Ž .The Science of the Total Environment 247 2000 175]187

Changes in mussel biometry on exposure to metals:implications in estimation of metal bioavailability in

‘Mussel-Watch’ programmes

Manu Sotoa,U, Mike P. Irelandb, Ionan Marigomeza´aBiologia Zelularra Atala, Zoologia eta Animali Zelulen Dinamika Saila, Zientzi Fakultatea, Euskal Herriko Unibertsitatea,

644 PK, E-48080 Bilbo, Basque Country, SpainbInstitute of Biological Sciences, Uni ersity of Wales, Aberystwyth SY23 3DA, Wales, UK

Abstract

The occurrence of changes in flesh and shell weights and in other biometric parameters of mussels, Mytilusgallopro¨incialis, has been related to different metal levels found in their soft tissues. The effects of clean andZn-polluted environments and laboratory experiments where Zn-polluted mussels were exposed to sublethalconcentrations of Zn, Cu and Cd were investigated. Zinc-polluted mussel shell weights increased significantly after a51-day depuration period. Exposure of Zn-polluted mussels to Zn or Cd, however, caused a slightly reduced shellgrowth in comparison with depurating mussel Cu-exposures not causing any reduction in growth. Apart from metalconcentrations, metalrshell weight indices have been used to assess metal bioavailability. Metal concentrationsrecorded in the soft tissues of depurating mussels increased without a source of ‘extra’ metals, while theZnrshell-weight index was reduced, as expected from a depuration process, Cu and Cdrshell-weight indicesremaining constant. Experimental exposure to Zn, Cu and Cd caused augmented values of Zn, Cu and Cdrshell-

Žweight indices, respectively. These different findings were attributed to changes in flesh weight related with gamete.spawning that would produce inconsistent estimates of whole metal concentration in soft tissues. Since changes in

the tissue composition and in growth rates do not affect ZnrCu ratios and metalrshell-weight indices, theseparameters are proposed as reliable indices of metal bioavailability for ‘Mussel-Watch’ monitoring programmes. Themost sensitive parameter is the metalrshell-weight index, which is, therefore, highly recommended to be used in‘Mussel-Watch’ monitoring programmes in order to determine metal bioavailability in seawaters. Q 2000 ElsevierScience B.V. All rights reserved.

Keywords: Biometry; AAS; Mussel-Watch; Metal bioavailability; Metal index

U Corresponding author. Tel.: q34-946012690; fax: q34-944648500.Ž .E-mail address: zopsolom@lg.ehu.es M. Soto

0048-9697r00r$ - see front matter Q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S 0 0 4 8 - 9 6 9 7 9 9 0 0 4 8 9 - 1

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187176

1. Introduction

The changes in normal patterns as a result ofthe exposure to environmental toxicants couldappear as differences in physiology and conditionbetween populations in ‘clean’ environments andthose under the additional stress produced by a

Ž .sublethal level of toxins Aldrich, 1991 . Exposureto sublethal concentrations of pollutants may pro-voke a stress response first evident at lower levels

Žof biological organisation Marigomez et al., 1989;´Vega et al., 1989; Cajaraville et al., 1991, 1992b,

.1993; Etxeberria et al., 1994 and subsequently atŽ .higher levels Cajaraville et al., 1992a, 1993 . Thus,

sublethal levels of environmental toxicants maycause changes in allometric ratios, condition in-dices and growth parameters in molluscs, whichare classical sentinel organisms, after intermedi-

Žate and long-term exposures Lobel and Wright,1982a,b; Stromgren, 1982, 1986; Sunila and Lind-¨strom, 1985; Marigomez et al., 1990; Marigomez¨ ´ ´

.and Ireland, 1990; Cajaraville et al., 1992a .Interestingly, metal content is generally nor-

malised to soft tissue weight of sentinel molluscsŽ .Phillips, 1976 . However, since weight has beendemonstrated to change as a result of sublethalexposures to the toxicants it seems unreliable touse flesh weight as a reference for estimations ofmetal bioavailability in the environmentŽ .Marigomez and Ireland, 1990 . In addition, mus-´sels change in flesh weight seasonally adding an-other variable to be taken into account whenusing metal concentrations in soft body to esti-mate metal bioavailability. In order to overcomethis problem, other parameters, such as the shell

Žweight, have been proposed Fischer, 1983, 1984,1986; Marigomez and Ireland, 1990; Soto et al.,´

. Ž1997a,b . Alternatively, other authors Carbonell.and Tarazona, 1993 have proposed indices such

Žas the ratio between essential metals ZnrCu.ratio .

In this work we investigate the effects of sub-Ž .lethal concentrations of three metals, zinc Zn ,

Ž . Ž .copper Cu and cadmium Cd on flesh and shellgrowth in mussels, as well as alternative measuresof metal bioavailability based on metalrshell-weight indices. The present experimental design

comprises field studies comparing clean andZn-polluted environments and laboratory experi-ments under controlled conditions where Zn-polluted mussels were exposed to sublethal con-centrations of Zn, Cu and Cd.

2. Materials and methods

2.1. Experimental procedure

Ž . ŽMussels 1000 , Mytilus gallopro¨incialis 25]35.mm shell length were collected from Menakoz,˜

Ž .Bizkaia 438249N, 28239W in March 1988 for vari-Žous experimental purposes Cajaraville et al.,

. Ž1992a,b,c . A subsample of their soft tissues ns.13 was analysed by atomic absorption spec-

trophotometry after nitric acid digestion and themussels contained a Zn concentration in the softtissues of 173.9"77.3 mg gy1. The group is re-ferred to as clean or healthy mussels.

A second group of 1000 mussels were collectedfrom the same mussel population in Menakoz in˜March 1989 and transferred to the laboratory.

Ž .After metal analysis of a subsample ns19 ,these mussels exhibited a Zn concentration in thesoft tissues of 501.5"194.2 mg gy1. The group is

Žreferred to as Zn-polluted mussels Soto et al.,.1997b .

Fifty Zn-polluted mussels were distributed intoeach of 20 polyethylene covered tanks in a ther-mostated semicontinuous water-flow system withactive charcoal and glass-wool filtered natural

Ž .seawater Zierbena, Bizkaia . Mussels were main-tained unfed at 15]168C for 10 days to allowacclimatisation. Afterwards, mussels were ex-posed in two replicate series for 41 days to threedifferent sublethal concentrations of each metal:

w y1 Ž .Cu, Cd and Zn 0.8 mg l , low dose LD ; 8 mgy1 Ž . y1l , intermediate dose ID ; 80 mg l , high doseŽ .x Ž .HD . The water 25 l per tank was totallyreplaced every 2 days. Two replicate controlseries were also carried out by maintaining Zn-polluted mussels in clean seawater. These are

Žreferred to as depurating mussels Soto et al.,.1997b .

Mussels were maintained under natural pho-

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187 177

toperiods and fed with a commercial filter feederŽfood Marine Invertebrate Diet, Hawaiian Ma-

.rine Imports, Inc. during the experimental pe-riod. Tanks were continuously aerated and watertemperature, recorded every day at 15.00 h, aver-aged 16.7"1.28C in April and 16.8"1.88C inMay.

2.2. Weight and shell dimension measurements

ŽAfter 41 days exposure period, length L, maxi-.mum measure along the anterior]posterior axis ,

Ž .height H, maximum dorsoventral axis , and widthŽ .W, maximum lateral axis of each mussel weremeasured to the nearest millimetre. Mussels were

Ž . Ž .then opened and wet WFW and dry DFWflesh weights and dry upper and lower valve

Ž .weights V were measured. WFW and DFW wereused to estimate the percentage of hydrationŽ .%H . The ratios of shell widthrshell heightŽ . Ž .WrH , shell width:shell length WrL , shell

Ž .length:shell height LrH , dry flesh wt.:shellŽ . Ž .height FrH , dry flesh wt.rshell length FrL ,

Ž .dry flesh wt.rshell width FrW were calculated.Ž .Flesh condition index FC was calculated accord-

Ž .ing to Lobel and Wright 1982a as dry flesh wt.Ž . Ž .mg rdry shell wt. g . Shell valve inequality indexŽ .SVI was calculated according to Lobel and

Ž . Ž .Wright 1982b as heaviest valve dry wt. mg =Ž .100rlightest valve dry wt. mg .

Shell thickness was measured indirectly as de-Ž .scribed by Cooke and Pollard 1973 by calculat-

Ž .ing the regression of log dry shell-weight g10Žagainst shell size factor SSF, height=width=

3.length, mm .

2.3. Metal analysis

Zinc, Cu and Cd analyses were performed onŽ .individual whole soft body tissues of mussels a

Ž .collected from the field, b after 41 days ofŽexposure to each of the three metals qprevious

. Ž .10 days of acclimation , and c after 51 days ofdepuration. Tissue samples were rinsed in dis-tilled water and dried at 1208C for 48 h, digested

Ž .with nitric acid Aristar , diluted with 0.1 M nitricacid and analysed by atomic absorption spec-

Ž .trophotometry Perkin Elmer 2280 with simulta-neous background correction and a sensitivity of0.36 mg ly1. The sensitivity and detection limit of

Ž y1 .the spoiler in mg l for each metal analysedare the following: Cd, 0.03 and 0.001; Cu, 0.08and 0.002; and Zn, 0.018 and 0.002. The sensitiv-

Žity and detection limit for the impact bead in mgy1 .l are the following: Cd, 0.02 and 0.005; Cu,

0.03 and 0.001; and Zn, 0.011 and 0.0008. Additio-nally, Merck standard solutions were also dilutedwith 0.1 M nitric acid for calibration.

2.4. Calculation of metal indices

ŽMetalrshell weight index mg metalrg dry shell.wt. was calculated as a measure of metal

Ž .bioavailability according to Fischer 1984 andŽ .Soto et al. 1997b :

Ž y1 . Ž .MCSB mg g =DFWST gMswIs Ž .DSW g

where MswI is metalrshell weight index, MCSBis the metal concentration in soft body as mgmetalrg dry flesh wt., DFWST is the dry flesh wt.

Ž .of soft tissues g , and DSW is the dry shell wt.Ž .g .

The ZnrCu ratio was calculated by simply di-viding Znrshell-weight index by Curshell-weight

Ž .index Soto et al., 1997b .

2.5. Statistics

Comparisons between Zn-polluted and depu-rating mussels were made using a one-wayANOVA. In order to check the effects of expo-sure to metals on growth and condition of Zn-polluted mussels, a two-way ANOVA was per-

Ž .formed to detect the effects of the metal-type TŽ .and the exposure-concentration C . These analy-

ses were complemented with Duncan’s multiple-Ž .range test based on one-way ANOVAs at a

significance level of P-0.05. All statistical analy-ses were made with the aid of the statistical

Ž .package SPSSrPCq SPSS Inc., Microsoft Co. .

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187178

3. Results

3.1. Growth and condition of Zn-polluted mussels

Environmental levels of zinc increased fromŽ .March 1988 to March 1989 in Menakoz Table 1 .˜

Zinc was concentrated in the soft tissues of mus-sels collected in March 1989 whilst Cu concentra-tions decreased. As a result, the ZnrCu ratioincreased significantly. Accordingly, the biometricparameters recorded in Zn-polluted mussels indi-

Žcated a poorer growth in 1989 than in 1988 Table.1 , indicated by a severe reduction in both flesh

and shell weights and biometric parameters butŽ .an unchanged flesh condition index Table 1 .

Ž y1The Znrshell-weight index 8.5"2.5 mg g iny1 .1988 and 25.1"10.6 mg g in 1989 and the

Ž y1Curshell-weight index 1.1"0.4 mg g in 1988y1 .and 0.4"0.2 mg g in 1989 showed the same

picture as the Zn and Cu tissue concentrationsŽ .Table 1 .

After a 51-day depuration period under con-trolled laboratory conditions, ZnrCu ratio de-creased but, both Zn and Cu concentrations in-creased in depurating mussels in comparison with

Ž .those collected in March 1989 Table 1 . Cadmiumconcentration was twice as high in depurating

Žmussels compared to Zn-polluted mussels Table.1 . The Znrshell-weight index was, however, low-

ered to 14.9"7.1 mg gy1 and the Cu andŽCdrshell-weight indices remained constant 0.3"

y1 y1.0.1 mg Cu g and 0.14"0.05 mg Cd gthroughout the depuration time. Statistical analy-ses carried out to check the significance of changesoccurring in biometric indices of Zn-pollutedmussels after 51 days of depuration are shown inTable 1. Significant differences have beenrecorded in all the parameters tested, with the

Table 1Ž . Ž .Metal concentrations mg metalrg dry flesh wt. , metalrshell-wt indices mg metalrg dry shell-wt. , ZnrCu ratios and biometric

Ž . Ž .parameters mean"S.D. of mussels collected from Menakoz Bizkaia in March 1988 and March 1989 and after 51 days of˜aŽ .depuration in the laboratory in clean sea water March]May 1989

Ž .Parameter March 1988 March 1989 May 1989 ns20Ž . Ž . Ž .ns13 ns19 51-day depuration

UŽ .Zn concentration 173.9"77.3 501.5"194.2 827.5"410.4UŽ .Cu concentration 22.0"12.3 7.3"3.0 14.5"3.5UŽ .Cd concentration 1.9"1.0 2.9"0.7 7.9"3.2UŽ .Znrshell-wt. index 8.5"2.5 25.1"10.6 14.9"7.1

Curshell-wt. index 1.1"0.4 0.4"0.2 0.3"0.1Cdrshell-wt. index 0.087"0.03 0.14"0.04 0.14"0.05

UŽ .ZnrCu ratio 8.72"5.07 97.94"107.97 57.16"22.38UŽ .Length 34.92"3.73 25.31"2.84 30.35"2.41UŽ .Width 20.07"2.90 15.87"1.42 17.42"1.70UŽ .Height 14.77"2.38 11.50"2.67 12.87"1.26

UŽ . Ž .SSF 10351.47 a 4713.10"1626.59 6837.8"1252.28Heightrlength 0.42"0.04 0.45"0.08 0.42"0.03Widthrheight 1.37"0.23 1.44"0.32 1.37"0.21

UŽ .Widthrlength 0.57"0.06 0.62"0.06 0.57"0.06SVI 1.03"0.02 1.03"0.02 1.03"0.03

ŽU .Dry flesh wt. 107.07"43.61 61.79"19.36 29.98"7.07UŽ .Shell wt. 2.32"0.78 1.22"0.36 1.67"0.39

UŽ .Total weight 2424.28"812.07 1277.89"378.20 1701.31"390.68UŽ .FC 46.26"11.38 51.32"9.61 18.45"4.43UŽ . Ž .Fleshrwidth 5.33 a 3.90"1.21 1.72"0.35UŽ . Ž .Fleshrheight 7.25 a 5.37"1.15 2.34"0.53UŽ . Ž .Fleshrlength 3.06 a 2.43"0.66 0.98"0.18

aŽU . Ž . Ž . Ž .Significant differences P-0.05 between Zn-polluted March 1989 and depurating May 1989 mussels according to theŽ .one-way ANOVAs carried out with as1 and bs37 d.f. SSF, shell size factor length =width=height ; SVI: shell valve inequality

Ž .index. FC, flesh condition index; a , estimated from mean values.

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187 179

exception of SVI, HrL and WrH ratios. The dryflesh wt. in March 1989 was half of that recordedin March 1988 but, surprisingly, after depurationa reduction in dry wt. greater than 50% of that

Ž .recorded in March 1989 was observed Table 1 .This weight reduction under laboratory condi-tions does not appear to be associated with mal-nutrition or with handling stress since, as theincreased shell weight and shell size factor indi-cated, mussels kept on growing in the laboratory.In agreement with these findings, a significantreduction in flesh condition occurred if the twogroups of mussels collected at Menakoz are com-˜

Ž .pared with depurating mussels Table 1 .

3.2. Growth and condition of Zn-polluted musselsexposed to Cu, Zn and Cd

No significant mortality was recorded after Znand Cd treatment. Mussels exposed to Cu, how-ever, were the most affected and no survivorsoccurred after 27 days of exposure to HD-Cu.

The experimental exposure to Zn led to higherZn concentrations in the soft tissues of musselsŽ .Tables 1 and 2 . The Zn concentrations recorded,however, did not differ significantly from those

Ž .recorded in depurating mussels Table 2 . Copperconcentrations in the soft tissues of Zn-exposedmussels were significantly depressed in Zn-HD

Ž .mussels Table 2 . The ZnrCu ratio did not showa clear trend. Cadmium decreased gradually atincreasing Zn concentrations of exposure. Theexposure under laboratory conditions to Cu pro-voked increased Cu and Cd and concomitant de-creased Zn concentrations in the soft tissues of

Ž .mussels Table 2 when compared with their metalconcentration at the beginning of the experiment

Ž .in March 1989 Table 1 . The ZnrCu ratio wasŽdepressed until a value close to 0 at Cu-ID Table

.2 . The exposure to Cd caused increasing Cdconcentrations and lowered Zn and Cu concen-

Ž .trations in the tissues of mussels Table 2 . TheZnrCu ratio was low at Cd-LD and Cd-ID exper-imental groups but at Cd-HD reached valuessimilar to those exhibited by depurating musselssince Cu was heavily depressed as a result of this

Ž .treatment Table 2 .As seen in Table 2, the ZnrCu ratio decreased

significantly in mussels exposed to Cu and whileZn increased. These findings are obvious sincethe ZnrCu ratio depends directly on Zn and Cuconcentrations. The sensitivity of the index hasbeen verified since, according to a two-way

Ž .ANOVA, the type of metal M , the metal con-Ž .centration C and the interaction C=M affect

significantly the ZnrCu ratio by at least P-0.0001. Moreover, changes in ZnrCu ratio werelinearly explained by changes in the Cdrshell-weight index recorded in Cd-exposed mussels:

Table 2Ž .Metal concentrations mg metalrg dry flesh wt."S.D. recorded in mussels experimentally exposed to Zn, Cu and Cd after being

acollected from a Zn-polluted site

Experimental Zn concentration Cu concentration Cd concentrationgroup in soft tissues in soft tissues in soft tissues

Ž .C ns20 827.5"410.4 14.5"3.5 7.9"3.2Ž .Zn-LD ns26 1198.4"624.8 12.1"2.2 2.3"3.3Ž .Zn-ID ns26 771.7"259.5 12.5"6.2 2.1"2.9Ž .Zn-HD ns26 841.9"367.4 8.2"3.2 1.7"2.4Ž .Cu-LD ns20 799.1"256.1 25.8"13.0 8.5"5.6Ž .Cu-ID ns20 56.9"16.6 796.9"229.2 7.33"2.9

U U UŽ . Ž . Ž . Ž .Cu-HD ns0Ž .Cd-LD ns20 669.6"297.8 26.1"7.3 13.0"3.6Ž .Cd-ID ns20 839.9"397.1 31.5"5.5 103.0"19.0Ž .Cd-HD ns20 355.1"441.7 6.3"10.3 433.5"145.7

a ŽU .C, depurating mussels; LD, Low dose; ID, Intermediate dose; HD, High dose. Mussels died before Day 41 when metalanalyses were carried out.

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187180

Table 3Summary of two-way ANOVAs made to analyse the effect of

Ž . Ž .the exposure concentration C as2 d.f. and the type ofaŽ . Ž .metal T bs3 d.f. with w)200 d.f. within groups

Ž . Ž .Parameters P T P C

UWet flesh wt.UUDry flesh wt.UUUWidthUSSFUUUHydration

U UUUFCUUUFleshrwidthUUFleshrweightUUUFleshrlengthUUWidthrheightUUUWidthrlength

a No effect on dry shell wt., total weight, length, height,U Ž . UU Ž . UUU Ž .SVI, or heightrlength. P r -0.05; P r -0.01; P r

-0.001.

ZnrCusy4234.757 Cdrshell wt. index

w Ž . Ž . xq40.902 P b and P a -0.01

Ž .rs0.4303 Ps0.0003 d.f.s64

The changes in growth and condition of mus-sels experimentally exposed to Zn, Cu and Cd aresummarised in Figs. 1]4 and Tables 3 and 4.

The wet flesh wt. and the dry flesh wt. did notshow major changes as a result of the experimen-

tal exposures to Zn and Cu but wet flesh wt. wassignificantly increased in mussels exposed to CdŽ .Fig. 1; Table 3 . Both wet and dry flesh wt.exhibited significant changes in only a few experi-mental groups. The relation between bothparameters given as the percentage of hydrationin the soft tissues, however, changed significantlyas a result of exposure to each of the three metalsŽ .Fig. 1; Table 3 . The changes, however, were ofdifferent nature with dehydration in the cases ofexposure to Zn and Cu and hydration in the case

Ž .of exposure to Cd Fig. 1 . These changes inhydration might suggest some significant alter-ations in the tissue composition of musselsbetween March and May 1989, which could be inagreement with the significant changes in flesh

Ž .condition index FC in Table 1 .Shell growth occurred during laboratory experi-

ments and, in general terms, it was similarbetween depurating mussels and most of thegroups of mussels exposed to Zn, Cu and CdŽ .Tables 3 and 4 . Data for the linear regression

Ž . Ž .lines Fig. 2 of log shell wt log SW against the10

SSF are given in Table 4. For comparison pur-poses, shell weights have been calculated for thenine regression lines at a SSF of 6797.09 mm3, theoverall mean, and are shown in Table 4. Zinc-,Cu- and Cd-exposures did not produce a signifi-

Table 4w Ž . xSummary of one-way ANOVAs as P F , probability of the F-ratios made for each metal exposure to assess the effect of the

aŽ .exposure-concentration on the biometric parameters b, d.f. between groups; w, d.f. within groups

Zn-exposure Cu-exposure Cd-exposureŽ . Ž . Ž .bs3; ws94 bs2; ws57 bs3; ws76

UUUWet flesh wt.UUDry flesh wt.

ULengthUUU UUUWidthUUSSFUUU UU UUUHydrationUU UUUFCU UUUFleshrwidthU UFleshrheightU UUFleshrlength

UUUWidthrheightUU UUUWidthrlength

a U Ž . UU Ž .No significant differences for dry shell weight, total weight, height, SVI, or heightrlength. P r -0.05; P r -0.01;UUU Ž .P r -0.001.

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187 181

Ž . Ž . Ž .Fig. 1. Flesh weight parameters of mussels experimentally exposed to zinc Zn , copper Cu and cadmium Cd after beingcollected from a zinc-polluted site. C, depurating mussels; L, low dose; I, intermediate dose; H, high dose. Significant differencesbetween all pairs of mean values calculated for each exposure metal are indicated in the upper triangular matrix by an asteriskŽ .P-0.05 . Statistical significance is based on the multiple range test of Duncan after a one-way ANOVA.

cant effect on growth in thickness andror a re-duction in shell density of mussels in comparisonto depurating animals. However, HD-Cd exerted

Ž .a thickening in shell density beyond that foundŽ .in depurating mussels Fig. 2; Table 4 .

Flesh condition exhibited significant changes inŽ .relation to the exposure to Zn and Cd Fig. 3 .

Unexpectedly, the experimental exposures to Zn-ID, Zn-HD and to Cd-ID led to FC valuessignificantly higher than those recorded in de-

Ž .purating mussels Fig. 3 . Although no significantdifferences were found in all the other experi-mental groups, a trend to increase FC wasobserved in mussels exposed to Zn-LD, Cu-LD,Cu-ID and Cd-LD while a trend to reduce FC

Žwas observed in mussels exposed to Cd-HD Fig..3; Table 3 .

Significant changes were recorded in Znrshell-weight index in both mussels exposed eitherto Zn and Cd, in Curshell-weight index in mus-sels exposed to each of the three metals and inCdrshell-weight index in Cd-exposed musselsŽ .Fig. 4 . The Znrshell-weight indices of Zn-polluted mussels decreased to approximately 40%after the depuration period, while Cu andCdrshell-weight indices did not change signifi-

Ž .cantly Fig. 4 . A correlation analysis was carriedout to check whether the changes in some of thestudied biometric parameters could be associatedwith metalrshell-weight indices recorded in mus-

Ž .sels Table 5 . Zinc-exposure appeared to be re-

lated to reduced shell growth and increased fleshcondition index. Copper-exposure appeared to berelated also to altered shell growth and increasedflesh condition index but, additionally, it was as-

Žsociated with dehydration of the soft tissues Ta-.ble 6 . Conversely, Cd-exposure is linearly related

to hydration of the soft tissues.Significant changes were recorded in ZnrCu

ratios in response to the experimental exposureŽ .to either Zn, Cu or Cd Fig. 4 . Nevertheless, this

index appeared to be less sensitive than themetalrshell-weight indices and significant differ-ences were only found between depurating mus-

Ž .sels with HD-Zn, HD-Cu and HD-Cd Fig. 4 .

4. Discussion

Estimating metal bioavailability by measuringthe metal concentration in the soft tissues ofsentinel organisms is not as reliable as expected

Ž .from pioneer reports Phillips, 1976 . Some re-searchers have reported that growth is reduced in

Žmussels after exposure to pollutants Brown andAhsanullah, 1971; Scott and Major, 1972; Phillips,

.1976; Stromgren, 1982; Cajaraville et al., 1992a .¨Still, flesh weight and, to a lesser extent, wetweight are commonly used as references to assessmetal bioavailability. The changes in flesh weightrelated to depressed growth in the presence ofmetals could then be a handicap in the use of

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187182

ŽFig. 2. Linear regression lines for the three doses L, low; I,. Ž .intermediate; H, high of exposure to metals Zn, Cu and Cd

Ž . Ž .and for depurating mussels C , log shell weight mg on10Ž . 3shell size factor width= length =height in mm . Regression

line for C series is shown as a dashed line. The arrow indicatesŽ .the overall mean for the three experiments see Table 5 .

metal concentration values as indices of bioavail-Ž .ability Marigomez and Ireland, 1990 . Alterna-´

tively, condition indices were proposed by LobelŽ .and Wright 1982a to normalise metal concentra-

tions to a standard weight of sentinel mussels.Unfortunately, condition indices may also changeinconsistently in the presence of pollutantsŽCrosby and Gale, 1990; Marigomez et al., 1990;´

.Cajaraville et al., 1992a . Allometric parametersŽ .Lobel and Wright, 1982a,b may be useful as a

basis for trace metal determinations in a compar-ison to metal levels in mussels populations fromwidely differing habitats where absolute size is apoor indicator of age or growth rate. We havepresently found that HrL was invariant and onlyLD-Cd and ID- Zn have a slight effect on WrHand WrL. Therefore, allometric ratios indicatethat all the mussels belonged to the same size-class of the same population and that exposure tosublethal concentrations of either Zn, Cu or Cdfor a few weeks does not affect allometric ratios.Other alternative indices of metal bioavailabilityexamined try to avoid the influence of the changesin the biometric characteristics of sentinel organ-isms: metalrmetal ratios and metalrshell weight

Žindices Fischer, 1983, 1984, 1986, 1988; Car-.bonell and Tarazona, 1993; Soto et al., 1997a,b .

Within this context, the present investigation sug-gests that the use of metalrshell-weight indicesŽ .Fischer, 1983 provide a simple approach to esti-mate metal bioavailability which is not affected bynatural or metal-induced changes in musselbiometry.

In agreement with a series of previous reportsŽBrown and Ahsanullah, 1971; Stromgren, 1982;¨

.Manley et al., 1984; Stromgren, 1986 , the occur-¨rence of bioavailable Zn in the environment has

Fig. 3. Flesh condition index of mussels experimentally ex-Ž . Ž . Ž .posed to zinc Zn , copper Cu and cadmium Cd after being

collected from a zinc-polluted site. Symbols and statisticalsignificance as indicated in Fig. 1.

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187 183

Ž . Ž . Ž .Fig. 4. Metal indices as mg Znrg shell-wt. Zn index , mg Curg shell-wt. Cu index and mg Cdrg shell-wt. Zn index and values ofŽ . Ž . Ž .ZnrCu ratio in mussels experimentally exposed to zinc Zn , copper Cu and cadmium Cd after being collected from a

Ž .zinc-polluted site in March 1989 F . Symbols and statistical significance as indicated in Fig. 1.

been presently related to reduced flesh and shellgrowth in mussels. We have also found that thesize of mussels collected in March 1989 in the

field was reduced suggesting the occurrence of astress response due to increased Zn-bioavailabil-ity. Reduced shell growth could also be related to

Table 53 aŽ . Ž .Detail of the regression lines, log shell weight mg on shell size factor mm , calculated for each experimental group10

2Ž .Treatment N a"S.E. b"S.E. S y yyS.D. yqS.D.

Ž .C-Series 20 2.8705"0.0961 0.0000500"0.0000138 0.00570 1561.06 1687.25Ž .Zn-LD 20 2.7353"0.0815 0.0000678"0.0000126 0.00649 1506.91 1637.23Ž .Zn-ID 20 2.9018"0.0468 0.0000420"0.0000064 0.00295 1496.79 1582.85Ž .Zn-HD 38 2.8184"0.0476 0.0000601"0.0000079 0.00373 1648.12 1725.08Ž .Cu-LD 20 2.7786"0.0643 0.0000598"0.0000092 0.00753 1464.40 1601.25Ž .Cu-ID 20 2.8139"0.0554 0.0000537"0.0000071 0.00427 1459.79 1561.38Ž .Cd-LD 20 2.8692"0.0854 0.0000417"0.0000119 0.00576 1366.66 1477.75Ž .Cd-ID 20 2.8498"0.0674 0.0000467"0.0000047 0.00588 1412.86 1528.97Ž .Cd-HD 20 2.8203"0.0520 0.0000637"0.0000081 0.00281 1743.49 1841.28

a C-series, maintained in clean sea water for 51 days; LD, low dose of metal; ID, intermediate dose of metal; HD, high dose ofmetal; N, sample size; a, Y-intercept; S.E., standard error; b, regression coefficient; S y2, residual variance of y on x; y, shell weightŽ . 3 Ž .mg when shell size factor is 6797.09 mm overall mean .

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187184

Table 6Ž .Significant correlation coefficients r between biometric parameters and Znrshell-wt, Curshell-wt and Cdrshell-wt indices for the

athree exposure metals studied

Growth and condition Zn-exposure Cu-exposure Cd-exposureŽ . Ž . Ž .parameters Znrshell-wt. ns78 Curshell-wt. ns60 Cdrshell-wt. ns66

UUWet flesh wt. 0.4741UDry shell wt. y0.3359U UTotal weight y0.3295 y0.3067

ULength y0.3462UU UHydration y0.5052 0.3232

SVIUU UUFC 0.3813 0.5425

a U Ž .No significant correlation coefficients: dry flesh wt., width, height, SSF, widthrheight, heightrlength, widthrlength. P r -0.05;UU Ž .P r -0.01.

many other natural factors, such as, for instance,variations in food-particle concentration, timingof valve closure, and reduced filtration-rate or

Ž .food intake Nielsen and Stromgren, 1991 . On¨the other hand, changes in allometric shell growth

Žor flesh condition Lobel and Wright, 1982a,b;.Crosby and Gale, 1990; Aldrich, 1991 have not

been recorded. Under such conditions, the in-creased Zn concentration in soft tissues is notaccomplished by increasing the concentrations ofother metals such as Cu or Cd and the valuesof metal concentration resemble those of metalrshell-weight indices. Both metal concentrationand metalrshell-weight index values may be thenconsidered equally useful to assess metal bioavail-ability since both flesh and shell growth weresimilarly reduced in 1989 when environmentallevels of Zn raised.

Zinc-polluted mussels significantly improveŽ .their shell weight growth after depuration for 51

Ž .days. In agreement, Manley et al. 1984 foundŽ .that 3 days exposure to copper 10 ppb had a

significant inhibitory effect on shell growth, andthat recovery occurred in 14 days under normalconditions. Thus, the present laboratory condi-tions did not affect shell growth. The exposure toCd, however, caused a slightly reduced shellgrowth in comparison with depurating mussels.

Ž . Ž .Accordingly, Phillips 1976 and Stromgren 1982¨found that Zn- and Cd-exposure may provoke areduction in growth rate in M. edulis. The re-duced growth recorded in this work might be the

Žconsequence of protective valve closure Nielsen

.and Stromgren, 1991 or the result of a stress¨response related to metal detoxificationŽ .Stromgren, 1986 . In addition, mussels exposed¨to HD-Cd produced a small size shell but it wasthickened, suggesting that valve closure couldhave been relevant. Curiously, the effects of Cd-exposure were less marked than those recordedfor Zn-exposure. Maybe, Cd might displace a partof extra Zn in Zn-polluted mussels and thus thetoxic effect of Cd at LD- and ID-Cd is lowered.

Ž .Alternatively, Stromgren 1982 found that Cd¨may be innocuous to mussels at certain low doses,and may induce a significant increase in length at2 mg Cd ly1. Unexpectedly, the Cu concentra-

Žtions used in the present experiment 0.8 mg Cuy1 y1.l and 8 mg Cu l did not affect shell growth.

Exposures to 3 mg Cu ly1 are known to provoke aŽ .retarded growth in M. edulis Stromgren, 1982 ,¨

which close their valves when the environmentaly1 ŽCu is increased over 5 mg Cu l Davenport and

.Manley, 1978 . Consequently, it appears that theCu concentrations used herein should have causeda reduced growth in mussels. The absence of suchan effect might be attributed to the fact thatpre-exposure to Zn could diminish the toxicity of

Žfurther exposures to low doses of Cu Roesijadi.and Fellingham, 1987 .

In summary, as shell growth is affected byexperimental metal exposure we could expect towithdraw a conclusion similar to that obtained inthe field study but, interestingly, metal concentra-tion values and metalrshell-weight indices arenot comparable. In this sense, while metal con-

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187 185

centrations recorded in the soft tissues of depu-rating mussels increased without a source of ‘ex-tra’ metals, the Znrshell-weight index was re-duced as expected from a depuration process Cuand Cdrshell-weight indices remaining constant.Experimental exposure to either Zn, Cu or Cd ledto augmented values of Zn, Cu and Cdrshell-weight indices, respectively. The increase inmetalrshell-weight indices was related to the am-bient concentration of each metal, even in thecase of exposure to Zn, where Zn concentrationsdid not reveal any change between treatmentgroups. Since shell growth is affected as in thecase of the previous field study, how could it bethen that metal concentration and metalrshell-weight indices do not behave similarly?

It seems that the most relevant changes duringthe laboratory experiments were not in shellgrowth but in flesh weight. As a result of labora-tory maintenance and of the gamete spawningusually occurring between March and May, the

Ž .flesh condition index FC was reduced in depu-rating mussels in comparison with Zn-pollutedmussels collected from the field in March 1989.The fact that laboratory handling may inducegamete spawning has been previously reportedŽ .Cajaraville et al., 1991, 1992a . With regards tothe experimental exposures to metals, several in-teresting comments must be outlined. FC in-creased with increased concentrations of Zn andin mussels exposed to LD- and ID-Cd. Con-versely, HD-Cd produced a strong decrease in FCvalues beyond those recorded in depurating mus-sels. In most cases of metal-exposed mussels thereduction of FC is smaller at increasing doses.This response might be related to a blockage ofthe spawning activity in order to use gonad tissueas an energy reservoir for extra demands in thepresence of pollutants. Gonad resorption is atypical mechanism of response against environ-

Ž .mental stress Cajaraville et al., 1991 . In agree-ment with the present results, Stromgren and¨

Ž .Nielsen 1991 found in mussels exposed to 10 mgCurl and 40 mg Curl a reduction in spawningactivity of approximately 20% and 13%, respec-tively, in relation to controls.

Ž .Thus, flesh weight either wet or dry is veryvariable as a result of reproductive activity and of

the effects of metallic pollutants on this activity.This contributes to an unpredictable variability inflesh weight and thus in the estimations of themetal concentration in soft tissues as a whole.Moreover, for most metallic pollutants the gonad

Žis not an accumulating tissue Soto and.Marigomez, 1997 . Alternatively, we could use the´

metal concentration in a particular tissue otherthan the gonad. Sometimes, the digestive gland

Žhas been used for this purpose Regoli and.Orlando, 1993; Soto et al., 1997b but, unfortu-

nately, in many sentinel molluscan species gonadfollicles and digestive diverticula are closely im-

Ž .bricated i.e. winkles Littorina littorea . Moreover,the amount of connective tissue surrounding thedigestive diverticula changes throughout the re-

Žproductive cycle in marine molluscs Marigomez´.et al., 1992 and this could account for the unpre-

dictable variability in metal concentration in thesoft tissues.

5. Conclusion

Since the changes in the tissue composition andthe changes in growth rates do not affect ZnrCuratio and metalrshell-weight indices, theseparameters could be proposed as reliable indicesof metal bioavailability for ‘Mussel-Watch’ moni-toring programmes. From the present data, themost sensitive of these is the metalrshell-weightindex proposed by Fischer in 1983 and rarely

Žapplied afterwards Fischer, 1984, 1986, 1988;Marigomez and Ireland, 1990; Marigomez et al.,´ ´

.1990; Regoli and Orlando, 1993 . For this reasonit has been applied in regional ‘Mussel-Watch’

Žprogrammes in the Biscay Bay Soto et al., 1995,.1996; Soto and Marigomez, 1997 . In addition,´

metalrshell-weight index offers a high sensitivityand is easily calculated. Moreover, by simply us-ing the weight calculated for biometric estima-tions in previous ‘Mussel-Watch’ monitoring pro-grammes, or by weighing stored shells, most ofthe previous records on metal bioavailability,based on the conventional use of metal concen-trations, can be transformed into metalrshell-weight indices for comparison purposes.

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187186

Acknowledgements

This work was funded by the University of theŽ .Basque Country Project E136r90 and the

Ž .CICYT Project NAT90-0280 . This study wasalso partially funded by the Ministry of Educationof the Spanish Government through a researchfellowship to M. Soto. Thanks are due to Mr I.Quincoces for his helpful assistance in computing.

References

Aldrich JC. The use of allometry and individuality in theexamination of field data for sublethal toxicological effects.Chemosphere 1991;22:747]767.

Brown VM, Ahsanullah M. Effects of heavy metals on mortal-ity and growth. Mar Pollut Bull 1971;2:182]188.

Cajaraville MP, Marigomez JA, Angulo E. Automated mea-´surements of lysosomal structure alterations in oocytes ofmussels exposed to petroleum hydrocarbons. Arch EnvironContam Toxicol 1991;21:395]400.

Cajaraville MP, Marigomez JA, Angulo E. Comparative ef-´fects of the water accommodated fraction of three oils onmussels. 1. Survival, growth and gonad development. CompBiochem Physiol 1992a;102C:103]112.

Cajaraville MP, Marigomez JA, Angulo E. Comparative ef-´fects of the water accommodated fraction of three oils onmussels. 2. Quantitative alterations in the structure of thedigestive tubules. Comp Biochem Physiol 1992b;102C:113]123.

Cajaraville MP, Marigomez JA, Angulo E. Comparative ef-´fects of the water accommodated fraction of three oils onmussels. 3. Quantitative histochemistry of enzymes relatedto the detoxification metabolism. Comp Biochem Physiol1992c;103C:369]377.

Cajaraville MP, Marigomez I, Angulo E. Correlation between´cellular and organismic responses to oil-induced environ-mental stress in mussels. Sci Total Environ 1993;Suppl:1353]1371.

Carbonell G, Tarazona JV. A proposed method to diagnoseacute copper poisoning in cultured rainbow trout, On-

Ž .corhynchus mykiss. Sci Total Environ 1993;Suppl part 2 :1329]1334.

Cooke AS, Pollard E. Shell and operculum formation byimmature Roman snails Helix aspersa L. when treated withpp9-DDT. Pest Biochem Physiol 1973;3:230]236.

Crosby MP, Gale L. A review and evaluation of bivalvecondition index methodologies with a suggested standardmethod. J Shellfish Res 1990;9:233]237.

Davenport J, Manley AR. The detection of heightened sea-water copper concentrations by the mussel Mytilus edulis. JMar Biol Assess UK 1978;58:843]850.

Etxeberria M, Sastre I, Cajaraville MP, Marigomez I. Diges-´tive lysosome enlargement induced by experimental expo-

Ž .sure to metals Cu, Cd, and Zn in mussels collected from a

zinc-polluted site. Archiv Environ Contam Toxicol 1994;27:338]345.

Fischer H. Shell weight as an independent variable in relationto cadmium content in molluscs. Mar Ecol Prog Ser1983;12:59]75.

Fischer H. Cadmium body burdenrshell weight of mussels: aprecise index for environmental monitoring. Coun Meet IntExplor Sea CM-ICESrE 1984;41:1]19.

Fischer H. Influence of temperature, salinity and oxygen onthe cadmium balance of mussels Mytilus edulis. Mar EcolProg Ser 1986;32:265]278.

Fischer H. Mytilus edulis as a quantitative indicator of dis-solved cadmium. Final study and synthesis. Mar Ecol ProgSer 1988;48:163]174.

Lobel PB, Wright DA. Relationship between body zinc con-centration and allometric growth measurements in themussel Mytilus edulis. Mar Biol 1982a;66:145]150.

Lobel PB, Wright DA. Total body zinc concentration andallometric growth ratios in Mytilus edulis collected fromdifferent shore levels. Mar Biol 1982b;66:231]236.

Manley AR, Gruffydd LLD, Almada-Villela PC. The effect ofcopper and zinc on the shell growth of Mytilus edulismeasured by a laser diffraction technique. J Mar BiolAssess UK 1984;64:417]427.

Marigomez JA, Ireland MP. A laboratory study of cadmium´exposure in Littorina littorea in relation to environmentalcadmium and exposure time. Sci Total Environ 1990;90:75]87.

Marigomez JA, Vega MM, Cajaraville MP, Angulo E. Quanti-´tative responses of the digestive lysosomal system of winklesto cadmium. Cell Mol Biol 1989;35:555]562.

Marigomez JA, Ireland MP, Angulo E. Correlation of´cadmium shell-weight index with environmental stress indi-cators at the cellular and organismic levels in Littorinalittorea. Mar Ecol Prog Ser 1990;67:171]176.

Marigomez JA, Soto M, Angulo E. Seasonal variability in the´quantitative structure of the digestive tubules of Littorinalittorea. Aquat Living Resour 1992;5:299]305.

Nielsen MV, Stromgren T. Shell growth response of mussels¨Ž .Mytilus edulis exposed to toxic microalgae. Mar Biol1991;108:263]267.

Phillips DJH. The common mussel Mytilus edulis as an indica-tor of pollution by zinc, cadmium, lead and copper. I.Effects of environmental variables on uptake of metals.Mar Biol 1976;38:56]59.

Regoli F, Orlando E. Mytilus gallopro¨incialis as a bioindicatorof lead pollution: biological variables and cellular respon-

Ž .ses. Sci Total Environ 1993;Suppl part 2 :1283]1292.Roesijadi G, Fellingham GW. Influence of Cu, Cd and Zn

pre-exposure on Hg toxicity in the mussel Mytilus edulis.Can J Fish Aquat Sci 1987;44:680]684.

Ž .Scott DM, Major CW. The effect of copper II on survival,growth, respiration and heart rate in the common bluemussel, Mytilus edulis. Biol Bull 1972;143:679]688.

Soto M, Marigomez I. Metal bioavailability assessment in´‘Mussel-Watch’ programmes by automated image analysis

( )M. Soto et al. r The Science of the Total En¨ironment 247 2000 175]187 187

of BSD in digestive cell lysosomes. Mar Ecol Prog Ser1997;156:141]150.

Soto M, Kortabitarte M, Marigomez I. Bioavailable heavy´metals in estuarine waters as assessed by metalrshell-weightindices in sentinel mussels, Mytilus gallopro¨incialis. MarEcol Prog Ser 1995;125:127]136.

Soto M, Kortabitarte M, Marigomez I. Bioavailable heavy´Ž .metals in the Urdaibai estuary Biscay Coast as assessed

by metalrshell-weight indices in Mytilus gallopro¨incialis.Polish J Environ Stud 1996;5:59]66.

Soto M, Ireland MP, Marigomez I. The contribution of´metalrshell weight index in target tissues to metal bodyburdens in sentinel marine molluscs. 1. Littorina littorea. SciTotal Environ 1997a;198:135]147.

Soto M, Ireland MP, Marigomez I. The contribution of´metalrshell weight index in target tissues to metal bodyburdens in sentinel marine molluscs. 2. Mytilus gallopro¨in-cialis. Sci Total Environ 1997b;198:149]160.

Ž .Stromgren T. Effects of heavy metals Zn, Hg, Cu, Cd, Pb, Ni¨on the length growth of Mytilus edulis. Mar Biol1982;72:69]72.

Stromgren T. The combined effect of copper and hydrocar-¨bons on the length growth of Mytilus edulis. Mar EnvironRes 1986;19:252]258.

Stromgren T, Nielsen MV. Spawning frequency, growth and¨mortality of Mytilus edulis larvae, exposed to copper anddiesel oil. Aquat Toxicol 1991;21:171]180.

Sunila I, Lindstrom R. Survival, growth and shell deformities¨Žof copper- and cadmium-exposed mussels Mytilus edulis

.L. in brackish water. Estuar Coast Shelf Sci 1985;21:555]565.

Vega MM, Marigomez JA, Angulo E. Quantitative alterations´in the structure of the digestive cell of Littorina littorea onexposure to cadmium. Mar Biol 1989;103:547]553.