, Supplement I * February 199722

AOUATIC ECOTOXICOLOGY

disturbances of varying severity. For exam-ple, a large increase or decrease in watertemperature can lead rapidly, given theheterothermal nature of the aquatic species,to a modification in the internal tempera-ture. This modification, depending on thedegree of thermal stress, can induce adap-tive responses at organism level (e.g., respi-ratory and circulatory functions) and at thecellular and molecular levels (enzymaticactivities, membrane fluidity, etc.). Thedegree of exchange between individualsand their surrounding environment can bemodified considerably, as can capacities ofuptake, sequestration, or biotransformationof toxic products at tissue and cell levels.

Thus, contaminant transfers betweenbiotopes and living organisms, togetherwith the toxicological effects, result fromactions and interactions between the threefundamental poles of ecotoxicology: abi-otic, contamination, and biotic factors (9).Interindividual and interspecies differencesaccount for the high diversity of responsesbetween ecosystems. These responses arebased on direct biological effects, includ-ing the mortality of different life stagesand disruption of reproductive cycles,etc., and indirect biological effects, essen-tially based on species-to-species interac-tions, which play a fundamental role inthe structure and function of the systems(trophic relationships via predator andprey interactions, disruption of habitats,etc.). Ecotoxicological impacts at theecosystem level can lead to new functionsat the community level, such as primaryand secondary production and nutrientcycling, with a marked delay between thesteps from organism to whole ecosystem(Figure 3)(12).

Principal Study Methodsin Aquatic EcotoxicologyThe complexity of the mechanismsinvolved in ecotoxicology and the breadthand diversity of the field of investiga-tion lead to the crucial issue of researchmethodologies and their representativity inrelation to the processes that occur in thenatural environment.

In Situ StudiesFor many research scientists, the only real-istic approach in aquatic ecotoxicologyconsists of studies carried out in situ on theecosystem or ecocomplex (a set of con-nected and interactive ecosystems), such asa lake and its catchment basin (13).Information from such studies bringstogether all the processes mentioned above,

Heavy metal sources(input from natural andanthropogenic origin)

Figure 2. Ecotoxicological approach to metal bioaccumulation and transfers at ecosystem level. Actions and inter-actions between the three fundamental sets of ecotoxicological factors: contamination, abiotic, and biotic factors.From Boudou and Ribeyre (10), with permission.

Time scaleImmediate to

days

Minutes toweeks

Days tomonths

Months toyears

Months todecades

Pollutant input

FBioaccumulation to effect threshold-Behavioral response Biochemical response

Physiological response Morphological response

Altered performance

Population impactEcological interactions

Community and ecosystem structure and dynamics

Ecosystem functionReduced organic decomposition, nutrient conservation, alteration of cyclesof essential nutients, reduced prmary productivity, increased energy cost,reduced ecosystem production, alteration offood web, functional regulation

I r of ecosystem processes.

Figure 3. Conceptual chronology of induced effects following exposure to toxic pollutants, emphasizing changes inecosystem functions. Modified from Sheehan (12).

which are highly representative and there-fore provide invaluable direct studies ofnature. They also form a frame of refer-ence, both for the distribution and fate ofcontaminants within the different abioticand biotic compartments of the systemsstudied and for certain effects produced inthe structure and the function of these sys-tems. Nevertheless, such studies inevitablyconfront the complexity of the phenomenainvolved and the reductionism in space(using a smaller number of sampling

stations) and in time (limiting samplingperiods) imposed by the constraints ofthese investigation methods. Hence it isoften very difficult to move beyond thedescriptive stage, as the diversity of factorsand their variability make it difficult to for-mulate interpretive hypotheses regardingthe mechanisms. "Capturing key signalsthat describe central processes or conse-quences will continue to be a major chal-lenge in the rapidly changing field ofaquatic ecotoxicology" (14). Induced

Environmental Health Perspectives * Vol 105, Supplement * February 1997 23

BOUDOU AND RIBEYRE

changes in the specific structure ofcommunities and in their structural prop-erties can be assessed at global levels, butstudies of this kind require major resourcesto set up sampling strategies and collectand interpret data (inventories of flora andfauna, quantification of biomasses andproductivities, etc.). Given these con-straints, many methodological compro-mises have been developed in order tolimit the area of investigation covered infield studies and to facilitate their estab-lishment. For example, the determinationof fairly simple indices, such as abun-dance, biomass, species richness, saprobicindex, biotic score, etc., or the selection ofindicator species make it possible to esti-mate the quality of the environments, toestablish interstation or interecosystemcomparisons, or to follow medium- orlong-term developments (12,15,16).

Listed below are biological indicators ofecological effects selected to optimize thedetection of potential or actual changes inselected ecological end points [modifiedfrom Kelly and Harwell (8).

Purposes for indicators* intrinsic importance: indicator is end

point- economic species

* early warning indicator: rapid indica-tion of potential effect- use when end point is slow or

delayed in response- minimal time lag in response to

stress; rapid response rate- signal-to-noise ratio low; discrimina-

tion low- screening tool; accept false positives

* sensitive indicator: reliability in predict-ing actual response- use when end point is relatively

insensitive- stress specificity- signal-to-noise ratio high- minimize false positives

* process/functional indicator: end pointis process- monitoring other than biota, e.g.,

decomposition ratescomplement structural indicators

Criteria for selecting indicators* signal-to-noise ratio

- sensitivity to stress- intrinsic stochasticity

* rapid response- early exposure, e.g., low trophic level- quick dynamics, e.g., short life span,

short life cycle phase* reliability/specificity of response

* ease/economy of monitoring- field sampling- lab identification- preexisting database, e.g., fisheries

catch data- easy process test, e.g., decomposi-

tion, chlorophyll* relevance to end point

- addresses "so what" question* monitoring of feedback to regulation

- adaptative managementDetection of biochemical responses to

pollutants from selected species collected inthe field-the biomarker concept-hasreceived considerable attention duringthe last decade in the assessment of func-tional effects in contaminated ecosystems.Biomarkers represent an organism's attemptto compensate for or tolerate stressors inthe environment (17). For example, mixed-function oxygenase (MFO) reactions in thewhole organism or in organs (liver, gills)may indicate accumulation of organic com-pounds such as oils or some halogenatedbiphenyls (18). Increased levels of metal-lothioneins (low molecular weight, metal-binding proteins rich in cysteine) appear tobe characteristic of exposure to several met-als (Cd, Cu, Zn). Their induction has beenproposed as tissular or cellular indicators inaquatic animals despite the large numberof other agents able to induce their biosyn-thesis, such as hormones and second mes-sengers, growth factors, vitamins, cytotoxicagents, stress-producing conditions, etc.(19,20). Bioprobes or biosensors withpotential application at field level wererecently developed with two essentialmonitoring applications: the detection ofspecific pollutants, using enzyme- or anti-body-based devices; and the detection ofunexpected changes in environmental

chemistry, using broad-spectrum whole cellbiosensors (21).

~nSilepesAppraches

Experimental studies using monospecificmodels at the laboratory level are at theother end of the methodology complexityscale (Figure 4). These studies fulfill twomain objectives. The first is linked tomechanistic approaches that, because theyuse lower biological levels of organizationand can monitor a large number of para-meters, provide a better understanding ofecotoxicological mechanisms. This researchgenerally complements toxicologicalstudies, especially those based on moreadvanced investigation methods, such assophisticated biophysical and biochemicaltechniques (22). It enables us to investi-gate, for example, fundamental linksbetween chemical speciation reactions ofheavy metals in the medium and biouptakeat the biological barrier level or structuraland functional toxic effects at cellular ormolecular levels through the use of isolatedorgans (e.g., perfused gills) or culture cells(23-25). The second objective, from amore applied viewpoint, is to set up stan-dardized toxicity tests based on one species,or a battery of single-species tests to pre-dict the toxicity of pure chemicals or efflu-ents on the laboratory level and undercarefully controlled conditions (qualityassurance/quality control requirementsunder standard operating procedures).Bioassays are widely used in aquatic eco-toxicology. Most regulatory studies, estab-lished to estimate the risk involved inmarketing a new chemical product or inassessing the quality of an environment, arebased on these methodologies (26). The

Complexity-representativity In situ studies

Outdoor ecotoxicological models(ponds, mesocosms, enclosures, artificial streams, etc.)

Indoor ecotoxicological modelsImicrocosms, experimental trophic chains, etc.)

Single species studies/ bioassays-mechanistic approaches)

Replicability-simplicity Mathematical models

Figure 4. Principal methodologies in aquatic ecotoxicology showing the relationship between representativity-complexity and reproducibility-simplicity.


AQUATIC ECOTOXICOLOGY

aquatic species selected for such method-ologies, apart from the universal cladoceranDaphnia magna and a few phytoplanktonicalgae (Chlorella, Scenedesmus), differ fromcountry to country but the biological mod-els are similar in position within the fresh-water communities: primary producers,zooplanktonic species, mollusks, and fish(27). Data from standardized tests are reli-able, repeatable, and rapidly collected; theyare generally used as screening tools to pro-vide information on acute and chroniceffects from a very large set of criteria: death,immobility, reproduction, growth, physio-logical functions (e.g., respiration andhematological parameters), behavior, geno-toxicity, etc. Standardization necessitatesrigorous experimental designs with a highdegree of control of the abiotic and bioticfactors. Such control can only be obtainedby simplification-sometimes drastically-of the experimental conditions. This leadsto major problems with the representativityof results and extrapolation in relation tooccurrences in the natural environment(Table 1). Accurate predictions of ecotoxiceffects from bioassays are extremely diffi-cult, if not impossible (28). For example,almost all monospecific tests in aquaticecotoxicology are carried out in artificialenvironments (synthetic river water) inorder to avoid variations in the main physi-cochemical factors and in their direct orindirect effects on the bioavailability of thecontaminants studied and on the organisms.Similarly, only the direct contaminationroute via the water column is considered inthe majority of freshwater bioassays, whereasfor many chemical products, trophic trans-fers represent the chief contamination route.However, despite uncertainties in predic-tion or risk estimation, these methods areessential tools and they provide indispens-able information; it is important to beaware of the limitations of these data andto apply them with care whenever it is nec-essary to extrapolate to a complex, naturalecosystem (Figure 5) (29-31). Accordingto Chapman (32) "the field does not neces-sarily validate the laboratory. Each providesa different viewpoint. Different views, inthe form of tools such as toxicity tests andfield studies, together provide the bestoverall perspective."

Multispedes ModelsMajor research programs have been devel-oped to devise ecotoxicological models thatoccupy an intermediate place between fieldstudies and the monospecific approach(Figure 4). These models are intended to

Table 1. Biological uncertainties in single-species bioassays that lead to an underestimation of the toxicity ofmetals to natural communities.

Source of uncertainty

Choice of species

Exposure time

Exposure route

Multigenerational life cycle

Higher order secondary effects

Interaction with natural disturbances

Source of insensitivity

UnrepresentedHighly sensitive speciesEcological keystone speciesLong-lived upper trophic levels

UnderestimatedEffects of lifelong exposure

Rarely consideredAdditive pathwaysRelevance of pathways

UnrepresentedProgressive accumulation of toxic effects through generations

Changed biological interactionsEffects on community structure when keystone species are lost

Cost of metal stress in tolerance to natural perturbation: individualsand populations

From Luoma (28), with permission.

facilitate predictions of community levelresponses and to reduce uncertaintieswhen extrapolating from indoor bioassaysto field conditions.

There are two main types of multi-species models. The first type is based onexperimental studies in outdoor conditionsin order to emulate the structural andfunctional properties of ecosystems whilestill retaining some of the advantages of theexperimental approach. Some of these

Laboratory

Acute effects

Chronic effects

advantages include monitoring certainabiotic or biotic factors, possibily setting upexperimental controls and replicates, andtesting the effects of scale. These method-ologies "attempt to bridge the gap betweenlaboratory experiments and field observa-tions" (33). The design of field test studiesin aquatic ecotoxicology depends on theirobjectives. The objectives may be todevelop and validate predictive models forchemical fate and/or effects, to evaluate

Z Rea world I

Indicatorspecies

Indicatorspecies

Key receptorspecies

Key receptorspecies

Key receptorcommunitiesChronic effects Multispeciessystems

Figure 5. Extrapolation factors for ecological effect assessment in the laboratory and from the laboratory to thefield. Modified from Persoone and Janssen (29). Abbreviations: EXP, experimental variability; L(E)C50, lethal envi-ronmental concentration, 50%; ENV, environmental variability; AC, acute to chronic extrapolation; SP-SP, species-to-species extrapolation; SP-COM, species-to-community extrapolation; COM-COM, community-to-communityextrapolation; LC50, lethal concentration inducing 50% mortality on the studied populations; NOEC, no-observed-effect concentration.

Environmental Health Perspectives * Vol 105, Supplement - February 1997 25

BOUDOU AND RIBEYRE

environmental quality standards derivedfrom laboratory toxicity data throughextrapolation, to study resilience of ecosys-tems in terms of time required for restora-tion after physicochemical disturbance,or to obtain data required for regulatorypurposes in order to assess fate and/oreffects in natural ecosystems (34). Manyoutdoor ecotoxicological models have beendeveloped over the past few decades, withmarked differences according to their scale(from liters to millions of liters) (Figure 6),their biological component complexity(from two selected species to complex nat-ural assemblages), their study environment(ponds, lakes, streams), their dependenceon natural reference systems, and also theiroperational time scales (from hours toyears). The terms mesocosm, outdoormicrocosm, pond, enclosure, and artificialstream have been used to describe such testsystems (36,37); most of these studies havebeen carried out on pesticides (insecticides,herbicides) (38). Since the early 1980s,freshwater field tests have been required byseveral countries (including the UnitedStates and the United Kingdom) to sup-port the registration of some pesticides;these tests are requested only if laboratorydata together with environmental exposureestimations suggested that aquatic ecosys-tems might be at an unacceptable level ofrisk. In 1987, the U.S. EnvironmentalProtection Agency (U.S. EPA) published atechnical guidance document for the con-duct of aquatic mesocosm tests (39).Several international workshops were orga-nized in the United States and Europe,including Society of EnvironmentalToxicology and Chemistry-RESOLVE inWintergreen, Virginia, in 1991, and theEuropean Workshop on Freshwater FieldTests (EWOFFT) in Potsdam, Germany,in 1992, which resulted in the publication

>250 m3, D-1 m3,24 25%

250 M3 ...........,,g.

9%

15m3,15-50 m3, 5-15 m3, 19%19% 4%

Figure 6. Size distribution of lentic outdoor freshwatermicro-/mesocosms in which structural responses underpesticide treatments have been studied (n = 85).Modified from Brock and Budde (35).

of proposals for the conduct of such fieldtests (40,41). A growing number of reportsindicate that outdoor mesocosms and arti-ficial streams are powerful tools for devel-oping predictions of the indirect effects oftoxicants or mixtures of toxicants on whole-community assemblages. Nevertheless, theU.S. EPA recently decided to discontinuemesocosm studies on pesticide risk assess-ment on the basis that "they do not providesubstantial information for making riskdecisions beyond that already revealed bylower tiered studies" (42). In the EuropeanCommunity, no real strategy is currentlydefined to integrate this type of methodol-ogy in the regulations governing authori-zations to market new chemical products.

The second type of multispecies modelis laboratory-based and aims to producemore complex, and therefore more repre-sentative, ecotoxicological models than sin-gle species tests, in relation both to bioticfactors (multispecies systems) and abioticfactors (indoor microcosms with a mixedbiotope: for example, natural sediment andwater column). Models vary according to

the purpose of the experiments and thelevel of control and artificiality of the sys-tems. The first well-defined experimentalsystem was the Metcalf model for the eval-uation of pesticide biodegradability andecological biomagnification (43). Morerecently, the standardized aquatic micro-cosm set up by Taub (44) made it possibleto demonstrate significant effects of toxi-cants on ecological interactions within andbetween two trophic levels-alternativefood chains were based on 10 species ofprimary producers and 5 species of grazers(Figure 7).

As shown in Figure 4, the approachesto aquatic ecotoxicology, varying from thefield to fairly simple laboratory tests, arecomplemented by theoretical approachesbased on mathematical models. Manymodels are currently available, with differ-ences according to their objectives andto the complexity and diversity of the eco-toxicological criteria. For example, quanti-tative structure-activity relationships(QSARs), essentially based on the physico-chemical properties of the contaminants

Figure 7. Principal components of a standardized aquatic indoor microcosm, based on an artificial medium with 10species of primary producers and 5 species of grazers. From Taub (44), with permission.

Environmental Health Perspectives - Vol 105, Supplement 1 * February 1997

r v

%b-L.-MI:

26


(hydrophilic/hydrophobic properties, Ko,tHenry's constant, molar volume, electronicparameters, etc.), are frequently used forthe prediction of the fate of chemicals inthe environment, their bioaccumulationcapacities, and sometimes their toxicologi-cal effects (45). At the other end of thecomplexity scale, deterministic simulationmodels incorporate the major processes oftransport and chemical fate of contami-nants in aquatic systems and enable quan-tification of their bioaccumulation andtrophic transfers. Certain models, such asthe mercury cycling model (46), take intoaccount the influence of key limnologicalparameters on the biogeochemical cycles atecosystem level.

Mercury Contaminationof Freshwater SystemsWe have selected mercury as an example toillustrate the fundamental links and com-plementarity between field and indoorstudies in freshwater ecotoxicology.

Mercury is a natural component of theecosphere with a low relative abundance(ranking 62nd), although for geochemicalreasons it tends to be concentrated in vastmercurous belts (47). Mercury occurs invarious physical and chemical forms in theenvironment: minerals (cinnabar, HgS),metallic Hg0, inorganic forms (mercurous,HgI, and mercuric, HgII), and organomer-curials (methylmercury, MeHg; dimethyl-mercury, Me2Hg), etc. These differentforms and their associations with com-plex ligands regulate transport pathways,residence times within the different com-partments of the biogeochemical cycle(atmosphere, geosphere, hydrosphere, bio-sphere), bioaccumulation and trophic trans-fers along the food webs, and toxic effectson living organisms. Humans use largeamounts of mercury-world production isestimated at 8,000 to 10,000 tons/year(48). Mercury's different chemical formsare used in diverse processes (catalysis,amalgamation, electrical/battery, chloral-kali, instruments, laboratory, etc.), thoughsome processes no longer use mercurybecause of the risk associated with its use(fungicides in agriculture, for example).Numerous secondary sources contribute alarge amount of mercury into the environ-ment: coal combustion, waste incineration,metal smelters, etc. (49). Current esti-mates for anthropogenic interference withthe natural cycle range from approximately50 to 75% of total annual Hg emissions tothe atmosphere; recent modeling suggeststhat the present atmospheric Hg burden

has increased by a factor of 3 during thelast 100 years, with a current rate ofincrease of about 0.6%-year-1 (about0.01 ngm-3oyear') (50,51).

As mercury has been used since ancienttimes (Hg was the seventh metal of antiq-uity and has been known and used for morethan 3500 years), cases of poisoning alsohave a long history: the first recorded casewas a mercury miner in the 15th century(52). Ecotoxicologically, the poisoning out-break in Japan in 1953 to 1960 (Minamatadisease) gave rise to an international aware-ness of the risks linked with the emission ofHg into aquatic systems and of the com-plex mechanisms that led to the poisoningof huge numbers of the population(899 cases were officially recognized, with143 deaths and numerous congenital andinfant cases)(53). A factory producingacetaldehyde and vinyl chloride had usedmercury as a catalyst and considerableamounts of metal had been directly dis-charged into the enclosed marine bay ofMinamata. After lengthy epidemiologicalstudies, the cause of the disease was identi-fied as high metal concentrations in fishand shellfish, which were the staple diet forthe families of the fishermen living on thebanks of Minamata Bay (54). Complex newphenomena were discovered on the globalecosystem scale, including the concept ofbiomagnification, whereby Hg concentra-tions increase considerably in successivelevels of the marine food webs, from pri-mary producers to terminal carnivorousconsumers (fish, marine mammals).

Since the beginning of the 1980s,ecotoxicological problems linked to thecontamination of aquatic systems by mer-cury have entered a new phase. High mer-cury levels were found in carnivorous fishliving in thousands of forest lakes in theNorthern Hemisphere-in areas remotefrom human activity-and in the hydro-electric reservoirs constructed in northernQuebec, Canada, close to James Bay. Thesefield studies demonstrate a new pattern ofHg pollution, with the food webs becom-ing a vital element linking the metal chem-ical fate in the environment and humanhealth risks. For example, boreal forestlakes in Sweden are characterized by verylow human population densities in theircatchment areas and most of the lakes havenever been affected by Hg discharge intotheir watersheds. In more than half of these83,000 lakes, it was estimated that totalmercury concentrations in 1-kg pikes (Esoxlucius) exceeded the health advisory limitfor safe human consumption (1 mg Hg/kg,

fresh weight); on the other hand, totalmercury levels in the water column werelower than 10 ng/liter (48).

Recent Field Studiesat Ecosystem LevelResults from the Mercury in TemperateLakes project (MTL), connected with theNational Acidic Precipitation Assessmentproject (NAPAP) in north-central Wisconsinin the United States, clearly demonstratethe progress made in field studies in aquaticecotoxicology. This progress is mainlylinked to the multidisciplinary nature ofthe research, which is able to benefit fromall the new forms of technology establishedby the various branches of research, such aslimnology, analytical chemistry, geochem-istry, and microbiology, as well as new ana-lytical procedures. By using ultracleantechniques during the sampling steps anddetermining total Hg by cold vapor atomicfluorescence spectrometry (CVAF) after adual preconcentration step (Hg amalgama-tion on gold traps), it is now possible toreach detection limits (DL) of 0.02 ngHg/liter or 10-13 mol/liter (55). Theseprocedures now challenge virtually allmeasurements carried out on earlier watersamples (precipitation, surface and under-ground water, porewater in sediments),revealing overestimations as high as, orgreater than, a factor of 500 (56). More-over, using new specific analytical proce-dures based on the combination of anethylation phase (NaBEt4), separation ofthe different volatile Hg chemical forms byisothermal gas chromatography, elec-trothermal atomization, and Hg° determi-nation by CVAF, it is now possible tosimultaneously determine the principalchemical forms of the metal (Hgo, HgII,MeHg, Me2Hg) in the different matrices(water, sediment, biota), with a DL closeto the picogram level (57).

Over a 3-year period, total Hg andMeHg mass balance were analyzed at theecosystem level in Little Rock Lake,Wisconsin (58). Figure 8 shows mercurydistribution in the principal abiotic andbiotic compartments and metal fluxesbetween atmosphere, water column, andsediments. The results indicate that atmos-pheric deposition was the major source ofmercury in the lake, but MeHg accountedfor only 1% of the input. About half of thetotal Hg burden estimated in the water col-umn compartment was in the biota; theMeHg pool in the fish corresponded to 73%of the total amount in the water column,including seston (mostly phytoplankton)


BOUDOU AND RIBEYRE

_ PGaseous Paculate1.6 0.4 ngm3 * 0.02 ngm3

Wet Dry_.mmmml. Evasion

S ± 2 Aghm2/yar - 3.5 ± 3gmlyar .<_0. Syea 0.3 )r

Total Mercury Methylmercury

Figure 8. Mass balances for mercury and methylmercury in the treatment basin of Little Rock Lake, Wisconsin. All concentrations are in ng Hg/liter or ng Hg/g, wet weight.Dashed line represents the oxic/anoxic boundary layer. Modified from Watras et al. (58), with permission.

and zooplankton. Most of the net atmos-

pheric Hg inputs into the lake (1 g Hg/year)accumulated in sediments, which were thelargest Hg reservoir in the Little Rock Lakesystem. The percentage of MeHg versus

total Hg in the water column was close to

10%. Mercury release as Hg0 from the lakeinto the atmosphere was about 5% of thedeposited input; this average value was

lower than in results obtained from theother seven Wisconsin lakes studied underthe MTL project (10-50%). The mass bal-ance showed that almost all MeHg formedwithin the lake system; endogenous methyl-ation reactions were a key process in metalbioaccumulation at the food web level.

MeHg biomagnification occurred inthe lake-the organic compound was bio-concentrated by a factor of 3 million timesin the whole fish (bioconcentration factor(BCF) = [Hg] organism/[Hg] water, esti-mated in 1-year-old yellow perch, Percaflavescens); BCF values increased by about0.5 log units per trophic level (59). Dataobtained on fish showed that almost allaccumulated Hg was in the monomethyl-ated form. Average percentages of MeHgin the phytoplankton and zooplankton

were dose to 15 and 30%, respectively (60).Comparative analysis of Hg concentrationsin fish from several lakes shows markedvariability between individuals of the same

species as well as between different species.For example, the lowest and highest valueswithin individuals of the same species andage can differ by a factor of 10 or more (48).Field studies have shown that not only age

but also size and weight may influence Hgbioaccumulation in fish either directly or

indirectly. The physicochemical character-istics of the biotopes also play a fundamen-tal role. Numerous data from lake surveys

in Scandinavia, the United States, andCanada indicate that Hg concentrations infish are negatively correlated to lake pH or

alkalinity (61,62); moreover, Hg concen-

trations in fish from low alkalinity lakeswithout anthropogenic metal sources oftenexceed those from Hg-contaminated waters

with high alkalinities. When a lake is acidi-fied, complex changes occur in the turnover

of substances and in the biotic structure ofthe ecosystem (48,63). High levels of Hgin biota from acidic forest lakes in Swedenmay be the result of low productivity ratherthan a direct influence of pH on the metal

biogeochemical cycle (48,64). After the first2-year stage of experimental acidification ofthe treatment basin (pH 5.6) of Little RockLake, total Hg determinations in yellowperch differed from the unacidified refer-ence basin (pH 6.1) by about 20%. Theseresults were linked to differences in within-basin processes that influenced the bioavail-ability of the metal-perhaps a greatermethylation of Hg at the lower pH (59).Recent data from acidified lakes in Ontario,Canada, show that Hg concentrations infish are positively correlated with dissolvedorganic carbon concentrations (DOC) inthe water (65). Acidification may alsomodify the abiotic and biotic processes

controlling transformations of the differentchemical forms of the metal (66). Anincreased reduction of HgII in high-pHlake waters may be an important mecha-nism for decreasing the supply of substratefor in-lake methylation (67). However,increases in the rates of methylation rela-tive to demethylation at low pH may alsocontribute to the increased MeHg concen-

tration in fish from acidified lakes (68).Interesting results were recently obtainedfrom planktonic samples collected in the

Environmental Health Perspectives * Vol 105, Supplement * February 1997

Wet

i.i00g/m2tyear- 0.01 g/year

.................... .........-. ......

28


two basins of the Little Rock Lake (60):Comparison of Hg burdens in individualzooplankton from the acidified and refer-ence basins shows that changes in total Hgconcentrations as a result of acidificationwere small and not statistically significant.However, MeHg constitutes more than 90%of the total Hg concentrations in cladocer-ans from the acidified basin, compared toless than 30% in the reference basin.

It is not possible to determine therespective proportions of the two contami-nation routes through field studies: thedirect route (DR), via Hg in the water, andthe trophic route (TR), via the ingestion ofcontaminated prey. This is because of theirsuperposition and the highly complex inter-species relationships involved. Nevertheless,empirical evidence indicates that diet is theprimary uptake route of MeHg by fish innatural conditions. According to Spry andWiener (61), the TR appears to representmore than 90% of Hg bioaccumulated incarnivorous fish. Several field studies showthat changes in the diet of carnivorous fish,because of the depletion of prey or theincrease in predator size, have a markedeffect on metal bioaccumulation, evenwhen the contamination pressure from theDR remains identical (48). Moreover,given the low concentrations of total Hgand MeHg in the water column in lakes orhydroelectric reservoirs, it is difficult toaccount for the levels of Hg bioaccumula-tion measured in fish after several monthsof direct exposure.

Field studies also have revealed thatsediments play an important role in thebiogeochemistry of mercury within aquaticsystems (69). The sediments represent amajor storage compartment: about 99% ofthe Hg in a typical forest lake is present inthis compartment (70). The sediments arealso a favorable site for transformations ofHg chemical forms, notably methylationreactions. Isotopic methods, wet chemistry,and mass-balance approach have recentlyconfirmed the important role played by sedi-ments or, more exactly, by layers within afew centimeters of the sediment surface inMeHg production. Sulfate-reducing bacte-ria are described as important Hg methyla-tors in anoxic sediments; Hg methylationmay also take place in sediments with ahigh decomposition of organic matter viaextracellular enzymes (71,72). Among thebenthic species, insect larvae may have animportant influence on Hg release fromthe sediment compartments (73) andstudies on hydroelectric reservoirs andlakes in northern Quebec, Canada, show

that these species constitute up to 90% ofthe diet ofmany fish.A recent review by Rudd (72) of the

sources of MeHg in freshwater systemshighlights some marked differences betweenlakes. For example, atmospheric MeHgdepositions in southern Sweden are tentimes higher than the average rates esti-mated in northwestern Ontario in theExperimental Lakes Area Reservoir project;similarly, terrestrial catchment areas thatcontain wetlands have been identified asimportant sources of MeHg. It has longbeen recognized that the in-lake produc-tion of MeHg was essentially based onmethylating bacteria in the superficial sedi-ment layers. Today, many studies haveshown that internal MeHg production alsooccurs in the water column via biotic andabiotic methylation (74,75). More recently,flooded terrestrial surfaces, as in the hydro-electric reservoirs in northern Quebec, havebeen shown to be important sources ofMeHg (76). Contradictory study resultshave been published on fish and Hgmethylation in the gut (72). Given thepredominance of the methylated form ofHg in terminal consumers in lake systems,it is important to define the mechanismsresponsible for methylation in the naturalenvironment, to localize them in the differ-ent compartments, and to quantify MeHgproduction when confronted with theantagonistic reactions of demethylation.Our present knowledge of these funda-mental processes is still very limited: "Thereason that our understanding of internalMeHg production is so vague is that thereare no methods for the measurement of nat-ural rates of Hg methylation or demethyla-tion in the field" (72). Isotopic methodsbased on 203HgII and 14CH3HgX do notgive quantitative rate information; they areuseful to determine methylation sites in thewater column or sediment compartmentsor to analyze factors influencing rates ofmethylation or demethylation (72).

Although considerable progress hasbeen made in the quantification of Hgcompound distribution and flux betweenthe different abiotic and biotic compart-ments in freshwater systems, a very limitedamount of data has been published on thetoxic effects induced under such exposureconditions in living organisms. Accordingto Wiener and Spry (77), neurotoxicityseems to be the most probable chronicresponse of wild adult fish to dietaryMeHg. Nevertheless, the primary effects, ifany, under these exposure conditionswould be reduced reproductive success in

fish populations as a result of the toxicity ofmaternally derived MeHg to the embryonicand larval stages.

Even in light of the mass of data col-lected from field studies, many questionsremain about the mechanisms involved andthe direct and indirect roles of the variousecotoxicological factors. It is at this levelthat experimental approaches can con-tribute arguments to assist interpretation.We have selected examples from laboratorystudies into bioaccumulation mechanismsfrom the trophic chain level to cellular andmolecular levels in order to illustrate howthe two approaches can be complementary.

Complementarity ofExpermentalApproaches: Analysis ofInorganicMercury and MethylmercuryBioaccumulation and TrophicTransfer MechanismsA detailed comparative approach to inor-ganic Hg and MeHg transfers was set upon an experimental four-link trophic chain(Figure 9). Phytoplanktonic algae were theprimary producers and carnivorous fish thethird-level consumers. Results for the pri-mary producer Chlorella vulgaris showedvery rapid and high bioaccumulation capac-ities for the two Hg chemical forms, leadingto similar BCF after 24-hr exposure (Figure9A) (78). For each consumer species, thetwo contamination routes were analyzed:direct uptake from the water and trophicuptake from living prey previously exposedto the two Hg compounds. Trophic trans-fers between contaminated algae and thezooplanktonic herbivorous species (Daphniamagna) revealed marked differences betweenthe two compounds: at 18°C, the esti-mated transfer rates were close to 6% forinorganic Hg and 58% for the methylatedform (Figure 9B). Optimized transfersbetween algae and daphnia, without over-abundant food supply, were greater than90% for MeHg (79). Recent data obtainedby Mason et al. (80) were in agreementwith these values: transfer rates for HgIIand MeHg between diatoms (Thalassiosiraweissflogii) and calanoid copepods (Acartiatonsa, Temora longicornis) were 15 and62%, respectively. Comparative studies ofMeHg bioaccumulation in the third levelof the experimental trophic chain, the mos-quito fish (Gambusia affinis), from the DRand the global route (direct and trophic),showed marked differences: the estimatedaverage BCFs after 30 days of exposurewere 2500 and 27,000, respectively (Figure9C)(81). A similar experimental approachwith rainbow trout alevins (Onchorynchus


BOUDOU AND RIBEYRE

mykiss), fed for 10 days with daphniapreviously contaminated with HgII orMeHg, showed average transfer rates of 41and 92%, respectively (82). The upper linkof the trophic chain was the adult rainbow

trout. Fish were fed living prey (3alevins/day per trout) for 30 days (Figure9D). The estimated trophic transfer rateswere 54% for inorganic Hg and 95% forMeHg; these were in agreement with

assimilation efficiencies for dietary uptakein several fish species (77).

Results from trophic transfers at theexperimental food chain level, from primaryproducer to terminal consumer, show a

A

16 24

Hours

100 r ----

75i~-0 50~

MeHg25

0Chlorellavulgaris

100 -

75 -

-- 50-HgIl -

Daphniamagna

30- ^DR +TR*DR /

24

18

12

os~~~~~~$ 0

0 4 8 12 16 20 24

Days

1: MeHg3 * HgIl/

2-

III

0 10 20 30

DaysDR

0.

0.

o.

0.

0.

.5- O MeHg

.4*- HgIl

3-

.2

.1-

0

Figure 9. Experimental freshwater trophic chain set up for the comparative study of inorganic mercury (HgIl) and methylmercury (MeHg) transfers, from the Direct route (DR)and the Trophic route (TR). From Boudou and Ribeyre (10). (A) Bioaccumulation of two Hg compounds by phytoplanktonic algae Chlorella vulgaris. (B) Trophic transfers of Hglland MeHg between Chlorella vulgaris and zooplanktonic cladocera Daphnia magna. (C) Bioaccumulation of MeHg in the mosquito fish Gambusia affinis, from the DR and theglobal route (DR + TR). (D) Bioaccumulation of Hgll and MeHg in the rainbow trout Onchorynchus mykiss.


a 600)

- 40

20

0

B100

75

d 50=25

0

100

75

Ci 50-

25

C

cmC,aLC,=

D

c,cmaLC,S

20 3010

DaysTR

I

0

30


marked difference overall between the twochemical forms of the metal: 1.3% for HgIIand 51 or 87% for MeHg, depending onthe transfer rate retained between the firsttwo levels (82-84). These results provide abasis for the interpretation of field studydata, especially the predominance of themethylated form of the metal in the termi-nal consumers and the empirical evidencethat diet is the primary route of MeHguptake by fish (48,61,64,77).

Through a detailed experimentalapproach to Hg bioaccumulation in fishbased on a separate analysis of the fourcomponents of this global process-directcontamination, trophic contamination, anddecontamination after direct or trophiccontamination-in the whole organism andin the main organs (liver, brain, gills, skele-tal muscle, posterior intestine, kidneys,spleen, blood), we were able to contributefurther to the interpretive analysis of resultsfrom in situ studies. After 30 days of directexposure via the surrounding water, withprecise control of the contamination pres-sure for the two Hg chemical forms, theratio between the average concentrations

measured in the whole organisms was closeto 6.0, in favor of methylmercury. For allthe organs, higher values were obtainedafter contamination with MeHg (Figure10) (83,85). After 250 days of depuration,decontamination rates at the whole-fishlevel were 28 and 37% for inorganic Hgand MeHg, respectively. During thisperiod, the average weight of the organismsincreased by a factor of 4, leading tomarked growth dilution effects on the con-centration criteria. Several organs, such asthe gills and liver, showed a rapid andmarked decrease in their Hg burdens.Other organs or tissues, on the other hand,such as the skeletal muscle after exposure toMeHg and the kidneys after exposure toHgII, showed a large increase in their Hgburdens during the decontamination phase.These fundamental processes were theresult of metal transfers between donor andreceiver compartments (86). The mostimportant conclusion was in relation to theskeletal muscle: although the fish were nolonger subjected to Hg contamination, thistissue received large amounts of Hg duringthe first 20 days of the decontamination

phase, leading to a 200% increase in theaverage metal burden, with no furthersignificant decrease up to the end of the250-day period. These processes are impor-tant as the skeletal muscle represents about55% of the fresh weight of the fish andconstitutes virtually all of the flesh eaten byhumans (83).

Trophic contamination of the trout wascharacterized by a marked difference in Hgconcentrations in the whole fish after 30-day exposure: the average ratio between thetwo Hg chemical forms was about 3, infavor of MeHg (87). Metal distribution inthe eight organs revealed considerabledivergence (Figure 10): For inorganic Hg,there were higher concentrations in the pos-terior intestine than in the other organs; forMeHg, the distribution was more homoge-nous (87). During the depuration phaseafter trophic contamination, inorganic Hgaccumulated in the intestine disappearedrapidly and almost totally during the first 2weeks, with no significant increase of theHg burdens in the other organs, notablythe blood and kidney; inorganic Hg waseliminated through the feces (82).

A30 -

0,j.e 20-

.0= 10

0 100 200 300Days

Bcm

=LD

.00,

30

20

10

0

C9 -

=L,0 3

0)-a0 20 40 60

Days

X Inorganic Hg* Methyl Hg

a0)

.2_

4-

am01)C.)

0

C.)cm

0,

L Br G Mu Int K Sp BI WF

Direct route

L Br 6 Mu Int K Sp BI WF

Trophic route

Figure 10. Bioaccumulation of inorganic mercury and methylmercury in the rainbow trout-whole fish and organ-after 30 days of exposure via the direct route (DR) (Hgadded to the water) or the trophic route (TR) (ingestion of living contaminated prey). Evolution of Hg burdens in the gills (A) and the skeletal muscle (B) during the 250-daydecontamination phase, after contamination by the DR. (C) Evolution of Hg burdens in the posterior intestine during the 60-day decontamination phase, after contamination bythe trophic route. From Boudou et al. (23). Abbreviations: fw, fresh water; BI, blood; Br, brain; G, gills; Int, posterior intestine; K, kidney; L, liver; Mu, skeletal muscle; Sp,spleen; WF, whole fish.

Environmental Health Perspectives * Vol 105, Supplement 1 * February 1997

Ia*rse0

co

(0

0

0)=)

3 1

BOUDOU AND RIBEYRE

These experimental studies clearlyrevealed the fundamental role played bythe biological barriers involved in Hguptake by fish. Particularly significant wasthe high specificity of the intestine wall toMeHg absorption in contrast to inorganicHg adsorption at the microvilli interfaceon the apical part of the enterocytes with avery low uptake rate. Interactions betweenHg chemical forms and species and biolog-ical barriers are of major interest in under-standing bioaccumulation mechanisms andcombined toxicological effects (23). What-ever the level of complexity of living organ-isms, these barriers control the uptake andaccessibility of contaminants to the internalcompartments: the cytosol at the cell level,the hemolymph or blood at the organismlevel. Although composed of a complexarrangement of epithelial cells, each barrieris in fact based on the same fundamentalstructure: the plasma membrane and itsphospholipidic bilayer, including proteins(88). Basic Hg uptake by the cells is bypassive diffusion through the membranes;facilitated transport mechanisms anduptake by carrier systems have also beendescribed for a limited number of celltypes, such as Hg-resistant bacteria or neu-ronal cells (75,89). A high MeHg bioaccu-mulation capacity is frequently attributedto the lipophilic character of this organiccompound. Indeed, the predominance ofMeHg burdens in the skeletal muscle offish rather than in the fatty tissue clearlyshows that bioaccumulation is not gov-erned solely by the liposolubility of thischemical form. In fact, neutral chemicalspecies of inorganic Hg have octanol/waterpartition coefficients (Ko4) similar to, indeedhigher than, those of MeHg: HgCl2, 3.3;HgOHCI, 1.2; Hg(OH)2, 0.05; CH3HgCl,1.7; CH3HgOH, 0.07 (80,90,91).

Biophysical studies based on lipidicmodel membranes (biomolecular lipidmembranes, liposomes) and several com-plementary techniques (nuclear magneticresonance, fluorescence polarization,203Hg flux) have demonstrated the funda-mental role of HgII chemical speciationreactions in the external medium in themetal's accessibility to the cell interfaces bybinding to proteins and phospholipid polarheads and in transport by diffusion throughthe hydrophobic core of the membranes(Figure 11) (90,92,93). For example, theneutral HgCI2 species binds specifically tothe primary amine groups on serine andethanolamine polar heads, inducing astrong rigidification of the phospholipidicbilayers because of the formation of bridges

Mercurychemicalspeciationin solution B

Aic

4

Inorganic mercury

so | HgC;}2gX

| HgOHCI HgCI3---..: .

4

Cellular membrane

Figure 11. Chemical speciation diagrams of inorganic mercury and methylmercury in solution: (A) varying pCI witha constant pH of 7.0; (B) varying pH with a constant pCI of 4.0; and (C) fundamental links with metal accessibilityto the cellular membrane interface, binding to proteins and phospholipid polar heads and transport through thehydrophobic bilayer. From Boudou and Ribeyre (10).

by Hg atoms between the adjacent polarheads. These effects were not significantlyinfluenced by the electrical charges at themembrane interface; they differed greatlyfrom results obtained with other di- ortrivalent cations-Ca2 , Cd2 , A13+-andalso with MeHg. Interactions of theorganic compound with the phospholipidicbilayers were essentially based on theCH3Hg+ species (94).

Recent experimental studies on mercurybioaccumulation in unicellular marinediatoms showed that passive uptake ofuncharged chloride chemical species-HgCl2 and CH3HgCl-was the principalbioaccumulation route of both inorganicand organic Hg. Analysis of mercury distri-bution at the subcellular level showed thatinorganic Hg was essentially sequesteredwithin the membranes (91%) because of itsrate of reaction with thiol ligands, whereasthere was a preferential accumulation ofthe methylated form in the cytoplasm.Differences in Hg assimilation by the her-bivorous copepods described aboveappeared to be essentially caused by thesespecificities in metal distribution at thealgal cell level: several zooplanktonicspecies digest the dissolved cytoplasmiccontent of the algae but simply eliminatethe membrane material (80).

Experimental studies into the toxiceffects of MeHg on the nervous system

have shown that this compound can crossthe cell membranes of the blood-brain bar-rier by forming a complex that mimics thestucture of amino acids. Thus, MeHgreacts with the thiol-containing cysteine inplasma; this complex is also formed in thehydrolysis of glutathione. The molecularstructure of the MeHg-cysteine complex isvery similar to that of the large neutralamino acid methionine. MeHg transportacross the membranes thus takes the indi-rect route via the methionine carrier. Inmammals, the fetal brain is more vulnera-ble to the neurotoxic effects of MeHg. Infact, amino acid transport is two or moretimes faster than in the adult, and induceshigh levels of metal bioaccumulation andsevere structural and functional distur-bances (89). Results from Ballatori (95)also show that the glutathione transportsystem may act as an Hg carrier throughthe canalicular membrane in rat liver cells.

ConclusionAquatic ecotoxicology covers a vast anddiverse field of investigation. In most cases,situations encountered in the natural envi-ronment are, for the most part, unique andever changing. Clearly, current researchmethodologies used in this field are ofprimary importance; only a close degreeof complementarity between in situstudies and experimental approaches in the



laboratory will provide a better knowledgeof the mechanisms involved and, at thesame time, enable us to make a betterassessment of the short-, medium- andlong-term risks. A multidisciplinary

approach is likewise essential. Finally, letus emphasize how important it is thatgreater effort be put into fundamentalstudies, with the aim of analyzing theecotoxicological mechanisms involved in

the various stages from the chemical fate ofcontaminants within biotopes, their bio-availability, their uptake across biologicalbarriers, and their ultimate bioaccumulationand toxicity.

REFERENCES

1. Butler GC. Principles of Ecotoxicology. Chichester, England:JWiley & Sons, 1978.

2. Seitz A. The concept of ecological stability applied to aquaticecosystems. In: Freshwater Field Tests for Hazard Assessment ofChemicals (Hill IR, Heimbach F, Leeuwangh P, Matthiessen P,eds). Boca Raton, FL:Lewis Publishers, 1994;3-18.

3. Capblancq J. Special features of lake ecosystems. In: AquaticEcotoxicology: Fundamental Concepts and Methodologies,Vol. 1 (Boudou A, Ribeyre F, eds). Boca Raton, FL:CRC Press,1989;21-34.

4. Capblancq J. Production primaire autotrophe. In: LimnologieGenerale (Pourriot R, Meybeck M, eds). Paris:Masson,1995;228-253.

5. Pourriot R, Meybeck M, eds. Limnologie Generale.Paris:Masson, 1995.

6. Carpenter SR. Complex Interactions in Lake Communities.NewYork:Springer-Verlag, 1988.

7. Shugart HH. Ecological models and the ecotone. In: TheEcology and Management of Aquatic-terrestrial Ecotones(Naiman RJ, Decamps H, eds). Parthenon:United NationsEducational, Scientific, and Cultural Organization, 1990;23-36.

8. Kelly JR, Harwell MA. Indicators of ecosystem response andrecovery. In: Ecotoxicology: Problems and Approaches (LevinSA, Harwell MA, Kelly JR, Kimball KD, eds). NewYork:Springer-Verlag, 1989;9-40.

9. Boudou A, Ribeyre F. Fundamental concepts in aquaticEcotoxicology. In: Aquatic Ecotoxicology: FundamentalConcepts and Methodologies, Vol 1, (Boudou A, Ribeyre F,eds). Boca Raton, FL:CRC Press, 1989;35-75.

10. Boudou A, Ribeyre F. Mercury in the food webs: accumulationand transfer mechanisms. In: Metal Ions in Biological Systems,Vol. 34, Mercury and Its Effects on Environment and Biology(Sigel H, Sigel A, eds). New York:Dekker, 1997;289-319.

11. Tessier A, Turner DR. Metal Speciation and Bioavailability inAquatic Systems. Chichester, England:J Wiley & Sons, 1995.

12. Sheehan PJ. Effects on community and ecosystem structureand dynamics. In: Effects of Pollutants at the Ecosystem Level(Sheehan PJ, Miller DR, Butler GC, Bourdeau P, eds).Chichester, England:J Wiley & Sons, 1984;5 1-101.

13. Moriarty F. Ecotoxicology: the Study of Pollutants inEcosystems. London:Academic Press, 1983.

14. Joern A, Hoagland KD. In defense of whole-community bio-assays for risk assessment. Environ Toxicol Chem 15:407-409(1996).

15. Herricks EE, Schaeffer DJ, Perry JA. Biomonitoring: closingthe loop in the environmental sciences. In: Ecotoxicology:Problems and Approaches (Levin SA, Harwell MA, Kelly JR,Kimball KD, eds). New York:Springer-Verlag, 1989;351-366.

16. Hellawell JM. Biological Indicators of Freshwater Pollutionand Environmental Management. Amsterdam:Elsevier, 1986.

17. Cormier SM, Daniel FB. Biomarkers: taking the science for-ward. Environ Toxicol Chem 13:1011-1012 (1994).

18. Livingstone DR, Forlin L, George SG. Molecular biomarkersand toxic consequences of impact by organic pollution inaquatic organisms. In: Water Quality and Stress Indicators inMarine and Freshwater Systems (Sutcliffe DW, ed). Ambleside,England:Freshwater Biology Association, 1994;1 54-171.

19. Roesijadi G, Robinson WE. Metal regulation in aquatic ani-mals: mechanisms of uptake, accumulation and release. In:Aquatic Toxicology: Molecular, Biochemical and Cellular

Perspectives (Malins DC, Ostrander GK, eds). Boca Raton, FL:Lewis Publishers, 1994;387-420.

20. Garvey JS. Metallothionein: a potential biomonitor of exposureto environmental toxins. In: Biomarkers of EnvironmentalContamination (McCarthy JF, Shugart LR, eds). Chelsea,England:Lewis, 1990;267-287.

21. Rawson DM. Bioprobes and biosensors. In: Handbook ofEcotoxicology (Calow P, ed). London:Blackwell, 1991 ;428-437.

22. Lu FC. Basic Toxicology: Fundamentals, Target Organs andRisk Assessment. Washington:Hemisphere PublishingCorporation, 1985.

23. Boudou A, Delnomdedieu M, Georgescauld D, Ribeyre F,Saouter E. Fundamental roles of biological barriers in mercuryaccumulation and transfer in freshwater ecosystems: analysis atthe organism, organ, cell and molecular level. Water Air SoilPollut 56:807-821 (1991).

24. Hinckle P, Osborne ME. Cadmium toxicity in rat pheochro-mocytoma cells: studies on the mechanism of uptake. ToxicolAppl Pharmacol 124:91-98 (1994).

25. Grouselle M, Boudou A, Oreja-Erroz B. Cadmium uptake in asingle MDCK cell evidenced by fura-2 titration: a fluorescencedigital imaging study. C R Acad Sci Paris 319:277-287 (1996).

26. Bartell SM, Gardner RH, O'Neill RV. Ecological RiskEstimation. Michigan: Lewis, 1992.

27. Pascoe D, Edwards RW. Single species toxicity tests. In: AquaticEcotoxicology: Fundamental Concepts and Methodologies,Vol. 2 (Boudlou A, Ribeyre F, eds). Boca Raton, FL:CRC Press,1989;93-126.

28. Luoma SN. Prediction of metal toxicity in nature from bioas-says: limitations and research needs. In: Metal Speciation andBioavailability in Aquatic Systems (Tessier A, Turner DR, eds).Chichester, England:Elsevier, 1995;609-661.

29. Persoone G, Janssen CR. Field validation of predictions basedon laboratory toxicity tests. In: Freshwater Field Tests forHazard Assessment of Chemicals (Hill IR, Heimbach F,Leeuwangh P, Matthiessen P, eds). Boca Raton, FL:LewisPublishers, 1994;379-397.

30. Giesy JP, Graney RL. Recent developments in and intercom-parisons of acute and chronic bioassays and bioindicators.Hydrobiology 188:21-60 (1989).

31. Cairns, JJ. What constitutes field validation of predictionsbased on laboratory evidence? In: Aquatic Toxicology andHazard Assessment (Adams WJ, Chapman GA, Landis WG,eds). Philadelphia:American Society for Testing and Materials,1988;361-368.

32. Chapman PM. Do sediment toxicity tests require field valida-tion? Environ Toxicol Chem 14:1451-1453 (1995).

33. Gearing JN. The role of aquatic microcosms in ecotoxicologicresearch as illustrated by large marine systems. In:Ecotoxicology: Problems and Approaches (Levin SA, HarwellMA, Kelly JR, Kimball KD, eds). New York:Springer-Verlag,1989;411-470.

34. Crossland NO, Heimbach F, Hill IR, Boudou A, LeeuwanghP, Matthiessen P, Persoone G. Summary and recommendationsof the EWOFFT. In: Freshwater Field Tests for HazardAssessment of Chemicals (Hill IR, Heimbach F, Leeuwangh P,Matthiessen P, eds). Boca Raton, FL:Lewis Publishers,1994;15-27.

35. Brock TCM, Budde BJ. On the choice of structural parametersand endpoints to indicate responses of freshwater ecosystems to

Environmental Health Perspectives - Vol 105, Supplement * February 1997 33

BOUDOU AND RIBEYRE

pesticide stress. In: Freshwater Field Tests for HazardAssessment of Chemicals (Hill IR, Heimbach F, Leeuwangh P,Matthiessen P, eds). Boca Raton, FL:Lewis Publishers,1994;19-56.

36. Odum EP. The mesocosm. Bioscience 34:558-562 (1984).37. Giesy JP, Odum EP. Microcosmology: introductory com-

ments. In: Microcosms in Ecological research (Giesy JP, ed).Department of Energy Symposium Series 52, Conference781101. Springfield, VA:National Technical InformationService, 1980;1-13.

38. Giddings JM, Rodgers JH. Workshop on Aquatic Microcosmsfor Ecological Assessment of Pesticides. Pensacola, FL:Societyfor Environmental Toxicology and Chemistry, 1992.

39. Touart LW. Aquatic Mesocosm Test to Support PesticideRegistrations. EPA 540/09-88-035. Washington:U.S.Environmental Protection Agency, 1988.

40. SETAC-RESOLVE. Proceedings of a Workshop on AquaticMicrocosms for Ecological Assessment of Pesticides. Society forEnvironmental Toxicology and Chemistry Foundation forEnvironmental Education and Resolve Program (WWF), 1992.

41. Hill IR, Heimbach F, Leeuwangh P, Matthiessen P. FreshwaterField Tests for Hazard Assessment of Chemicals. Boca Raton,FL:Lewis Publishers, 1994.

42. U.S. EPA. Questions and Answers: Improvements to EPA'sProgram to Prevent Adverse Environmental Effects ofPesticides. EPA-H7506. Washington:U.S. EnvironmentalProtection Agency, 1993.

43. Metcalf RL, Sangha GK, Kapoor IP. Model ecosystem for eval-uation of pesticide biodegradability and ecological magnifica-tion. Environ Sci Technol 5:709-713 (1971).

44. Taub F. Standardized aquatic microcosm: development andtesting. In: Aquatic Ecotoxicology: Fundamental Concepts andMethodologies, Vol. 1 (BouTou A, Ribeyre F, eds). BocaRaton, FL:CRC Press, 1989;47-92.

45. Donkin P. Quantitative structure-activity relationships. In:Handbook of Ecotoxicology (Calow P, ed). London:Blackwell,1991; 321-347.

46. Hudson RJM, Gherini SA, Watras CJ, Porcella DB. Modelingthe biogeochemical cycle of mercury in lakes: the MercuryCycling Model (MCM) and its application to the MTL studylakes. In: Mercury Pollution: Integration and Synthesis (WatrasCJ, Huckabee JW, eds). Boca Raton, FL:Lewis Publishers,1994; 473-523.

47. Andren AW, Nriagu JO. The global cycle of mercury. In: TheBiogeochemistry of Mercury in the Environment (Nriagu JO,ed). Amsterdam:Elsevier, 1979;1-21.

48. Lindqvist 0, Johansson K, Aastrup M, Andersson A,Bringmark L, Hovsenius G, Hakanson L, Iverfeldt A, Meili M,Timm B. Mercury in the Swedish Environment.Dordrecht:Kluwer, 1991.

49. Hudson RJM, Gherini SA, Fitzgerald WF, Porcella DB.Anthropogenic influences on the global mercury cycle: amodel-based analysis. Water Air Soil Pollut 80:265-272(1995).

50. Fitzgerald WF. Is mercury increasing in the atmosphere? Theneed for an atmospheric mercury network (AMNET).WaterAir Soil Pollut 80:245-254 (1995).

51. Mason RP, Fitzgerald WF. The distribution and biogeochemi-cal cycling of mercury in the equatorial Pacific Ocean. DeepSea Res 40:1897-1924 (1993).

52. Nriagu JO. Production and uses of mercury. In: TheBiogeochemistry of Mercury in the Environment (Nriagu JO,ed). Amsterdam:Elsevier, 1979;23-40

53. Ellis D. Environments at Risk. Berlin:Springer-Verlag, 1989.54. Chang LW. Pathological effects of mercury poisoning. In: The

Biogeochemistry of Mercury in the Environment (Nriagu JO,ed). Amsterdam:Elsevier, 1979;519-580.

55. Bloom NS, Horvat M, Watras CJ. Results of the internationalaqueous mercury speciation intercomparison exercise. WaterAir Soil Pollut 80:1257-1268 (1995).

56. Porcella DB. Mercury in the environment: biogeochemistry.In: Mercury Pollution: Integration and Synthesis (Watras CJ,

Huckabee JW, eds). Boca Raton, FL:Lewis Publishers,1994;3-19.

57. Bloom NS. Determination of picogram levels of methylmer-cury by aqueous phase ethylation, followed by cryogenic gaschromatography with cold vapor atomic fluorescence detection.Can J Fish Aquat Sci 46:1131-1140 (1989).

58. Watras CJ, Bloom NS, Hudson RJM, Gherini S, Munson R,Claas SA, Morrison KA, Hurley J, Wiener JG, Fitzgerald WF,et al. Sources and fates of mercury and methylmercury inWisconsin lakes. In: Mercury Pollution: Integration andSynthesis (Watras CJ, Huckabee JW, eds). Boca Raton,FL:Lewis Publishers, 1994;1 53-177.

59. Wiener JG, Fitzgerald WF, Watras CJ, Rada RG. Partitioningand bioavailability of mercury in an experimentally acidifiedWisconsin lake. Environ Toxicol Chem 9:909-918 (1990).

60. Watras CJ, Bloom NS. Mercury and methylmercury in indi-vidual zooplankton: implications for bioaccumulation. LimnolOceanogr 37:1313-1318 (1992).

61. Spry DJ, Wiener JG. Metal bioavailability and toxicity to fishin low-alkalinity lakes: a critical review. Environ Pollut71:243-304 (1991).

62. Wren CD, Scheider WA, Wales DL, Muncaster BW, Gray IM.Relation between mercury concentrations in walleye and north-ern pike in Ontario lakes and influence of environmental fac-tors. Can J Fish Aquat Sci 48:132-139 (1991).

63. Nelson ON, Campbell PGC. The effects of acidification on thegeochemistry of Al, Cd, Pb and Hg in freshwater environ-ments: a literature review. Environ Pollut 71:91-131 (1991).

64. Meili M. Aqueous and biotic mercury concentrations in boreallakes: model predictions and observations. In: MercuryPollution: Integration and Synthesis (Watras CJ, HuckabeeJW, eds). Boca Raton, FL:Lewis Publishers, 1994;99-106.

65. Richardson M, Egyed E, Currie DJ. Human exposure to mer-cury may decrease as acidic deposition increases. Water Air SoilPollut 80:31-39 (1995).

66. Winfrey MR, Rudd JWM. Environmental factors affecting theformation of methylmercury in low pH lakes. Environ ToxicolChem 9:853-869 (1990).

67. Fitzgerald WF, Mason RP, Vabdal GM. Atmospheric cyclingand air-water exchange of mercury over mid-continental lacus-trine regions. Water Air Soil Pollut 56:745-751 (1991).

68. Xun L, Campbell NER, Rudd JWM. Measurements of specificrates of net methylmercury production in the water columnand surface sediments of acidified and circumneutral lakes. CanJ Fish Aquat Sci 44:750-757 (1987).

69. Campbell PGC, Lewis AG, Chapman PM, Crowder AA,Fletcher WK, Imber B, Luoma SN, Stokes PM, Winfrey M.Biologically Available Metals in Sediments. Ottawa, Canada:National Research Council of Canada, 1988.

70. Verta M, Matilianen T. Methylmercury distribution and parti-tion in stratified Finnish forest lakes. Water Air Soil Pollut80:585-588 (1995).

71. Gilmour CC, Henry EA, Mitchell R. Sulfate stimulation ofmercury methylation in freshwater sediments. Environ SciTechnol 26:2281-2287 (1992).

72. Rudd JWM. Sources of methylmercury to freshwater ecosys-tems: a review. Water Air Soil Pollut 80:697-713 (1995).

73. Hare L. Aquatic insects and trace metals: bioavailability, bioac-cumulation and toxicity. Crit Rev Toxicol 22:327-369 (1992).

74. Weber JH. Review of possible paths for abiotic methylation ofmercury(II) in the aquatic environment. Chemosphere26:2063-2077 (1993).

75. Pelletier E. Environmental organometallic chemistry of mer-cury, tin and lead: present status and perspectives. In: MetalSpeciation and Bioavailability in Aquatic Systems (Tessier A,Turner DR, eds). Chichester, England:Elsevier, 1995;103-149.

76. Verdon R, Brouard D, Demers C, Lalumiere R, Laperle M,Schetagne R. Mercury evolution in fishes of the La Grandehydroelectric complex, Canada. Water Air Soil Pollut56:405-412 (1991).

77. Wiener JG, Spry DJ. Toxicological significance of mercury infreshwater fish. In: Environmental Contaminants in Wildlife:

34 Environmental Health Perspectives * Vol 105, Supplement 1 * February 1997


Interpreting Concentrations of Environmental Contaminantsin Wildlife Tissues (Heinz G, Beyer N, eds) Chelsea:Lewis,1994;42-71.

78. Ribeyre F, Boudou A. Study of the dynamics of the accumula-tion of two mercury compounds (HgCI2 and CH3HgCl) byChlorella vulgaris effects of temperature and pH. Int J EnvironStud 20:35-40 (1982).

79. Boudou A, Ribeyre F. Comparative study of the trophic transferof two mercury compounds (HgCI2 and CH3HgCl) betweenChlorella vulgaris and Daphnia magna Influence of temperature.Bull Environ Contam Toxicol 27:624-629 (1981).

80. Mason RP, Reinfelder JR, Morel FMM. Bioaccumulation ofmercury and methylmercury. Water Air Soil Pollut80:915-921 (1995).

81. Boudou A, Delarche A, Ribeyre F, Marty R. Bioaccumulationand bioamplification of mercury compounds in a second levelconsumer, Gambusia affinis-temperature effects. Bull EnvironContam Toxicol 22:813-818 (1979).

82. Boudou A, Ribeyre F. Processes of contamination of aquaticbiocenoses by mercury compounds. In: Aquatic Toxicology(Nriagu JO, ed). New York:J Wiley & Sons, 1983;74-98.

83. Boudou A, Ribeyre F. Fish as biological models in aquatic eco-toxicology. In: Aquatic Ecotoxicology: Fundamental Conceptsand Methodologies, Vol. II (Boudou A, Ribeyre F, eds). BocaRaton, FL:CRC Press, 1989;127-162.

84. Ribeyre F, Delarche A, Boudou A. Transfer of CH3HgCl in anexperimental freshwater trophic chain. Environ Pollut1:259-268 (1980).

85. Ribeyre F, Boudou A. Bioaccumulation et repartition tissulairedu mercure (HgCl2 et CH3HgC3 ) chez Salmo gairdneri aprescontamination par voie directe. Water Air Soil PoIlut23:169-186 (1984).

86. Ribeyre F, Boudou A. Etude experimentale des processus dedecontamination chez Salmo gairdneri apres contamination parvoie directe avec deux derives du mercure (HgCl2 etCH3HgCl)-Analyse des transferts aux niveaux organisme etorgane. Environ Pollut 35:203-228 (1984).

87. Boudou A, Ribeyre F. Experimental study of trophic contami-nation of Salmo gairdneri by two mercury compounds (HgCI2and CH3HgCl)-analysis at the organism and organ levels.Water Air Soil Pollut 26:137-148 (1985).

88. Simkiss K, Taylor MG. Transport of metals across membranes.In: Metal Speciation and Bioavailability in Aquatic Systems(Tessier A, Turner DR, eds). Chichester, England:Elsevier,1995;2-44.

89. Clarkson TW. The toxicology of mercury and its compounds.In: Mercury Pollution: Integration and Synthesis (Watras CJ,Huckabee JW, eds). Boca Raton, FL:Lewis Publishers,1994;631-642.

90. Bienvenue E, Boudou A, Desmazes JP, Gavach C,Georgescauld D, Sandeaux J, Seta P. Transport of mercurycompounds across bimolecular lipid membranes: effect of lipidcomposition, pH and chloride concentration. Chem BiolInteract 48:91-101 (1991).

91. Faust BC. The octanol/water distribution coefficients ofmethylmercury species: the role of aqueous-phase chemical spe-ciation. Environ Toxicol Chem 11:1373-1376 (1991).

92. Delnomdedieu M, Boudou A, Desmazes JP, Georgescauld D.Interaction of mercury chloride with the primary amine groupof model membranes containing phosphatidylserine and phos-phatidylethanolamine. Biochim Biophys Acta 986:191-199(1989).

93. Delnomdedieu M, Boudou A, Georgescauld D, Dufourc EJ.Specific interactions of mercury chloride with membranes andother ligands as revealed by Hg-NMR. Chem Biol Interact81:243-269 (1992).

94. Girault L, Lemaire P, Boudou A, Dufourc EJ. Inorganic mer-cury interactions with lipid components of biological mem-branes: 31P-NMR study of HgII binding to headgroups ofmicellar phospholipids. Water Air Soil Pollut 80:95-98(1995).

95. Ballatori N. Mechanisms of metal transport across liver cellplasma membranes. Drug Metab Rev 23:83-95 (1991).

Environmental Health Perspectives * Vol 105, Supplement I * February 1997 35

aquatic ecotoxicology: from the ecosystem to ... · system. ecosystems are not under steady-state...

Documents