- Email: [email protected]
Reproductive Impairment in the Zebrafish, Danio rerio, upon Chronic Exposure to 1,2,3-Trichlorobenzene
Ecotoxicology and Environmental Safety 48, 196}201 (2001) Environmental Research, Section B doi:10.1006/eesa.2000.2029, available online at http://www.idealibrary.com on
Reproductive Impairment in the Zebrafish, Danio rerio, upon Chronic Exposure to 1,2,3-Trichlorobenzene Erwin W. M. Roex, MarieK lla Giovannangelo, and Cornelis A. M. van Gestel Vrije Universiteit, Institute of Ecological Science, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands Received March 30, 2000
Most organic pollutants are supposed to act via the mechanism of nonpolar narcosis upon acute exposure. Because the chronic e4ects of these compounds are still relatively unknown, in this study a chronic toxicity experiment was performed with zebra5sh, Danio rerio, exposed to 1,2,3-trichlorobenzene (123TCB), a nonpolar narcotic. Fish were exposed in a 6owthrough system for 68 and 147 days. Parameters measured are survival, growth, reproduction, and glycogen and protein content. The only parameter which was in6uenced was the number of eggs produced per female, resulting in an EC50 of 40 lg/L. Using this value and acute toxicity data for 123TCB, an acute to chronic ratio (ACR) of 80 was calculated, which is larger than ACRs for other species exposed to nonpolar narcotics. This 5nding might indicate that compounds acting by nonpolar narcosis in acute tests can have completely di4erent e4ects upon chronic exposure. 2001 Academic Press
Key Words: reproduction; chronic.
zebra5sh;
nonpolar
narcotic;
INTRODUCTION
Environmental risk assessment of chemicals is mostly based on standardized toxicity tests, usually lasting for only a few days and performed with a limited number of selected organisms. The results of these standardized tests are extrapolated to safety limits by using "xed factors (Okkerman et al., 1991). These extrapolation factors depend on the amount and the quality of the data available, but do not take into account the mode of action of the compound, the life history, or the physiology of the test species. Four classes of chemicals can be distinguished by their mode of action (Verhaar et al., 1992): (1) nonpolar narcotics, (2) polar narcotics, (3) reactive chemicals, and (4) speci"cally acting chemicals. This classi"cation is based on results of To whom correspondence should be addressed at present address: National Institute of Public Health and the Environment, P.O. Box 1, 3720 BA Bilthoven, The Netherlands. Fax: #31302744401. E-mail: erwin. [email protected].
acute toxicity tests. Because mortality is the most important parameter in these tests, relatively high concentrations must be used. In the environment, however, organisms are usually exposed to sublethal concentrations of toxicants for a long time. Because of processes like bioactivation and induction of metabolizing enzymes, other modes of action can become active upon chronic exposure, which do not appear in acute tests. Kenaga (1982) stated that the formation of more toxic metabolites can result in large di!erences between acute and chronic e!ect concentrations for certain compounds. These di!erences may exceed the extrapolation factors, which are used in risk assessment. Furthermore, important life-cycle parameters of the test organism may be a!ected after chronic exposure, such as growth, development, and reproduction. These parameters are usually not taken into account in acute tests. Small adverse e!ects on di!erent life-history parameters could seriously damage the stability of a population in the "eld. Chronic life-cycle toxicity tests, in which these parameters were included, would be more ecologically relevant than the standardized short-term toxicity tests (Van Straalen and Kammenga, 1998). Because of time and cost-consuming aspects, mostly small invertebrate species are chosen as test organisms in chronic tests. Only few toxicity tests have been performed which investigate the chronic e!ects of compounds on aquatic vertebrates, like "sh species, and the compounds tested in chronic "sh toxicity tests mostly belong to the classes of polar narcotics and speci"cally acting chemicals. Although the data are rather scarce, "sh species demonstrate the largest di!erences between acute and chronic toxicity, independent on the mode of action upon acute exposure (Roex et al., 2000). However, most organic chemicals are supposed to act via the mechanism of nonpolar narcosis (Veith et al., 1983). This mechanism has drawn a lot of attention and a lot of e!ort has been put into modeling acute toxicity, for example, by means of QSARs (KoK nemann, 1981; Veith et al., 1983; Klopman et al., 1999). Chronic e!ect levels of compounds
196 0147-6513/01 $35.00 Copyright 2001 by Academic Press All rights of reproduction in any form reserved.
EFFECTS OF 1,2,3-TRICHLOROBENZENE TO ZEBRAFISH
197
Zebra"sh (D. rerio) were obtained from a commercial supplier and were approximately 2 months old at arrival in our laboratory. The "sh were acclimated in glass aquaria containing copper-free Amsterdam tap water for 2 weeks before use in the experiment. During the acclimation, "sh were fed daily with commercial "sh food (Tetraphyll). The length of the "sh at the start of the experiment was 25.3$2.6 mm, weight was 164.6$48.2 mg (n"7).
solution with a higher concentration was prepared to reach the desired concentrations in the water phase, i.e., 30, 90, 300, and 900 lg/L. Every treatment was performed in duplicate, including a control and a solvent control, resulting in a total of 12 aquaria. The highest amount of acetone that was used in the concentration range, 0.1 ml/L, was added to the solvent control. The two duplicates of each treatment were positioned on the left and the right side of the climate chamber and were designated as treatments I and II. The experiments were started with 20 animals per aquarium. Fish were fed commercial "sh food (Tetraphyll) at a level of approximately 1% of their body weight per day. They were fed twice daily during workdays and once a day during the weekend. During exposure the following water parameters were checked frequently: #ow rate, pH, hardness, oxygen, and nitrate and nitrite concentrations. After 68 days of exposure half the "sh of every group were sacri"ced. Animals were killed in liquid nitrogen and stored at !203C before analysis. From Days 110}125, the number of eggs per aquarium were counted. This was established by placing a nylon net between the "sh and the bottom of the aquaria, thus preventing the adult "sh from predating on the eggs. After 147 days, the remaining "sh were sampled. After sampling, the animals were weighed individually, their length was measured, and the sex was determined. Reproduction was expressed as the number of eggs produced per female per 10 days. The reproduction data were plotted in a dose-response relationship, from which an EC was calculated by curve "tting according to Haanstra et al. (1985).
Chemicals
Physiological Measurements
1,2,3-Trichlorobenzene was obtained from Aldrich Chemical Company (purity 99%). Hexane ('99.5%) and acetone ('99.5%) were purchased from J. T. Baker. Pentachlorobenzene ('99%) was obtained from Fluka AG.
Half of every group was used for determination of protein and glycogen. Individual "sh were homogenized on ice in 9 ml of 0.05 M Tris/HCl, pH 8.0, bu!er, using a motordriven glass homogenizer. Glycogen determination was done according to the method of Roe and Daily (1966) with minor modi"cations. It is known that some tissue components may interfere with the glycogen measurements (Roe and Daily, 1966; Lavy and Verhoef, 1996). To preclude this interference, the homogenates were "rst centrifuged at 10,000 g for 10 min at 43C. Tests revealed that this procedure had no signi"cant e!ect on the amount of glycogen extracted. The remaining supernatants were stored at !803C until further analysis. The same supernatant was used for both protein and glycogen determination. For glycogen determination, duplicates of 250 or 650 ll supernatant were digested in 1 ml KOH (30%) for 15 min at 1003C. Subsequently glycogen was precipitated with 50 ll of a saturated Na SO solution and 2 ml of ethanol (96%). After 15 min of centrifugation at 2000 rpm, the pellet was washed twice with 200 ll H O and precipitated again using 400 ll ethanol and centrifugation. The remaining pellet was
acting via nonpolar narcosis are lacking (Woltering, 1984) and the relationship with acute toxicity is unclear. The objectives of this study were to investigate the chronic e!ects of a compound, acting via the mechanism of nonpolar narcosis, on an aquatic vertebrate. As a model organism the zebra"sh, Danio rerio, was used, because it is easy to maintain and breed in the laboratory (Laale, 1977). During a period of 147 days, groups of zebra"sh were exposed to di!erent concentrations of 1,2,3-trichlorobenzene (123TCB). In addition to the life-history parameters, survival, growth, and reproduction, some physiological parameters were investigated. Stress, chemical or otherwise, may cause a depletion of energy reserves, such as glycogen (DangeH , 1986) or protein (De Boeck et al., 1997; Sancho et al., 1997) because of an enhanced energy demand. The way in which narcotic compounds, like 123TCB, a!ect energy reserves is not yet known. Therefore, whole-body glycogen and protein contents of the "sh were measured. MATERIALS AND METHODS
Animals
Exposure Fish were exposed in a #owthrough system, which was set up in a climate chamber. The temperature was 243C and the light regime was 12/12 (L/D). Aliquots of 1 ml of stock solution of 123TCB in acetone were added to 10-L vessels, which contained 10 L of copper-free Amsterdam tap water. Solutions were prepared freshly every day and were stirred for at least 4 h before use. A peristaltic pump was employed to deliver the desired volumes of the contaminated water as well as clean tap water to the exposure aquaria, which had a volume of 10 L. In this way four di!erent concentrations were obtained. Preliminary experiments indicated that approximately 30% of the 123TCB in the water was adsorbed by the glass wall. It is also known that 123TCB is a relatively volatile compound. Because of these "ndings, a stock
198
ROEX, GIOVANNANGELO, AND VAN GESTEL
diluted in 100 ll H O an 1 ml of anthrone reagent was added. After 15 min at 903C, glycogen was determined spectrophotometrically at 620 nm. Glycogen content was expressed as milligram per gram wet weight. Duplicates of 10 ll supernatant were used for protein determination of the individual samples according to the method of Bradford (1976). Total protein was determined spectrophotometrically at 595 nm, using bovine serum albumin (BSA) as a standard. Protein content was expressed as milligram per gram wet weight.
Extraction and Analysis of Water Samples Actual concentrations of 123TCB were checked every 5 days. For this purpose 20}25 ml of the test solutions was extracted with 5 ml of hexane. Before extraction, 100 ll of pentachlorobenzene in acetone (10 lg/ml) was added to each water sample. Pentachlorobenzene served as an internal standard to determine the recovery of the extraction procedure. Results were corrected for recovery. After extraction, the hexane fractions were analyzed on a Hewlet Packard 5890 gas chromatograph, equipped with a 50 m HP Ultra 2 column and an electron capture detector (ECD) in the splitless mode. The injection volume was 1 ll. The temperatures of the detector and the injector were 275 and 3003C, respectively. Oven temperature was programmed at 903C for 3 min, raised by 103C per min to 2003C, and kept at 2003C for 15 min.
Statistical Analysis Results of weight, protein, and glycogen determinations were analyzed with a nested analysis of variance, to account for &&climate chamber side'' e!ects (see under Exposure). When the data were not normally distributed or when variances were not homogeneous, data transformation was included. Tukey's test for multiple comparison of means was applied, when the F value for the treatment was signi"cant. All statistical analyses were done using the SYSTAT 5.2.1 software package (SYSTAT, 1992). The signi"cance level in all tests was 0.05. RESULTS
Test Conditions The measured water parameters did not reveal any significant deviations. The average parameter values were as follows (means and standard deviations): pH, 7.9$0.2; O , 6.6$1.7 mg/L; CaCO , 193$24 mg/L. Nitrate concentra tion was below 10 mg/L and nitrite concentration never exceeded a concentration of 0.1 mg/L during the time of exposure.
TABLE 1 Average Aqueous Concentrations of 123TCB (lg/L) and Standard Deviations in the Di4erent Treatments during the Chronic Toxicity Study with Danio rerioa Nominal concentrations 30-I 30-II 90-I 90-II 300-I 300-II 900-I 900-II
Actual concentrations 47.8$30.0 (24) 33.3$22.0 (22) 54.8$43.9 (24) 49.2$37.8 (26) 110$87.2 (26) 86.7$61.7 (27) 226$221 (22) 276$233 (23)
? Number of replicates is given in parentheses. The series designated as I and II were positioned on di!erent sides of the climate chamber.
In Table 1, the actual aqueous concentrations of 123TCB during the total exposure period are given. It is obvious that the actual concentrations #uctuated over time. However, despite these #uctuations, a range of concentrations could still be demonstrated. A few days after the beginning of the exposure, a slimy whitish cover started to form on the glass wall in the aquaria, and later on also in the stock vessels. This cover is most likely caused by the acetone, because this phenomenon did not take place in the control treatments. Because of the presence of this cover the aquaria had to be cleaned regularly. Only after this cleaning, the actual concentrations reached the desired levels. The average recovery of the internal standard, added in the chemical analysis, was 79$12%. Further on, average actual concentrations will be used as exposure concentrations. Survival, Weight, and Reproduction Analysis of variance for growth, reproduction, and physiological parameters demonstrated that there was no di!erence between aquaria of the same treatment positioned on the left and the right side of the climate chamber. The aquaria were therefore considered as true duplicates. Because no di!erences were observed between controls and solvent controls, the results of these groups were pooled before further statistical analysis. No signi"cant di!erences between the survival rates of the treatments or the controls were observed. Overall survival was between 85 and 100%, except for treatment 300-II, where survival was 65%. At Day 130 all the animals in treatment 90-I died. Because hardly any mortality was observed in one of the other treatments, this e!ect is supposed to have a di!erent cause than exposure to 123TCB. At both sampling dates, the average wet weight of the exposed "sh was lower than that in the control groups, but this reduction in growth was not signi"cant, and no doseresponse relation was observed. Mean weight of the control
EFFECTS OF 1,2,3-TRICHLOROBENZENE TO ZEBRAFISH
199
DISCUSSION
FIG. 1. Number of eggs per female over a period of 10 days, after exposure for 110 days to mean actual 123TCB concentrations. The line represents the dose-response curve "tted according to the log-logistic model from Haanstra et al. (1985).
"sh after 68 days was 235$52 mg, and of the exposed "sh 201$45 mg. After 147 days, the average weight of the control "sh was 275$56 mg, of the exposed "sh 247$ 51 mg. In Fig. 1 the average egg production per female is plotted against the average aqueous concentration of 123TCB in the di!erent treatments. A clear dose-response relationship was observed for the e!ect of 123TCB. The EC on the basis of actual measured concentrations was calculated to be 40 lg/L; the 95% con"dence interval was 31}51 lg/L. Physiological Parameters Glycogen and protein contents at the start of the experiment were 2.3$1.0 and 76.6$19.5 mg/g wet weight, respectively. Treatment 90-1 indicated a signi"cantly reduced protein and glycogen content at the end of the experiment, most probably related to the abrupt lethality in this group. No signi"cant di!erences in glycogen and protein content were observed between treatments or treatments and controls for the two exposure times, but a strong decrease in overall glycogen content during the exposure period was observed. The average protein content of the control "sh after 68 days was 109$13 mg/g, of the exposed "sh was 103$10 mg/g wet weight. After 147 days, both protein contents were 95$18 and 87$15 mg/g wet weight for control, and exposed "sh, respectively. The average glycogen content of the control "sh after 68 days of exposure was 0.201$0.100 mg/g, of the exposed "sh was 0.203$0.081 mg/g wet weight. After 147 days, both contents were 0.071$0.025 and 0.096$0.038 mg/g wet weight for control, and exposed "sh, respectively.
123TCB concentrations in the aquaria #uctuated over time, and actual concentrations almost never reached the nominal ones. The decrease of aqueous concentrations of 123TCB, or related compounds (124TCB), has also been found in other studies with #owthrough tests (KoK nemann and van Leeuwen, 1980; Smith et al., 1991; van Eck et al., 1997). KoK nemann and van Leeuwen (1980) demonstrated that the whitish cover in their aquaria consisted of bacteria, whose growth was stimulated by acetone. The biomass of bacteria in their experiment contained a substantial amount of 123TCB. The fact that aqueous concentrations were highest directly after cleaning of the aquaria indicates that in this experiment the bacteria also absorbed a signi"cant part of the 123TCB. Loss of the compound due to evaporation will also have contributed to the concentration decrease observed (van Eck et al., 1997). The whole-body glycogen and protein contents were not signi"cantly a!ected due to exposure, but a time-dependent decrease in the case of glycogen was observed for all treatments, including the controls. Two explanations are reasonable for this time-dependent change in carbohydrate metabolism. First, there is an increasing demand for energy, because of the enhancing reproductive output of the animals during the experiment. In three-spined sticklebacks glycogen content decreases when the reproductive season starts, in males as well in females (Wootton et al., 1980; Chellappa et al., 1989). A second explanation are the suboptimal conditions in which the organisms remained in the experiment. It is known that, for example, handling has a negative e!ect on the glycogen content of teleosts (Reubush and Heath, 1996). This could also be the case in the current experiment. Both total protein and glycogen contents are probably not subtle enough to elucidate sublethal e!ects of the chronic exposure to 123TCB. No e!ects on survival were observed. This "nding is in agreement with other studies. Mosquito "sh (Gambussia a.nis) exposed to the same concentration range of 123TCB did not exhibit any dose-dependent mortality either (Chaisuksant et al., 1998). In that study, however, 123TCB, as well as other chlorobenzenes, was found to reduce the growth of mosquito "sh. In the present study no signi"cant reduction in growth was found. This could be due to di!erences in growth rates between "sh species. Woltering (1984) argued that the growth response in "sh toxicity tests is not suitable in the process of risk assessment, because it is an inconsistent endpoint and not as sensitive as other parameters, as reproduction for example. This is consistent with these current "ndings. To present knowledge, detrimental e!ects of nonpolar narcotic compounds on reproductive parameters in "sh species have not been reported before. As no signi"cant e!ects on growth, energy reserves, or onset of reproduction
200
ROEX, GIOVANNANGELO, AND VAN GESTEL
were found, it is concluded, according to Kooijman and Bedaux (1996), that the reduced egg production is a speci"c e!ect of exposure to 123TCB and does not involve an allocation in energy reserves from reproduction to assimilation, growth, or maintenance. Several studies have found that chlorobenzenes can be metabolized by several "sh species. Di!erent metabolites of chlorobenzene were detected in the water phase after 48 h of exposure of zebra"sh to this compound, although the amounts were relatively low (Kasokat et al., 1989). In vitro metabolism of 123TCB in rainbow trout (Oncorhynchus mykiss) and swordtail (Xiphophorus helleri) liver homogenates was also relatively low (De Wolf et al., 1993). The exposure time in these experiments was short in comparison with the current study. It can, therefore, not be excluded that after long-term exposure the formation of metabolites can play a substantial role in the chronic toxcity of 123TCB and related compounds. This can lead to more speci"c e!ects of these compounds than the e!ects found upon acute exposure. Calamari et al. (1983) established an LC for 123TCB of 3.1 mg/L (95% con"dence limits: 2.3}4.1 mg/L) in the zebra"sh. This value was con"rmed in this laboratory, where an LC of 1.9 mg/L was determined (unpublished results). Relating the LC value of Calamari et al. (1983) to our EC for reproduction, an acute to chronic ratio (ACR) of 80 can be calculated. This ACR is relatively high for the class of nonpolar narcotics, but also for the other classes. In a literature study an average ACR of approximately 3 was found for nonpolar narcotics. This ACR, however was solely based on data obtained from studies with aquatic invertebrates (Roex et al., 2000). On the other hand, ACR in this present study is smaller than the ACRs observed in the literature for "sh species, including zebra"sh, exposed to compounds with other modes of action, i.e., acetylcholinesterase inhibitors and polar narcotics. Chronic tests in those studies were all more than three orders of a magnitude more sensitive than acute tests (Roex et al., 2000). Although only based on one species exposed to one compound, the results of this study corroborate the hypothesis that, in general, teleost exhibit a larger ACR than aquatic invertebrates. However, when comparing results from di!erent studies with "sh species, the mode of action of the compound, and not the life history of the test organism, seems to be the most important factor in determining the relationship between acute and chronic toxicity. CONCLUSION
Although the acute toxicity of nonpolar narcotics can very well be explained by their nonspeci"c mode of action, e.g., hydrophobicity, upon chronic exposure these compounds may exhibit other modes of action. In this way e!ects can occur that cause large ACRs. These ACRs may exceed the extrapolation factors that are used in risk assess-
ment. This makes extrapolation from results of acute toxicity tests to environmental quality criteria, based on "xed extrapolation factors, doubtful. ACKNOWLEDGMENTS The authors thank Frans van der Wielen from the Department of Environmental Toxicology and Chemistry of the University of Amsterdam and Kees Swart of the Institute of Environmental Studies of the Vrije Universiteit for their contribution to the GC analyses and interpretation. Prof. N. M. van Straalen is acknowledged for critically reading the manuscript. This study was "nancially supported by the Dutch Ministry of Transport, Public Works, and Water Management, National Institute for Coastal and Marine Management.
REFERENCES Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248}254. Calamari, D., Galassi, S., Setti, F., and Vighi, M. (1983). Toxicity of selected chlorobenzenes to aquatic organisms. Chemosphere 12, 253}262. Chaisuksant, Y., Yu, Q., and Connell, D. W. (1998). E!ects of halobenzenes on growth rate of "sh (Gambusia a.nis). Ecotoxicol. Environ. Saf. 39, 120}130. Chellappa, S., Huntingford, F. A., Strang, R. H. C., and Thomson, R. Y. (1989). Annual variation in energy reserves in male three-spined stickleback, Gasterosteus aculeatus L. (Pisces, Gasterosteidae). J. Fish. Biol. 35, 275}286. DangeH , A. D. (1986). Changes in carbohydrate metabolism in tilapia, Oreochromis (sarotherodon) mossambicus, during short-term exposure to di!erent types of pollutants. Environ. Pollut. 41, 165}177. De Boeck, G., Vlaeminck, A., and Blust, R. (1997). E!ects of sublethal copper exposure on copper accumulation, food consumption, growth, energy stores, and nucleic acid content in common carp. Arch. Environ. Contam. ¹oxicol. 33, 415}422. De Wolf, W., Seinen, W., and Hermens, J. L. M. (1993). N-Acetyltransferase activity in rainbow trout liver and in vitro biotransformation of chlorinated anilines and benzenes in "sh. Xenobiotica 23, 1045}1056. Haanstra, L., Doelman, P., and Oude Voshaar, J. H. (1985). The use of sigmoidal dose response curves in soil ecotoxicological research. Plant Soil 84, 293}297. Kasokat, T., Nagel, R., and Urich, K. (1989). Metabolism of chlorobenzene and hexachlorobenzene by the zebra "sh, Brachydanio rerio. Bull. Environ. Contam. ¹oxicol. 42, 254}261. Kenaga, E. E. (1982). Predictability of chronic toxicity from acute toxicity of chemicals in "sh and aquatic invertebrates. Environ. ¹oxicol. Chem. 1, 347}358. Klopman, G., Saiakhov, R., Rosenkranz, H. S., and Hermens, J. L. (1999). Multiple computer-automated structure evaluation program study of aquatic toxicity 1: Guppy. Environ. ¹oxicol. Chem. 18, 2497}2505. KoK nemann, H., and Van Leeuwen, K. (1980). Toxicokinetics in "sh: Accumulation and elimination of six chlorobenzenes by guppies. Chemosphere 9, 3}19. KoK nemann, H. (1981). Quantitive structure-activity relationships in "sh toxicity studies. Part I: Relationship for 50 industrial pollutants. ¹oxicology 19, 209}221. Kooijman, S. A. L. M., and Bedaux, J. J. M. (1996). ¹he Analysis of Aquatic ¹oxicity Data. VU Univ. Press, Amsterdam.
EFFECTS OF 1,2,3-TRICHLOROBENZENE TO ZEBRAFISH Laale, H. W. (1977). The biology and use of zebra"sh Brachydanio rerio, in "sheries research. A literature research. J. Fish Biol. 10, 121}173. Lavy, D., and Verhoef, H. A. (1996). E!ects of food quality on growth and body composition of the collembolan Orchesella cincta. Physiol. Entomol. 21, 64}70. Okkerman, P. C., Van der Plassche E. J., Sloo!, W., Van Leeuwen, C. J., and Canton, H. (1991). Ecotoxicological e!ects assessment: A comparison of several extrapolation procedures. Ecotoxicol. Environ. Saf. 21, 182}193. Reubush, K. J., and Heath, A. G. (1996). Metabolic responses to acute handling by "ngerling inland and anadromous striped bass. J. Fish Biol. 49, 830}841. Roe, J. H., and Dailey, R. E. (1966). Determination of glycogen with the anthrone reagent. Anal. Biochem. 15, 245}250. Roex, E. W. M., van Gestel, C. A. M., van Wezel, A. P., and van Straalen, N. M. (2000). Ratios between acute aquatic toxicity and e!ects on population growth rates in relation to toxicant mode of action. Environ. ¹oxicol. Chem. 19, 685}693. Sancho, E., Ferrando, M. D., and Andreu, E. (1997). Sublethal e!ects of an organophosphate insecticide on the European eel, Anguilla anguilla. Ecotoxicol. Environ. Saf. 36, 57}65.
201
Smith, A. D., Bharath, A., Mallard, C., Orr, D., Smith, K., Sutton, J. A., Vukmanich, L. S., McCarty, L. S., and Ozburn, G. W. (1991). The acute and chronic toxicity of ten chlorinated organic compounds to the american #ag"sh (Jordanello #oridae). Arch. Environ. Contam. ¹oxicol. 20, 94}102. SYSTAT, Inc. (1992). SYSTAT 5.2.1. Evanston, IL. Van Eck, J. M. C., Koelmans, A. A., and Deneer, J. W. (1997). Uptake and elimination of 1,2,4-trichlorobenzene in the guppy (Poecilia reticulata) at sublethal and lethal aqueous concentrations. Chemosphere 34, 259}270. Van Straalen, N. M., and Kammenga, J. E. (1998). Assessment of ecotoxicity at the population level using demographic parameters. In Ecotoxicology (G. SchuK uK rman and B. Markert, Eds.), pp. 621}644. Wiley, Chichester, UK. Veith, G. D., Call, D. J., and Brooke, L. T. (1983). Structure-toxicity relationships for the fathead minnow, Pimephales promelas: Narcotic industrial chemicals. Can. J. Fish Aquat. Sci. 40, 743}748. Verhaar, H. J. M., van Leeuwen, C. J., and Hermens, J. L. M. (1992). Classifying environmental pollutants. Chemosphere 25, 471}491. Woltering, D. M. (1984). The growth response in "sh chronic and early life stage toxicity tests: A critical review. Aquat. ¹oxicol. 5, 1}21. Wootton, R. J., Allen, J. R. M., and Cole, S. J. (1980). Energetics of the annual reproductive cycle in female sticklebacks, Gasterosteus aculeatus L. J. Fish Biol. 17, 387}394.
Reproductive Impairment in the Zebrafish, Danio rerio, upon Chronic Exposure to 1,2,3-Trichlorobenzene Erwin W. M. Roex, MarieK lla Giovannangelo, and Cornelis A. M. van Gestel Vrije Universiteit, Institute of Ecological Science, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands Received March 30, 2000
Most organic pollutants are supposed to act via the mechanism of nonpolar narcosis upon acute exposure. Because the chronic e4ects of these compounds are still relatively unknown, in this study a chronic toxicity experiment was performed with zebra5sh, Danio rerio, exposed to 1,2,3-trichlorobenzene (123TCB), a nonpolar narcotic. Fish were exposed in a 6owthrough system for 68 and 147 days. Parameters measured are survival, growth, reproduction, and glycogen and protein content. The only parameter which was in6uenced was the number of eggs produced per female, resulting in an EC50 of 40 lg/L. Using this value and acute toxicity data for 123TCB, an acute to chronic ratio (ACR) of 80 was calculated, which is larger than ACRs for other species exposed to nonpolar narcotics. This 5nding might indicate that compounds acting by nonpolar narcosis in acute tests can have completely di4erent e4ects upon chronic exposure. 2001 Academic Press
Key Words: reproduction; chronic.
zebra5sh;
nonpolar
narcotic;
INTRODUCTION
Environmental risk assessment of chemicals is mostly based on standardized toxicity tests, usually lasting for only a few days and performed with a limited number of selected organisms. The results of these standardized tests are extrapolated to safety limits by using "xed factors (Okkerman et al., 1991). These extrapolation factors depend on the amount and the quality of the data available, but do not take into account the mode of action of the compound, the life history, or the physiology of the test species. Four classes of chemicals can be distinguished by their mode of action (Verhaar et al., 1992): (1) nonpolar narcotics, (2) polar narcotics, (3) reactive chemicals, and (4) speci"cally acting chemicals. This classi"cation is based on results of To whom correspondence should be addressed at present address: National Institute of Public Health and the Environment, P.O. Box 1, 3720 BA Bilthoven, The Netherlands. Fax: #31302744401. E-mail: erwin. [email protected].
acute toxicity tests. Because mortality is the most important parameter in these tests, relatively high concentrations must be used. In the environment, however, organisms are usually exposed to sublethal concentrations of toxicants for a long time. Because of processes like bioactivation and induction of metabolizing enzymes, other modes of action can become active upon chronic exposure, which do not appear in acute tests. Kenaga (1982) stated that the formation of more toxic metabolites can result in large di!erences between acute and chronic e!ect concentrations for certain compounds. These di!erences may exceed the extrapolation factors, which are used in risk assessment. Furthermore, important life-cycle parameters of the test organism may be a!ected after chronic exposure, such as growth, development, and reproduction. These parameters are usually not taken into account in acute tests. Small adverse e!ects on di!erent life-history parameters could seriously damage the stability of a population in the "eld. Chronic life-cycle toxicity tests, in which these parameters were included, would be more ecologically relevant than the standardized short-term toxicity tests (Van Straalen and Kammenga, 1998). Because of time and cost-consuming aspects, mostly small invertebrate species are chosen as test organisms in chronic tests. Only few toxicity tests have been performed which investigate the chronic e!ects of compounds on aquatic vertebrates, like "sh species, and the compounds tested in chronic "sh toxicity tests mostly belong to the classes of polar narcotics and speci"cally acting chemicals. Although the data are rather scarce, "sh species demonstrate the largest di!erences between acute and chronic toxicity, independent on the mode of action upon acute exposure (Roex et al., 2000). However, most organic chemicals are supposed to act via the mechanism of nonpolar narcosis (Veith et al., 1983). This mechanism has drawn a lot of attention and a lot of e!ort has been put into modeling acute toxicity, for example, by means of QSARs (KoK nemann, 1981; Veith et al., 1983; Klopman et al., 1999). Chronic e!ect levels of compounds
196 0147-6513/01 $35.00 Copyright 2001 by Academic Press All rights of reproduction in any form reserved.
EFFECTS OF 1,2,3-TRICHLOROBENZENE TO ZEBRAFISH
197
Zebra"sh (D. rerio) were obtained from a commercial supplier and were approximately 2 months old at arrival in our laboratory. The "sh were acclimated in glass aquaria containing copper-free Amsterdam tap water for 2 weeks before use in the experiment. During the acclimation, "sh were fed daily with commercial "sh food (Tetraphyll). The length of the "sh at the start of the experiment was 25.3$2.6 mm, weight was 164.6$48.2 mg (n"7).
solution with a higher concentration was prepared to reach the desired concentrations in the water phase, i.e., 30, 90, 300, and 900 lg/L. Every treatment was performed in duplicate, including a control and a solvent control, resulting in a total of 12 aquaria. The highest amount of acetone that was used in the concentration range, 0.1 ml/L, was added to the solvent control. The two duplicates of each treatment were positioned on the left and the right side of the climate chamber and were designated as treatments I and II. The experiments were started with 20 animals per aquarium. Fish were fed commercial "sh food (Tetraphyll) at a level of approximately 1% of their body weight per day. They were fed twice daily during workdays and once a day during the weekend. During exposure the following water parameters were checked frequently: #ow rate, pH, hardness, oxygen, and nitrate and nitrite concentrations. After 68 days of exposure half the "sh of every group were sacri"ced. Animals were killed in liquid nitrogen and stored at !203C before analysis. From Days 110}125, the number of eggs per aquarium were counted. This was established by placing a nylon net between the "sh and the bottom of the aquaria, thus preventing the adult "sh from predating on the eggs. After 147 days, the remaining "sh were sampled. After sampling, the animals were weighed individually, their length was measured, and the sex was determined. Reproduction was expressed as the number of eggs produced per female per 10 days. The reproduction data were plotted in a dose-response relationship, from which an EC was calculated by curve "tting according to Haanstra et al. (1985).
Chemicals
Physiological Measurements
1,2,3-Trichlorobenzene was obtained from Aldrich Chemical Company (purity 99%). Hexane ('99.5%) and acetone ('99.5%) were purchased from J. T. Baker. Pentachlorobenzene ('99%) was obtained from Fluka AG.
Half of every group was used for determination of protein and glycogen. Individual "sh were homogenized on ice in 9 ml of 0.05 M Tris/HCl, pH 8.0, bu!er, using a motordriven glass homogenizer. Glycogen determination was done according to the method of Roe and Daily (1966) with minor modi"cations. It is known that some tissue components may interfere with the glycogen measurements (Roe and Daily, 1966; Lavy and Verhoef, 1996). To preclude this interference, the homogenates were "rst centrifuged at 10,000 g for 10 min at 43C. Tests revealed that this procedure had no signi"cant e!ect on the amount of glycogen extracted. The remaining supernatants were stored at !803C until further analysis. The same supernatant was used for both protein and glycogen determination. For glycogen determination, duplicates of 250 or 650 ll supernatant were digested in 1 ml KOH (30%) for 15 min at 1003C. Subsequently glycogen was precipitated with 50 ll of a saturated Na SO solution and 2 ml of ethanol (96%). After 15 min of centrifugation at 2000 rpm, the pellet was washed twice with 200 ll H O and precipitated again using 400 ll ethanol and centrifugation. The remaining pellet was
acting via nonpolar narcosis are lacking (Woltering, 1984) and the relationship with acute toxicity is unclear. The objectives of this study were to investigate the chronic e!ects of a compound, acting via the mechanism of nonpolar narcosis, on an aquatic vertebrate. As a model organism the zebra"sh, Danio rerio, was used, because it is easy to maintain and breed in the laboratory (Laale, 1977). During a period of 147 days, groups of zebra"sh were exposed to di!erent concentrations of 1,2,3-trichlorobenzene (123TCB). In addition to the life-history parameters, survival, growth, and reproduction, some physiological parameters were investigated. Stress, chemical or otherwise, may cause a depletion of energy reserves, such as glycogen (DangeH , 1986) or protein (De Boeck et al., 1997; Sancho et al., 1997) because of an enhanced energy demand. The way in which narcotic compounds, like 123TCB, a!ect energy reserves is not yet known. Therefore, whole-body glycogen and protein contents of the "sh were measured. MATERIALS AND METHODS
Animals
Exposure Fish were exposed in a #owthrough system, which was set up in a climate chamber. The temperature was 243C and the light regime was 12/12 (L/D). Aliquots of 1 ml of stock solution of 123TCB in acetone were added to 10-L vessels, which contained 10 L of copper-free Amsterdam tap water. Solutions were prepared freshly every day and were stirred for at least 4 h before use. A peristaltic pump was employed to deliver the desired volumes of the contaminated water as well as clean tap water to the exposure aquaria, which had a volume of 10 L. In this way four di!erent concentrations were obtained. Preliminary experiments indicated that approximately 30% of the 123TCB in the water was adsorbed by the glass wall. It is also known that 123TCB is a relatively volatile compound. Because of these "ndings, a stock
198
ROEX, GIOVANNANGELO, AND VAN GESTEL
diluted in 100 ll H O an 1 ml of anthrone reagent was added. After 15 min at 903C, glycogen was determined spectrophotometrically at 620 nm. Glycogen content was expressed as milligram per gram wet weight. Duplicates of 10 ll supernatant were used for protein determination of the individual samples according to the method of Bradford (1976). Total protein was determined spectrophotometrically at 595 nm, using bovine serum albumin (BSA) as a standard. Protein content was expressed as milligram per gram wet weight.
Extraction and Analysis of Water Samples Actual concentrations of 123TCB were checked every 5 days. For this purpose 20}25 ml of the test solutions was extracted with 5 ml of hexane. Before extraction, 100 ll of pentachlorobenzene in acetone (10 lg/ml) was added to each water sample. Pentachlorobenzene served as an internal standard to determine the recovery of the extraction procedure. Results were corrected for recovery. After extraction, the hexane fractions were analyzed on a Hewlet Packard 5890 gas chromatograph, equipped with a 50 m HP Ultra 2 column and an electron capture detector (ECD) in the splitless mode. The injection volume was 1 ll. The temperatures of the detector and the injector were 275 and 3003C, respectively. Oven temperature was programmed at 903C for 3 min, raised by 103C per min to 2003C, and kept at 2003C for 15 min.
Statistical Analysis Results of weight, protein, and glycogen determinations were analyzed with a nested analysis of variance, to account for &&climate chamber side'' e!ects (see under Exposure). When the data were not normally distributed or when variances were not homogeneous, data transformation was included. Tukey's test for multiple comparison of means was applied, when the F value for the treatment was signi"cant. All statistical analyses were done using the SYSTAT 5.2.1 software package (SYSTAT, 1992). The signi"cance level in all tests was 0.05. RESULTS
Test Conditions The measured water parameters did not reveal any significant deviations. The average parameter values were as follows (means and standard deviations): pH, 7.9$0.2; O , 6.6$1.7 mg/L; CaCO , 193$24 mg/L. Nitrate concentra tion was below 10 mg/L and nitrite concentration never exceeded a concentration of 0.1 mg/L during the time of exposure.
TABLE 1 Average Aqueous Concentrations of 123TCB (lg/L) and Standard Deviations in the Di4erent Treatments during the Chronic Toxicity Study with Danio rerioa Nominal concentrations 30-I 30-II 90-I 90-II 300-I 300-II 900-I 900-II
Actual concentrations 47.8$30.0 (24) 33.3$22.0 (22) 54.8$43.9 (24) 49.2$37.8 (26) 110$87.2 (26) 86.7$61.7 (27) 226$221 (22) 276$233 (23)
? Number of replicates is given in parentheses. The series designated as I and II were positioned on di!erent sides of the climate chamber.
In Table 1, the actual aqueous concentrations of 123TCB during the total exposure period are given. It is obvious that the actual concentrations #uctuated over time. However, despite these #uctuations, a range of concentrations could still be demonstrated. A few days after the beginning of the exposure, a slimy whitish cover started to form on the glass wall in the aquaria, and later on also in the stock vessels. This cover is most likely caused by the acetone, because this phenomenon did not take place in the control treatments. Because of the presence of this cover the aquaria had to be cleaned regularly. Only after this cleaning, the actual concentrations reached the desired levels. The average recovery of the internal standard, added in the chemical analysis, was 79$12%. Further on, average actual concentrations will be used as exposure concentrations. Survival, Weight, and Reproduction Analysis of variance for growth, reproduction, and physiological parameters demonstrated that there was no di!erence between aquaria of the same treatment positioned on the left and the right side of the climate chamber. The aquaria were therefore considered as true duplicates. Because no di!erences were observed between controls and solvent controls, the results of these groups were pooled before further statistical analysis. No signi"cant di!erences between the survival rates of the treatments or the controls were observed. Overall survival was between 85 and 100%, except for treatment 300-II, where survival was 65%. At Day 130 all the animals in treatment 90-I died. Because hardly any mortality was observed in one of the other treatments, this e!ect is supposed to have a di!erent cause than exposure to 123TCB. At both sampling dates, the average wet weight of the exposed "sh was lower than that in the control groups, but this reduction in growth was not signi"cant, and no doseresponse relation was observed. Mean weight of the control
EFFECTS OF 1,2,3-TRICHLOROBENZENE TO ZEBRAFISH
199
DISCUSSION
FIG. 1. Number of eggs per female over a period of 10 days, after exposure for 110 days to mean actual 123TCB concentrations. The line represents the dose-response curve "tted according to the log-logistic model from Haanstra et al. (1985).
"sh after 68 days was 235$52 mg, and of the exposed "sh 201$45 mg. After 147 days, the average weight of the control "sh was 275$56 mg, of the exposed "sh 247$ 51 mg. In Fig. 1 the average egg production per female is plotted against the average aqueous concentration of 123TCB in the di!erent treatments. A clear dose-response relationship was observed for the e!ect of 123TCB. The EC on the basis of actual measured concentrations was calculated to be 40 lg/L; the 95% con"dence interval was 31}51 lg/L. Physiological Parameters Glycogen and protein contents at the start of the experiment were 2.3$1.0 and 76.6$19.5 mg/g wet weight, respectively. Treatment 90-1 indicated a signi"cantly reduced protein and glycogen content at the end of the experiment, most probably related to the abrupt lethality in this group. No signi"cant di!erences in glycogen and protein content were observed between treatments or treatments and controls for the two exposure times, but a strong decrease in overall glycogen content during the exposure period was observed. The average protein content of the control "sh after 68 days was 109$13 mg/g, of the exposed "sh was 103$10 mg/g wet weight. After 147 days, both protein contents were 95$18 and 87$15 mg/g wet weight for control, and exposed "sh, respectively. The average glycogen content of the control "sh after 68 days of exposure was 0.201$0.100 mg/g, of the exposed "sh was 0.203$0.081 mg/g wet weight. After 147 days, both contents were 0.071$0.025 and 0.096$0.038 mg/g wet weight for control, and exposed "sh, respectively.
123TCB concentrations in the aquaria #uctuated over time, and actual concentrations almost never reached the nominal ones. The decrease of aqueous concentrations of 123TCB, or related compounds (124TCB), has also been found in other studies with #owthrough tests (KoK nemann and van Leeuwen, 1980; Smith et al., 1991; van Eck et al., 1997). KoK nemann and van Leeuwen (1980) demonstrated that the whitish cover in their aquaria consisted of bacteria, whose growth was stimulated by acetone. The biomass of bacteria in their experiment contained a substantial amount of 123TCB. The fact that aqueous concentrations were highest directly after cleaning of the aquaria indicates that in this experiment the bacteria also absorbed a signi"cant part of the 123TCB. Loss of the compound due to evaporation will also have contributed to the concentration decrease observed (van Eck et al., 1997). The whole-body glycogen and protein contents were not signi"cantly a!ected due to exposure, but a time-dependent decrease in the case of glycogen was observed for all treatments, including the controls. Two explanations are reasonable for this time-dependent change in carbohydrate metabolism. First, there is an increasing demand for energy, because of the enhancing reproductive output of the animals during the experiment. In three-spined sticklebacks glycogen content decreases when the reproductive season starts, in males as well in females (Wootton et al., 1980; Chellappa et al., 1989). A second explanation are the suboptimal conditions in which the organisms remained in the experiment. It is known that, for example, handling has a negative e!ect on the glycogen content of teleosts (Reubush and Heath, 1996). This could also be the case in the current experiment. Both total protein and glycogen contents are probably not subtle enough to elucidate sublethal e!ects of the chronic exposure to 123TCB. No e!ects on survival were observed. This "nding is in agreement with other studies. Mosquito "sh (Gambussia a.nis) exposed to the same concentration range of 123TCB did not exhibit any dose-dependent mortality either (Chaisuksant et al., 1998). In that study, however, 123TCB, as well as other chlorobenzenes, was found to reduce the growth of mosquito "sh. In the present study no signi"cant reduction in growth was found. This could be due to di!erences in growth rates between "sh species. Woltering (1984) argued that the growth response in "sh toxicity tests is not suitable in the process of risk assessment, because it is an inconsistent endpoint and not as sensitive as other parameters, as reproduction for example. This is consistent with these current "ndings. To present knowledge, detrimental e!ects of nonpolar narcotic compounds on reproductive parameters in "sh species have not been reported before. As no signi"cant e!ects on growth, energy reserves, or onset of reproduction
200
ROEX, GIOVANNANGELO, AND VAN GESTEL
were found, it is concluded, according to Kooijman and Bedaux (1996), that the reduced egg production is a speci"c e!ect of exposure to 123TCB and does not involve an allocation in energy reserves from reproduction to assimilation, growth, or maintenance. Several studies have found that chlorobenzenes can be metabolized by several "sh species. Di!erent metabolites of chlorobenzene were detected in the water phase after 48 h of exposure of zebra"sh to this compound, although the amounts were relatively low (Kasokat et al., 1989). In vitro metabolism of 123TCB in rainbow trout (Oncorhynchus mykiss) and swordtail (Xiphophorus helleri) liver homogenates was also relatively low (De Wolf et al., 1993). The exposure time in these experiments was short in comparison with the current study. It can, therefore, not be excluded that after long-term exposure the formation of metabolites can play a substantial role in the chronic toxcity of 123TCB and related compounds. This can lead to more speci"c e!ects of these compounds than the e!ects found upon acute exposure. Calamari et al. (1983) established an LC for 123TCB of 3.1 mg/L (95% con"dence limits: 2.3}4.1 mg/L) in the zebra"sh. This value was con"rmed in this laboratory, where an LC of 1.9 mg/L was determined (unpublished results). Relating the LC value of Calamari et al. (1983) to our EC for reproduction, an acute to chronic ratio (ACR) of 80 can be calculated. This ACR is relatively high for the class of nonpolar narcotics, but also for the other classes. In a literature study an average ACR of approximately 3 was found for nonpolar narcotics. This ACR, however was solely based on data obtained from studies with aquatic invertebrates (Roex et al., 2000). On the other hand, ACR in this present study is smaller than the ACRs observed in the literature for "sh species, including zebra"sh, exposed to compounds with other modes of action, i.e., acetylcholinesterase inhibitors and polar narcotics. Chronic tests in those studies were all more than three orders of a magnitude more sensitive than acute tests (Roex et al., 2000). Although only based on one species exposed to one compound, the results of this study corroborate the hypothesis that, in general, teleost exhibit a larger ACR than aquatic invertebrates. However, when comparing results from di!erent studies with "sh species, the mode of action of the compound, and not the life history of the test organism, seems to be the most important factor in determining the relationship between acute and chronic toxicity. CONCLUSION
Although the acute toxicity of nonpolar narcotics can very well be explained by their nonspeci"c mode of action, e.g., hydrophobicity, upon chronic exposure these compounds may exhibit other modes of action. In this way e!ects can occur that cause large ACRs. These ACRs may exceed the extrapolation factors that are used in risk assess-
ment. This makes extrapolation from results of acute toxicity tests to environmental quality criteria, based on "xed extrapolation factors, doubtful. ACKNOWLEDGMENTS The authors thank Frans van der Wielen from the Department of Environmental Toxicology and Chemistry of the University of Amsterdam and Kees Swart of the Institute of Environmental Studies of the Vrije Universiteit for their contribution to the GC analyses and interpretation. Prof. N. M. van Straalen is acknowledged for critically reading the manuscript. This study was "nancially supported by the Dutch Ministry of Transport, Public Works, and Water Management, National Institute for Coastal and Marine Management.
REFERENCES Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248}254. Calamari, D., Galassi, S., Setti, F., and Vighi, M. (1983). Toxicity of selected chlorobenzenes to aquatic organisms. Chemosphere 12, 253}262. Chaisuksant, Y., Yu, Q., and Connell, D. W. (1998). E!ects of halobenzenes on growth rate of "sh (Gambusia a.nis). Ecotoxicol. Environ. Saf. 39, 120}130. Chellappa, S., Huntingford, F. A., Strang, R. H. C., and Thomson, R. Y. (1989). Annual variation in energy reserves in male three-spined stickleback, Gasterosteus aculeatus L. (Pisces, Gasterosteidae). J. Fish. Biol. 35, 275}286. DangeH , A. D. (1986). Changes in carbohydrate metabolism in tilapia, Oreochromis (sarotherodon) mossambicus, during short-term exposure to di!erent types of pollutants. Environ. Pollut. 41, 165}177. De Boeck, G., Vlaeminck, A., and Blust, R. (1997). E!ects of sublethal copper exposure on copper accumulation, food consumption, growth, energy stores, and nucleic acid content in common carp. Arch. Environ. Contam. ¹oxicol. 33, 415}422. De Wolf, W., Seinen, W., and Hermens, J. L. M. (1993). N-Acetyltransferase activity in rainbow trout liver and in vitro biotransformation of chlorinated anilines and benzenes in "sh. Xenobiotica 23, 1045}1056. Haanstra, L., Doelman, P., and Oude Voshaar, J. H. (1985). The use of sigmoidal dose response curves in soil ecotoxicological research. Plant Soil 84, 293}297. Kasokat, T., Nagel, R., and Urich, K. (1989). Metabolism of chlorobenzene and hexachlorobenzene by the zebra "sh, Brachydanio rerio. Bull. Environ. Contam. ¹oxicol. 42, 254}261. Kenaga, E. E. (1982). Predictability of chronic toxicity from acute toxicity of chemicals in "sh and aquatic invertebrates. Environ. ¹oxicol. Chem. 1, 347}358. Klopman, G., Saiakhov, R., Rosenkranz, H. S., and Hermens, J. L. (1999). Multiple computer-automated structure evaluation program study of aquatic toxicity 1: Guppy. Environ. ¹oxicol. Chem. 18, 2497}2505. KoK nemann, H., and Van Leeuwen, K. (1980). Toxicokinetics in "sh: Accumulation and elimination of six chlorobenzenes by guppies. Chemosphere 9, 3}19. KoK nemann, H. (1981). Quantitive structure-activity relationships in "sh toxicity studies. Part I: Relationship for 50 industrial pollutants. ¹oxicology 19, 209}221. Kooijman, S. A. L. M., and Bedaux, J. J. M. (1996). ¹he Analysis of Aquatic ¹oxicity Data. VU Univ. Press, Amsterdam.
EFFECTS OF 1,2,3-TRICHLOROBENZENE TO ZEBRAFISH Laale, H. W. (1977). The biology and use of zebra"sh Brachydanio rerio, in "sheries research. A literature research. J. Fish Biol. 10, 121}173. Lavy, D., and Verhoef, H. A. (1996). E!ects of food quality on growth and body composition of the collembolan Orchesella cincta. Physiol. Entomol. 21, 64}70. Okkerman, P. C., Van der Plassche E. J., Sloo!, W., Van Leeuwen, C. J., and Canton, H. (1991). Ecotoxicological e!ects assessment: A comparison of several extrapolation procedures. Ecotoxicol. Environ. Saf. 21, 182}193. Reubush, K. J., and Heath, A. G. (1996). Metabolic responses to acute handling by "ngerling inland and anadromous striped bass. J. Fish Biol. 49, 830}841. Roe, J. H., and Dailey, R. E. (1966). Determination of glycogen with the anthrone reagent. Anal. Biochem. 15, 245}250. Roex, E. W. M., van Gestel, C. A. M., van Wezel, A. P., and van Straalen, N. M. (2000). Ratios between acute aquatic toxicity and e!ects on population growth rates in relation to toxicant mode of action. Environ. ¹oxicol. Chem. 19, 685}693. Sancho, E., Ferrando, M. D., and Andreu, E. (1997). Sublethal e!ects of an organophosphate insecticide on the European eel, Anguilla anguilla. Ecotoxicol. Environ. Saf. 36, 57}65.
201
Smith, A. D., Bharath, A., Mallard, C., Orr, D., Smith, K., Sutton, J. A., Vukmanich, L. S., McCarty, L. S., and Ozburn, G. W. (1991). The acute and chronic toxicity of ten chlorinated organic compounds to the american #ag"sh (Jordanello #oridae). Arch. Environ. Contam. ¹oxicol. 20, 94}102. SYSTAT, Inc. (1992). SYSTAT 5.2.1. Evanston, IL. Van Eck, J. M. C., Koelmans, A. A., and Deneer, J. W. (1997). Uptake and elimination of 1,2,4-trichlorobenzene in the guppy (Poecilia reticulata) at sublethal and lethal aqueous concentrations. Chemosphere 34, 259}270. Van Straalen, N. M., and Kammenga, J. E. (1998). Assessment of ecotoxicity at the population level using demographic parameters. In Ecotoxicology (G. SchuK uK rman and B. Markert, Eds.), pp. 621}644. Wiley, Chichester, UK. Veith, G. D., Call, D. J., and Brooke, L. T. (1983). Structure-toxicity relationships for the fathead minnow, Pimephales promelas: Narcotic industrial chemicals. Can. J. Fish Aquat. Sci. 40, 743}748. Verhaar, H. J. M., van Leeuwen, C. J., and Hermens, J. L. M. (1992). Classifying environmental pollutants. Chemosphere 25, 471}491. Woltering, D. M. (1984). The growth response in "sh chronic and early life stage toxicity tests: A critical review. Aquat. ¹oxicol. 5, 1}21. Wootton, R. J., Allen, J. R. M., and Cole, S. J. (1980). Energetics of the annual reproductive cycle in female sticklebacks, Gasterosteus aculeatus L. J. Fish Biol. 17, 387}394.