Short-Term Toxicity of Lindane, Hexachlorobenzene, and Copper Sulfate to Tubificid Sludgeworms (Oligochaeta) in Artificial Media

ECOTOXICOLOGY AND ENVIRONMENTAL SAFETY

39, 10—20 (1998)

ENVIRONMENTAL RESEARCH, SECTION B ARTICLE NO.

ES971603

Short-Term Toxicity of Lindane, Hexachlorobenzene, and Copper Sulfate to Tubificid Sludgeworms (Oligochaeta) in Artificial Media Michael Meller,*,1 P. Egeler,* J. Ro¨mbke,* H. Schallnass,* R. Nagel,- and B. Streit‡ * ECT Oekotoxikologie GmbH, Bo( ttgerstrasse 2-14, D-65439 Flo( rsheim/M, Germany; -Technische Universita( t Dresden, Institut fu( r Hydrobiologie, D-01062 Dresden, Germany; and ‡ J. W. Goethe-Universita( t Frankfurt, Abt, O$ kologie und Evolution, D-60054Frankfurt/M, Germany Received April 23, 1997

have been standardized on a national level (EPA, 1994; ASTM, 1995; BBA, 1995), and none have been accepted internationally, e.g., as an OECD guideline. Important criteria for choosing the test species include (1) the ecological relevance of the organism with respect to its local and global distribution and its functional role within the ecosystem; (2) its tolerance to a wide range of abiotic sediment characteristics; (3) easy handling and culturing (Hill et al., 1993). It is well known that tubificids fulfill these requirements (e.g., Wachs, 1967; Hill et al., 1993). Another major selection criterion is the sensitivity of the test species. Tubificids do have a reputation for being very tolerant of chemical stress (Chapman and Brinkhurst, 1984), but it has been found that this is not the case regarding sublethal effects (Keilty et al., 1988a, b; Reynoldson et al., 1991). Furthermore, as sediment-ingesting endobenthic animals, they are subject to exposure to sediment-bound substances by all potential routes (overlying water, interstitial water, and ingestion of sediment). Taking into consideration all of these characteristics, tubificids are regarded as suitable tools for ecotoxicological research. Depending on the source of natural sediment, the presence of micropollutants as well as indigenous organisms can influence toxicity tests (Reynoldson et al., 1994; Suedel and Rodgers, 1994). In addition, comparison of results from various laboratories is further complicated by the wide variety of abiotic sediment characteristics. It is therefore recommended that an artificial test medium be used to standardize whole-sediment toxicity tests. The purpose of this study was to develop a test method to assess not only the lethal effects but more notably the sublethal effects caused by chemicals. The acute toxicity test with ¹ubifex tubifex described by Ammon (1985) was used as the basis for this work. Particular attention was paid to quick and simple performance of the test as well as the use of an artificial sediment. For a better evaluation of the variations between different species, two tubificid species (¹. tubifex, ¸imnodrilus hoffmeisteri) were used in all of the experiments carried out in this study.

The toxicity of lindane, hexachlorobenzene, and copper sulfate to Tubifex tubifex and Limnodrilus hoffmeisteri was determined using an easily applicable and standardizable 72-h short-term test system. It was designed for the quick assessment of sublethal and lethal effects of sediment-associated chemicals on the worms. An artificial sediment based on the Artificial Soil according to OECD Guideline No. 207 was used as test medium. The data confirm the common view that oligochaetes are highly tolerant of lethal effects. However, sublethal effects were detected at considerably lower concentrations than found for lethal effects. The EC50 values for autotomy (172 mg/kg dry wt sediment) and sediment avoidance (217 mg/kg) for T. tubifex exposed to lindane-contaminated sediment were, for example, more than five '1000 mg/kg). The no-obtimes lower than the LC50 value (' served-effect concentration for reworking activity (8 mg/kg) was more than 125 times lower than the LC50 . Tubificids thus turned out to represent useful test organisms for the assessment of the ecotoxicological hazard potential of chemicals in the sediment compartment, because the sublethal effects not only affect the individual, but can influence the population levels and, consequently, the composition of the benthic community. ( 1998 Academic Press

Key Words: artificial sediment; tubificids; autotomy; sediment avoidance; reworking activity; lindane; hexachlorobenzene; copper sulfate.

INTRODUCTION

Over the last few years, many representatives of freshwater benthos fauna were examined for their suitability as test organisms for the testing of chemicals, as well as for the ecotoxicological assessment of sediment contamination. Most of the tested species were insects (e.g., Lydy et al., 1990; Fleming et al., 1994), oligochaetes (e.g., Ammon, 1985; Reynoldson et al., 1991; Phipps et al., 1993), and amphipods (e.g., Becker et al., 1995). To date, only a few test methods 1 To whom correspondence should be addressed. 0147-6513/98 $25.00 Copyright ( 1998 by Academic Press All rights of reproduction in any form reserved.

10

SHORT-TERM TOXICITY TO TUBIFICIDS

METHODS

Culture Conditions Two common tubificid species were selected for laboratory culture: ¹. tubifex (Mu¨ller) and ¸. hoffmeisteri (Clapare´de). The animals have been continuously kept in the laboratory of ECT Oekotoxikologie GmbH (Flo¨rsheim, Germany) since March 1994. They were originally supplied by FEE Fischfutter Etzbach (Mechernich-Bergheim, Germany). According to the supplier, the animals originated from the River Mass, including its tributaries in Belgium. The worms were kept in semistatic single-species cultures using an artificial sediment and reconstituted water according to OECD Guideline No. 203 (1983) as the overlying medium. Prior to setting up the cultures, mature animals were identified according to Wachs (1967) and Brinkhurst (1971). The identification also was confirmed by P. Rodriguez, Universidad del Pais Vasco, Bilbao. The culture procedure is described in detail by Egeler et al. (1997b).

Artificial Sediment An artificial sediment based on the Artificial Soil according to OECD Guideline No. 207 (OECD, 1984) was used as culture and test medium. Artificial Soil was slightly modified for use as sediment for tubificids by Egeler et al. (1995, 1997a). The substrate was composed (percentages refer to dry weight) of 2% sphagnum peat; 22% kaolinite clay (kaolinite content '30%); 76% quartz sand (grain size: more than 50% of the particles in the range 0.05—0.2 mm); approximately 0.05%. CaCO (pulverized, chemically pure). 3 The air-dried peat was shredded in a chaff-cutter (grain size 41 mm). A suspension of the required amount of peat powder in demineralized water (11.5]dry weight of peat) was prepared using a high-performance homogenizing device. The pH of this suspension was adjusted to 5.5$0.5 with CaCO . To establish a stable microorganism compon3 ent, the suspension was gently stirred for 48 h at room temperature. After this conditioning period the pH was 6.0$0.5. To obtain a homogenous sediment with a water content of approximately 46% of the dry weight of the sediment, the suspension was mixed with the other constituents and demineralized water. The pH of the complete artificial sediment studied be 6.0$0.5. More details and a complete description of the characteristics of this artificial sediment are published elsewhere (Egeler et al., 1997b).

Test Substances The main criteria for selecting test substances were a certain tendency to associate with sediments and ubiquitous occurrence in freshwater sediments. Therefore, two organic chemicals, one with a moderate and the second with a high lipophilicity, and a metal were chosen. Lindane (c-hexa-

11

chlorocyclohexane) represents an insecticide of global distribution in aquatic systems with a moderate low P of 3.63 08 (Rippen, 1991). It was supplied by Sigma Chemical Company (St. Louis, MO; 99% c-hexachlorocyclohexane). The second organic chemical, hexachlorobenzene (HCB) ('99% GC, Fluka Chemika, Buchs, Germany) is a widely distributed hydrophobic pollutant that was used as a fungicide, and is generated as a by-product of chlorinated hydrocarbons. It has a log P of 5.72 (Rippen, 1991). Copper 08 sulfate (CuSO ) 5H O, p.a. quality, Roth, Karlsruhe, Ger2 4 many) was chosen as the metal compound, as it is fairly water soluble [230.5 g/liter, 25°C (Royal Society of Chemistry, 1994)]. Nevertheless, copper is known to be highly particle associated in aquatic systems (e.g., van de Plassche, 1994).

Test System The test system was designed as a 72-h static wholesediment system using artificial sediment and reconstituted water according to OECD Guideline 203 (OECD, 1983) as the overlying medium. The ratio of sediment : overlying water was 1 : 4, following Hooftman et al. (1993).

Spiking Procedure and Test Vessel Setup Test substances were introduced into the system by spiking a bulk sediment for each test concentration, from which different replicates were subsampled before the overlying water was added. Lindane was dissolved in n-hexane and subsequently diluted with n-hexane to prepare an application solution for each required concentration. The quartz sand fraction of the sediment was moistened with a defined volume of this application solution in a glass vessel. After the solvent had evaporated, the coated quartz sand was thoroughly mixed with the other sediment constituents. Application of HCB followed the same spiking procedure, where cyclohexane replaced n-hexane as a solvent. Application of copper sulfate was performed using demineralized water as a solvent. The quartz sand and kaolinite fractions of each bulk sediment were mixed with the peat suspension. Then, the aqueous application solutions were added to this slurry. Concentrations of application solutions and added volumes were calculated to obtain a water content of 46% of dry weight of the sediment. To disperse the test substances homogeneously within the sediment, the spiked bulk sediments were gently stirred for 1 h at room temperature using a magnetic stirring device. Subsequently the sediment was immediately subsampled to the test vessels (100-ml glass tubes, height 14.5 cm, H 3.4 cm), and the overlying water was added using an antiturbation device to minimize turbation of sediment particles. Each test vessel contained a 2-cm layer of spiked artificial sediment and 62.5 ml of reconstituted water. Before

12

MELLER ET AL.

the test organisms were added, the test vessels were incubated for a 24-h equilibration period in a climatic chamber under the described test conditions (Table 1). Range-finding tests with test substance concentrations between 0.01 and 1000 mg/kg (dry weight of the artificial sediment) spaced by a constant factor of 10 according to the OECD Guideline No. 207 (1984) were performed to determine the concentrations for the definitive tests. Range-finding tests with lindane indicated a wide range between sublethal and lethal effects. Since the aim of the study was to compare all described endpoints within one experiment, it was necessary to use a constant factor of 5 for the geometrical concentration series in the definitive test with lindane. In the cases of lindane and HCB the test design included a solvent control; this was prepared by adding only nhexane and cyclohexane, respectively.

Exposure Period Test organisms were sampled from the cultures by sieving the sediment through a 1-mm mesh which retained the adult worms. Using the negative thermotaxis of the tubificids according to the method of Wachs (1965), the animals were removed from the sieves into a vessel containing reconstituted water. Only undamaged, actively creeping adults (with fully developed clitellum) of uniform size (5$2 cm) were selected as test organisms (Ammon, 1985). The worms were transferred randomly into the test vessels using a soft steel forceps or a pipette. An acclimation period was not

required, since the test conditions were identical to culture conditions. The test vessels were then incubated under the described test conditions, in a climatic chamber (Table 1). After 72 h the animals were removed from the sediment. The sediment was suspended with the overlying water by shaking the test vessels. The test organisms could be easily sieved from this sediment suspension using a 1-mm mesh. According to the recommendations of the SETAC Workshop on Sediment Toxicity Assessment (Hill et al., 1993), the sediment—water system was characterized at the beginning and the end of exposure period (see Table 1). To assess sublethal effects on the behavior of the tubificids, the vessels were checked visually after approximately 0.5—1 h and again after 24 and 48 h. At the end of the test, sublethal and lethal effects were recorded. Morphological changes of the worms were determined using a binocular microscope.

Endpoints and Data Analysis Table 2 gives an overview of endpoints and their classifications. The effect rates in precentages (mortality, autotomy, sediment avoidance) for each concentration (i) were calculated according to the following equations: + dead animals in i ]100. Mortality in " i + animals exposed in i

TABLE 1 Test Conditions for Conducting Short-Term Whole-Sediment Toxicity Test Using Tubifex tubifex and Limnodrilus hoffmeisteri Parameter Test system Temperature Photoperiod Test vessel Sediment Overlying water Application Test design/concentrations

Test organisms Loading Sediment quality Overlying water quality Equilibration period Exposure period Endpoints Validity Evaluation

Conditions Static short-term whole-sediment test 20$2°C 16L : 8D; 4100 lx 100-ml glass tubes, height 14.5 cm, H 3.4 cm containing a 2-cm layer sediment and 8 cm overlying water (ratio 1 : 4) Artificial sediment based on Artificial Soil according to OECD Guideline No. 207 (Egeler et al., 1997b) Reconstituted water according to OECD Guideline No. 203 The test substance is applied into the sediment Range-finding test: (0.01), 0.1, 1, 10, 100, 1000 mg/kg (sediment dry weight) including a control and, if necessary, a solvent control using 1 replicate per treatment; definitive test: concentrations based on results of range finding using 4 replicates per treatment Adults with fully developed clitellum of uniform size (5$2 cm), undamaged and actively creeping 10 animals per test vessel pH and redox potential at the beginning and end of test pH, redox potential, hardness, conductivity, ammonia, and dissolved oxygen at the beginning and end of test 24 h referring to test conditions 72 h, no feeding, no aeration Reworking activity, sediment avoidance, autotomy, mortality Sediment avoidance, autotomy, mortality (10% in the control; reworking activity in the lowest concentration should not differ from control Usual statistical treatment to calculate EC and LC , e.g., probit analysis 50 50

SHORT-TERM TOXICITY TO TUBIFICIDS

13

TABLE 2 Description of Endpoints for Short-Term Whole-Sediment Toxicity Test Using Tubifex tubifex and Limnodrilus hoffmeisteri Endpoint

Description

Reworking activity

Sediment avoidance Autotomy

Mortality

The animals leave traces while digging through the sediment (so-called galleries). To estimate the reworking activity, a qualitative comparison on replicate level was performed. The reworking activity of all animals in one test vessel was defined as reduced when the visible number of galleries was distinctly lower than in the control vessels (Fig. 1). According to Keilty et al. (1988a) a worm was considered unburrowed if more than an estimated 75% of its body was visible on the sediment surface. Autotomy starts with a local constriction of the circular muscles, which can be seen macroscopically. The segments behind the constriction are completely autotomized (Kaster, 1979). Autotomy was defined as all animals showing either constriction and/or loss of segments. Animals were recorded as dead when they did not respond to a gentle mechanical stimulus to the front end.

Autotomy in " i + dead

animals in #+ animals showing Autotomy in i i + animals exposed in i

]100. Sediment avoidance in " i +

animals showing Sediment avoidance in i]100. + animals surviving in i

The LC and the EC and their corresponding 95% 50 50 confidence intervals were determined using probit analysis (Finney, 1971). To perform probit analysis, concentrations with 0% effect were set as 1 of n (n"number of exposed or surviving animals) and concentrations exhibiting 100% effect as (n!1) of n. If fewer than three data points between 0 and 100% effect were available, or if there was no linear

dose response after probit transformation, an arcsinus transformation of the data was performed. The LC or 50 EC was then calculated by nonlinear interpolation be50 tween the two concentrations that bracketed the LC /EC value. In this case the 95% confidence interval 50 50 was determined using an independent binomial test (Peltier and Weber, 1985). Since the n of the endpoints mortality, autotomy, and sediment avoidance was small, a no-observed-effect concentration (NOEC) for each endpoint was defined as the highest concentration exhibiting an effect (10%, and lowest-observed-effect concentration (LOEC), as the lowest concentration with an effect 510%. The animals leave traces while digging through the sediment (so-called galleries). The quantity of galleries depends on the activity of the worms in the sediment. To estimate this ‘‘reworking activity,’’ a qualitative comparison on replicate level was performed. The reworking activity of all animals in one test vessel was defined as reduced when the visible number of galleries was distinctly lower than in the control vessels (Fig. 1).

FIG. 1. Schematic figure of a test vessel showing reduced reworking activity and a control vessel. The reworking activity of all animals of one test vessel was defined as reduced when the visible number of galleries was distinctly lower than in the control vessels.

14

MELLER ET AL.

The maximum effect of the endpoint ‘‘reworking activity’’ was met when all replicates of a specific concentration level indicated reduced reworking activity (see Table 2, Fig. 1). NOEC ‘‘reworking activity’’ was defined as the highest concentration with no replicate demonstrating reduced reworking activity, and LOEC reworking activity, as the next higher concentration. RESULTS

In both range-finding and definitive tests, sedimentbound HCB did not cause any sublethal or lethal effects on the two tubificid species up to 1000 mg/kg. Table 3 gives an overview of EC and LOEC/NOEC values obtained from 50 the experiments with lindane and copper sulfate. Lindane influences the reworking activity of ¹. tubifex and ¸. hoffmeisteri at sediment concentrations of 40 and 8 mg/kg, respectively. From a threshold value of 200 mg/kg onward, lethal effects as well as morphological changes (autotomy) and sediment avoidance were observed (see Fig. 2). Since the mortality rates, even at the highest concentration, were far below 50%, no LC values were calculated (Table 3). 50 In the experiments with copper sulfate-spiked artificial sediment, ¸. hoffmeisteri appeared to be more sensitive to the toxicant than ¹. tubifex (Fig. 3). As demonstrated with lindane, copper sulfate caused sublethal effects in both species, although the range between sublethal and lethal effects was a lot smaller (Table 3). Since the number of dead animals was only recorded once, at the end of the test, the rate of sediment avoidance could only be calculated for 72 h. The test design did not allow an estimation of lethal effects during the exposure period. Copper sulfate caused sediment avoidance in ¹. tubifex at several concentrations below the

LOEC ‘‘mortality’’; therefore, it was possible to calculate rates of sediment avoidance for these concentrations during the exposure period. Figure 4 indicates that most animals initially burrowed into the spiked sediment and then some of them returned to the sediment surface during the following 24 h. In all presented experiments, lethal as well as sublethal effects demonstrated a dose—response relationship. Only in one case was this relationship not found: the rate of autotomy of ¸. hoffmeisteri as a response to lindane did not increase at concentrations higher than the threshold value (Fig. 2). The presented EC value (Table 3) therefore rep50 resents the threshold concentration. DISCUSSION

There was no obvious difference between the two tubificid species with respect to the toxicity of sediment-bound lindane. In accordance with Wiederholm et al. (1987) ¸. hoffmeisteri seems to be slightly more sensitive to copper sulfate than ¹. tubifex. HCB caused no adverse effects in the wholesediment test (this study), even though it is known that tubificids quickly and strongly accumulate HCB (Oliver, 1987; Egeler et al., 1997a). Referring to Nebeker et al. (1989), HCB is also not toxic to different benthic invertebrates in water-only tests at concentrations in the range of its water solubility. These authors presumed that the cause of the lack of acute effects on aquatic invertebrates is that HCB, based on its high lipophilicity, is localized in lipids within the organism and may not be available. Although the sensitivity of tubificids to chemical stress is a matter of controversy, they fit most criteria for test species selection, e.g., ecological relevance, tolerance of a wide range

TABLE 3 Overview of Sublethal and Lethal Effects of Lindane or Copper Sulfate Associated with Artificial Sediment on the Tubificids T. tubifex and L. hoffmeisteri a Lindane

Copper sulfate

Endpoints

E(L)C

50

95% Clb

LOEC

NOEC

¹. tubifex Reworking activity Sediment avoidance Autotomy Mortality

— 217 172 '1000

— 156—309 40—1000 —

40 200 200 200

8 40 40 40

¸. hoffmeisteri Reworking activity Sediment avoidance Autotomy Mortality

— 224 200 '1000

— 164—314 c —

8 200 200 200

1.6 40 40 40

aAll data refer to nominal concentrations in mg/kg sediment dry weight. b95% confidence interval. cCI not determined, because EC value is just an approximation. 50

E(L)C

50

95% Clb

LOEC

NOEC

— 547 601 '1000

— 250—1000 500—1000 —

125 250 500 1000

62.5 125 250 500

— 392 349 516

— 250—500 294—403 458—581

125 500 125 500

62.5 250 62.5 250

SHORT-TERM TOXICITY TO TUBIFICIDS

15

FIG. 2. Dose—response relationship of lindane. Plots on the left-hand side show toxicity of lindane to ¹. tubifex and plots on the right to ¸. hoffmeisteri. Error bars represent the SD of the four replicates in definitive tests.

16

MELLER ET AL.

FIG. 3. Dose—response relationship of copper sulfate. Plots on the left-hand side show toxicity of copper sulfate to ¹. tubifex and plots on the right to ¸. hoffmeisteri. Error bars represent the SD of the four replicates in definitive tests.

17

SHORT-TERM TOXICITY TO TUBIFICIDS

FIG. 4. Time-dependent sediment avoidance of ¹. tubifex in copper sulfate-spiked artificial sediment.

of sediment characteristics, simple handling and culturing. Therefore, the following discussion focuses on four areas: (1) the sensitivity of sublethal effects; (2) the power and interpretation of the different endpoints; (3) the reproducibility of the results and (4) the definition of validity criteria for a standardized performance of the test. The suitability of artificial sediment for ecotoxicological tests, as well as the advantages and disadvantages of breeding and keeping the two tubificid species, is discussed in detail by Egeler et al. (1997b).

the action of lindane as an insecticide, the 20 or 100 times respectively lower NOEC for chironomids in lindane-spiked sediment is not unexpected (Table 4).

Endpoints Based on its body size and resistance to mechanical stress ¹. tubifex is easier to handle than ¸. hoffmeisteri. ¸. hoffmeisteri is more inclined to injury, usually accompanied by bruising. To avoid autotomy from injury (Kaster, 1979) ¸. hoffmeisteri must be handled with extreme care. Autotomy

Sensitivity Oligochaetes are well known for being highly tolerant of lethal effects induced by chemicals (e.g., Wiederholm et al., 1987). Distinct physiological mechanisms are discussed as a possible explanation. The chloragog tissue, typically found in oligochaetes, seems to have a key function with respect to the availability of toxicants within the organisms (e.g., Hagens and Westheide, 1987; Klerks and Bartholomew, 1991; Fischer and Molna´r, 1992). Nevertheless, the detection of sublethal effects reveals that tubificids are affected by chemicals far below lethal threshold levels (e.g., Keilty et al., 1988a, b; Reynoldson et al., 1991). Larvae of Chironomus riparius (Meigen) are well known as sensitive test organisms (e.g., Hill et al., 1993). A comparison of toxicity data from this study and results of experiments with copper sulfate-spiked artificial sediment using first-instar larvae of C. riparius (Meller et al., 1997) gives a first impression of the tubificids’ high sensitivity (Table 4). Because of

TABLE 4 No-Observed-Effect Concentration (NOEC) Values Obtained from 72-h Whole-Sediment Testsa Test organism Copper sulfate ¹. tubifex ¸. hoffmeisteri C. riparius (first-instar larvae) Lindane ¹. tubifex ¸. hoffmeisteri C. riparius (first-instar larvae)

NOECb

Reference

62.5 62.5 125.0

This study This study Meller et al. (1997)

8.00 1.60 0.08

This study This study Meller et al. (1997)

aTests with larvae of Chironomus riparius were conducted using the same artificial sediment and the same spiking procedure used in this study (Meller et al., 1997). Presented values represent the NOEC ‘‘reworking activity’’ (tubificids) and the NOEC ‘‘body length’’ (chironomids). bRefers to nominal concentration in mg/kg sediment dry weight.

18

MELLER ET AL.

as a morphological response to chemical exposure is also known for other oligochaetes (e.g., Roberts and Dorough, 1984; Rombke and Knacker, 1989). However, until now, this reaction has not been assessed as a quantitative endpoint in toxicity tests. The data presented prove that autotomy for both species of tubificids is a valid endpoint. From the investigations of McMurtry (1984), it is known that tubificids have the ability to avoid stress induced by chemicals associated with the sediment. Keilty et al. (1988a) observed in experiments with endrin-spiked sediment that ¸. hoffmeisteri and Stylodrilus heringianus initially burrowed into the sediment and later returned to the sediment surface. Such avoidance behavior is neither restricted to endrin nor confined to both species of aquatic oligochaetes examined by Keilty and co-workers (1988a). It was possible to prove this behavior also with ¹. tubifex in the present experiments using copper sulfate (see Fig. 4). Based on the presumption that autotomy leads to death of the single individual after prolonged exposure, mortality and autotomy directly affect the population density of tubificids. Also, the composition of species within the benthic community can be indirectly changed below lethal threshold levels, e.g., as a result of sediment avoidance, via migration and a greater threat of being eaten (Ahlf, 1995). Additionally there is the risk of increased transmission of chemicals within the food chain if the animals accumulate the substances and are more easily preyed on because they are on the sediment surface. Keilty et al. (1988b) noticed a change in the reworking behavior of tubificids during longterm studies, using a labor-intensive 137Cs marker layer burial technique, at concentrations three to five magnitudes lower than the LC (96 h). However, the experiments with lindane- or 50 copper sulfate-contaminated sediment described here demonstrate that the reworking activity of the examined tubificids was already reduced within 72 h. Using this method, no complicated or costly equipment was necessary to assess changes in reworking activity. For further investigations it should be easy to establish a method to quantify reworking activity (e.g., using digital video analysis). Then a NOEC could be determined by a usual statistical treatment (e.g., ANOVA). Keilty and co-workers (1988b) also observed that decreased reworking activity was accompanied by a reduction of worm biomass. They therefore presumed this effect to reflect decreased feeding rates. Lotufo and Fleeger (1996) recently demonstrated in experiments with pyrene- and phenanthrene-contaminated sediments that reproduction of ¸. hoffmeisteri was reduced at those concentrations at which they had observed a decrease in the egestion rate. As a result, a reduction in reworking activity affects the population dynamics of tubificids and should therefore not be ignored in the risk assessment of chemicals.

Reproducibility The reproducibility of the results is sufficiently ensured. All of the definitive tests performed met the expectations of the range-finding tests. This assumption was also confirmed in a second experiment using copper sulfate, where only the mortality rate was recorded (data not provided). In that experiment, the LC for ¸. hoffmeisteri was almost exactly 50 reproduced, whereas the LC for ¹. tubifex with 627 mg/kg 50 was approximately two times lower than the results described here. These variations, which remain within the range of biological variability, could also be clarified by possible diverse ages of the test organisms. A synchronized culture would be essential to perform toxicity tests on animals of the same age and physiological state. However, at the time of the investigations, such a culture had not been established.

Validity Criteria In addition to common criteria of toxicity tests, the following validity criteria for the standardized performance of whole-sediment tests using tubificids are required: (1) With reference to the endpoint autotomy, the test organisms should not exhibit any signs of injury at the beginning of the test. (2) Rates of sediment avoidance, autotomy, and mortality must be less than 10% in the control. (3) To determine a NOEC, the reworking activity at the lowest concentration should not differ from that of the control. CONCLUSIONS

Oligochaetes are commonly known to be highly tolerant of chemically induced lethal effects (e.g., Wiederholm et al., 1987) and the results presented here confirm this for the two tubificid species ¹. tubifex and ¸. hoffmeisteri. However, based on the tubificids’ wide spectrum of potential responses to chemical stress, the sublethal endpoints are far more suitable for assessment of the ecotoxicological hazard potential of chemicals and qualify tubificids as a useful tool in sediment ecotoxicology, especially because these sublethal effects not only affect the individual, but can influence population levels and, consequently, the composition of benthic community in the medium term. This test system ensures quick and simple handling. It allows the evaluation of sublethal effects, which are far more sensitive than lethal effects. In addition, it has already been successfully used in distinguishing the effects of various storage conditions on the toxicity of natural sediment spiked with lindane (Hanne, 1997) and artificial sediment spiked with copper sulfate (Walther, 1997).

SHORT-TERM TOXICITY TO TUBIFICIDS

ACKNOWLEDGMENTS This study was sponsored by the Federal Environmental Agency (Berlin, Germany), R&D Project 106 03 106. Thanks to Dr. Pilar Rodriguez, Universidad del Pais Vasco, Bilbao, who confirmed the identity of the worms, and to Susan and Denis Squires and Rachel Gallagher, University of Cardiff, who improved the English. A special thanks to all the friends of Michael Meller for a great farewell.

REFERENCES Ahlf, W. (1995) Aquatische O®kotoxikologie: Biotests an Sedimenten. In Handbuch Angewandte ¸imnologie. (C. Steinberg, H. Bernhardt, and H. Klapper, Eds.). Ecomed Verlagsgsellschaft, Landsberg. Ammon, H. U. (1985) Worm toxicity tests using ¹ubifex tubifex. FAO Workshop, Versailles. Colloq. INRA 31, 303—317. American Society for Testing and Materials (ASTM) (1995) Standard test methods for measuring the toxicity of sediment-associated contaminants with freshwater invertebrates. E1706-95a. In Annual Book of AS¹M Standards, Vol. 11.05, pp. 1204—1285. ASTM, Philadelphia. BBA (1995) Effects of Plant Protection Products on the Develoment of Sediment Dwelling Larve of Chironomus riparius in a Water—Sediment System. BBA Guideline (Final Draft), April 5, 1995, Biologische Bundesanstalt fu¨r Land- und Forstwirtschaft, Branuschweig, Germany. Becker, D. S., Rose, C. D., and Bigham, N. (1995) Comparison of the 10-day freshwater sediment toxicity tests using Hyalella azteca and Chironomus riparius. Environ. ¹oxicol. Chem. 14, 2089—2094. Brinkhurst, R. O. (1971) A Guide for the Identification of British Aquatic Oligochaeta. Sci. Publ. No. 22. Freshwater Biol. Assoc. Chapman, P. M., and Brinkhurst, R. O. (1984) Lethal and sublethal tolerances of aquatic oligochaetes with reference to their use as a biotic index of pollution. Hydrobiologia 115, 139—144. Egeler, P., Ro¨mbke, J., Knacker, T., Schallnass, H., and Nagel, R. (1995) Bioaccumulation of hexachlorobenzene by ¹ubificid Sludgeworms (Oligochaeta) from spiked Artificial Sediments. Poster, Fifth SETACEurope Congress, 25—28 June 1995, Copenhagen. Egeler, P., Ro¨mbke, J., Meller, M., Knacker, T., Franke, C., Studinger, G., and Nagel, R. (1997a) Bioaccumulation of lindane and hexachlorobenzene by tubificid sludgeworms (Oligochaeta) under standardised laboratory conditions. Chemosphere, in press. Egeler, P., Ro¨mbke, J., Knacker, Th., Schallnass, H., Meller, M., and Nagel, R. (1997b) Entwicklung und Erprobung eines Bioakkumulationstests mit endobenthischen Organismen. Forschungsbericht 106 03 106. Umweltbundesamt Berlin, Germany. Environmental Protection Agency (EPA) (1994) Procedures for Assessing the Toxicity and Bioaccumulation of Sediment-Associated Contaminants with Freshwater Invertebrates. EPA 600/R-94/024. USEPA, Duluth, MN. Finney, D. J. (1971) Probit Analysis, 3rd ed. Cambridge Univ. Press, Cambridge, UK. Fischer, E., and Molna´r, L. (1992) Environmental aspects of the chloragogenous tissue of earthworms. Soil Biol. Biochem. 24, 1723—1727. Fleming, R., Crane, M., Van de Guchte, C., Grootelar, L., Smaal, A., Ciarelli, S., Karbe, L., Borchert, J., Westendorf, J., Vahl, H., Holwerda, D., Looise, B., Guerra, M., Vale, C., Castro, O., Gaudencio, M. J., and Van de Hurk, P. (1994) Sediment ¹oxicity ¹ests for Poorly ¼aterSoluble Substances. Final report to the European Commission. Report No. 3738. Hagens, M., and Westheide, W. (1987) Subletale Scha( digungen bei Enchytraeus minutus (Oligochaeta, Annelida) durch das Insektizid Parathion:

19

Vera( nderungen in der Ultrastruktur von Chloragog- und Darmzellen in Abha( ngigkeit von der Belastungsdauer. Poster zu Verhandlungen der GfO® (Giessen 1986), Band XVI. Hanne, C. (1997) Einflu( sse der Lagerungsbedingungen auf die Toxizita( t von dotierten Freilandsedimenten: Untersuchungen mit einer o( kotoxikologischen Testbatterie. Diploma Thesis, Johann Wolfgang Goethe-Universita¨t, Frankfurt/M. Hill, I. R., Matthiesen, P., and Heimbach, F. (Eds.) (1993) Guidance Document on Sediment ¹oxicity ¹ests and Bioassays for Freshwater and Marine Environments. From the Workshop on Sediment Toxicity Assessment, 8—10 November 1993, SETAC-Europe. Hooftman, R. N., Van de Guchte, K., and Roghair, C. J. (1993) Aucte and (Sub)chronic ¹ests with the Model Compound Chlorophyrios. Rijksinstituut Voor Volksgezondheid en Milieuhygiene, RIVM-719102022, Joint Report no. 1, The Netherlands. Kaster, J. L. (1979) Morphological development and adaptive significance of autotomy and regeneration in ¹ubifex tubifex Mu¨ller. ¹rans. Am. Microsc. Soc. 98, 473—477. Keilty, T. J., White, D. S., and Landrum, P. F. (1998a) Short-term lethality and sediment avoidance assays with endrin-contaminated sediment and two oligochaetes from Lake Michigan. Arch. Environ. Contam. ¹oxicol. 17, 95—101. Keilty, T. J., White, D. S., and Landrum, P. F. (1988b) Sublethal responses to endrin in sediment by ¸imnodrilus hoffmeisteri (Tubificidae), and in mixed-culture with Stylodrilus heringianus (Lumbriculidae). Aquat. ¹oxicol. 13, 227—250. Klerks, P. L., and Bartholomew, P. R. (1991) Cadmium accumulation and detoxification in a Cd-resistant population of the oligochaete ¸imnodrilus hoffmeisteri. Aquat. ¹oxicol. 19, 97—112. Lotufo, G. R., and Fleeger, J. W. (1996) Toxicity of sediment-associated pyrene and phenanthrene to ¸imnodrilus hoffmeisteri (Oligochaeta: Tubificidae). Environ. ¹oxicol. Chem. 15, 1508—1516. Lydy, M. J., Brunner, K. A., Fry, D. M., and Fisher, S. W. (1990) Effects of sediment and the route of exposure on the toxicity and accumulation of neutral lipophilic and moderately water-soluble metabolisable compounds in the midge, Chironomus riparius. In Aquatic ¹oxicology and Risk Assessment (W. G. Landis and W. H. Schalie, Eds.), 13th vol., ASTM STP 1096. pp. 140—164. ASTM, Philadelphia. McMurtry, M. J. (1984) Avoidance of sublethal doses of copper and zinc by tubificid oligochaetes. J. Great ¸akes Res. 10, 267—272. Meller, M., Ro¨mbke, J., Egeler, P., Schallnass, H., and Streit, B. (1997) A short-term sediment toxicity test using first instar larvae of the non-biting midge Chironomus riparius (Meigen): Sublethal and lethal effects of lindane, hexacholorobenzene and copper sulfate. In preparation. Nebeker, A. V., Griffis, W. L., Wise, C. M., Hopkins, E., and Barbitta, J. A. (1989) Survival, reproduction and bioconcentration in invertebrates and fish exposed to hexachlorobenzene. Environ. ¹oxicol. Chem. 8, 601—611. Organization for Economic Cooperation and Development (OECD) (1983) Guideline for Testing of Chemicals, ¹G 203 Fish, Acute ¹oxicity ¹est. OECD Paris. Organization for Economic Cooperation and Development (OECD) (1984) Guideline for Testing of Chemicals, ¹G 207 Earthworm, Acute ¹oxicity ¹est. OECD, Paris. Oliver, B. G. (1987) Biouptake of chlorinated hydrocarbons from laboratory-spiked and field sediments by oligochaete worms. Environ. Sci. ¹echnol. 21, 785—790. Peltier, W. H., and Weber, C. I. (1985) Methods for Measuring the Acute Toxicity of Effluents to Freshwater and Marine Organisms, 3rd ed. Appendix E, Toxdat Multi-Method Program. USEPA, Cincinnati, OH.

20

MELLER ET AL.

Phipps, G. L., Ankley, G. T., Benoit, D. A., and Mattson, V. R. (1993) Use of the aquatic oligochaete ¸umbriculus variegatus for assessing the toxicity and bioaccumulation of sediment-associated contaminants. Environ. ¹oxicol. Chem. 12, 269—274. Reynoldson, T. B., Thompson, S. P., and Bamsey, J. L. (1991) A sediment bioassay using the tubificid oligochaete worm ¹ubifex tubifex. Environ. ¹oxicol. Chem. 10, 1061—1072. Reynoldson, T. B., Day, K. E., Clarke, C., and Milani, D. (1994) Effect on indigenous animals of chronic endpoints in freshwater sediment toxicity tests. Environ. ¹oxicol. Chem. 13, 973—977. Rippen, G. (1991) Handbuch der ºmweltchemikalien, 10. Erg. Lfg. 6/91, Ecomed Verlagsgesellschaft, Landsberg. Roberts, B. L., and Dorough, H. W. (1984) Relative toxicities of chemicals to the earthworm Eisenia foetida. Environ. ¹oxicol. Chem. 3, 67—78. Ro¨mbke, J., and Knacker, Th. (1989) Aquatic toxicity test for Enchytraeids. Hydrobiologia 180, 235—242. Royal Society of Chemistry (1994) ¹he Agrochemicals Handbook, 3rd ed. update 5—January 1994. Royal Soc. of Chem., Cambridge, UK.

Suedel, B. C., and Rodgers, J. H. (1994) Development of formulated reference sediments for freshwater and estuarine sediment testing. Environ. ¹oxicol. Chem. 13, 1163—1175. Van de Plassche, E. J. (1994) ¹owards Integrated Environmental Quality Objectives for Several Compounds with a Potential for Secondary Poisoning, Report No. 679101012. RIVM National Institute of Public Health and Environmental Protection, Bilthoven, The Netherlands. Wachs, B. (1965) Die Oligochaetenfauna der Fliessgewa¨sser unter besonderer Beru¨ck sichtigung der Beziehung zwischen Tubificidenbesiedlung und dem Substrat. Arch. Hydrobiol. 63, 310—386. Wachs, B. (1967) Die ha¨ufigsten ha¨moglobinfu¨hrenden Oligochaeten der mitteleuropa¨ischen Binnengewa¨sser. Hydrobiologia 30, 225—247. Walther, S. (1997) Entwicklung von Kontrolltests eines dotierten/undotierten Kunstsediments zur qualitativen Beurteilung von Sedimentbiotests. Diploma thesis, Johann Wolfgang Goethe-Universita¨t, Frankfurt/M. Wiederholm, T., Wiederholm, A. M., and Milbrink, G. (1987) Bulk sediment bioassays with five species of freshwater oligochaetes. ¼ater Air Soil Pollut. 36, 131—154.