Research Determining Genetic Variability in the Distribution of Sensitivities to Toxic Stress among and within Field Populations of Daphnia magna C A R L O S B A R A T A , * ,†,‡ DONALD J. BAIRD,‡ AND AMADEU M. V. M. SOARES† Departamento de Biologia, Universidade de Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal, and Institute of Aquaculture, University of Stirling, Stirling FK9 4LA, United Kingdom
To extrapolate credibly from individuals in the laboratory to field populations, it is essential to account for genetic differences in susceptibility to toxic stress and thus incorporate genetic variability into ecological risk estimates. In this study, the distribution of sensitivities across two toxic chemicals among and within field populations of Daphnia magna were used to quantify genetic variability. The study employed 30 D. magna clones from three geographically separate European populations. The sensitivity of each population studied and its constituent clones was estimated in terms of the concentrations of λ-cyhalothrin and cadmium impairing individual fitness by 10 and 50% (EC10-50). Results revealed that differences in tolerance among clones within populations were large when compared with differences between populations and that the genetic range of sensitivities to toxic stress within populations was log-normally distributed. Furthermore, reported variation in sensitivity values to toxic stress among different laboratory species, populations, and clones was similar to that observed among and within field populations of Daphnia. These results suggest that it is possible to estimate genetic variability by estimating the tolerance distribution of laboratory populations and clones and that extrapolation approaches currently used in ecological risk assessment should explicitly incorporate genetic variability in tolerance into risk estimates.
Introduction Quantifying uncertainty in susceptibility to toxic stress is an important aspect of environmental risk assessment, and genetic factors can play an important role (1-3). The occurrence of differences in tolerance among populations and among clones or morphotypes within populations is a well-known phenomenon (3-10). Such variability in toxic responses complicates attempts to extrapolate effects of toxicants from standardized toxicity tests to natural systems. * Corresponding author phone: (34) 93 400 61 00; fax: (34)93 204 59 04; e-mail:
[email protected]. Present address: Department of Environmental, Chemistry, IIQAB-CSIC, Jordi Girona 18-26, 08034 Barcelona, Spain. † Universidade de Aveiro. ‡ University of Stirling. 10.1021/es0158556 CCC: $22.00 Published on Web 06/11/2002
2002 American Chemical Society
A major concern in ecotoxicology is that toxicity testing is done without clear understanding of how selecting laboratory populations with reduced genetic diversity can bias test results (11). Indeed, standardized toxicity tests performed with laboratory populations may underprotect certain populations, since loss of sensitive genotypes may not be accounted for (1, 11). Therefore, to extrapolate credibly from individuals in the laboratory to field populations, it is essential to account for genetic differences in susceptibility to toxic stress and incorporate this genetic variability into risk estimates. It has been suggested that the distribution of sensitivities to toxic stress among and within natural populations could be used to estimate genetic variability (2, 3). Thus it is important to determine the extent and nature of sensitivities in natural populations. For example, if differences in tolerance among genotypes are large compared with those of populations, and genotype sensitivities are normally distributed, genetic variability could be estimated from the variance of the tolerance distribution of a random sample of genotypes. Hence toxic responses of few laboratory populations could be used to predict environmental risks in natural populations. Alternatively if the range of sensitivities within and among populations are large and are not normally distributed, genetic variability would vary among populations. As a result toxic responses based on a limited number of laboratory populations could not be used to predict risks. Assessing the susceptibility range of sensitivities of natural populations is not an easy task. Wild populations of laboratory-cultured species are often subject to strong selection when brought into the lab, and thus, the genetic variation present in lab cultures of a species may be a poor representation of that of the source population (5, 11). Nevertheless in a recent study we showed that it was possible to gauge population responses of outbred populations by using ephippial eggs of the freshwater zooplanktonic cladoceran, Daphnia magna (9). Ephippial eggs are sexually produced dormant eggs, which can represent the complete gene pool of the population in certain parts of the year. These resting eggs can be sampled and hatched in the laboratory to study population responses of outbred populations. Furthermore, in the laboratory D. magna usually reproduces asexually by ameiotic parthenogenesis. Thus, single genotypes (hereafter referred as clones) can be tested across a range of toxic concentrations. In this study, reproductive responses of ephippial clones from three different field populations of D. magna to the toxicants λ-cyhalothrin and cadmium were used to assess the distribution of tolerances within and among populations. This allowed us to compare the distribution of sensitivities of two differently acting chemicals with strong sublethal effects on D. magna and with different degrees of environmental persistence. The pyrethroid insecticide λ-cyhalothrin is a sodium channel blocker and aminobutyric (GABA) inhibitor currently used to control arthropod pests (12) and also impairs feeding in Daphnia (13). λ-cyhalothrin is applied at low rates, is very insoluble in water, and is therefore expected to occur at very low concentrations in the aquatic environment (14). In contrast, cadmium is an abundant nonessential metal which accumulates in the aquatic environment as a result of industrial practices (15). Cadmium toxicity to organisms is related with the production of reactive oxygen species that disrupt physiological processes (16). In Daphnia, cadmium impairs feeding and hence inhibits growth and reproduction (17). Reproductive reVOL. 36, NO. 14, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3045
sponses to sublethal toxic exposure levels were selected to assess tolerance to toxic stress in D. magna for two reasons: D. magna reproduction tests are among the few sensitive ecotoxicological tests that provide the basis for legislative decisions (18), and in the field toxic chemicals usually occur at low concentrations, thus sublethal responses to low exposure levels of toxic chemicals are more likely to govern population responses than lethal responses measured at high toxic exposure levels (17, 19).
Material and Methods Study Populations and Culture Conditions. Clonal lineages were established from D. magna populations from three habitats located in northern Germany and southern Spain. German clones originated from two populations near Plo¨n: a small fishless eutrophic pond from Rixdorfer Pferdertra¨nke (20) and a large eutrophic semipermanent pond from Lebrader Schleswig-Holstein used to rear fish (21). Spanish clones originated from a small fishless eutrophic pond located in San Fernando, Ca´diz. All populations studied were located in pristine environments with no known history of pollution although their habitats differed in climate, predation regime, and husbandry. Therefore, these populations could be considered as a representative of unpolluted European habitats. All experimental clones were hatched from ephippia isolated from sediment samples collected at the start of the growing season for each habitat. Ephippia were hatched in the laboratory at various temperatures and photoperiod regimes to ensure a nonselective regime. All hatchlings were isolated to initiate clonal lineages, with a success rate of 95%, and each female was followed until it released the second or third brood. From each of these females, two neonates were isolated in separate vessels as replicates of the clonal line. After five generations a total of 10 clonal lines from Rixdorfer, Ca´diz, and Lebrader were randomly chosen for inclusion in the experiments. It was assumed that each hatchling was genetically distinct as a result of sexual reproduction (9). Subsequent allozyme analysis by standard cellulose acetate allozyme electrophoresis (22) at five different polymorphic loci (MDH, GOT, PGI, MPI, PGM) confirmed that in each population from 6 to 7 out of 10 clones were electrophoretically distinct. Single Daphnia females were maintained in 120 mL of ASTM hard water (23), in 150 mL screw-top glass, with the addition of a standard organic extract (24). Animals were fed every other day with Chlorella vulgaris Beijerinck (5 × 105 cells/mL, corresponding to 1.8 µg C/mL). Photoperiod was maintained to a 14:10 h light:dark cycle and temperature to 20 ( 1 °C.
Experimental Section Design Tolerance to toxic stress was determined from reproduction responses of D. magna individuals exposed to four concentrations of λ-cyhalothrin (0, 0.062, 0.125, 0.25 µg/L) and cadmium (0, 0.5, 1, 2 µg/L). Reproduction responses were assessed using a shortened version of the Daphnia reproduction test (9), in which animals were held individually from birth until first reproduction at which point the experiment was terminated. During the assays, reproduction (age and size of first brood) of experimental animals was assessed at 8 h intervals. Experiments were started with < 8 h neonates, originated from second or third brood females previously acclimated to the laboratory environment for at least five generations. Test solutions and algae were changed every other day. Responses to λ-cyhalothrin and cadmium were determined in two separate experiments. In each experiment, 10 clones from each of the three populations studied, hereafter designated as Rixdorfer, Ca´diz, and Lebrader, were exposed 3046
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 14, 2002
to each chemical simultaneously to assess differences in tolerance among clones and populations. For each concentration and clone three individual replicates were performed. Replicates were randomized, and their location was changed every other day after feeding. To better control aqueous physicochemical parameters and hence to minimize differences in toxicity across experiments, D. magna exposures were conducted in artificial hard ASTM water of known composition (23, 25). Water Chemistry. Saturated aqueous stock solutions of λ-cyhalothrin (Labor Dr Ehrenstorfer-Scha¨fers, Augsburg, Germany; 98.6% purity) were prepared following the method described by Barata and Baird (17). Prior to the preparation of λ-cyhalothrin test concentrations, the stock solution was filtered through a glass microfiber filter (0.2 µm) to obtain only the maximum soluble fraction of the chemical. In water at 20 °C, this amount corresponds to 3 µg L-1 (26). Stock solutions of cadmium (expressed as total ion concentrations) were prepared by adding analytical reagent-grade (BDH Anar grade, BDH, Poole, UK) salt (3CdSO4‚8H2O) to deionized water (Milli-Q, Bredford, MA, U.S.A.; 18 MΩ/cm resistibility). Nominal test concentrations of λ-cyhalothrin and cadmium were prepared by adding appropriate aliquots of stock solutions to ASTM hard water. Stocks were kept in darkness at 4 °C during the experiments. Duplicated water samples of freshly prepared and old (2 days) test solutions were collected at the beginning and end of the tests to measure toxicant concentrations, oxygen levels, and pH. Analysis of λ-cyhalothrin was restricted to the two highest concentrations (0.125, 0.25 µg L-1). λ-Cyhalothrin test solutions were extracted from pooled water samples of 400 mL and determined by HPLC using the method described by Barata and Baird (17). Water samples were pretreated with propan-2-ol and nitric acid and then λ-cyhalothrin test solutions extracted from water and preconcentrated in a solid-phase, Bond Elut C18 extraction column (Varian, Phenomenex, Macclesfield, UK). λ-Cyhalothrin was eluted from the solid-phase cartridges with a mixture of acetonitrile: toluene of 3:1.5. The obtained solutions were then evaporated, and the residue was resuspended in 2 mL of mobile phase. A final sample extract of 100 µL was injected onto a SUPELCOSIL 59142 LC-ABZ 5 µm 2.5 cm × 4.6 mm column (SUPELCO Inc. Bellefonte, PA). A mobil phase of acetonitrile: water 75:25, a flow rate of 1.0 mL/min, and wavelengths of 210 nm were set to analyze λ-cyhalothrin. Recovery of λ-cyhalothrin from water samples was determined from duplicated standard solutions of known concentration, which were extracted and eluted at the same time as test solutions. Mean ( 1 SD retention time and recovery (%) rates from water samples were 9.58 ( 0.01 min and 75 ( 5%. Effective total ionic concentrations of cadmium were measured from 10 mL water samples using an ATI Unicam Model 939graphite furnace atomic absorption spectrophotometer following the procedure described by Barata et al. (25). Data Analysis. For D. magna the criterion normally used to assess tolerance in reproduction tests is a significant drop in mean fecundity relative to unstressed controls (i.e. NOEC, EC10-50 (17, 18)). Here clonal and population sensitivities to stress (EC10-50) were derived from concentration-individual fitness response curves. For each individual, fitness was estimated as rc ) loge (clutch size)/(time to first reproduction). We use individual fitness, characterized as rc, instead of reproductive output, since fitness integrates toxicant effects on two key demographic parameters: clutch size and age at first reproduction (24). Individual fitness responses varied linearly with log transformed cadmium and λ-cyhalothrin concentrations (9, 10). This allowed effect concentration values (EC) to be determined by fitting individual fitness responses to a linear regression model. Prior to fitting the data to the regression model, toxicant concentrations (X)
were loge (X+1) transformed to include control treatments (10, 17). Model accuracy was assessed by using adjusted coefficients of determination r2 and by analyzing residual distribution. Significance of the entire regression and regression coefficients were determined by F and t tests, respectively. EC values and confidence intervals (95% C.I.) were then estimated by inverse prediction of individual fitness responses (27). To test whether the range of sensitivities within populations was normally distributed, clonal EC10-50 values obtained within each of the studied population and toxicant were tested for normality using a Shapiro-Wilks tests (27). To increase the sample size and hence the statistical power of normality tests (27), Shapiro-Wilks tests were repeated considering data from the three populations.
Results Chemical Analysis. Measured λ-cyhalothrin concentrations (mean ( SD) of the two highest freshly prepared solutions (0.076 ( 0.006, 0.145 ( 0.01 µg L-1) were 40% lower than expected (0.125, 0.25 µg L-1) and decreased again by 50% after 48 h (0.037 ( 0.003, 0.079 ( 0.01µg L-1). Thus, results reported for λ-cyhalothrin were based on the interpolated midpoint value between initial and final measured concentrations. Actual concentrations of freshly prepared and 48 h-old test solutions of cadmium were within 10% of nominal concentrations. Thus, responses to cadmium were based on nominal concentrations. Oxygen levels in old test solutions were within 91% of saturation levels. Mean pH values were 8.3, and differences between initial and final values never exceeded 0.2 pH units. Responses to λ-Cyhalothrin and Cadmium. Individual fitness responses of the studied populations and clones across λ-cyhalothrin and cadmium concentrations are depicted in Figure 1. Individual fitness responses were dramatically affected by the studied λ-cyhalothrin concentrations (Figure 1A). Indeed, within the three studied populations only a small number of individuals from a few clones survived to reproduce at the highest λ-cyhalothrin concentration (Figure 1A). For cadmium, however, the studied concentrations had moderate effects on fitness (Figure 1B). Regression analyses denoted that individual fitness responses varied linearly across logarithm transformed λ-cyhalothrin and cadmium concentrations. Under exposure to λ-cyhalothrin, except in one clone from Ca´diz, the concentration-individual fitness regression models obtained were significant (P < 0.05) and on average explained more than 80% of variability (r2 ) 0.88). Under cadmium exposure, except in three clones (one from each population), the regression equations obtained were significant (P < 0.05) and on average explained more than 60% of variability (r2 ) 0.65). For each studied population and clone, predicted concentrations of λ-cyhalothrin and cadmium impairing individual fitness 10 and 50% (EC10-50) were derived from the concentration-response curves given in Figure 1. The studied concentrations of cadmium only caused small impairments on fitness responses ( 0.05, Shapiro-Wilks tests). Only within Rixdorfer and Ca´diz populations, the existence
FIGURE 1. Concentration-individual fitness response curves of the studied populations (top panels) and clones (bottom panels) of D. magna under exposure to λ-cyhalothrin (A) and cadmium (B). Symbols and error bars represent means and SD values. For the sake of clarity, error bars are only depicted for population responses (top panels) and within each concentration; the mean values of the studied populations are shifted slightly to one side. Toxicant concentrations are depicted in log scale. of few clones with high levels of tolerance to cadmium (one clone in each population was at least 10 times more tolerant than average, Figure 2B) prevented mean clonal sensitivities (EC10) to follow a normal distribution. Thus, when those clones were removed from the analysis, all tests including those performed on pooled data from the three populations denoted normality, and within all populations the range of clonal sensitivities was within a 6-fold range.
Discussion The results reported here showed that differences in tolerance to toxicants among clones were large compared with that of populations. This means that a strong potential exists to evolve resistance to toxic stress within these natural populations, and, as a result, genetic variability in susceptibility to toxic exposure should be included in environmental risk assessment. Moreover, our results also indicate that despite the presence of a few clones with high levels of tolerance to cadmium, clonal sensitivities to the two studied chemicals tend to be log-normally distributed. Indeed, within most of the studied field populations differences in tolerance between the most and least sensitive clones were e 6-fold. This means that genetic variability could be estimated from the variance of the tolerance distribution of a random sample of clones. Hence toxic responses of laboratory clones or populations may be used to predict environmental risks in natural populations. To recommend the use of laboratory populations to estimate genetic variability, it would be necessary to compare VOL. 36, NO. 14, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3047
FIGURE 2. Sensitivities to λ-cyhalothrin (A) and cadmium (B) of clones within populations (left panels) and among the three studied populations (right panels). Sensitivities are depicted as predicted concentrations of λ-cyhalothrin and cadmium impairing individual fitness 10 and 50% (EC10-50). Symbols and error bars represent mean and 95% CI values. In same clones and populations, error bars are lower than symbol size. EC10-50 values are depicted in log scale. the range of sensitivities of natural populations with those reported in laboratory populations. Studies conducted with laboratory populations and clones have also reported relative small differences in tolerance to sublethal toxic stress levels (3). Barata et al. (10) studying feeding responses of laboratory D. magna clones to copper, cadmium, and fluoranthene reported differences in EC10 and EC50 of about 4-fold between tolerant and sensitive clones. Long-term reproduction D. magna toxicity studies performed with cadmium, DCA, sodium bromide, and parathion reported that in all but one of the studied chemicals the range of sensitivities (i.e. LOEC) among laboratory clones spanned over 4-fold, the exception reported 6-fold differences between extreme genotypes (4, 6, 10). Similar results have been reported for other population species. In the gastropod Potamopyrgus atipodarum, the collembolan Folsomia candida, the crustacean species Artemia, and the duckweed Lemna gibba, differences in chronic tolerance to toxic substances (i.e. LOEC, EC10-50) among clones or populations varied from 2 to 4-fold (7, 8, 29, 30). Therefore, it is reasonable to assume that the magnitude of variation in sensitivity values to toxic stress between laboratory and field populations and across species is similar. Current procedures of environmental risk assessment use extrapolation models to account for a variety of uncertainty factors, which may result in laboratory test results leading to underprotective environmental quality standards. When only a limited number of toxicity tests are available, concentration effects are divided by an application factor that varies from 10 to 1000 depending on the number of species tested and the endpoint used. In situations where 3048
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 14, 2002
Daphnia reproduction tests form the basis to derive environmental hazards (i.e. PNEC), a safety margin of division of 10 is recommended (1). When a large number of tests is available, distribution-based extrapolation models are used to estimate environmental risks. Derived concentration effects (i.e. NOEC) are fitted to a statistical frequency distribution, and the parameters of the distribution are used to estimate the toxicant concentration that protects most species (i.e. 95%). This concentration is considered to be protective with respect to the ecosystem. Both the application factor approach and the distribution-based extrapolation models have been criticized for not considering genetic variability among other factors (1-3). According to Forbes (2) the range and distribution of sensitivities in field populations could be used to quantify the extent to which genetic variability should be incorporated into risk estimates. In the present study we showed that in 7 out of 9 tolerance distributions examined, the genetic range of sensitivities to toxic stress among D. magna clones did not exceed 6-fold, the exceptions showed only 10-fold differences in tolerance. Similar ranges of variation in sensitivity among clones or populations have been subsequently obtained in other studies performed with different Daphnia clones (3, 4, 6, 9, 10) and species (7, 8, 29, 30). This means that safety factors g 10 should be included into risk estimates to account for genetic variability. Thus, current application factors (10 for Daphnia reproduction tests) for prediction of no-effect concentration from test data should be increased (1), and distribution-based extrapolation models should include sensitivity distributions within and among populations of the studied species (2). By using a random and representative sample of field populations and clones of D. magna and two different toxic chemicals, our results showed that the nature and magnitude of variation in susceptibility to toxic stress of field and laboratory populations was similar, indicating that genetic uncertainty could be estimated from the tolerance distribution of laboratory populations. This means that ecological risk estimates derived from toxic responses of laboratory populations are likely to be representative of field populations. Overall the results obtained in this study and those reported by Forbes et al. (19) suggest that extrapolation approaches currently used in ecological risk assessment need to incorporate safety margins of at least an order of magnitude. Nevertheless it is important to notice that the results and conclusions obtained here are limited to populations of Daphnia originating from relatively pristine environments. The genetic range of tolerances among and within populations previously exposed to toxic stress is likely to be large and non-normally distributed due to the presence of individuals with high levels of resistance (31). This means that studies of this kind on a range of other species and populations differing in life-cycle and origin are needed to be certain that current ecological risk assessment procedures protect field populations.
Acknowledgments Carlos Barata was supported by a Portuguese Fundac¸ a˜o para a Cieˆncia e Tecnologia grant PRAXIS XXI/1629/98 while carrying out this research.
Literature Cited (1) Baird, D. J.; Barata, C. Arch Toxicol. Suppl. 1997, 20, 399406. (2) Forbes, V. E. In Genetics in Ecotoxicology; Forbes, V. E., Ed.; Taylor & Francis: Ann Arbor, MI, 1998. (3) Baird, D. J.; Barata, C. In Genetics in Ecotoxicology; Forbes, V. E., Ed.; Taylor & Francis: Ann Arbor, MI, 1998. (4) Baird, D. J.; Barber, I.; Calow, P. Funct. Ecol. 1990, 4, 399-407. (5) Baird, D. J.; Barber, I.; Bradley, M.; Soares, A. M. V. M.; Calow, P. Ecotox. Environ. Safe 1991, 21, 257-265.
(6) Soares, A. M. V. M.; Baird, D. J.; Calow, P. Environ. Toxicol. Chem. 1992, 11, 1477-1483. (7) Crommentuijn, T.; Stab, J. A.; Doornekamp. A.; Estoppey, O.; Van Gestel, C. A. M. Funct. Ecol. 1995, 9, 734-742. (8) Mazzeo, N.; Dardano, B.; Marticorena, A. Ecotoxicology 1998, 7, 151-160. (9) Barata, C.; Baird, D. J.; Amat, F.; Soares, A. M. V. M. Funct. Ecol. 2000, 14, 513-523. (10) Barata, C.; Baird, D. J.; Min ˜ arro, A.; Soares, A. M. V. M. Environ. Toxicol. Chem. 2000, 19, 2314-2322. (11) Baird, D. J. Funct. Ecol. 1992, 6, 616-619. (12) Hill, I. R.; Shaw, J. L.; Maund, S. J. In Freshwater Field tests for Hazard Assessments of Chemicals; Hill, I. R., Heimbach, F., Leeuwangh, P., Matthiessen, P., Eds.; Lewis Publishers: CRC Press: London, 1994. (13) McWilliam, R. A.; Baird, D. J. Environ. Toxicol. Chem. 2002, 21, 1198-1205. (14) Maund, S. J.; Hamer, M. J.; Warinton, J. S.; Kedwards, T. J. Pestic. Sci. 1998, 54, 408-417. (15) Hare, L. Crit. Rev. Toxicol. 1992, 22, 327-369. (16) Stohs, S. J.; Bagghi, D. Free Rad. Biol. Med. 1995, 18, 321-336. (17) Barata, C.; Baird, D. J. Aquat. Toxicol. 2000, 48, 195-209. (18) O. E. C. D. Daphnia sp. 14-day Reproduction Test (including an acute immobilisation test). Guidelines for Testing of Chemicals; No 202; OECD: Paris, 1981. (19) Forbes, V. E.; Calow, P.; Sibly, R. M. Environ. Toxicol. Chem. 2001, 20, 442-447. (20) Boersma, M.; De Meester, L.; Spaak, P. Limnol. Oceanog. 1999, 44, 393-402. (21) Vanoverbeke, J.; De Meester, L. Hydrobiologia 1997, 360, 135142.
(22) Hebert, P. D. N.; Beaton, M. J. Methodologies for Allozyme Analysis Using Cellulose Acetate Electrophoresis; Helena Laboratoires: Windsor, Ontario, 1989. (23) A. S. T. M. Standard Practice for Conducting Toxicity Tests with Fishes, Microinvertebrates and Amphibians; E 729-90; Annual Book of ASTM Standards; ASTM: Philadelphia, PA, 1998; Vol. 11.4. (24) Barata, C.; Baird, D. J. Funct. Ecol. 1998, 12, 412-419. (25) Barata, C.; Markich, S. J.; Baird, D. J. Aquat. Toxicol. 1998, 42, 115-137. (26) Worthing, C. R. The Pesticide Manual. A World Compendion, 8th ed.; The Lavenham Press Limited: Lavenham, Suffork, 1987. (27) Zar, J. H. Biostatistical Analysis, 3rd ed.; Prentice Hall: Upper Saddle River, NJ, 1996. (28) Forbes, V. E.; Forbes, T. L. Ecotoxicology in Theory and Practice; Chapman and Hall: London, 1994. (29) Jensen, A.; Forbes, V. E.; Parker, E. D. Environ. Toxicol. Chem. 2001, 20, 2503-2513. (30) Forbes, V. E.; Moller, V.; Browne, R. A. In Genetics in Ecotoxicology; Forbes V. E., Ed.; Taylor & Francis: Ann Arbor, MI, 1988. (31) Forbes, V. E.; Depledge, M. H. In Ecotoxicology: Ecological Dimensions; Baird, D. J., Maltby, L., Greig-Smith, P. W., Douben, P. E. T., Eds.; Chapman and Hall: London, 1996.
Received for review December 19, 2001. Revised manuscript received May 7, 2002. Accepted May 13, 2002. ES0158556
VOL. 36, NO. 14, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3049
