Development of a Reproductive Performance Test for Endocrine

Jun 14, 2000 - J. E. Harries , T. Runnalls , E. Hill , C. A. Harris , S. Maddix , J. P. ... However, given the large number of chemicals that have now...
29 downloads 0 Views 203KB Size
Environ. Sci. Technol. 2000, 34, 3003-3011

Development of a Reproductive Performance Test for Endocrine Disrupting Chemicals Using Pair-Breeding Fathead Minnows (Pimephales promelas) J. E. HARRIES,† T. RUNNALLS,† E. HILL,‡ C. A. HARRIS,† S. MADDIX,† J . P . S U M P T E R , † A N D C . R . T Y L E R * ,† Department of Biological Sciences, Brunel University, Uxbridge, Middlesex, UB8 3PHUK and The School of Chemistry, Physics and Environmental Sciences, University of Sussex, Brighton, BN1 9OJ U.K.

Existing in vivo tests (with the exception of the full lifecycle test) are not adequate for assessing the reproductive effects of endocrine disrupting chemicals (EDCs) on fish, and hence the need for partial life-cycle tests has been recognized internationally. In this paper we describe the development of a short-term (6 week) reproductive performance test for EDCs using pair-breeding fathead minnows (Pimephales promelas). In the test, reproductive performance in paired fish is assessed over two 3 week periods, one with exposure to the test chemical and one without. The test is highly integrative and measures effects of exposure to chemicals on fecundity, gonadosomatic index (GSI), vitellogenin (VTG) induction, and secondary sexual characteristics (fat pad and tubercles in males). In this test, exposure to butyl benzyl phthalate (BBP) at a nominal concentration of 100 µg/L (measured concentration between 69 µg/L and 82 µg/L) had no discernible effects on reproductive performance. In contrast, all reproductive parameters measured were affected by exposure to 4-NP, albeit some (e.g. VTG induction and reduction in the prominence of secondary sexual characteristics - lowest effective dose between 0.65 µg/L and 8.1 ( 1 µg/L [measured]) were more sensitive than others (e.g. number of eggs and spawnings, where the lowest effective dose was between 8.1( 1 µg/L and 57.7 ( 3 µg/L [measured]). Concentrations of 4-NP at or above 48 µg/L [measured] inhibited reproduction completely.

Introduction It has now been clearly established that there are a number of natural and man-made chemicals in the aquatic environment that mimic either estrogens, androgens, or thyroid hormones, or their antagonists (1-5). Furthermore, evidence from wildlife populations has indicated that exposure to environmental hormones can disrupt the endocrine system and cause alterations in both reproduction and development (mammals (6); birds (7); reptiles (8); fish (9-11)). Most of the * Corresponding author phone: +44 1985 247000; fax: +44 1895 274348; e-mail: [email protected]. † Brunel University. ‡ University of Sussex. 10.1021/es991292a CCC: $19.00 Published on Web 06/14/2000

 2000 American Chemical Society

hormone mimics identified to date are weak compared with the endogenous hormones (12). Nevertheless, exposure of animals in the laboratory to some of these chemicals at environmentally relevant concentrations has shown that they may cause alterations in reproductive function (13). Concern about the prevalence of endocrine-disrupting chemicals (EDCs) in the environment has led to the development of a wide variety of in vitro screening assays (recombinant yeast assays and various hormone-responsive cell-lines) for endocrine mimics [reviewed in ref 14]. It has been established, however, that a number of EDCs bioaccumulate (15) and/or result from metabolism/environmental degradation of the parent compound (3, 16) and that these features are not assessed adequately in in vitro assays. Furthermore, some EDCs are now known to have multiple target sites and in vitro screening/testing systems exclude affects that are not mediated via the relevant receptor(s). It is now recognized, therefore, that to show a cause-effect relationship between exposure to EDCs and a reproductive/ developmental effect, it is necessary to conduct whole animal studies. However, as yet, there are very few validated in vivo screening or short-term testing systems available. Most of the evidence for endocrine disruption in wildlife populations has been derived from organisms living in, or closely associated with, the aquatic environment [reviewed in ref 4], and, therefore, fish have been recommended for the development of tests for EDCs (17, 18). The fish full life-cycle (FFLC, or multigeneration) protocol provides an integrative approach for evaluating the effects of EDCs on development and reproduction in fish. However, given the large number of chemicals that have now been reported to be endocrineactive, there is a practical need to develop more pragmatic partial life-cycle tests before moving to the very challenging FFLC protocol. A priority for these short-term in vivo tests is that they are capable of assessing effects of EDCs on reproductive/developmental function, rather than necessarily be indicative of a specific mechanism of action. There is also a need to relate commonly used biomarkers for endocrine disruption (e.g. induction of vitellogenin) with parameters of physiological fitness; in the case of reproduction, these would be parameters such as fecundity and fertility. The fathead minnow, Pimephales promelas, a member of the cyprinid family, is one of the most widely used fish species in ecotoxicology, and it is readily induced to breed in captivity, facilitating the development of assays that monitor the effect of water borne chemicals on reproduction. There are already test protocols in place that use fathead minnows for chemical toxicity assessments (e.g. OECD Protocol 204) (19) and protocols for assessing toxic effects of chemicals on growth/ development/reproduction (e.g. U.S. EPA fish life-cycle) (20). The aim of this study was to develop a relatively short-term (6 week) integrative reproductive performance test for EDCs using pair-breeding fathead minnows (Pimephales promelas). The chemicals chosen for the development of the reproductive performance test were butyl benzyl phthalate (BBP) and 4-nonylphenol (4-NP), for the following reasons: they have been shown to be weakly active as endocrine mimics (compared with estradiol-17β, 4-NP is between 10-4 and 10-5 times less potent in vitro (12, 21) and between 10-3 and 10-4 times less potent in vivo (13, 22) and BBP 10-6 less potent in vitro (23)); they have only low- to midrange abilities to bioaccumulate (10-1000-fold (15, 24-26)); both chemicals are present widely in the aquatic environment (4-NP (27, 28), BBP (29-31)); and they are EDCs that potentially have multiple target sites (BBP and 4-NP have been shown to interact with both the estrogen (as agonists) and androgen VOL. 34, NO. 14, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3003

(as antagonists) receptors in hormone responsive cells and/ or reporter genes assays (2, 12-13, 21, 23, 32)), and 4-NP has also been shown to affect some of the cytochrome p450 mixed function oxidase enzymes (22).

Materials and Methods General Experimental Apparatus and Water Quality. Pairbreeding fathead minnow were housed in all-glass 30 L aquaria (0.6 m × 0.3 m × 0.3 m) that were aerated. All pipe work and casings were either glass, silicone, or durapipe, to avoid potential contamination with known environmental estrogens, including phthalates and bisphenol A. Dechlorinated tap water was supplied at a flow rate to the tanks of 0.5 L/min (i.e. there was one complete change of water every hour). The parameters monitored routinely within the test tanks throughout the studies were dissolved oxygen (6.28.2 mg per L), pH (6.44-7.85), and temperature (25 ( 0.5 °C). The water supply had an excess of 100 mg of CaCO3/L. Concentrations of metals (including zinc and copper, that can affect the adhesion of eggs to the spawning tiles) were undetectable. The room used for these experiments contained two series of identical tanks. Acclimation of the pair-breeding fish and assessments of reproductive performance in the 3 week period prior to chemical exposure were conducted in the first series of tanks, and then the pairs of fish were transferred to the second series of tanks, which had been predosed and equilibrated with the test chemical. Maintenance and Acclimation of Pair-Breeding Adult Fathead Minnows. Fathead minnows (Pimephales promelas) were supplied from breeding stocks maintained at AstraZeneca Limited, Brixham, U.K. Fish were fed four times each day (three times with frozen, γ-irradiated brine shrimp and once with the dry flake food, Tetramin). The photoperiod was maintained at 16 h light:8 h dark throughout, with 20 min dawn:dusk transition periods. Fathead minnows reproduce under specific conditions, and in their native environment the spawning season can extend over a period of up to 5 months. In captivity, conditions can be manipulated to induce reproduction, and under optimal conditions the spawning cycle may be maintained for many months. Females deposit eggs onto the surface of a spawning substrate/structure, which are then fertilized by the male. The male then expels the female from the area and guards the eggs from predation until hatching. Pilot studies in our laboratory established that, providing that the eggs are removed daily from the tank, each pair spawned every 3-4 days. The spawning substrate consisted of sections of black plastic gutter pipe. The pipe sat on a perforated stainless steel grid that allowed eggs that did not adhere to the pipe to fall into the Petri dish. This apparatus was designed to ensure that all the eggs were collected (both those that adhered to the spawning substrate and those that did not). Egg collection was completed by 12:00 a.m., and the tanks were then siphoned of waste food, faeces, and any other debris, and spawning tiles were carefully cleaned. Minimal disturbance of the fish was important, as in our pilot studies it was established that well fed fish only ate spawned eggs as a response to disturbance or threat. General Experimental Protocol. Pairs of fish that were breeding successfully during an acclimation period of at least 4 weeks were selected. Reproductive performance of each breeding pair was measured for 3 weeks prior to exposure to the test chemical and then for 3 weeks during exposure to the test chemical. There were four breeding pairs (two tanks of fish, each containing two breeding pairs that were separated by a perforate stainless) for each treatment. Only data from pairs of fish that continued to spawn throughout the full 3 week “preexposure” were subsequently used for analysis of reproductive performance. 3004

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 14, 2000

Dosing. Stock solutions of BBP (Greyhound, Cheshire, U.K. - 98% pure) and 4-NP (Technical grade - branched chain isomers - ACROS Organics, Leicester, U.K. - 99% pure) were made in Analar grade methanol as a carrier. A continuous flow-through system was employed. Each stock solution was dosed using a Watson-Marlow 205U peristaltic pump equipped with six channels of 0.63 mm diameter silicone rubber tubing. Stock solution was pumped continuously at 0.1 mL per min into glass aspirator jars receiving an influent of heated aquarium water at 500 mL per min. The aspirator jars were stirred continually using magnetic stirrers, to ensure adequate mixing of diluent water and chemical stock solutions. Fresh stock solutions were made up every 3 days. Tanks were dosed with the test chemicals for 10 days prior to the introduction of the pair breeding fish. In each experiment, dilution water, and dilution water with methanol vehicle were run as controls. Analysis of Water Concentrations of Test Chemicals. In experiment 1, concentrations of BBP and 4-NP were determined in the stock solution and in the tank water after 10 days of chemical dosing, prior to deployment of the fish into the exposure tanks. In experiment 2, concentrations of 4-NP were determined in the tank prior to fish exposure and for some doses, postexposure too. One liter water samples were collected from the tanks, acidified to pH 2.5 with 2 mL of HCl, and stored in the darks at 4 °C. Prior to analysis of BBP, 100 µg/L of diethylhexyl phthalate (DEHP, Fluka, Dorset) was added to act as an internal standard. For quantification of 4-NP, 4-tert-octylphenol (4-OP, Fluka, Dorest) was added as an internal standard, at a concentration of one-quarter of the nominal tank concentrations of 4-NP. HPLC grade methanol (5% final concentration) was added to the water sample, and 0.1-1.0 L aliquots were subjected to solid-phase extraction on 100 mg C18 Sep-pak cartridges (Waters Ltd., Watford, U.K.). The cartridges were washed with de-ionized water, and alkylphenols and phthalates were eluted with 2 mL of methanol. The eluent was reduced to approximately 0.2 mL under vacuum, diluted with 0.2 mL of water (pH 4.0), and 0.2 mL were injected into the HPLC for analysis. BBP and 4-NP were quantified by reverses phase HPLC using a 5 µm C18 column (4.6 × 150 mm; Waters Nova-Pakl, MA, U.S.A.). Gradient elution was carried out at 1 mL/min from methanol/water/acetic acid (70:29.9:0.1, 10 min) to 100% methanol (20 min). Phthalates and alkylphenols were detected at 278 nm using a photodiode array detector (Waters, M. A., U.S.A.). DEHP eluted at 5.0 min, BBP and 4-NP as broad peaks between 14.5 and 16 min. BBP and 4-NP were quantified using response factors from the HPLC analyses of the standard samples. Analyses of the tanks water alone, or tank water with methanol carrier, were run to check for interference and showed that none of these chemicals were present in the water supply (limits of detection of BBP and 4-NP were 0.05 µg/L and 0.1 µg/L, respectively). Recoveries of spiked samples were between 90 and 98%. Measurement of Estrogenic Activity in the Tank Water. Estrogenic activity in the tanks dosed with BBP and 4-NP was measured using an in vitro assay based on a recombinant yeast expressing the human estrogen receptor (33). Water samples were first extracted using solid phase, according to Janbakhsh (34). Briefly, water samples were collected in solvent rinsed bottles, treated with 10 mL of 38% formaldehyde/L water to prevent bacterial degradation of the test chemical, and stored in the dark at 4 °C until extraction. Aliquots of water were extracted onto ethyl-bonded silica, using a 12-port vacuum extraction method. For the given protocol, the detection limit for estradiol-17β in the tank water was approximately 2 ng/L. In addition, distilled water samples were spiked with BBP and 4-NP at the same nominal concentrations as the dosed tanks, and then they were extracted onto ethyl-bonded silica. The chemicals bound to

the cartridges were eluted with 5 mL of methanol, and the resulting eluent was evaporated to dryness under a stream of nitrogen. The chemicals were then redissolved in a small volume (100 µL in experiment 1, 200 µL in experiment 2) of ethanol. The extracts were tested for estrogenic activity in the recombinant yeast screen (33). Measurement of Reproductive Performance. Number of Spawnings, Fecundity, and Egg Size. Each day, spawning events were recorded, and the eggs were collected. The eggs were counted, and the diameters of 20 were measured from each spawning, using a low-power dissecting microscope fitted with an eyepiece graticule. The eggs were not necessarily spherical in shape, and hence the diameters determined were the mean of two measurements taken across opposite axes of the egg. Gonadosomatic Index and Measurement of Plasma Vitellogenin. At the end of each experiment, fish were anaesthetized in MS-222 (Sigma, Poole, U.K.), and the length and weight of the fish were measured. Blood samples were collected into ice-chilled, heparinized (500 units heparin/ mL) vessels (syringes or capillary tubes) and centrifuged at 3000 g for 10 min, and the resulting plasma were withdrawn and deep-frozen at -20 °C, until required. Fish were then dispatched and pithed, and the gonadosomatic index was determined (GSI ) gonad weight/[body weight - gonad weight] × 100). Plasma VTG concentrations were determined using a carp-VTG ELISA that has been validated for measuring VTG in fathead minnow (35, 36). Secondary Sexual Characteristics in Males. During the breeding season, male fathead minnows develop a soft, mucus-secreting dorsal pad (fat pat) associated with changes in the epidermis thickness and a series of rows of tubercles across the snout (37). The development of these secondary sexual characteristics in males is androgen dependent and can be induced in males and females by exposure to androgens (38). The number of tubercles on the snout and the size (surface area occupied; experiment 1) and thickness (at its widest margin; experiment 2) of the dorsal fat pad in the males were measured at the end of the experiments, to assess effects of chemical exposure on these secondary sexual characteristics. When counting tubercles, only prominent ones (defined as those that were raised above the surface of the skin) were recorded. To avoid stressing pairs of successfully breeding fish, secondary sexual characteristics were not quantified before experiments were initiated. Experiment I. Fish used in the first experiment were 6 months old. In this experiment, reproductive performance was assessed in individual breeding pairs exposed to BBP or to 4-NP, both at a nominal concentration of 100 µg/L. Water and methanol were run as controls. All fish in the solvent control and chemical exposure tanks received 0.2 mL of methanol/L. Experiment II. Fish used in experiment 2 were 4 months old. In this experiment reproductive performance was assessed in breeding pairs exposed to three doses of 4-NP (nominal concentrations - 1 µg/L, 10 µg/L, and 100 µg/L). Water and methanol were run as controls. All fish in the solvent control and chemical exposure tanks received 0.2 mL of methanol/L. Statistical Analysis. The statistical analyses were carried out using STATVIEW and SUPERANOVA statistical programs (Abacus Concepts Inc. Berkeley, CA). Prior to analysis, data were transformed where necessary to improve normality and homogeneity of variance. In each of the two experiments, for each treatment, there were two tanks each containing two pairs of fish. To investigate possible intertank differences within a treatment, therefore, for both the pre- and duringexposure periods nested ANOVAs were carried out on the data sets. With the exception of egg batch size in the preexposure period for the NP treatment group in experiment

TABLE 1. Nominal and Measured Concentrations of Butyl Benzyl Phthalate (BBP) and 4- Nonylphenol (4-NP)a Experiment 1 actual concn (µg/L)

treatment

nominal tank water concn (µg/L)

stock bottle (µg/L)

tank water (µg/L)

controls (a and b) solvent controls (a and b) butyl benzyl phthalate-a butyl benzyl phthalate-b nonylphenol-a nonylphenol-b

0 0 100 100 100 100

0 0 104 82 102 97

0 0 69 82 82 60

Experiment 2

treatment

nominal tank water concn (µg/L)

controls (a and b) solvent controls (a and b) nonylphenol-1a nonylphenol-1b nonylphenol-10a nonylphenol-10b nonylphenol-100a nonylphenol-100b

0 0 1 1 10 10 100 100

actual tank water concn (µg/L) prepostexposure exposure 0 0 0.62 0.68 6.3 8.2 48.1 58.6

nd nd nd nd nd 9.9 61.7 62.7

a In experiment 1, BBP and 4-NP were measured in the stock bottle and in the tank water immediately prior to chemical exposure of the fish. In experiment 2, 4-NP was measured in the tank water immediately prior to chemical exposure of the fish and in some tanks, at the end of the experiment too. a and b represent the duplicate tanks. BBP and 4-NP were measured using GCMS. nd ) not done.

1 (which was significant at the 1% level only; P ) 0.004), there were no such intertank differences for any treatment for any time period. As a consequence each pair of spawning fish was considered as a single replicate (i.e. there were four replicates for a full compliment of fish). One way analysis of variance (ANOVA) was then performed on transformed data, and differences in the weights, lengths, GSIs, and plasma concentrations of VTG of fish exposed to BBP and 4-NP were tested using Scheffes test. Effects of BBP and 4-NP on the total number of eggs spawned, the number of spawnings mean egg batch size and egg size were analyzed for pairs of fish pre- and during-chemical exposure, between pairs within a treatment and between the groups of pairs across treatments using one way ANOVA, followed by Fischers PLSD test. Differences in the secondary sexual characteristics in males (tubercle number and fat pad dimensions) exposed to 4-NP were analyzed using one way ANOVA followed by Fishers PLSD test.

Results and Discussion Concentrations and Estrogenic Activity of the Test Chemicals in the Dosed Tanks. Measured concentrations of BBP and 4-NP in the tank water in the two experiments are given in Table 1. Measured concentrations of BBP were between 69% and 82% of nominal (100 µg/L), with a mean of 75.5 µg/L (n ) 2). In experiment 1, measured concentrations of 4-NP were between 60% and 82% of nominal (100 µg/L), with a mean of 71 µg/L (n ) 2). In experiment 2, measured concentrations of 4-NP were between 48% and 99% of nominal, with means of 0.65 µg/L [n ) 2, nominal concentration 1 µg/L), 8.1 ( 1 µg/L (n ) 3, nominal concentration 10 µg/L), and 57.7 ( 3 (n ) 4, nominal concentration 100 µg/L). To obtain an accurate estimate of the actual chemical exposure over the full test period, more regular chemical samplings would be needed. VOL. 34, NO. 14, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3005

Estrogenic activity in the tanks dosed with BBP and 4-NP was demonstrated using an in vitro assay (Figure 1). Extracts from tanks treated with BBP produced dose-response curves in the yeast assay that mimicked the response curve for the BBP standard (BBP produces a characteristic “half” (partial agonist-like) dose-response curve in the estrogen yeast assay (23)), confirming that the estrogen activity measured in the tank was due to BBP. Extracts from tanks dosed with 4-NP in both experiments produced full dose-response curves, with a potency of approximately 1/10 000 compared with estradiol-17β (based on ED50s). This potency of 4-NP is similar to that determined for 4-NP in other studies using estrogen responsive cell-lines (21, 33). A previous study on spawning fathead minnows indicated that mature fish exude steroid estrogens into the water (measured concentrations of estradiol-17β in the water in their control tanks were between 3.5 and 14.8 ng/L (39)). Indeed, it is well established that some species of sexually mature cyprinid fish secrete steroids hormones, some of which act as pheromones (40). However, in our experiments no “additional” estrogenic activity was detected in the exposure tanks above that which could be accounted for by the concentrations of BBP and 4-NP present. In the studies by Kramer et al. (39), the stocking density of the fathead minnows in the tanks was approximately 2-fold higher than in this study, and the rate of exchange of water in their tanks was considerably lower (approximately 10-fold lower); these factors may account for the presence of detectable concentrations of endogenously produced steroid estrogens in the water in their studies but not in ours. The possible presence of endogenously derived steroids in the tank water is an important consideration when designing and conducting experiments to test EDCs; the use of low stocking densities and relatively high throughput of water in the tanks appear be prerequisites for studies of this nature. Fish Survival and Growth. There was no effect of chemical treatment (either BBP or 4-NP) on fish survival. In experiment 1, all four pairs of fish for each treatment survived the trial. In experiment 2, during the first 3 weeks of the trial, one fish in a breeding pair in the dilution water controls and three pairs in the solvent controls (presolvent exposure) stopped breeding, either because a fish died (the male of one pair) or because females suddenly became unable to release their eggs. The results on reproductive performance in experiment 2 were, therefore, analyzed with n ) 3 pairs in the dilution water controls, n ) 1 in solvent control, and the full complement of n ) 4 pairs for all the 4-NP treatment groups. Having only one pair of successfully reproducing fish in the solvent controls in experiment 2 limited the powers of the analyses for the exposure to 4-NP in this experiment. Experiment 1, however, had established that there was no effect of the solvent vehicle on somatic growth or fecundity. At the end of experiment 1, there were no differences in the size (length or weight) of the females or the males between the controls and the BBP or 4-NP-exposed fish. In experiment 2, there were significant differences in weight (but not length) between the males in all the NP-treated groups and the solvent control males (P < 0.05) but not compared with the dilution water control males. There were no differences in size (length or weight) of the females in either of the control groups compared with the 4-NP-exposed fish (at any dose) in experiment 2. In summary, there was no effect of BBP on growth over a 3 week exposure period, and no dose-related effect of exposure to 4-NP on growth in either males or females. These data are similar to that obtained from a limited number of other studies in which fish have been exposed to 4-NP at similar doses (e.g. male rainbow trout (13); female rainbow trout (41)), in which growth was not affected. Longer term exposure of fish to 4-NP, however, has been shown to 3006

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 14, 2000

FIGURE 1. Estrogenic activity of the water in the tanks dosed with BBP (experiment 1) and 4-NP (experiments 1 and 2). Water samples were concentrated via solid phase extraction. Data are plotted according to their nominal value multiplied by the concentration factor, such that water samples in which measured chemical concentrations equaled nominal would overlay the chemical standard curve for the respective chemical. The displacement curves for water samples from the 4-NP exposures tend to be displaced to the right of the standard curve in the figure, showing that the measured concentrations of the test chemical in the samples were less than the nominal. Data are presented for one of the duplicate exposure tanks only (the dilution curves for the duplicate exposure tanks overlaid each other in each treatment).

FIGURE 2. Number of eggs spawned by pairs of fathead minnow (Pimephales promelas) exposed to BBP (experiment 1) and 4-NP (experiments 1 and 2). Data are presented as the mean number of eggs spawned ( SEM for 3 week periods prior to (open bars) and during (solid bars) exposure. Significant differences between the two periods are shown (* P < 0.05). For experiment 1, n ) 4 pairs for all treatments. For experiment 2, n ) 3 pairs for the dilution water controls, n ) 1 pair for the solvent controls, and n ) 4 pairs for all the 4-NP treatments. affect growth in fish (rainbow trout exposed to concentrations of 4-NP between 1 µg/L and 30 µg/L) (42). Fecundity. Number of Eggs Spawned. The consistency in size of the female fish, both within and between the treatment groups enabled comparisons of absolute fecundity (number of eggs per female), rather than relative fecundity (number of eggs per gram of body weight), within an experiment. The numbers of eggs spawned during the two 3 week periods in fish exposed to BBP and 4-NP are shown in Figure 2. In the control females in experiment 1, the mean number of eggs spawned per female over the 3 week periods ranged between 1527 and 2774 (dilution water controls) and between 719 and 2427 (solvent controls). Control females in experiment 2 had a lower fecundity compared with control females in experiment 1, and this was probably due to their smaller size (they were only 4 months old, compared with 6 months old, respectively); it has been shown in other fish species that fecundity is positively correlated with body size (43). Analyzed on a weekly basis, there were no changes in fecundity with time in the control fish throughout either experiment, indicating that the ability to produce eggs did not diminish over the 6 week test period (P < 0.05). Fecundity of female fathead minnow in this study in both experiments was higher than that reported for similarly sized fathead minnows (which received a similar food ration) in some other studies (see refs 39 and 44) but was similar to that reported by Gale and Buynak (45). A common feature of the studies of Gale and Buynak and our studies was the

FIGURE 3. Number of spawnings in fathead minnow (Pimephales promelas) exposed to BBP (experiment 1) and 4-NP (experiments 1 and 2). Data are presented as mean number of spawnings ( SEM for 3 week periods prior to (open bars) and during (solid bars) exposures. Significant differences between the two periods are shown (* P < 0.05). For experiment 1, n ) 4 pairs for all treatments. For experiment 2 (panel b), n ) 3 pairs for the dilution water controls, n ) 1 pair for the solvent controls, and n ) 4 pairs for all the 4-NP treatments. placement of spawning trays in the tanks to collect eggs that did not adhere to the spawning tiles. There were no significant differences in fecundity during the initial 3 week period of the tests between any of the groups (in either experiment). Similarly, there were no differences in the number of eggs spawned during the two 3 week periods in the dilution water or solvent control groups in either of the two experiments (i.e. there was no effect of the solvent vehicle on fecundity). Exposure to BBP had no effect on fecundity (comparison between pre- and during-exposure). In contrast, exposure to 4-NP resulted in a reduction in fecundity. In fish exposed to mean measured concentrations of 0.65 µg and 8.1 ( 1 µg/L, the mean number of eggs spawned was 60% and 45%, respectively, of that in the same breeding pairs before chemical exposure. These apparent reductions in fecundity, however, were not statistically significant. Exposure of fathead minnows to a nominal concentration of 100 µg of 4-NP/L dramatically reduced egg production; it ceased within the first week of exposure in two of the four females in experiment 1, and in three of the four females in experiment 2. Frequency of Spawning. There were no significant differences in the number of spawnings between the two successive 3 week periods for any of the control groups in either experiment (Figure 3). The frequency of spawnings in the control fish was consistent with that reported in other studies on the fathead minnow (mean every 3.9 days, mode every 3 days) (45). In contrast, exposure to BBP caused a VOL. 34, NO. 14, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3007

FIGURE 4. Egg batch size in fathead minnows (Pimephales promelas) exposed to BBP (experiment 1) and 4-NP (experiments 1 and 2). Data are presented as mean egg batch size ((SEM) spawned for 3 week periods prior to (open bars) and during (solid bars) exposures. Significant differences between the two periods are shown (* P < 0.05). For experiment 1, n ) 4 pairs for all treatments. For experiment 2, n ) 3 pairs for the dilution water controls, n ) 1 pair for the solvent controls, and n ) 4 pairs for all the 4-NP treatments. significant reduction in the number of spawnings (mean for the pre- and during-exposure periods: 6.25 ( 0.75 and 4.75 ( 0.5, respectively (P < 0.05)). The altered spawning frequency in response to BBP did not affect fecundity, because with a decrease in the number of spawnings there was a reciprocal increase in egg batch size (see below). Exposure to 4-NP resulted in a significant reduction in the number of spawnings in a dose-dependent manner. Exposure to 4-NP at a concentration of 8.1 ( 1 µg/L halved the number of spawnings (4.0 ( 0.7 to 2. ( 0.9, P < 0.05). Exposure to the highest dose of 4-NP, caused a reduction in the mean number of spawnings from 5.5 ( 0.86 to 1.5 ( 0.3 in experiment 1 and from 4.75 ( 0.25 to 0.25 ( 0.25 in experiment 2 (in which only one female spawned once on the first day of the 3 week exposure period to 4-NP). Number of Eggs per Spawning. The mean egg batch size of the control fish in experiment 1 ranged between 223 ( 13 and 441 ( 20 (dilution water controls) and between 103 ( 23.5 and 450 ( 26 (solvent controls; Figure 4). In experiment 2, the mean egg batch size was lower compared with females in experiment 1 and ranged between 154 ( 37 and 266 ( 34 in the different control groups. This difference in egg batch size is probably due to the difference in size between females in the two experiments (see above). There were no differences in egg batch size between the two successive 3 week periods in the dilution water control fish in both experiments. There was an increase in the egg batch size in the solvent control and on exposure to BBP (from 216 ( 19 to 282 ( 37.5; P < 0.05). The variation in egg batch size between females of a 3008

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 14, 2000

similar size emphasizes the need to compare reproductive performance within individual pairs (pre- and duringexposure). In experiment 1, exposure to 4-NP/L (mean measured concentration 71 µg/L) resulted in a significant reduction in egg batch size, which dropped from 287 ( 31 to 46 ( 22. Similarly in experiment 2, for fish exposed to 57.7 ( 3 µg 4-NP/L egg batch size dropped from 251 ( 37 to 55 ( 50 (only one of the four females spawned during exposure to this concentration of 4-NP). Egg Size. The mean egg size (diameter) in the control fish (dilution water and solvent) ranged between 693.5 ( 2.1 µm and 801.5 ( 1.7 µm in experiment 1 and between 749 ( 2.2 µm and 830.6 ( 5.5 µm in experiment 2. In the control fish, there were significant differences in egg size between egg batches laid by the same females, both within a 3 week assessment period and between the two 3 week periods (P < 0.05). The variability in egg size between batches of eggs laid by individual females and between females within a treatment group could make eggs size difficult to use as a marker for assessing effects of chemicals on reproductive performance in the fathead minnow. Gonadosomatic Index. At the end of experiment 1, GSIs in control groups ranged between 0.55 and 2.1 in males and between 11.2 and 19 in females (Figure 5). There were no differences in the sizes of the testes or ovaries in the control fish at the end of experiment 1, compared with the GSIs in a sample of eight males and eight females from the same population of fish taken at the start of the experiment (student “t”-test, p < 0.05). Maintenance of gonad size in the control fish during the 6 week experiment supported the fecundity data in illustrating that the ability of the fathead minnows to reproduce was not diminished over the test period. In experiment 2, GSIs in the control groups ranged between 0.92 and 2.15 in males and between 9.2 and 23.4 in females (Figure 5). These GSIs of mature fathead minnows are similar to those reported in other studies (1-2%, males; 10-20%, females) (46). There was no observed effect of exposure to BBP on testis or ovary size. The GSIs in males exposed to the highest concentration of 4-NP/L were 46% and 62% of the controls in experiments 1 and 2, respectively, but they were not statistically significantly different from them. The GSIs in females exposed to 4-NP at concentrations of 8.1 ( 1.0 µg/L and 57.7 ( 3 µg/L (experiment 2) were 72% (significant reduction, P < 0.05) and 66% (not significant, P > 0.05) of the controls. Other studies on fish have shown that exposure to 4-NP can affect gonad growth. For example, exposure of male rainbow trout to NP for a 3 week period caused a suppression of testis growth (effective concentration 54.3 µg/L) (13). However, this inhibitory effect of 4-NP on testis growth in the rainbow trout occurred only in male fish during the most active period of gonadal growth and not in adult, fully mature males (13). In batch spawning fish (like the fathead minnow), the gonads undergo rapid cyclical changes over short periods of time (every few days in the fathead minnow (46)), as successive batches of eggs or sperm are produced, and thus means the size of the gonads in breeding adults can vary considerably between individuals at any one point in time, and this makes identifying any alterations in the GSI, as a result of exposure to chemicals, difficult. Recent studies in the male fathead minnow suggest that other features of testis growth and development, including alterations in the number and sizes of Sertoli cells and germ cell syncytia, may be more sensitive than the GSI as indicators of exposure to EDCs; effects on these cells occurred at 4- NP concentrations between 1.1 and 3.4 µg/L (42 day exposure (47)). Vitellogenin Induction. Plasma concentrations of VTG in male fathead minnow in the control groups were 23 ( 6 ng/mL (dilution water) and 53 ( 8ng/mL (solvent) in experiment 1 and 29 ( 6 ng/mL (dilution water) and 62 (

FIGURE 5. Gonadosomatic index and plasma vitellogenin concentrations in fathead minnow (Pimephales promelas) after 3 weeks exposure to BBP (experiment 1) and 4-NP (experiments 1 and 2). Data are presented as means ( SEM. Significant differences between treatments and dilution water (*) and solvent (L) controls are shown (P < 0.05). For experiment 1, n ) 4 pairs for all treatments. For experiment 2 (panel b), n ) 3 pairs for the dilution water controls, n ) 1 pair for the solvent controls, and n ) 4 pairs for all the 4-NP treatments. 3 ng/mL (solvent) in experiment 2. Plasma concentrations of VTG in the control groups of females in experiments 1 and 2 ranged between 107 000 ( 17 500 ng/mL and 345 000 ( 68 000 ng/mL. These VTG concentrations are similar to those reported in other studies in mature male and female fathead minnows (35). There was no effect of BBP on the plasma VTG concentration in either males or females. BBP has been shown to induce VTG synthesis in fish in vivo but only at an extremely high dose (500 µg/g body weight via intraperitoneal injection), and even at this dose there was only a low level of induction (a 3-fold increase (48)). 4-NP induced a dose-related induction of VTG. The effective concentration for induction of VTG in males was between 0.65 µg/L and 8.1 ( 1 µg/L. In males, at the highest exposure concentration to 4-NP (nominal 100 µg/L), there was approximately a 4000-fold (experiment 1) and 45 000fold (experiment 2) increase in plasma VTG above the controls. In females, the effective concentration for induction of VTG was between 8.1 ( 1 µg NP/L and 57.7 ( 3.1 µg/L (P < 0.05), and at the highest 4-NP exposure concentration, there was a 2- and 10-fold increase in plasma VTG in experiments 1 and 2, respectively. Plasma titers of VTG in both males and females exposed to the highest concentration of 4-NP were at or close to the maximum response for this species (35)). In the rainbow trout, induction of plasma VTG in response to NP (for a 3 week exposure period) occurred at an exposure concentration between 5 and 20 µg/L (13). Lech et al. (49) demonstrated that exposure of rainbow trout to 10 µgNP/L induced synthesis of VTG mRNA after 72 h. A very recent study has reported that exposure to only 0.1 µgNP/L for prolonged periods of time induced VTG synthesis

in male fathead minnows (50, cited from ref 47). In that study, it is reported that VTG induction in response to NP was four times more sensitive than histological changes in the gonads. Thus, from the published information and the data presented in this study, it has been established that the vitellogenic response in the fathead minnow is a very sensitive biomarker for environmental estrogens. In this study, although induction of VTG occurred as a result of exposure to 4-NP and in a dose-dependent manner, we do not know whether the effects of 4-NP on reproductive function/performance were the result of its estrogenic activity or as a result of activity through a different mechanism of action. Secondary Sexual Characteristics in Males. The number of tubercles on control males varied between 13 and 18 in experiment 1 and between 11 and 14 in experiment 2. BBP had no effect on the number of tubercles. In contrast, 4-NP caused a reduction in the number of tubercles in males exposed to the highest dose; in experiment 1, male fish had no tubercles, and in experiment 2 there were between 0 and 4 tubercles per fish (P < 0.05). The size (surface area) of the fat pad in males in experiment 1 varied between 37.4 mm2 and 107 mm2 (mean 83 ( 15.5 mm2) and between 71.7 mm2 and 139 mm2 (mean 94.9 ( 14.7 mm2) for the dilution water and methanol control males, respectively. Neither BBP or 4-NP had an effect on the surface area of the fat pad. There was, however, a dose-dependent effect (a reduction) in the thickness of the fat pad on exposure to 4-NP. The thickness of the fat pad in the dilution water and methanol controls varied between 2.5 and 3.5 mm (overall mean 2.83 ( 0.33 mm), whereas the mean thicknesses of the fat pads in males exposed to 0.65, 8.1 ( 1 and 57.7 ( VOL. 34, NO. 14, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3009

3.1 µg 4-NP/L were 1.75 ( 0.32 mm, 0.5 ( 0.35 mm, and 0.75 ( 0.43 mm, respectively (P < 0.05 in all cases). In a previous study (47) reporting the effects of 4-NP and NPEO on the secondary sex characteristics in male fathead minnows, no effects on the tubercles or fat pad were seen for measured exposure concentrations of up to 3.4 µg/L and 5.5 µg/L, respectively. The differences between our findings on the effects of 4-NP on the fat pad and those reported by MilesRichardson et al. (47) may be due to differences in the age/ state of maturity of the males used. Our findings would indicate that the thickness of the fat pad is a sensitive biomarker for exposure to the antiandrogenic effects of chemicals, such as 4-NP. An antiandrogenic effect of 4-NP in vivo is consistent with recent findings in vitro where 4-NP was shown to be antagonistic in a androgen reporter gene assay (32). Histological analysis of the fatpad would allow effects of chemicals on the number of mucus cells present and the thickness of the connective tissue dermal layer (Stratum laxium) to be assessed, both of which are androgen dependent (38, 46). Indeed, the androgen dependency of these aspects of fat pad development may provide useful parameters for assessing androgenic or antiandrogenic effects of EDCs.

A Summary of the Test Protocol Developed and Effects of BBP and 4-NP on Reproductive Performance Successful breeding in pairs of fathead minnow involves an interplay between behavioral and physiological responses that are controlled by the hypothalamic-pituitary-gonadalaxis and any chemical that interferes with these control processes, at any level in the hierarchy, is likely to affect reproductive performance. The reproductive performance test developed here, therefore, is a highly integrative test. The test system has been developed for use in assessing the effects of exposure to chemicals on fecundity (egg number, egg batch size, and number of spawnings), gonadosomatic index, induction of plasma vitellogenin, and secondary sexual characteristics (in males). The design of the test would also allow other endpoints to be incorporated, if desired, for example, histology of the gonad and measurement of plasma sex steroids (estradiol-17β, 11-keto testosterone etc.; see ref 50). A very valuable addition to the reproductive performance test would be to assess gamete viability; this could be determined by simply maintaining the fertilized eggs until they hatch 4 days later. Results from the reproductive performance test indicate that a relatively short-term exposure to a high concentration of BBP (71 µg/L) did not cause any adverse effects on reproduction in fathead minnows. These data support the very limited information available in the literature, where BBP and other phthalates, such as di-n-butyl-phthalate (DBP), have been shown to have deleterious effects in fish only at very high concentrations (24). Given that the concentrations of BBP in most aquatic environments are only a few µg/L or less (reviewed in ref 4), with the possible exception of “hotspots”, it would appear that BBP is unlikely to pose a threat to reproduction in wild populations of fish. In contrast to BBP, 4-NP caused adverse effects on reproduction. Threshold concentrations for effects of 4-NP on vitellogenin and secondary sexual characteristics were between 0.65 µg/L and 8.1 ( 1 µg/L, which are at least as low (and in many case lower) as that reported for effects of NP on gonadal growth/development in other fish (13, 51). Effects on spawning performance in the fathead minnow occurred at concentrations of 4-NP between 8.1 ( 1 µg/L and 57.7 ( 3.1 µg/L. Effluents discharged into rivers may often contain 4-NP in tens of µg/L and can exceed 100 µg/L (52). In most rivers in the U.K., 4-NP concentrations are less than 10 µg/L (27), but in a few U.K. rivers, concentrations of NP exceed 3010

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 14, 2000

100 µg/L (e.g. River Aire (27)). In other European countries (Switzerland and Belgium), NP has been measured in rivers at concentrations between 0.3 µg/L and 42 µg/L (28, 52). In contrast to Europe, in the U.S.A., a survey of 30 rivers found that 70% had concentrations of NP that did not exceed 0.1 µg/L (53). Given the concentrations of NP in some aquatic environments, the data presented here would suggest that 4-NP may offer a threat to reproduction in some wild populations of fish.

Acknowledgments Jule Harries was funded by a grant from ECETOC. Tamsin Runnalls was funded by Brunel University. We would like to thank AstraZeneca (Brixham Environmental Laboratory) for the supply of fathead minnows and Steve Pash (Brunel) for his valuable help with the care of the fish.

Literature Cited (1) Sumpter, J. P.; Jobling, S.; Tyler, C. R. In Toxicology of Aquatic Pollution; Taylor, E. W., Ed.; Cambridge University Press: U.K., 1996; pp 205-224. (2) Gray, L. E.; Ostby, J. S.; Kelce, W. R. Toxicol. Appl. Pharmacol. 1994, 129, 46-52. (3) Kelce, W. R.; Stone, C. R.; Laws, S. C.; Gray, L. E.; Kemppainen, J. A.; Wilson, E. M. Nature 1995, 375, 581-585. (4) Tyler, C. R.; Jobling, S.; Sumpter, J. P. Crit. Rev. Toxicol. 1998, 28, 319-361. (5) Brouwer, A.; Ahlborg, U. G.; Vandenberg, M.; Birnbaum, L. S.; Boersma, E. R.; Bosveld, B.; Denison, M. S.; Gray, L. E.; Hagmar, L.; Holene, E.; Huisman, M.; Jacobson, S. W.; Jacobson, J. L.; Koopmanesseboom, C.; Koppe, J. G.; Kulig, B. M.; Morse, D. C.; Muckle, G.; Peterson, R. E.; Sauer, P. J. J.; Seegal, R. F.; Smitsvanprooije, A. E.; Touwen, B. C. L.; Weisglaskuperus, N.; Winneke, G. Eur. J. Pharmacol. Environ. Toxicol. Pharmacol. 1995, 293(1), 1-40. (6) Facemire, C. F.; Gross, T. S.; Guillette, L. J. Environ. Health Perspect. 1995, 103, 79-86. (7) Fry, D. M.; Toone, C. K. Science 1981, 922-924. (8) Guillette, L. J.; Gross, T. S.; Mason, G. R.; Matter, J. M.; Percival, H. F.; Woodward, A. R. Environ. Health Perspect. 1994, 102, 680-688. (9) Purdom, C. E.; Hardiman, P. A.; Bye, V. J.; Eno, N. C.; Tyler, C. R.; Sumpter, J. P. Chem. Ecol. 1994, 8, 275-285. (10) Folmar, L. C.; Denslow, N. D.; Rao, V.; Chow, M.; Crain, D. A.; Enblom, J.; Marcino, J.; Guillette, L. J. Environ. Health Perspect. 1996, 104, 1096-1101. (11) Lye, C. M.; Frid, C. L. J.; Gill, M. E.; McCormick, D. Mar. Pollut. Bull. 1997, 34, 34-41. (12) White, R.; Jobling, S.; Hoare, S. A.; Sumpter, J. P.; Parker, M. G. Endocrinology 1994, 135, 175- 182. (13) Jobling, S.; Sheahan, D. A.; Osborne, J. A.; Matthiessen, P.; Sumpter, J. P. Environ. Toxicol. Chem. 1996, 15, 194-202. (14) Zacharewski, T. Environ. Sci. Technol. 1997, 31, 613-623. (15) Ekelund, R.; Bergman, A.; Granmo, A.; Berggren, M. Environ. Pollut. 1990, 64, 107-120. (16) Tyler, C. R.; Beresford, N.; van der Woning, M.; Sumpter, J. P.; Thorpe, K. Environ. Toxicol. Chem. 2000, 18, 801-809. (17) Organisation for Economic Co-orperation and Development. Fish Expert Consultation Meeting, London; Paris, Cedex 16, France, 1999. (18) USEPA. Endocrine Disruptor Screening Programme; Proposed Statement of Policy; 1998; Vol. 63, pp 71542-71568. (19) Organisation for Economic Co-orperation and Development. OECD Guidelines for the Testing of Chemicals. Guideline 204: Fish, Prolonged Toxicity Test; 14-day Study; OECD: Paris, Cedex 16, France, 1993. (20) USEPA. Fish life-cycle toxicity tests; Hazard Evaluation Division Standard Procedure EPA 540/9-86-137; Office of Pesticide Programmes: Washington, U.S.A., 1986. (21) Soto, A. M.; Justicia, H.; Wray, J. W.; Sonnenschein, C. Environ. Health Perspect. 1991, 92, 167-172. (22) Lee, P. C.; Patra, S. C.; Stelloh, C. T.; Lee, W.; Struve, M. Biochem. Pharmacol. 1996, 52, 885-889. (23) Harris, C. A.; Henttu, P.; Parker, M. G.; Sumpter, J. P. Environ. Health Perspect. 1997, 105, 802-811. (24) Woodward, K. N. Phthalate Esters: Toxicity and Metabolism; CRC Press: Boca Raton, FL, 1988; Vol. 2. (25) Ahel, M.; McEvoy, J.; Giger, W. Environ Pollut. 1993, 79, 243248.

(26) Liber, K. G.; J. A. Corry, T. D.; Heinis, I. J.; Stay, F. S. Environ. Toxicol. Chem. 1999, 18, 394-400. (27) Blackburn, M. A.; Waldock, M. J. Water Res. 1995. (28) Ahel, M.; Scully, F. E.; Hoigne, J.; Giger, W. Chemosphere 1994, 28, 1361-1368. (29) Hites, R. A.; Biemann, K. Science 1972, 178, 158-160. (30) Fatoki, O. S.; Vernon, F. Sci. Tot. Environ. 1990, 95, 227-232. (31) Suffet, T. H.; Brenner, L.; Cairo, P. R. Water Res. 1980, 14, 853. (32) Sohoni, P.; Sumpter, J. P. J. Endocrinol. 1998, 158, 327-339. (33) Routledge, E. J.; Sumpter, J. P. Environ. Toxicol. Chem. 1996, 15, 241-248. (34) Jambaksh, A. Masters Thesis, Brunel University, Middlesex, U.K., 1996. (35) Tyler, C. R.; van Aerle, R.; Hutchinson, T. H.; Maddix, S.; Trip, H. Environ. Toxicol. Chem. 1999, 18, 337-347. (36) Tyler, C. R.; Van der Eerden, B.; Jobling, S.; Panter, G.; Sumpter, J. P. J. Comput. Physiol. 1996, 166, 418-426. (37) Smith, R. J. F.; Murphy, B. D. Trans. Am. Fish. Soc. 1974, 103, 65-72. (38) Smith, R. J. F. Can. J. Zool. 1974, 52, 1031-1038. (39) Kramer, V. G.; Miles-Richardson, S.; Pierens, S.; Giesy, J. P. Aquat. Toxicol. 1997, 40, 311-322. (40) Stacey, N.; Zheng, W. B.; Cardwell, J. Gen. Comput. Endocrinol. 1994, 96, 288-297. (41) Thorpe, K. L.; Hutchinson, T. H.; Hetheridge, M. J.; Sumpter, J. P.; Tyler, C. R. Environ. Toxicol. Chem. 2000, In Press. (42) Ashfield, L. A.; Pottinger, T. G.; Sumpter, J. P. Environ. Toxicol. Chem. 1998, 17, 679-686.

(43) Bagenal, T. J. Fish Biol. 1969, 1, 167-182. (44) Weber, D. N. Neurotoxicology 1993, 14, 347-358. (45) Gale, W. F.; Buynak, G. L. Trans. Am. Fish. Soc. 1982, 111, 3540. (46) Smith, R. J. F. Can. J. Zool. 1978, 56, 2103-2109. (47) Miles-Richardson, S. R.; Pierens, S. L.; Nichols, K. M.; Snyder, E. M.; Snyder, S. A.; Render, J. A.; Fitzgerald, S. D.; Giesy, J. P. Environ. Res. Section A 1999, s122-s137. (48) Christiansen, T.; Korsgaard, B.; Jespersen, A. J. Exp. Biol. 1998, 201, 179-192. (49) Lech, J. J.; Lewis, S. K.; Ren, L. F. Fundam. Appli. Toxicol. 1996, 30, 229-232. (50) Giesy, J. P.; Pierens, S. L.; Miles-Richardson, S.; Kramer, V. J.; Snyder, S. S.; Nichols, K. M.; Snyder, E.; Villenueve, D. A. Environ. Toxicol Chem. In press. (51) Gray, M. A.; Metcalf, C. D. Environ. Toxicol. Chem. 1997, 16, 1082-1086. (52) Tanghe, T. D. G.; Verstraete, N. J. Environ. Qual. 1999, 28, 702709. (53) Naylor, C. G. Soap Cosmetics Specialties 1992, 72, 27-31.

Received for review November 16, 1999. Revised manuscript received April 20, 2000. Accepted April 24, 2000. ES991292A

VOL. 34, NO. 14, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3011