Gerhardt, A. & C. Palmer (1998): Copper Tolerances of Adenophlebia

The Science of the Total Environment 215Ž. 1998 217᎐229

Copper tolerances of Adenophlebia auriculata Ž.Eaton 1884 Ž.Insecta: Ephemeroptera and Burnupia stenochorias Cawston 1932Ž. Gastropoda: Ancylidae in indoor artiﬁcial streams

U A. Gerhardt , C. Palmer

Rhodes Uni¨ersity, Institute for Water Research, Box 94, 6140 Grahamstown, South Africa

Received 14 November 1997; accepted 15 January 1998

Abstract

Copper tolerances of two potential African model species for toxicity testing, the mayfly Adenophlebia auriculata Ž.Eaton 1884 Ž exposed to 10 different concentrations in the range of 0.02᎐65 mg ly1. and the limpet Burnupia stenochorias Cawston 1932Ž exposed to eight different concentrations in the range of 0.02᎐3 mg ly1. were determined in continuous flow-through indoor artificial streams. The following parameters were monitored daily: Cu, pH, temperature, oxygen, survival and location of each individual in the channels. Behavioural effects after 96 h exposure y1 on the mayflies were measured with impedance conversion technique. B. stenochorias ŽLC50 : 0.36 mg lŽ. 24 h , 0.10 y1 y1 y1 mg lŽ. 48 h , 0.07 mg lŽ.. 72 h was more sensitive to copper than A. auriculata ŽLC50 : 1.78 mg lŽ. 24 h , 0.79 mg ly1 Ž. 48 h , 0.18 mg ly1Ž.. 96 h . A. auriculata occurred most frequently under the stones, however moved up to the top of the stones just before death at G0.5 mg Cu ly1. With increasing exposure time the mayflies significantly preferred the stones close to the outflow Ž.Ps0.016 . B. stenochorias reacted to Cu-exposure by moving out of the water Ž.Ps0.009 and towards the inflow or outflow of the channels at G0.06 mg ly1 . After 96 h of exposure, Cu-exposed mayflies ventilated more than the controls Ž.P-0.05 . A dose-dependent sequence of different behaviours was seen with increasing Cu-concentrations: increased abdomen undulations ŽG0.046 mg ly1.Ž. , increased locomotion escape at G0.231 mg ly1 and increased ventilation ŽG0.277 mg ly1. combined with increased variance in the data sets. In both species the Cu body burdens increased proportionally with copper exposure. ᮊ 1998 Elsevier Science B.V.

Keywords: Copper; Tolerance; Artiﬁcial streams; Adenophlebia auriculata; Burnupia stenochorias

U Corresponding author: Oststrasse 24, D-49477 Ibbenburen,¨ Germany. Tel.rfax: q49 5451 2065.

1. Introduction Sublethal effects of Cu include the valve closure of juvenile Unionidae at 20 ␮gly1 and In South Africa, aquatic toxicology is a rather shortening of activity phases at 100 ␮gly1 new research field. The development of water and Ž.Jacobson et al., 1993 , reduced filtration rate of sediment quality criteria is based on toxicity mussels at 94 ␮g ly1 Ž Blaxter and Ten Hallers bioassays with standard and indigenous test Tjabbes, 1992. , fewer movements and reproduc- species. The aim of the present paper is to reveal tive success in Lymnaea truncatula at G1mgly1 short-term lethal and sublethal toxicity of Cu for Ž.Rondelaud, 1988 , changes in macroinvertebrate two new potential African test species, the mayfly species composition and abundance at G6 ␮g Adenophlebia auriculata Ž.Eaton 1884 and the gas- ly1 Ž. Clements et al., 1989, 1990 and reduced tropod Burnupia stenochorias Cawston 1932. emergence of Clistoronia magnifica Ž.Trichoptera Short-term toxicity tests using survival and sub- Ž.Nebeker et al., 1984 . lethal responses of macroinvertebrates are an im- The variety in tolerances for survival and other portant component of the routine battery of tests toxicity endpoints indicates that it is necessary to to evaluate toxicity of chemicals, support registra- perform several toxicity bioassays with different tion of pesticide products for outdoor application, parameters, such as survival, behaviour and monitor effluents, establish water quality criteria growth for each species to be considered as in- and provide aquatic safety assessments for chemi- digenous test species. calsŽ. Bitton et al., 1995 . Benthic macroinvertebrates are particularly suited as model species for 2. Materials and methods such tests due to the relative sedentary habits of some taxaŽ. e.g. molluscs and their wide range of 2.1. Test species sensitivities to contaminantsŽ. e.g. mayflies ŽPontasch and Cairns, 1991; Kifney and Cle- Compared to standard toxicity test species such ments, 1994. . as Daphnia magna, toxicity assessment with in- CopperŽ. Cu is an essential micronutrient for digenous species has the advantage of increased most living organisms, being incorporated in at ecological relevance. The disadvantage is, how- least 30 different enzymes mediating metabolic ever, the results are restricted to the area of processes and oxygen transportŽ e.g. haemo- distribution of the test species. Apart from geo- cyanin.Ž Flemming and Trevors, 1989 . . Copper graphical distribution, abundance, availability and has been considered as a priority pollutant by the ecological importance are key factors for the USEPAŽ. Suedel et al., 1996 . Its wide use as an choice of indigenous toxicity test speciesŽ APHA, algicide, aquatic herbicide, fungicide and mollus- 1992.Ž . Adenophlebia auriculata Ephemeroptera: cicide reflects selective toxicity at high concentra- Leptophlebiidae. has been chosen because it is tions. There are numerous short-term toxicity easily naturally available throughout the year studies which reveal a wide range of tolerances to Ž.O’Keeffe and Palmer, 1995 , comparatively large ᎐ ␮ Cu amongst aquatic invertebrates: LC50 : 5 86 g Ž.up to 3 cm length , and easy to feed and rear in ly1 for Daphnia sp.Ž. Crustacea , 17᎐98 ␮g ly1 for the laboratoryŽ. Haigh and Davies-Coleman, 1996 . Chironomus tentans Ž.Diptera , 10᎐890 ␮g ly1 for The species is abundant in ‘deadwater’ areas of Tubifex sp.Ž.Ž Oligochaeta Flemming and Trevors, many headwater streams such as behind big stones 1989. , 44 ␮g ly1 for Villosa iris Ž.Unionidae , 83 and pools in the Eastern Cape Province, South ␮g ly1 for Anodonta grandis Ž.ŽUnionidae Jacob- AfricaŽ. Palmer et al., 1994 . A. auriculata is classi- son et al., 1993. , 30᎐64 000 ␮g ly1 for caddisflies, fied as an omnivorous brusher collectorŽ Palmer 37 ␮g ly1 for Gammarus pulex Ž.ŽCrustacea Taylor et al., 1993. . It survived well between 15 and 25ЊC et al., 1991. , 1.7 mg ly1 for Dugesia tigrina in still water and in recirculating systems with Ž.Ž.Turbellaria See, 1976 and 0.075 to 1.1 mg ly1different food sources such as detritus, commer- for different species of freshwater fishŽ Atchison cial fish food, leaves, and algaeŽ O’Keeffe and et al., 1987. . Palmer, 1995. . The taxonomy is rather easy as A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229 219

South African Leptophlebiidae comprise three Table 1 genera, of which Adenophlebia contains two Water quality parameters of the sampling sites and of the tapwater used for the experiments speciesŽ. Mc Cafferty, 1990 . Burnupia stenochorias Ž.Gastropoda: Ancylidae Parameter Water body Ž.y1 is sedentary and easy to breed in the laboratory mg l Blaaukrantz PalmietrBerg Tapwater on different sources of artificial food. A starving period of 4 days can be survived without negative pH 7.8 6.7 7.6 Ž NH4 -N 0.76 0.03 0.04 effects Davies-Coleman, personal communica- r . NO32NO -N 0.8 0.01 0.1 tion . The species is found on stones in the cur- PO -P 0.49 0.008 0.43 rent in many streams of the Eastern Cape 4 SO4 56 37 28 ProvinceŽ. O’Keeffe and Palmer, 1995 . The An- Na 282 23 54 cylidae are distributed worldwide, comprising K 6.7 0.9 3.9 seven generaŽ. Hubendick, 1964 , three of which Ca 66 1 14 Mg 50 3 10 occur in Africa and one of them, Burnupia being F 0.4 0.1 0.1 restricted to Africa, however little is known about Cl 426 16 95 their ecology and ecotoxicologyŽ. Brown, 1980 . ECŽ. mSrm 188.3 16.3 47.9 The ancylid freshwater limpets have lost nearly all traces of pulmonary cavity and instead have developed well-vascularized, extensive neomor- rent and that their behaviour might be seriously phic gills, resulting in cutaneous respirationŽ Rus- affected by being kept in lentic conditions. Exper- sel-Hunter, 1978. . The limpets seem to have an imental systems with current maintain the largest especially high oxygen demand, as they are con- species diversity and numbers of organisms fined almost entirely to small, stony streams and Ž.Pontasch and Cairns, 1989 . All toxicity tests the wave-washed shores of lakesŽ. Brown, 1980 . were therefore performed in a flow-through system. Dechlorinated tapwater was pumped from 2.2. Sampling sites 25-l aerated reservoirs into white PVC channels Ž.1.5 m length at 10 mlrmin with peristaltic A. auriculata was collected from the pumps. The water passed from each channel via BergrPalmiet River system, ;12 km south west an overflow pipe, covered with nylon nettingŽ 500 ␮ of GrahamstownŽ.Ž 33Њ22Ј S, 26Њ28Ј E Eastern m. and was collected in 100-l containers for Cape. using hand nets in June 1996. The river is treatment and safe disposal. Four hand-sized r unpollutedŽ. Table 1 , has a stony bed and flows stones from the Berg Palmiet River system and r through natural Valley Bush vegetation. B. steno- two Whatman GF C filterpapers covered with chorias was collected from the Blaaukrantz River, fine detritus according to GerhardtŽ. 1992a were ;13 km outside Grahamstown,Ž 33Њ20Ј26Љ S, added in each channel as substrate and food for 26Њ40Ј45Љ E. , partially shaded by trees on grassed the mayflies. The substrate and food for the lim- banks. The limpets’ abundance was higher than in pets were two filter papers and a plastic sheet the BergrPalmiet River system. The bed con- marked with a gridŽ. 2.54 cm spacing . The plastic sisted of silt and boulders and current speeds sheet allowed for fast and easy determination of ranged from 0.06᎐0.42 cm sy1. The stream re- the survival of the animals without touching them ceives domestic effluents from the townships and as dead limpets fell off when it was lifted out of from the municipal waste water purification plant the water. of GrahamstownŽ. Table 1 . 2.4. Short-term toxicity tests 2.3. Flow-through systems The animals were distributed randomly between FrutigerŽ. 1984 has stated that stream-dwelling the channels and evenly over the area of each organisms were physiologically dependent on cur- channel. They were all of similar dry weight Ž A. 220 A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229 auriculata: 0.55Ž. S.D. 0.1 mg; B. stenochorias: 1.39 tion, abdomen-undulations, resting and swim- Ž..S.D. 0.76 mg . After 4 days of acclimation to the mingrlocomotion were quantified. dechlorinated tap waterŽ. Table 1 and the new environmentŽ 16"1ЊC, 12 h: 12 h daily photope- 2.5. Element analyses riod. , 20 A. auriculata and separately 25 B. stenochorias in each channel were exposed to different Element analyses for the chemical characteri- nominal Cu-levels for 4 daysŽ. Table 2 . Based on sation of the sampling sites and the water used experience from the experiment with the mayflies for the experiments were done with ICP-AES for and the greater sensitivity of molluscs to Cu, Na, K, Ca, Mg, F, Cl; Cutot -analyses of the water maximal Cu-concentrations of 3 mg ly1 were and individual animals were performed with flame used for B. stenochorias. and graphite furnace AAS. After dryingŽ 48 h at The concentrations of copper for each channel 80ЊC. and weighing of the individual animals were prepared daily by dilutions from a freshly Ž.limpets with shell they were prepared by wet 2qy1 prepared stock solutionŽ 100 mg Cu l. at pH digestion in 30% HNO322Ž. 1 night and 30% H O 4 and mixed with 25 l tapwater of circumneutral Ž.1 night in an aluminium blockheater at 80ЊC. pH. The following chemical parameters were The Cu-analyses of MA-A-3rTM shrimp homo- measured daily: Cutot at inflow and outflow of genate which served as reference material reached each channel, pH, oxygen and temperature. 90% of the certified values. Survival was recorded daily. The filling of the gut was recorded daily on a semiquantitative basis 2.6. Statistical analyses Ž.Ž.-50%, )50% Gerhardt, 1992a . Different parameters of behaviour were recorded daily af- Stephan and RogersŽ. 1985 recommended a ter the acclimation period. B. stenochorias: posi- regression instead of ANOVA design for toxicity tion of each individual in the grid system painted data as regression analysis can accommodate un- on the plastic sheet, at the inflow, at the outflow, expected inversions in the data, it gives clear on or under the plastic sheet; A. auriculata: loca- trends and it does not require treating experi- tion on, under the stones, on food-plates or in the mental units as replicates if in fact they are not, channel, longitudinal distribution within the e.g. the actual toxicant concentrations in several channel, i.e. on which of the four stones. experimental units are seldom exactly the same.

After 4 days of exposure, the behaviour of nine Survival data were used to calculate LC50 values mayﬂies from each Cu-treatment with at least after probit transformation from a linear regres- nine survivors was measured with a quadrupole sion modelŽ. Weber, 1986 . As the probit analysis impedance conversion techniqueŽ Gerhardt et al., is restricted to cumulative binomial data, no at-

1994. . Placed in a water-filled chamber, an organ- tempt was made to calculate EC50 values from ism functions as resistance in a circuit of alternat- the quantitative behavioural data. Regression ing current generated by a pair of electrodes on analysis was also used to describe the dependency the opposite chamber walls. Movements of the between Cu-levels in the water and Cu body animal change the impedance between a second loads. non-current carrying pair of electrodes and thus Concentration-related differences in different generate specific electrical signals for different behavioursŽ e.g. location of animals in the chan- behavioursŽ Gerhardt, 1995, 1996; Gerhardt and nels. over time were evaluated with non-paramet- Janssens de Bisthoven, 1995; Gerhardt et al., ric repeated measurement analysisŽ Friedman 1998. . Each organism was placed individually in a test. followed by non-parametric post-hoc tests test chamberŽ 2=1=1 cm3. filled with water of Ž.Siegel and Castellan, 1988 . Concentration-re- its treatment. After 10 min of acclimation a lated differences in behaviour after 96 h exposure recording of 120 s was performed with a sampling to CuŽ locomotion, ventilation, number of activity frequency of 50 Hz using Super Scope Software and ventilation phases, ventilation frequency. Ž.GW Instruments . Behaviours such as ventila- were analysed by using Kruskal Wallis one-way A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229 221

Table 2 Table 3 y1 Exposure conditions for A. auriculata and B. stenochorias for Cu-LC50 valuesŽ. mg l for A. auriculata and B. stenochorias y1 Cutot Ž. mg l Parameter Adenophlebia Burnupia CH A. auriculata B. stenochorias auriculata stenochorias Mean S.D. Mean S.D. LC50 Ž. 95% CI Ž.᎐ Ž.᎐ 1a 0.017Ž. 0.0002 0.029 Ž. 0.015 24 h 0.66 0.16 17.7 0.36 0.29 1.1 Ž.᎐ Ž.᎐ 1b 0.017Ž. 0.0002 0.025 Ž. 0.011 48 h 0.50 0.25 1.12 0.10 0.06 5.4 Ž.᎐ 2a 0.050Ž. 0.02 0.030 Ž 0.0008 . 72 h 0.07 0.06 6 Ž.᎐ 2b 0.041Ž. 0.02 0.030 Ž 0.001 . 96 h 0.18 0.05 0.25 3a 0.070Ž. 0.02 0.060 Ž 0.006 . 3b 0.068Ž. 0.02 0.059 Ž 0.006 . 4a 0.240Ž. 0.07 0.099 Ž 0.007 . vival with exposure time was observed for both 4b 0.210Ž. 0.08 0.098 Ž 0.007 . speciesŽ. Figs. 1 and 2 . The limpets were more 5a 0.296Ž. 0.015 0.518 Ž. 0.038 sensitive to Cu than the mayﬂiesŽ. Table 3 . 5b 0.258Ž. 0.019 0.496 Ž. 0.034 6a 0.644Ž. 0.007 0.747 Ž. 0.031 6b 0.613Ž. 0.05 0.787 Ž 0.065 . 3.3. Cu body burdens 7a 2.940Ž. 0.69 2.000 Ž. 0.35 7b 2.030Ž. 0.12 1.500 Ž. 0.14 8a 5.410Ž. 1.76 3.000 Ž. 0.20 A concentration-dependent increase in Cu body 8b 3.626Ž. 0.07 2.700 Ž. 0.50 loads was found after 24, 48 and 120 hŽ. Fig. 3 . 9a 22.030 Regressions were signiﬁcant after 24 h Ž Ps0.001, 9b 21.635 R2 s0.9 for both species.Ž and after 48 h Ps 10a 67.259 2 s 10b 65.712 0.034, R 0.6 for A. auriculata..

CH, artificial stream channel,Ž. a at inflow, Ž. b at outflow. The values represent means of four measurements and standard 3.4. Distribution in the channels deviationsŽ. S.D. . Irrespective of treatment, between 60 and 80% analysis of variance as there were not enough of the mayflies were found sitting under the data left for a reliable regression analysis. stones, indicating a negative phototactic behaviour during the day. During the course of the 3. Results experiment, there was a significant trend of A. auriculata to prefer the stones close to the outlet s s 3.1. Chemical parameters of the channel Ž P 0.06 after 48 h, P 0.016 after 96 h.Ž Table 4 . . There was a slight trend of the mayflies to move onto the stonesŽ increase by The pH values during both experiments were 20% after 3 days of exposure. before they died at between 6.5 and 7.4 with a slight increase over y Cu-concentrations G0.5 mg l 1. Irrespective of time irrespective of Cu-treatment. Oxygen satura- Cu-treatment, more limpets moved under the tion of the water was G70% and no significant plastic sheet with exposure time Ž Ps0.019, dfs difference between the Cu-treatments was found. 4.Ž Table 5 . . There was a dose-dependent in- Precipitation of copper salts occurred at concen- crease of limpets moving to the waterrair line of tration levels G20 mg ly1 within 24 h and there the channels, starting in the higher concentration was no gradient in Cu-concentrations along the levels already after 24 h and subsequently increas- channelsŽ. Table 2 . ing in the lower Cu-treatments Ž.Ps0.009, dfs4 Ž.Table 5 . Once they reached these places, the ¨ ¨ 3.2. Sur i al number of immobile limpets increased linearly from 20% at 0.02 mg Cu ly1 to 50% at 0.1 mg Cu A significant dose-dependent decrease in sur- ly1. 222 A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229

Fig. 3. Cutot -body burdens in Adenophlebia auriculata Ž.Ad and Burnupia stenochorias Ž.Bu after 24, 48 and 120 h of

exposure in relation to actual Cutot -concentrations in the water. Fig. 1. Survival of the mayfly Adenophlebia auriculata exposed to different concentrations of copper as described in Table 2. types of behaviourŽ including resting and abdo- The beginning of Cu-exposure is marked with an arrow. men-undulations. more evenly distributed over the whole recording periodŽ. Fig. 4 . Cu-exposed 3.5. Behaïoural effects mayflies ventilated significantly more than control animals Ž.Ps0.05, dfs5 . However, there was no The mayflies showed different types of be- significant difference in ventilation frequency haviour such as swimmingrlocomotion, sittingr between the Cu-treatments; the mayflies venti- resting, ventilation and abdomen-undulations. lated within a range of 3.2Ž.Ž S.D. 0.8 to 4.7 S.D. With increasing Cu-concentrations more organ- 0.5. Hz. With increasing Cu-concentrations the isms either ventilated or swamŽ ‘escape’, ‘avoi- following sequence of behavioural effects could dance’. during the whole recording period, which be seen: increased abdomen undulations ŽG0.046 y resulted in increased variance of the data com- mg l1 .Ž. , increased escape swimming at G0.23 y pared to the control animals, which showed all mg l1 followed by increased ventilation ŽG0.27 mg ly1 .Ž Fig. 5 . .

4. Discussion

4.1. Water chemistry

The Cu-concentrations in the streams are in the range of freshwater background concentra- tionsŽ. Naimo, 1995 . The pH of the tapwater used for the experiments was between 6.5 and 7.5. According to speciation models, copper exists

mostly as CuŽ. OH232 -solid and Cu Ž CO . at pH values above 6Ž. Stumm and Morgan, 1981 . The copper stock solution was prepared at pH 4 to keep the CuŽ. II ions in solution. After dilution with tapwater of circumneutral pH to the Fig. 2. Survival of the limpet Burnupia stenochorias exposed to different concentrations of the treatments, how- different concentrations of copper as described in Table 2. ever, precipitation of CuŽ. OH2 , CuO and The beginning of Cu-exposure is marked with an arrow. Cu223Ž.Ž.Ž. OH CO green occurred in the reser- A. Gerhardt, C. Palmer rThe Science of the Total En¨ironment 215() 1998 217᎐229 223

Table 4 Distribution of A. auriculata Ž.% on each of the four stones in the channels

y1 Parameter Mean Cutot conc.Ž. mg l 0.017 0.046 0.07 0.231 0.277 0.629 2.486 4.482 21.8 66.48

Day Stone 0 1 24 60 19 64 52 20 42 50 21 61 2 3624278118 0421719 3 24 8 27 20 22 60 50 33 33 4 4 164270150 8122511

1 1 13 29 54 64 0 56 26 17 35 21 2 4317251241292108 3 22812812121702617 4 13 38 0 12 58 20 39 58 13 21

2 1 14 26 39 54 32 24 33 36 2 5993044457 3 4 17 22 17 32 8 22 7 4 9 44 9 17 23 56 33 36

3 1 522174735250 2 42262206100 3 16417012150 4 324844244150100

4 1 23 9 25 37 0 0 2 6 10 20 0 7 22 3 473810211433 4 243830377944

Day 0: after 48 h acclimation; day 1᎐4: under Cu exposure; stones 1, 2, 3 and 4: stones placed in the channel longitudinally from inﬂow to outﬂow.

␮ y1 voirs, the tubes of the pumps and the channels LC50 -96 h of 0.04 g l were foundŽ Rodgers et G within 24 h in the treatments of 20 mg Cu tot al., 1980. . CuCl2 is used as a molluscicide ly1. The tapwater contained a number of cations Ž.Rondelaud, 1988 and mussels have been found Ž.Na, K, Ca, Mg as potential ligands for precipi- to be more sensitive to Cu than to other metals tates. such as Hg, Cd, Pb and ZnŽ. Salanki, 1992 . y1 The LC50 s for A. auriculata Ž48 h: 0.50 mg l , 4.2. Surïäl 96 h: 0.18 mg ly1 . were higher than for e.g. y1 Gammarus pulex ŽLC50 -48 h: 0.047 mg l. and y1 In both experiments, survival decreased in a daphnid crustaceansŽ LC50 -48 h: 0.019 mg l , dose-dependent way with increasing Cu-con- Bitton et al., 1995. , but lower than that for Chi- y1 centrations. Adenophlebia auriculata was more ronomus riparius ŽLC50 -48 h: 1.2 mg l , Taylor et tolerant to Cu than Burnupia stenochorias. The al., 1991. , similar to those reported for different y1 ᎐ y1 LC50 values for the limpetŽ 0.36 mg l after 24 h, fish speciesŽ LC50 s: 0.1 1.1 mg l , Atchison et y1 0.07 mg l after 72 h. are in the range of those al., 1987.Ž , Chironomus tentans LC50 -96 h: 0.63 mg ␮ y1 y1 found for other molluscs, e.g. LC50 -24 h: 83 g lor.ŽHyalella azteca LC50 -96 h: 0.65 mg l . ly1 for Anodonta grandis,44␮gly1 for Villosa Ž.Suedel et al., 1996 . Leptophlebiid mayflies have iris, both UnionidaeŽ. Jacobson et al., 1993 . For earlier been reported to be metal tolerant, e.g. ␮ y1 Corbicula fluminea LC50 -24 h of 0.59 gl and Cd-LC50 -96 h for Leptophlebia marginata were 224 A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229

Table 5 Distribution of B. stenochorias Ž.% in the channels

y1 Parameter Mean Cutot conc.Ž. mg l Day 0.027 0.03 0.06 1.00 0.50 0.70 2.00 3.00

% Limpets under the plastic sheet with coordinate system 0 19 24 38 6 35 22 28 25 1 1718631545313344 2 372053178330 353257776 467415032

% Limpets at the waterrair interface on the channel walls 0 0 0 000000 1 0 0 0 0 0 17 10 11 2 105053671740 3 47 63 100 62 4 46 94 100 67

Day 0: after 48 h acclimation; day 1᎐4: under Cu-exposure.

G4.4 mg ly1 depending on pH and test design after 28 days: Janssens de Bisthoven et al., unpub- G ␮ y1 Ž.Gerhardt, 1992b , Pb-LC50 -96 h were 1.1 mg lished.Ž and Gammarus pulex 0.66 g g after 35 y1 ᎐ y1 y1 l and Fe-LC50 -96 h were 89.5 106.3 mg l days to 0.05 mg l : Gerhardt, 1995. . The biocon- depending on pHŽ. Gerhardt, 1994 . The lep- centration factor of A. auriculata and B. steno- tophlebiid Atalophlebia australis showed an LC50 chorias was in the range of 1000 to 10 000, similar for Cd of 0.84 mg ly1 Ž. Thorp and Lake, 1974 . to those reported for Daphnia magna ŽWinner Adenophlebia auriculata tolerated G31.6 mg Zn and Gauss, 1986. and freshwater plankton y1 lŽ. LC50 -96 h depending on data-analysis Ž.Jørgensen et al., 1991 . Indications of a threshold methodŽ. Probit, Spearman Karber¨ and test de- value for Cu regulation have been given for signŽ.Ž size of aquaria and aeration Everitt, perso- Anodonta cygnea Ž.Salanki and Balogh, 1989 and nal communication. . Leptophlebiid mayflies seem Palaemon elegans of about 120 ␮g ly1 Ž Rainbow to be less sensitive to metal toxicity than crus- and White, 1989. . taceans and molluscs; their tolerance is similar to that of chironomids and some caddisflies 4.4. Distribution in the channels Ž.Gerhardt, 1992a, 1994 . Both test species changed their distribution in 4.3. Cu body loads the channels during exposure to copper. A. auriculata tended to crawl on top of stones instead of Both species accumulated copper to the same the negative phototactic and cryptic behaviour of extent and proportionally with increasing expo- sitting under stones. A loss of positive phototactic sure concentration. The Cu body loads of the behaviour was found for calanoid copepods ex- limpets were similar to those mentioned in the posed to pCu: 13Ž. Stearns and Sharp, 1994 . A literature such as Asellus meridianus Ž200᎐800 ␮g slowing down of the crawling speed in Dugesia gy1 dry weight after 14 days to 0.57 mg ly1 : tigrina as a sign of decreased negative phototactic Brown, 1977.Ž , Asellus aquaticus 250 ␮g gy1 dry behaviour, was found at G0.5 mg Cu ly1 Ž See, weight after 50 days to 0.25 mg ly1 : van Hattum 1976. . After 4 days of exposure, the mayflies et al., 1993.Ž , Chironomus decorus 450 ␮ggy1preferred stones close to the outlet instead of after 24 h to 0.2᎐0.8 mg Cu ly1 : Kosalwat and being evenly distributed on the four stones in the Knight, 1987.Ž , Chironomus riparius 44.5 ␮ggy1channel. B. stenochorias moved to the channel A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229 225 . 1 y 1 y 1 y tot exposed to different Cu -concentrations 0.017 mg l , 0.28 mg l and 0.6 mg l Adenophlebia auriculata Ž. Ž Fig. 4. Examples of typical behavioural records 120 s of and signal magnifications for abdomen-undulation and ventilation. 226 A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229

NOEC for Brachionus calyciflorus:50␮gly1 Ž.Juchelka and Snell, 1994 , pumping rate-LOEC for Anodonta cygnea:5␮g ly1 Ž. Salanki, 1992 and life-cycle parameter-NOEC for Clistoronia magnifica: 8.3 ␮g ly1 Ž. Nebeker et al., 1984 . A plating industry effluent containing Cr, Ni and Fe pro- duced an avoidance reaction in rainbow trout at - - 1:300 x 1:75 LC50 Ž Hadjinicolaou and Spraggs, 1989. .

4.5. Beha¨ioural effects after 96 h of exposure

Besides the migrations of both species in the channels, different toxic effects on the behaviour of the mayfly A. auriculata were found after 96 h of exposure to Cu, such as increased abdomen undulations at G0.046 mg ly1, increased locomotion and at G0.231 mg ly1, increased ventilation Fig. 5. Changes in the behavioural patterns at G0.277 mg ly1 and variances in behaviour. r r Žlocomotion swimming escape, abdomen-undulation, ventila- Increased ventilation was also found in Gam- tion. of the mayfly Adenophlebia auriculata after 96 h exposure marus pulex exposed to 50 ␮gCuly1 and to to different Cutot -concentrations in the water. G0.01 mg Pb ly1 Ž. Gerhardt, 1995 . Increased ventilation might be a trial to get rid of Cu-ions edges and accumulated significantly more at the bound to the gill membranes, thus blocking the waterrair line of the channels treated with Cu Ca-binding sitesŽ. Winner and Gauss, 1986 . Hy- ŽG0.03 mg ly1 . compared to the control. With peractivity was found for estuarine copepods ex- increasing exposure time the limpets moved un- posed to sublethal pCu after 24 hŽ Sullivan et al., der the plastic sheet, which was placed in the 1983. . With increasing Cu-concentrations, more channel. All these migrations might be seen as mayflies either swam or ventilated during the avoidance or escape behaviour. Similar sensitive whole recording period, resulting in an increasing avoidance reactions in molluscs e.g. the valve variance of the behaviour. Similar results have closure of musselsŽ Blaxter and Ten Hallers been found for the chironomid Glyptotendipes pal- Tjabbes, 1992; Salanki, 1992; Jacobson et al., lens exposed to Cd for 96 hŽ. Heinis et al., 1990 . 1993. , indicate that Cu can be sensed by This kind of ‘bimodal’ or ‘dichotomous’ behaviou- chemoreceptorsŽ. Morton, 1971 . Changes in mi- ral response, described by SchererŽ. 1992 impedes grations due to Cu were found at G0.03 mg ly1 statistical analyses based on normal distributions for B. stenochorias and for A. auriculata at G0.3 and analyses of means as often no significant mg ly1. The values are in the range of those cited results can be found with these traditional meth- in the literature, e.g. valve closure-EC50 -24 h for ods. There is a need to develop methods to deal Unionidae: 27᎐33 ␮g ly1 Ž. Jacobson et al., 1993 , with interindividual variability in ecotoxicology LOEC for valve closure in D. polymorpha 0.03 mg Ž.Depledge, 1990 . y1 lŽ. Sloof et al., 1983 , swimming-EC50 -3 h for Brachionus calyciflorus:15␮g ly1 Ž Janssen et al., 5. Conclusions ␮ y1 1994. , filtration rate-EC50 for mussels: 94 gl Ž.Blaxter and Ten Hallers Tjabbes, 1992 , feeding Both species showed sublethal behavioural and ␮ y1 rate-EC50 -6 h for Ceriodaphnia dubia: 14.5 gl lethal responses to copper within 96 h of expo- Ž.Bitton et al., 1995 , avoidance-LOEC for fish: sure at concentration levels which are 10᎐100 0.1᎐6.4 ␮g ly1 Ž. Atchison et al., 1987 , ingestion- times higher than the background levels in the A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229 227

field. The limpet was more sensitive to Cu than of Southern Africa. West Lafayette, IN, USA: Dept. of the mayfly, the bioconcentration factor and the Entomology, Purdue University, 1990. LC value for the mayfly fell in the range of Clements WH, Farris JL, Cherry DS, Cairns J, Jr. The influ- 50 ence of water quality on macroinvertebrate community those reported for standard toxicity test species responses to copper in outdoor experimental streams. Aquat such as daphnids and chironomids. Thus, the two Toxicol 1989;14:249᎐262. species proved to be appropriate indigenous toxic- Clements WH, Cherry DS, Cairns J, Jr. Macroinvertebrate ity test species for the evaluation of freshwater community responses to copper in laboratory and field ecosystems where they occur and where the stan- experimental streams. Arch Environ Contam Toxicol ᎐ dard organisms are lacking. However, more toxic- 1990;19:361 365. ity tests with these new species and simultane- Depledge MH. New approaches in ecotoxicology: can interindividual physiological variability be used as a tool to ously with standard organisms have to be per- investigate pollution effects? Ambio 1990;19:251᎐252. formed with different contaminants to further Flemming CA, Trevors JT. Copper toxicity and chemistry in support this conclusion. the environment: a review. Water Air Soil Pollut 1989;44: 143᎐158. Acknowledgements Frutiger A. A versatile flow channel for laboratory experiments with running water macroinvertebrates. Schweiz Z Hydrol 1984;46:301᎐305. We thank Terry Butterworth for constructing Gerhardt A. Effects of subacute doses of ironŽ. Fe on Lep- the PVC channels, Kekere Soxujawa for collect- tophlebia marginata Ž.Insecta: Ephemeroptera . Freshwater ing mayflies and Heather Davies-Coleman for Biol 1992a;27:79᎐84. collecting limpets in the field. For skilful practical Gerhardt A. Acute toxicity of Cd in stream invertebrates in assistance during the experiments we thank Patri- relation to pH and test design. Hydrobiologia 1992b; 239:93᎐100. Ž. cia Goetsch Adenophlebia exp. and Nikita Gerhardt A, Clostermann M, Fridlund B, Svensson E. Moni- Muller¨ Ž.Burnupia exp. . Chemical analyses of the toring of behavioural patterns of aquatic organisms with an stream water and tapwater were performed by the impedance conversion technique. Environ Int 1994;20: Institute for Water Quality StudiesŽ Pretoria, S. 209᎐219. Africa.Ž and Tommy Olsson Lund University, Gerhardt A. Short term toxicity of ironŽ. Fe and lead Ž. Pb to Sweden. . T. Olsson also performed the Cu- the mayfly Leptophlebia marginata Ž.ŽL. Insecta . in relation to freshwater acidification. Hydrobiologia 1994;284: analyses in the animals and water with AAS. A 157᎐168. Rhodes University post-doc fellowship financed Gerhardt A, Janssens de Bisthoven L. Behavioural, develop- my stay in South Africa. mental and morphological responses of Chironomus gr. thummi larvaeŽ. Diptera, Nematocera to aquatic pollution. ᎐ References J Aquat Ecosystem Health 1995;4:205 214. Gerhardt A. Monitoring behavioural responses to metals in Gammarus pulex Ž.ŽL. Crustacea . with impedance conver- APHA. Standard methods for the examination of water and sion. Environ Sci Pollut ResŽ. ESPR 1995;2:15᎐23. waste water, 18th ed. Washington, DC: American Public Gerhardt A. Behavioural early warning responses to polluted Health Association, 1992. water: performance of Gammarus pulex Ž.ŽL. Crustacea . Atchison GJ, Henry MG, Sandheinrich MB. Effects of metals and Hydropsyche angustipennis Ž.Ž.Curtis Insecta to a com- on fish behavior: a review. Environ Biol Fish 1987;18:11᎐25. plex industrial effluent. Environ Sci Pollut Res Bitton G, Rhodes K, Koopman B, Cornejo M. Short-term 1996;3:63᎐70. toxicity assay based on daphnid feeding behavior. Water Environ Res 1995;67:290᎐293. Gerhardt A, Carlsson A, Ressemann C, Stich KP. In-situ Blaxter JHS, Ten Hallers Tjabbes CC. The effect of pollutants on-line biomonitoring: behavioural responses of Gammarus on sensory systems and behaviour of aquatic animals. Neth pulex Ž.ŽL. Crustacea . to changes in water quality. Environ J Aquat Ecol 1992;26:43᎐58. Sci Technol 1998:in press. Brown BE. Uptake of copper and lead by a metal-tolerant Hadjinicolaou J, Spraggs LD. Avoidance tests with a plating isopod Asellus meridianus Rac. Freshwater Biol industrial effluent. In: Nriagu JO, Lakshminarayana JSS, 1977;7:235᎐244. editors. Aquatic toxicology and water quality management. Brown DS. Freshwater snails of Africa and their medical New York: John Wiley, 1989:111᎐125. importance. London: Taylor and Francis, 1980. Haigh E, Davies-Coleman H. The final report of the standard Mc Cafferty WP. A preliminary guide to the ephemeroptera laboratory organisms for water quality studies programme. 228 A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229

Report number K5r545, Water research Commission, Pre- tory-based microcosms of riffle insect communities: their toria, S-Africa, 1996. potential for multispecies toxicity tests. Hydrobiologia Hattum B, van Korthals G, van Straalen NM, Govers HAJ, 1989;175:49᎐60. Joosse ENG. Accumulation patterns of trace metals in Pontasch KW, Cairns J, Jr. Multispecies toxicity tests using freshwater isopods in sediment bioassays ᎏ influence of indigenous organisms: predicting the effects of complex substrate characteristics, temperature and pH. Water Res effluents in streams. Arch Environ Contam Toxicol 1993;27:669᎐684. 1991;20:103᎐112. Heinis F, Timmermans KR, Swain WR. Short-term sublethal Rainbow PS, White SL. Comparative strategies of heavy metal effects of Cd on the filter-feeding chironomid larva Glypto- accumulation by crustaceans: zinc, copper and cadmium in tendipes pallens Ž.Ž.Meigen Diptera . Aquat Toxicol a decapod and amphipod and a barnacle. Hydrobiologia 1990;16:73᎐86. 1989;174:245᎐262. Hubendick B. Studies on Ancylidae, the sub group. Medd Rodgers JH, Cherry DS, Graney RL, Dickson KL, Cairns J, Jr. Goeteborgs Mus Zool Avd 1964;137:1᎐72. Comparison of heavy metal interactions in acute and arti- Jacobson PJ, Farris JL, Cherry DS, Neves RJ. Juvenile fresh- ficial bioassay techniques for the Asiatic clam ŽCorbicula water musselŽ. Bivalvia: Unionidae responses to acute toxi- fluminea.. In: Eaton JG, Parish PR, Hendricks AC, editors. city testing with copper. Environ Toxicol Chem 1993; Aquatic toxicology. ASTM STP 707, Philadelphia, 12:879᎐883. 1980:266᎐280. Janssen CR, Ferrando MD, Persoone G. Ecotoxicological Rondelaud D. Les effects d’une concentration subletale de studies with the freshwater rotifer Brachionus calyciflorus. molluscicideŽ. CuCl2 sur l’activite reproductrice et les de- IV. Rotifer behavior as a sensitive and rapid sublethal test placements du mollusque hote Lymnaea truncatula Muller.¨ criterion. Ecotoxicol Environ Safety 1994;28:244᎐255. Ann Rech Vet 1988;19:273᎐278. Juchelka CM, Snell TW. Rapid toxicity assessment using Russel-Hunter WD. Ecology of freshwater pulmonates. In: rotifer ingestion rate. Arch Environ Contam Toxicol Freiter V, Peake J, editors. Pulmonates, vol. 2A. New York: 1994;26:549᎐554. Academic Press, 1978:334᎐383. Jørgensen SE, Nielsen SN, Jørgensen LA. Handbook of eco- Salanki J. Heavy metal induced behviour modulation in mus- logical parameters and ecotoxicology. Amsterdam: Elsevier, sels: possible neural correlates. Acta Biol Hung 1992;43: 1991. 375᎐386. O’Keeffe J, Palmer CC. Standard laboratory organisms for Salanki J, Balogh KV. Physiological background for using water quality studies programme. Internal report for Water freshwater mussels in monitoring copper and lead pollu- Research Commission Pretoria, South Africa, 1995. tion. Hydrobiologia 1989;188r189:445᎐454. Kifney PM, Clements WH. Effects of heavy metals on a Scherer E. Behavioural responses as indicators of environ- macroinvertebrate assemblage from a Rocky Mountain mental alterations: approaches, results, developments. Appl stream in experimental microcosms. J North Am Benthol Ichthyol 1992;8:122᎐131. SocŽ. NAPS 1994;13:511᎐523. See CL. The effect of sublethal concentrations of selected Kosalwat P, Knight AW. Acute toxicity of aqueous and subs- toxicants on the negative phototactic response of Dugesia trate-bound copper to the midge, Chironomus decorus. Arch tigrina. PhD thesis, Virginia Polytechnic Institute, Blacks- Environ Contam Toxicol 1987;16:275᎐282. burg, 1976:76 p. Morton B. Studies on the biology of Dreissena polymorpha Siegel S, Castellan NJ, Jr. Nonparametric statistics for the Pall. II. Correlation of the rhythms of adductor activity, behavioural sciences. New York: McGraw Hill, 1988:399 p. feeding, digestion and excretion. Proc Malac Soc London Sloof W, de Zwart D, Marquenie JM. Detection limits of a 1971;38:401᎐404. biological monitoring system for chemical water pollution Naimo TJ. A review of the effects of heavy metals on freshwa- based on mussel activity. Bull Environ Contam Toxicol ter mussels. Ecotoxicology 1995;4:341᎐362. 1983;30:400᎐405. Nebeker AV, Savonen C, Baker RJ, Mc Cardy JK. Effects of Stephan CE, Rogers JW. Advantages of using regression anal- copper, nickel and zinc on the life cycle of the caddisfly ysis to calculate results of chronic toxicity tests. Am Soc Clistoronia magnifica Ž.Limnephilidae . Environ Toxicol Testing and Materials, Special Technical Publication 891, Chem 1984;3:645᎐649. 1985. Palmer C, O’Keeffe J, Palmer AR. Macroinvertebrate functio- Stearns DE, Sharp AA. Sublethal effects of cupric ion activity nal feeding groups in the middle and lower reaches of the on the phototaxis of three calanoid copepods. Hydrobiolo- Buffalo River, Eastern Cape, South Africa. II. Functional gia 1994;292r293:505᎐511. morphology and behaviour. Freshwater Biol 1993;29: Stumm W, Morgan JJ. Aquatic chemistry. An introduction 455᎐462. emphasizing chemical equilibria in natural waters. New Palmer C, Palmer A, O’Keeffe J, Palmer R. Macroinverte- York: J. Wiley, 1981:780 p. brate community structure and altitudinal changes in the Suedel BC, Deaver E, Rodgers JH, Jr. Experimental factors upper reaches of a warm temperate southern African river. that may affect toxicity of aqueous and sediment-bound Freshwater Biol 1994;32:337᎐347. copper to freshwater organisms. Arch Environ Contam Pontasch K, Cairns J, Jr. Establishing and maintaining labora- Toxicol 1996;30:40᎐46. A. Gerhardt, C. Palmer rThe Science of the Total Enïronment 215() 1998 217᎐229 229

Sullivan BK, Buskey E, Miller DC, Ritacco PJ. Effects of freshwater invertebrates and interaction of Cd and Zn on copper and cadmium on growth, swimming and predator the freshwater shrimp P. tasmaniensis. Aust J Mar Fresh- avoidance in Eurytemora afﬁnis Ž.Copepoda . Mar Biol water Res 1974;25:97᎐104. 1983;77:299᎐306. Weber E. Grundriss der biologischen Statistik. Jena: Gustav Taylor EJ, Maund SJ, Pascoe D. Toxicity of four common Fischer Verlag, 1986. pollutants to the freshwater macroinvertebrates Chirono- Winner RW, Gauss JD. Relationship between chronic toxicity mus riparius MeigenŽ. Insecta: Diptera and Gammarus and bioaccumulation of copper, cadmium and zinc as af- pulex Ž.ŽL. Crustacea: Amphipoda . . Arch Environ Contam fected by water hardness and humic acid. Aquat Toxicol Toxicol 1991;21:371᎐376. 1986;8:149᎐161. Thorp VJ, Lake PS. Toxicity bioassays of Cd on selected