Ecotoxicology and Environmental Safety 52, 221}226 (2002)

Environmental Research, Section B

doi:10.1006/eesa.2002.2193, available online at http://www.idealibrary.com on

Copper Speciation in the Gill Microenvironment of Carp (Cyprinus carpio)at Various Levels of pH1

Shu Tao, Aimin Long, Fuliu Xu, and R. W. DawsonDepartment of Urban and Environmental Sciences, Peking University, Beijing 100871, China

Received May 15, 2001

The 5sh gill microenvironment of Cyprinus carpio under stressof copper exposure was investigated. pH and other parametersincluding free copper activity, alkalinity, and inorganic and or-ganic carbons in the surrounding water (inspired water) and inthe gill microenvironment (expired water) were measured orcalculated at various levels of pH and varying total copperconcentrations. The chemical equilibrium calculation (fromMINEQA2) and complexation modeling (mucus+copper) werecoupled to calculate both species distribution. The results indi-cate that the pH in the 5sh gill microenvironment was di4erentfrom that in the surrounding water with a balance point around6.9. The secretion of both CO2 and mucus was a4ected in bothlinear and nonlinear ways when the 5sh were exposed to elevatedconcentrations of copper. The complexation capacity of the gillmucus was characterized by a conditional stability constant(log kCu+mucus) of 5.37 along with a complexation equivalent con-centration (LCu+mucus) of 0.96 mmol Cu /mg C. For both the 5shmicroenvironment and the surrounding water, the dominant cop-per species shifted from Cu2� to CuCO

�0 and to Cu(OH)2

0 whenthe pH of the surrounding water changed from 6.12 to 8.11. Thechange in copper speciation in the gill microenvironment issmaller than that in the surrounding water due to the pH bu4er-ing capacity of the 5sh gills. � 2002 Elsevier Science (USA)

Key Words: 5sh; gill; microenvironment; mucus; copper; speci-ation; pH.

INTRODUCTION

The concentration of a particular metal species in naturalwater systems is much more important than total concen-tration relative to a metal's mobility, bioavailability, ortoxicity (Mayer et al., 1994). The most signi"cant processescontrolling metal speciation are the adsorption of the metalon the surface of aquatic particles and any complexationwith naturally occurring organic molecules (Koretsky, 2000;Linnik, 1996). Both the particles and the organic ligands are

�Funding was provided by International Copper Association (TPT0604)and The Scienti"c Foundation (40024101).

22

capable of extensively reducing the uptake of metals byaquatic organisms including "sh (Tao et al., 1999a; Welshet al., 1996).The gills of "sh constitute a sieve-like structure providing

a huge charged surface area and serve as a major route formetal uptake in "sh (McKim, 1994). Metal speciation can becomplicated at the "sh's gills where pH and ionic composi-tion are often changed and gill mucus released. The pH inthe gill microenvironment of "sh is di!erent from that of thesurrounding water due to the release of ammonia, carbondioxide, and other compounds (Lin and Randall, 1990;Playle and Wood, 1989; Tao et al., 2000; Miller and Mackay,1982). The direction of such a change depends on the pH ofthe surrounding water, as well as the bu!ering capacity ofthe water (Lin and Randall, 1990). It was suggested thatsuch a pH change can alter metal speciation or solubility.Accordingly, this may need to be considered when modelingthe physiological and toxicological e!ects of some metals(Gordon and Pagenkopf, 1983).Metal speciation in the "sh gill microenvironment can be

altered by secretion of gill mucus. Gill epithelia are usuallycovered with an extracellular matrix containing a variety ofglycoproteins. This polyanionic matrix may function as anion exchange system, with di!erent a$nities for di!erentmetals. Handy and Eddy studied aluminum concentrationin rainbow trout and reported that the gill tissue and bodymucus are the major sites of aluminum absorption (Handyand Eddy, 1989). The speciation of other metals can also bealtered by binding to "sh mucus from either body or gills(Varanasi and Markey, 1978; Part and Lock, 1983). Al-though it is generally assumed that during the process ofgill uptake, metals are adsorbed onto sites in cell wallsand cell membranes, the processes actually occurringare more complicated. At the least the translocation ofthe metals from water to mucus layer then subsequentlyonto the binding sites occurs and the gill mucus actuallyserves as an intermediate storage for metal uptake (Hudson,1998). It is likely that the mucus can compete withother ligands to form mucus}metal complexes in the gillmicroenvironment.

1

0147-6513/02 $35.00� 2002 Elsevier Science (USA)

All rights reserved.

FIG. 1. Apparatus for gill microenvironment study.

222 TAO ET AL.

Since a large number of ligands are generally presentedtogether, a computer program is often required to solve thelarge number of simultaneous equilibrium equations formetal speciation modeling. For such calculations, com-plexation stability constants for all the major ligandsinvolved are required as inputs to the program. The com-plexation stability constant for metal}mucus binding,however, is usually unavailable. Tao and colleagues haveestablished an iterative algorithm to calculate mucus com-plexation parameters and metal speciation distributionsimultaneously using both a chemical equilibrium programand complexation modeling (Tao et al., 2000).It is assumed, therefore, the change of metal speciation

and bioavailability in the "sh gill microenvironment is basi-cally the results of pH change and mucus complexation,both of which response to change in ambient pH. In a pre-vious study, the in#uence of mucus on copper speciationwas investigated at "xed pH (Tao et al., 2000). In this article,by using "sh microenvironment study apparatus and theiterative algorithm modeling approach, pH change in thegill microenvironment of Cyprinus carpio, mucus secretionfrom the gills, and copper speciation in particular wereinvestigated under various pH conditions.

EXPERIMENTAL

Water and Fish

The experiments were conducted in a medium syn-thesized by adding reagent-grade salts to deionized water.The ionic composition of the major rivers of China wasadopted for the synthetic water base used in this study (Taoet al., 1999b). The major cation and anion concentrations ofthe water were Ca�� 0.710, Mg�� 0.324, Na� 0.144, Cl�1.733, K�0.009, and SO��

�0.324 mmol/L.

Carp, C. carpio, were obtained from the Zhangjiawan"sh market in Beijing. Upon arrival in the lab, the"sh was placed in the synthetic water for a one-week accli-mation period before utilization in the exposure experi-ments. Six carp were used for the gill microenvironmentstudy.

Exposure Apparatus

A standard apparatus for "sh microenvironment studywas adopted for this study (Fig. 1). The "sh was placed inthe tank and a latex membrane was sewn around the mouth.In this manner the tank was separated into two chambersholding inspired and expired water, respectively. Waterlevels in both chambers of the tank were maintainedthrough over#ow from two outlets. Tubing (PE-90, 6 mm indiameter) was attached through one side of opercula tosiphon water from near the gills. The apparatus was cleanedand re"lled after each set of the experiment.

Exposure Experiments

Two sets of the exposure experiments were conducted forpH measurement and copper speciation study, respectively.To test the pH change in the gill microenvironment of the"sh, pH levels of both inspired and expired waters were mea-sured simultaneously while the inspired water was spiked witheither alkaline or acid to cover the pH range from 5 to 11.For the copper speciation study, the exposure was con-

ducted by adding copper to the left chamber with continu-ous stirring. The total copper concentrations in the leftchamber were 0.5, 1.0, 2.5, 5.0, 10.0, and 20.0 �mol/L, respec-tively. The exposure temperature was measured as 233C.After 20 min of copper addition, the activities of free copperions, alkalinity, pH, inorganic carbon (TIC), and total or-ganic carbon (TOC) were determined for both inspired andexpired water samples. The contents of the mucus in theexpired water were represented by the measured TOC. Theexposure experiments were repeated at four pH levels. ThepH in the inspired water was adjusted by spiking with eitherNaOH or HCl.Six "sh were used in the exposure experiments as duplic-

ates at beginning. It was later recognized that, for the copperspeciation study, to control the pH of the inspired water at"xed levels among duplicate experiments within a reason-able short period of time was very di$cult, given very slowresponse of water pH to acid or alkaline spiking. Therefore,the pH levels for the copper speciation study varied fromone duplicate to another slightly. As a result, the data forthese experiments cannot be pooled together for ISE calib-ration and copper speciation modeling. Only the resultsfrom one experiment ("sh with weight of 235 g and theinspired water pH were 6.12, 6.70, 7.17, and 8.11, respective-ly) were used. For data compatibility, results for pH changeof the same "sh (similar to those of other "sh) were used inthis study.

Measurement

A copper ion selective electrode (Cu-ISE) (10��}0.1 M)and a reference electrode were calibrated using the inspired

FIG. 2. pH in the "sh gill microenvironment. Bulk solution pH wasadjusted through adding either NaOH or HCl to the inspired water whilepH in the gill microenvironment was measured directly on the expiredwater from the gills.

223COPPER SPECIATION IN FISH GILLS

water. A linear response of the Cu-ISE was obtained. TheCu-ISE was activated in 1 N Cu(NO

�)�for 2 h and rinsed

with deionized water before use. The activity of free copperions at each calibration point was calculated using theMINEQA2 speciation model based on the total copperconcentration together with other parameters including pH,alkalinity, and major ion concentrations constituting inputsto the model. Calibration equations were established foreach of the four pH conditions: 6.12, 6.70, 7.17, and 8.11,respectively,

E�(mV)"0.2040 M

�(pCu)#1.5016, pH 6.12, r�"0.9450

E�(mV)"0.2348 M

�(pCu)#1.5820, pH 6.70, r�"0.9656

E�(mV)"0.4506 M

�(pCu)#3.3405, pH 7.17, r�"0.9797

E�(mV)"0.0255 M

�(pCu)#2.3313, pH 8.11, r�"0.9583,

where E�is the response of the electrode and M

�represents

free copper activity (mol/L) in log units. The equations werethen applied in order to calculate the free copper activity ateach of the corresponding pH levels.Alkalinity of the samples was titrated with calibrated HCl

solution. The pH was determined using a F-13 ph combina-tion electrode calibrated in standard solutions of pH at 6.86and 9.18. A Shimadzu 5000A TOC analyzer was used formeasurement of total inorganic and organic carbons. Thedetection limit of the TOC analyzer is below 0.2 mgC/L andthe TOC in the inspired water measured was negligible. Allreagents employed in the experiment were of analyticalgrade or better. Deionized water was used throughout forcopper determinations. All glassware was soaked in 10%nitric acid (v/v) for 24 h and rinsed with deionized waterbefore use.

Copper Speciation Calculation

A single site complexation model was used to describe theinteraction between copper and mucus,

k"ML/[M�(¸

�!ML)], (1)

where k is the conditional complexation stability constant;ML (mol/L) represents the concentration of the cop-per}mucus complex; and M

�(mol/L) and ¸

�(mol/L) are con-

centrations of free metal and total mucus, with the latterbeing presented in the normal concentration equivalent tothe copper complexed. ¸

�is presented as the product of the

mass concentration of mucus in organic carbon (TOC,mgC/L) and the complexation equivalent concentration ofunit mucus (¸

�, mol Cu/mg C). The complexation para-

meters were calculated based on in situ measurement of freecopper in the expired water along with the metal speciation

simultaneously using a trial-and-error approach (Tao et al.,2000). MINTEQA2 was applied to model the copper speci-ation in the system.

RESULTS

pH Change and Mucus Secretion

Speciation and bioavailability of metals could be a!ectedby pH changes at "sh gills which were measured over pHlevels ranging from 5 to 11. The pH levels of the gill micro-environment (expired water from the gills) of the "sh wereplotted against the pH of the bulk solution (inspired water)and are presented in Fig. 2.On the other hand, exposure to elevated metal concentra-

tions in the surrounding water may cause changes in therelease rates of CO

�and HCO

�� (Lloyd and Herbert 1990),

which, in turn, can induce changes in pH in the "sh gillmicroenvironment. It was demonstrated that when exposedto various concentrations of copper, the concentration ofinorganic carbon (IC) in the expired water from the gillmicroenvironment of C. carpio increased linearly (Tao et al.,2000). In this study, the IC change was measured under twodi!erent pH conditions: 8.11 and 6.12, respectively. Figure 3depicts the measured results in inorganic carbon contentchange in the expired water from the gills of a "sh exposedto various levels of copper.Of equal signi"cance to the microenvironment of "sh gill

is the secretion of mucus. The protecting mucus is continu-ously washed away by the water stream and replaced bysecretions from mucus cells. It was found that the release oftotal organic carbon from the gills of C. carpio increasedexponentially with an increase in copper concentration (Taoet al., 2000). Two di!erent pH conditions (6.12 and 8.11)were tested in this study to explore the in#uence of copperexposure on mucus secretions from the "sh gills. Mucus inthe expired water from the gills was measured as the ratio of

FIG. 3. Inorganic carbon (IC) in the gill microenvironment exposedto various levels of copper (�) pH 8.11; (�) pH 6.12.

FIG. 5. Copper speciation in the surrounding water at various pHlevels (species 1}6: Cu��, CuHCO

��, CuCO

��, CuSO

��, Cu(OH)

��,

Cu(OH)�).

224 TAO ET AL.

TOC over copper exposure concentrations ranging from0 to 20 �mol/L. The results are plotted against the copperconcentrations in Fig. 4.

Copper Speciation at Various pH Levels

The distribution of copper species in the surroundingwater (inspired water in the experiment) without mucus wascalculated based on the measured parameters including pH,alkalinity, and concentrations of various ions. The resultsare presented in Fig. 5. The total concentrations of the sixdominant species in Fig. 5 accounted for more than 99% ofthe total copper in the system.Since the highest pH reached during the experiment was

8.11, precipitation of Cu(OH)�in the system might occur.

For this reason, a solubility product of 5.6�10��� wastaken into consideration during the chemical equilibriumcalculation. The result indicated that a trace amount ofprecipitation of the species would be produced when thetotal copper concentration reached 20 �mol/L or above.For the expired water representing the water in the gill

microenvironment, copper speciation was computed for thefour scenarios with di!erent pH values in the surroundingwater. The distribution of major copper species at these pHconditions is found in Fig. 6.

FIG. 4. Total organic carbon (TOC) in the gill microenvironmentexposed to various levels of copper (�) pH 8.11; (�) pH 6.12.

DISCUSSION

Changes of pH and Mucus Level in the Gill Microenvironment

pH changes at the "sh gills can be clearly seen in Fig. 2.The balance point is somewhere around 6.9. When the pHof the inspired water was less than 6.9, the water at the gillswas made more alkaline through release of more NH3 (Linand Randall, 1990). The change occurred in the other direc-tion making the pH more acidic when alkaline was added tothe inspired water. In that case, it was found that more CO

�was released from the gills.Playle has reported such changes for rainbow trout and

fathead minnows (Playle and Wood, 1989; Playle et al.,1990). The patterns of the changes are more or less the same.The di!erences are found in the balance points and themagnitudes of the change, both of which depend on "shspecies and the composition of the water. For instance, thebalance points for rainbow trout and fathead minnows, asreported by Playle, were 5.8 and 6.2, respectively, with some

FIG. 6. Copper speciation in gill microenvironment at various pHlevels (species 1}7: Cu��, CuHCO

��, CuCO

��, CuSO

��, Cu(OH)

��,

Cu(OH)�, Cu}mucus).

225COPPER SPECIATION IN FISH GILLS

0.3 pH units di!erence being observed (Playle and Wood,1989; Playle et al., 1990). Using the same "sh species as thisstudy (C. carpio) but di!erent water, the balance pointshifted from 6.9 to 8.0 (Tao et al., 2000). One major reasonthat the water layer near the surfaces of gills is markedlydi!erent from the surrounding water lies in the fact that "shrelease CO

�, HCO

��, NH

�and NH

��, at their gills. By

controlling the #ux of these compounds, the pH of the gillmicroenvironment can remain relatively stable given alter-ations in the surrounding water (Lloyd and Herbert, 1990).A linear increase or decrease in inorganic carbon in the

expired water was detected with increases in exposure con-centration of copper in the inspired water (Fig. 3). At a pHof 8.11 for the inspired water, the pH in the expired waterwas more acid (Fig. 2) and the in#uence of copper exposureon CO

�release tended to act in the same direction (the

results of an F test on a hypothesis of �"0 was signi"cant).This result is similar to that reported previously (Tao et al.,2000). Under acidic conditions with a pH of 6.12 (inspiredwater), however, the in#uence of copper exposure on CO

�release was negative. Secretion of CO

�and HCO

�� obvi-

ously increased or decreased under stress of copper expo-sure. There was, however, no noticeable pH changeobserved in this study.As with CO

�, secretion of mucus from the gills was

a!ected by exposure to elevated copper in the inspiredwater. The TOC in the expired water increased as theexposure level went from 0 to 20 �mol/L (Fig. 4). This resultis consistence with the "ndings of Randall and Lin, whoreported that the rate of mucus secretion increases in "shexposed to a number of toxic chemicals (Randall et al.,1991). However, the observed response of the mucus secre-tion to the exposure concentration was not linear. Instead,an exponential curve may best "t the results. In theory,increases in both IC and TOC will a!ect metal speciation.Within the range of copper concentration used in this study,however, the change in IC did not cause measurable alter-ation in the pH level of the expired water. The only signi"-cant in#uence possibly caused by the metal speciation isan increase in complexation capacity of the gill micro-environment.

Complexation Capacity of Gill Mucus for Copper

A trial-and-error procedure for simultaneous calculationof complexation and speciation was developed for an expo-sure experiment at a "xed pH level of 6.7 (Tao et al., 2000).For this study, the in situ calculation was performed for fourdi!erent pH conditions using conditional complexationstability constants of mucus for copper in log units(log k

��}�����). The complexation equivalent concentrations

of the mucus for copper (¸��}�����

) were calculated for theexpired water from the gill microenvironment. The conver-gence of the trial-and-error algorithm was tested by specify-

ing various initial values (log k�and ¸

�). Identical results

could always be reached.For the four pH conditions, the calculated log k

��}�����values are more or less the same, ranging from 5.3 to 5.5with a di!erence of less than 0.2 log units. The average valuewas 5.4, which was taken as the "nal result of the calculationand was used for purposes of the copper speciation calcu-lation in the "sh gill microenvironment. For the complexa-tion equivalent concentration, the results were very similarto that of log k

��}�����. A mean value of ¸

��}�����was

computed as 0.96 mmol Cu/mg C (ranging from 0.95 to0.97). When the #uctuation in free copper measurementusing the selective electrode is taken into account, the vari-ation is within the expected tolerance limit.

Copper Speciation in the Bulk Solution and the Fish GillMicroenvironment at Various pH Levels

With no mucus in the inspired water in the "rst place, theconcentrations of mucus complexed copper at various pHconditions were all zero. The distribution of copper speciesvaried dramatically as the pH changed from 6.12 to 8.11 andthe dominant species shifted from acidic to alkaline condi-tions (Fig. 5). Free copper ions (Cu��) were the most abun-dant species accounting for more than 45% of the totalmetal concentration at a pH of 6.12 and reduced quickly asthe pH of the solution increased. A similar trend of changein the concentration of copper bicarbonate complex(CuHCO

��) was observed. At pH levels of 6.70 and 7.17, the

dominant species was CuCO�� which was relatively lower

at either acidic or alkaline conditions. The hydroxy-coppercomplex (Cu(OH)

��) increased from very low at a pH of 6.12

to more than 90% at pH 8.11 where some precipitationoccurred as well. Over the pH range studied, Cu(OH)� aswell as all other species included in the MINEQA2 packagecould hardly be detected.The di!erence in copper speciation in the gill microen-

vironment (Fig. 6) among the four pH conditions is verysimilar to that in the surrounding water shown (Fig. 5). Ingeneral, the dominant species changed from free copper(Cu��) under acidic conditions to copper carbonate com-plex (CuCO

��) at a neutral condition and to hydroxy-

copper complex (Cu(OH)��) under alkaline conditions.

Compared to the surrounding water without mucus (in-spired water), the magnitude of change in speciation amongthe four di!erent pH conditions is relatively smaller thanthat in the surrounding water. The pH adjustment capabil-ity of the gill microenvironment contributed to this bu!er-ing phenomenon. Another profound di!erence of copperspeciation in the "sh gill microenvironment compared tothat of the surrounding water is the presence of mucus,especially under acidic conditions. The mucus}copper com-plex accounted for almost 20% of the total copper when thepH of the surrounding water was 6.12 and the proportion

226 TAO ET AL.

decreased gradually as the pH increased due to strongcompetition from the hydroxy ligand. Mucus can provideprotection for "sh by accumulating and sloughing o! poten-tially toxic copper during continuous secretion and release(Varanasi and Markey, 1978). According to the calculatedresults, the role of the protection is much stronger at lowerpH levels than that at higher levels.

CONCLUSION

When the pH in the surrounding water ranged from 5 to11, the pH in the expired water from the gill microenviron-ment of the "sh was generally di!erent from that in thesurrounding water. A balance point was found to exist nearpH 6.9. The secretion rates of both CO

�and mucus were

a!ected by elevating the concentration of copper in theinspired water. The conditional complexation stability con-stant and the complexation equivalent concentration of gillmucus for copper were calculated as 5.4 (log units) and0.96 mmol Cu/mg C, respectively. The dominant Cu speciesin the surrounding water were Cu��, CuCO

��, and

Cu(OH)�� at pH 6.12, 7, and 8.11, respectively. A similar

trend was observed for the "sh gill microenvironment, withthe exception that mucus complexed copper became animportant species under low pH conditions.

REFERENCES

Gordon, K., and Pagenkopf, G. K. (1983). Gill surface interaction modelfor trace metal toxicity to "sh: Role of complexion, pH and waterhardness. Environ. Sci. ¹echnol. 17, 342}347.

Handy, R. D., and Eddy, F. B. (1989). Surface absorption of aluminum bygill tissue and body mucus of rainbow trout, Salmo gairdneri, at the onsetof episodic exposure. J. Fish Biol. 34, 865}874.

Hudson, R. J. M. (1998). Which aqueous species control the rates of tracemetal uptake by aquatic biota? Observations and predictions of non-equilibrium e!ects. Sci. ¹otal Environ. 219, 95}115.

Koretsky, C. (2000). The signi"cance of surface complexation reactions inhydrologic systems: A geochemist's perspective. J. Hydrol. 230, 127}171.

Lin, H., and Randall, D. J. (1990). The e!ect of varying water pH onthe acidi"cation of expired water in rainbow trout. J. Exp. Biol. 149,149}160.

Linnik, P. M. (1996). Heavy metals in the rivers Dnieper and Danube andthe role of intra-reservoir processes in the migration and transformationof their compounds. Arch. Hydrobiol. Suppl. 113, 541}547.

Lloyd, R., and Herbert, D. W. M. (1990). The e!ect of varying water pH onthe acidi"cation of expired water in rainbow trout. J. Exp. Biol. 149,149}160.

Mayer, F. L., Jr., Marking, L. L., Bills, T. D., and Howe, G. E. (1994).Physiochemical factors a!ecting toxicity in freshwater: Hardness, pH,and temperature. In Bioavailability, Physical, Chemical, and BiologicalInteractions (J. L. Hamelink, P. F. Landrum, H. L. Bergman, and W. H.Benson, Eds.), pp. 5}22. Lewis, Boca Raton, FL.

McKim, J. M. (1994). Physiological and biochemical mechanisms thatregulate the accumulation and toxicity of environmental chemicals in"sh. In Bioavailability, Physical, Chemical, and Biological Interactions(J. L. Hamelink, P. F. Landrum, H. L. Bergman, and W. H. Benson,Eds.), pp. 179}202. Lewis, Boca Raton, FL.

Miller, T. G., and Mackay, W. C. (1982). Water chemistry changes in thegill microenvironment of rainbow trout. Bull. Environ. Contam. ¹oxicol.28, 68}74.

Part, P., and Lock, R. A. C. (1983). Di!usion of calcium, cadmium andmercury in a mucous solution from rainbow trout (Salmogairdneri), Comp. Biochem. Physiol. C Comp. Pharmacol. Toxicol. 76,259}263.

Playle, R. C., Gensemer, R. W., and Sixon, D. G. (1990). Copper accumula-tion on gills of fathead minnows: in#uence of water hardneww, com-plexation and pH of the gill micro-environment. Environ. ¹oxicol. Chem.11, 381}391.

Playle, R. C., and Wood, C. M. (1989). Water chemistry changes in the gillmicro-environment of rainbow trout: Experimental observations andtheory. J. Comp. Physiol. B 159, 527}537.

Playle, R. C., and Wood, C. M. (1990). Is precipitation of aluminum fastenough to explain aluminum deposition on "sh gills? Can. J. Fish. AquaSci. 47, 1558}1561.

Randall, D., Lin, H., and Wright, P. A. (1991). Gill water #ow and thechemistry of the boundary layer. Physiologic. Zool. 64, 26}38.

Tao, S., Liang, T., Liu, C. F., and Xu, S. P. (1999a). Uptake of copper byneon tetras (Paracheirodon innesi) in the presence and absence of partic-ulate and humic matter. Ecotoxicology 8, 269}275.

Tao, S., Dawson, R., Liang, T., Cao, J., and Liu, C. F. (1999b). Synergistice!ect of copper and lead uptake by "sh. Ecotoxicol. Environ. Saf. 44,190}195.

Tao, S., Long, A. M., and Dawson, R. (2000). In#uence of mucus on copperspeciation in the gill micro-environment of carp (Cyprinus carpio).Ecotoxicol. Environ. Saf. 47, 59}64.

Varanasi, U., and Markey, D. (1978). Uptake and release of lead andcadmium in skin and mucus of coho salmon (Onchorhynchus kisutch).Comp. Biochem. Physiol. C 60, 187}194.

Welsh, P. G., Parrott, J. L., Dixon, D. G., Hodson, P. V., Spry, D. J., andMierle, G. (1996). Estimating acute copper toxicity to larval fatheadminow (Pimephales promelas) in soft water from measurements ofdissolved organic carbon, calcium, and pH., Can. J. Fish. Aqua Sci.53, 1263}1271.