QUASIMEME laboratory performance study of the biological eVects of tributyltin (imposex and intersex) on two marine gastropod molluscs† I. M. Davies,*a A. Minchin,a B. Bauer,b M. J. H. Hardinga and D. E.Wellsa aFRS Marine Laboratory, PO Box 101, Victoria Road, Torry, Aberdeen, UK AB11 9DB. E-mail: daviesim@marlab.ac.uk, minchina@marlab.ac.uk, hardingm@marlab.ac.uk, wells@marlab.ac.uk; Fax:+44 1224 295511 bInstitut fur Spezielle Zoologie and Vergleichende Embriologie, HueVerstrasse 1, D 48149, Muenster, Germany Received 8th April 1999, Accepted 27th April 1999 The disruption of the endocrine systems of gastropod molluscs and consequential physiological changes (imposex and intersex) are used as biomarkers for environmental contamination by tributyltin compounds.The first international laboratory performance study on the determination of imposex and intersex in neogastropod molluscs, Nucella lapillus and Littorina littorea has been undertaken by the QUASIMEME (Quality Assurance of Information for Marine Environmental Monitoring in Europe) project.Samples of live gastropods were distributed and participants were asked to record shell height and sex, together with penis length and vas deferens sequence stage (VDS) in Nucella or the intersex stage (IS) and prostate length in Littorina.Calculations were made of vas deferens sequence index (VDSI) and the relative penis size index (RPSI) in Nucella and of intersex stage index (ISI ) and the average female prostate length (FPrL) in Littorina. Thirteen (87%) of the 15 participating laboratories returned data.The remaining two laboratories asked to participate in later exercises. For Nucella, seven laboratories reported sex ratios significantly diVerent from the reference laboratory data. DiVerences in penis length measurements between laboratories were largely random, although there were indications of systematic errors aVecting the data from three laboratories.Seven laboratories reported satisfactory data (Z-score |Z|<2) for VDSI. The inclusion of a high proportion of sub-adults in the Nucella samples may have made separation of the sexes more diYcult than in mature adults. The sub-adults will have had smaller pene than mature adults in the same population, and therefore any errors (random or systematic) in the measurement of penis length or observation of reproductive organs would have a potentially greater impact on the final reported values of the summary imposex indices.The Littorina sample did not show a high degree of intersex (ISI=0.41). The laboratories could determine the sex of Littorina reliably and only one laboratory reported data significantly diVerent from the reference laboratory.All except two laboratories reported satisfactory data for ISI. Introduction the Sea (ICES). A Training Workshop on the measurement It has been recognised for some years that the presence of of imposex and intersex was held at the FRS (Fisheries tributyltin (TBT) compounds in the marine environment, Research Services) Marine Laboratory, Aberdeen primarily derived from antifoulants, can lead to disruption (24–26 September, 1997) with the dual purposes of providing of the endocrine systems and characteristic physiological training in measurement techniques and to review and revise responses in gastropod molluscs.1,2 These alterations to the the JAMP Guideline document.10 Participants in the workshop female genital systems are termed imposex [superimposition recognised the need for a series of international laboratory of penis and/or vas deferens on female prosobranch snails, performance studies of the determination of imposex and notably the dogwhelk Nucella lapillus (L.)] and intersex (patho- intersex in marine snails, using the revised JAMP Guideline.logical alterations in the oviduct of littorinids and replacement The QUASIMEME (Quality Assurance of Information for of female by male organs), have proved to be sensitive Marine Environmental Monitoring in Europe) project provides biomarkers for the degree of organotin (especially TBT) a series of laboratory performance studies for chemical pollution in coastal waters.3–6 measurements in marine matrices.11 Part of the programme The inclusion of the biological eVects of tributyltin has developed towards providing quality assurance (QA) for compounds in Nucella lapillus and Littorina littorea (imposex measurements of polyaromatic hydrocarbon (PAH) metaband intersex respectively7,8) in the international marine Joint olites and organotin compounds in water and biota with a Assessment and Monitoring Programme (JAMP) of the Oslo view towards linking chemical measurements and biological and Paris Commission (OSPAR) immediately raised the need eVects studies.Conveners of the workshop collaborated with for quality assurance procedures including standardisation of the QUASIMEME project in the organisation of the first analytical techniques, training of scientists, and interlaboratory international exercise of this nature for the biological eVects studies.9 Detailed guidelines for sample site selection, sample of tributyltin compounds.12,13 This report presents an account collection, storage and data reporting were developed by of the exercise and of the lessons learned for future exercises.working groups under the OSPAR system, incorporating advice from the International Council for the Exploration of Materials and methods Samples of Nucella lapillus (sample code Q15001BT) and Littorina littorea (Q15002BT) were prepared and supplied by †©Crown copyright. J.Environ. Monit., 1999, 1, 233–238 233the FRS Marine Laboratory, Aberdeen. Nucella were collected described.3,8,14,15 Two reference laboratories were used in this study. Nucella were analysed by the FRS Marine Laboratory, from Skatey Shore, Stonehaven on 30 January and Littorina from Aberdeen harbour entrance on 2 February, 1998.The Aberdeen, Scotland and Littorina by the Institut fu� r Spezielle Zoologie und Vergleichende Embryologie, Muenster, snails were stored at the FRS laboratory aquarium until the morning of 3 February, the day of dispatch. The bulk samples Germany. These laboratories examined 424 Nucella and 241 Littorina respectively.of snails were placed on a tray, coned and quartered into samples of 50+, enclosed in nylon mesh bags, and placed into Styrofoam boxes lined with paper. Additional paper was added Results as packing material to minimise the movement of the animals The gastropod samples were generally received at UK during transit. A domestic freezer ice pack was placed on top destinations on 4 February and outside the UK on of the paper packing to keep the snails cool and the boxes 5/6 February and all these were live and in good condition.were sealed. All samples were packaged on the day of dispatch. Some samples suVered long delays, e.g., by customs authorities, Samples were distributed by overnight post (UK) or courier.and replacement samples were sent on 19 February. Participants in the exercise were provided with the samples of snails, the protocol for the exercise, the JAMP Guidelines Imposex in Nucella lapillus for Contaminant-specific Biological EVects Monitoring for TBT, as amended at the TBT EVects Training Workshop, Participants recorded their data on the data collector diskette September, 1997 and the data collector diskettes with from which the information (Table 2) was transferred directly instructions for their installation and use.10 by the QUASIMEME project oYce to the QUASIMEME The protocol for the exercise stated that: ‘This study covers data base.The initial step in data assessment was to check the the determination of imposex in Nucella lapillus (QIS001BT) calculations of the summary parameters [relative penis size and intersex in Littorina littorea (QIS002BT).Forty snails index (RPSI) and vas deferens sequence index (VDSI)] using should be chosen at random and examined from each sample, the formulae in the OSPAR guidelines. In some cases, particiand the observations on each individual snail should be pating laboratories had made errors in calculations.The reported using the QUASIMEME data collector and the datae-calculated values for each laboratory (Table 2) were used sheets included with this protocol. The observations listed in subsequent data analysis. below (Table 1) should be made, following the enclosed OSPAR JAMP guidelines. Nucella—homogeneity of samples. All laboratories reported Samples should be stored cool prior to analysis, preferably individual shell height measurements for Nucella and these in sea water, but it is advisable to analyse the samples as soon measurements (Fig. 1) were used to assess the homogeneity of as possible after receipt. You are requested to carry out the the samples. Visual examination of data indicated that two of observations following Technical Annex 3 of the OSPAR the participating laboratories (Q327P, Q158P) had reported Guidelines for Biological EVects Monitoring particularly for lower values for shell height than the other laboratories, and stage descriptions.The guidelines indicate that Nucella should this was confirmed by analysis of variance (ANOVA). The be examined live without narcotisation and that Littorina ANOVA was repeated omitting these two samples and indishould be narcotised with 7% magnesium chloride solution.cated that significant diVerences ( p<0.05) still existed between Parasitised animals should be discarded and alternative snails samples. The mean shell height of the remaining samples selected until a sample size of 40 has been obtained’. ranged from 22.79 to 24.60 mm (Table 2) and the ‘between The OSPAR Guidelines require that assessments of imposex sample’ sum of squares was 7% of the ‘within sample’ sum of and intersex be carried out according to the methods squares.The organisers decided that the remaining diVerences between these samples were small and would not compromise the exercise. Table 1 Measurements requested for the determination of imposex (Nucella lapillus) and intersex (Littorina littorea)a Nucella—penis length.Since there were significant diVerences in shell height between samples, the data for penis length for Nucella lapillus (QIS001BT) Shell height/mm both males and females were examined in relation to shell Sex coded (male=1, female=0) height. In both cases a positive correlation (Fig. 2) was found Penis length (eye piece graticule units) between mean penis length and the mean shell height. In order Penis length/mm to make comparisons between samples of diVerent mean shell Vas deferens stage height it was therefore necessary to correct mean penis length Calculated indices: Vas deferens stage index (VDSI) measurements to a standard shell height. Regression lines were Relative penis size index (RPSI) fitted to the relationships between penis length and shell height Littorina littorea (QIS002BT) Shell height/mm and used to correct the mean penis length measurements Sex coded (male=1, female=0) (Table 3) to the mean shell height (23.26 mm) for all Prostate length (graticule units) laboratories (omitting laboratories Q327P and Q158P).Prostate length/mm A bi-plot (Fig. 3) of the mean penis length for males and Intersex stage Calculated indices: Intersex stage index (ISI ) females (corrected for shell height) for all laboratories suggests Average female prostate length that diVerences between most laboratories arise from random (FPrL) errors, however there are indications that some degree of aThe indices referred to in Table 1 are calculated as follows: Vas systemic error has aVected the data for three (Q154P, Q158P deferens sequence index (VDSI) in Nucella=sum of imposex stage and Q104P) of the participating laboratories.values of all females sampled/number of females. Relative penis size index (RPSI) in Nucella=[(cube of average length of female Nucella—sex identification. The reference laboratory penis)/(cube of average length of male penis)]×100%.Intersex index examined 424 individual Nucella from the bulk sample. The (ISI ) in Littorina=sum of intersex stage values of all females sampled/ number of females. Average female prostate length (FPrL) in proportions of males and females found (47.2 and 52.8% Littorina=sum of prostate lengths (mm) of all females sampled/ respectively) were taken as the best estimates of the sex number of females.In addition, the relative penis length index (RPLI) composition of the bulk sample. There was no significant in Nucella was calculated as [(average length of female penis)/(average diVerence between the mean shell heights of males and females. length of male penis)]×100%. The numbers of males reported by the participating labora- 234 J.Environ. Monit., 1999, 1, 233–238Table 2 Summary of data received from participating laboratories and calculations undertaken by QUASIMEME for Nucella lapillusa Mean penis lengthb/mm Mean shell Recalculated RPSI Recalculated heightb/mm Male Female VDSI VDSIb (%) RPSIb (%) Q104P 22.79 2.95 1.91 3.95 3.95 27.14 27.14 Q128P 24.36 2.47 1.35 3.46 3.46 26.20 16.18 Q131P 24.57 2.46 1.31 4.00 4.00 15.10 15.10 Q136P 23.70 2.22 1.47 3.92 3.92 3.00 29.19 Q137P 22.87 2.27 1.50 3.95 3.95 28.57 28.57 Q154P 23.92 2.88 1.82c 2.70 2.70 16.80 25.22 Q156P 23.98 2.46 1.77 2.67 2.67 37.21 37.41 Q158P 20.57 1.14 0.65 4.00 4.00 18.80 18.80 Q247P 24.60 2.18 1.79 —d 3.23 —d 55.36 Q327P 19.53 1.48 0.86 1.81 1.81 19.60 19.60 Q328P 23.69 1.95 1.64 1.70 2.57 1.88 59.52 Q329P 24.47 2.38 1.37 3.95 3.95 19.07 19.15 Q330P 24.37 2.69 1.42 3.90 3.90 14.84 14.74 aFigures in bold indicate that calculation errors were made by the participating laboratories.bCalculations by QUASIMEME. cConversion from graticule units was miscalculated by laboratory. dNo calculated values were received for this index. Fig. 2 Mean penis length for males (a) and females (b) as a function of shell height. 52–66%, RPSI 14–29%) than the values for the other laboratories (RPLI 52–85%, RPSI 14–61%). Fig. 1 Box and whisker plots of shell height for Nucella lapillus (a) Nucella—vas deferens sequence index (VDSI). The reference and Littorina littorea (b). laboratory examined 224 female Nucella and obtained an overall VDSI value of 3.8. Based on these data, the predicted standard error of the VDSI on samples of 40 snails is 0.1 VDS tories in their sub-samples of 40 individuals ranged from 19 (47.5%) to 39 (97.5%).Chi-square statistics indicated that the units. The individual laboratory data may be interpreted in terms of Z-scores, following standard ISO/IUPAC16 data from seven (53.8%) laboratories were significantly diVerent from the reference laboratory data suggesting that some procedures where: misidentification of females as males had occurred at these Z=(observed value-target value)/total allowable error laboratories.The data in Table 3 are ordered by the proportion of Nucella identified as males. The application of this procedure within QUASIMEME is detailed by Wells and Cofino.17 If 0.1 VDS units is taken as The data for penis length and summary parameters were examined in two groups according to whether the sex ratio the total allowable error, then seven laboratories (Q158P, Q131P, Q136P, Q137P, Q329P, Q330P and Q104P) reported was significantly diVerent from the target value.T-tests indicated that there were no diVerences in male or female mean satisfactory data, i.e., -2<Z<2.The remaining six laboratories reported unsatisfactory data for VDSI. The OSPAR penis lengths between the two groups. However the calculated values of relative penis length index (RPLI) and RPSI for the Guideline (OSPAR 1997) indicates that in field surveys, VDSI values estimated by interpolation between sampling stations group of laboratories which had not shown significant diVerences from the reference laboratory were less variable (RPLI should be within 0.5 units of the true value with 90% confidence J.Environ. Monit., 1999, 1, 233–238 235Table 3 The recalculated summary values for Nucella lapillus that were used in data analysis, ordered by the proportion of snails identified as males Sex Mean penis length/mm Mean shell RPLI RPSI Male Female Chi-square height/mm Male Female VDSI (%) (%) Q158P 39 1 <0.001 20.57 1.59 0.97 4.00 83.68 58.60 Q154P 31 9 0.001 23.92 1.68 1.03 2.70 55.03 16.67 Q156P 31 9 0.001 23.98 2.77 1.52 2.67 72.12 37.52 Q131P 27 13 0.010 24.57 2.24 1.16 4.00 51.73 13.84 Q128P 27 13 0.010 24.36 2.28 1.22 3.46 53.52 15.33 Q136P 27 13 0.010 23.70 2.14 1.42 3.92 66.44 29.33 Q247P 26 14 0.024 24.60 2.34 1.69 3.23 84.70 60.76 Q327P 24 16 0.104 19.53 1.95 1.63 1.81 56.51 18.06 Q137P 20 20 0.720 22.87 2.34 1.55 3.95 66.15 28.95 Q329P 19 21 0.967 24.47 2.17 1.23 3.95 60.73 22.40 Q330P 19 21 0.967 24.37 2.50 1.29 3.90 61.27 23.00 Q104P 19 21 0.967 22.79 3.03 1.96 3.95 64.52 26.86 Q328P 19 21 0.967 23.69 2.12 1.30 2.57 51.60 13.74 the formulae as outlined in the revised (1997) JAMP guidelines.As for Nucella, some calculations for Littorina summary parameters showed errors and the re-calculated values are shown in Table 4. Littorina—homogeneity of samples. As for Nucella, the individual shell height measurements for Littorina were used to assess the homogeneity of the samples. Visual examination of data (Fig. 1) suggested that the two participating laboratories which had reported low values for shell height in Nucella (Q327P, Q158P) again reported lower shell heights compared to the other laboratories.This was confirmed by ANOVA. Fig. 3 Bi-plot of mean penis lengths (corrected for shell height) for The ANOVA was repeated omitting these two samples and male and female Nucella lapillus. indicated that significant diVerences still existed between samples.The mean shell height of these samples ranged from in the VDSI range 2–6. Depending upon the sampling density 19.13 to 20.37 mm (Table 4). The ‘between sample’ sum of and the gradient of eVect, this may equate to a less demanding squares was 5% of the ‘within sample’ sum of squares. The value for the total allowable error. For example, if 0.25 VDS organisers decided that the remaining diVerences between these units is taken as the total allowable error, then the data from samples were small and would not compromise the exercise.one further laboratory (Q128P) would be assessed as satisfactory and from four laboratories as unsatisfactory. Variation of the total allowable error between 0.1 and 0.25 units in this Littorina—sex identification.The reference laboratory exercise does not greatly alter the assessments of individual examined 241 individual Littorina from the bulk sample. The laboratory performance. proportion of males and females found (44.8 and 55.2% respectively) were taken as the best estimates of the sex Intersex in Littorina littorea composition of this bulk sample. There was no significant diVerence between the shell heights of males and females. The All data received were entered into the QUASIMEME data base and are summarised in Table 4.The initial step in the numbers of males reported by the participating laboratories, in their sub-samples of 40 individuals ranged from 12 (30%) data assessment was to re-calculate the summary parameters [interstage sex (ISI ) and female prostate length (FPrL)] using to 26 (65%).Chi-square statistics indicated that the data from Table 4 Summary of data received from participating laboratories and calculations undertaken by QUASIMEME for Littorina littoreaa Sexb Mean shell height/mm Recalculated Recalculated Male Female Chi-squareb Shell heightb ISI ISIb FPrL FPrLb Q104P 21 19 0.3290 19.13 0.63 0.63 Q128P 19 21 0.7337 19.82 0.24 0.24 0.14 Q131P 16 24 0.5395 19.40 0.04 0.04 Q136P 18 22 0.9822 20.37 0.23 0.23 Q137P 16 24 0.5395 18.72 0.04 0.04 Q154P 16 24 0.5395 19.73 0.75 0.75 0.76 0.76 Q156P 15 25 0.3516 19.75 0.72 1.00 Q158P 26 14 0.0103 16.33 1.50 1.50 0.13 0.13 Q247P —c —c —c —c —c —c —c —c Q327P 23 17 0.1070 15.14 0.12 0.12 Q328P 18 22 0.9822 19.66 0.44 0.46 Q329P —c —c —c —c —c —c —c —c Q330P 12 28 0.0594 19.67 0.75 0.72 aFigures in bold indicate that calculation errors were made by the participating laboratories.bCalculations by QUASIMEME. cNo data received for Littorina. 236 J. Environ. Monit., 1999, 1, 233–238Table 5 List of laboratories participating in the 1998 QUASIMEME laboratory performance study for imposex and intersex in gastropod molluscs Halifax Fisheries Research Fisheries and Oceans Canada 1707 and 1727 Lower Nova Scotia B3J 2S7 Canada Laboratory Water Street Halifax National Environmental Research PO Box 358 Frederiksborgvej 399 DK-4000 Denmark Institute Institut fur Spezielle Zoologie und HueVerstrasse 1 D 48149 Muenster Germany Vergleichende Embriologie Limnomar Bei der Neuen Munze 11 D-22145 Hamburg Germany Institute of Biology University of Iceland Grensasvegi 12 108 Reykjavik Iceland Fisheries Research Centre Marine Abbotstown Castleknock Dublin 15 Ireland Institute Martin Ryan Science Institute National University of Ireland Galway Ireland Netherlands Institute for Sea PO Box 59 1790 AB Den Burg Texel Netherlands Research Norwegian Institute for Water PO Box 173 Kjelsaas 0411 Oslo Norway Research University of Algarve Campus de Gambelas 8000 Faro Portugal Fisheries Research Services, Marine P O Box 101 Victoria Road Aberdeen AB11 9DB UK Laboratory Scottish Environment Protection Heriot-Watt Research Park Avenue North Riccarton Edinburgh EH14 4AP UK Agency (East), Clearwater House Scottish Environment Protection 5 Redwood Crescent Peel Park East Kilbride Glasgow G74 5PP UK Agency (West) Industrial Research and Technology 17 Antrim Road Lisburn Co Antrim BT28 3AL UK Unit Environment Agency-North East Skinnerburn Road Newcastle-upon-Tyne Tyne and Wear NE4 7AR UK only one (9%) laboratory (Q158P) were significantly diVerent The exercise was made more diYcult by collecting Nucella during February/March. In the east of Scotland, adult Nucella from the reference laboratory data.are normally concealed in deep fissures between rocks at this time of year. Consequently most of the Nucella used in Littorina—intersex sequence index (ISI ). The reference laboratory examined 133 female Littorina and obtained an this exercise were sub-adults and only a small proportion were the recommended toothed adults. This may have had various overall ISI value of 0.41. Based on these data, the predicted standard error of the ISI on samples of 40 snails is 0.08 ISI consequences for this exercise.Firstly, laboratories may have found it more diYcult to determine the sex of the sub-adults units. As for VDSI in Nucella, the individual laboratory data may be interpreted in terms of Z-scores. The OSPAR guidelines than of toothed adults.Secondly, the sub-adults probably had smaller pene than toothed adults, and consequently any errors state that ISI values should be within 0.5 ISI units of the true value (with 90% confidence). This is approximately equivalent (random or systematic) in the measurement of penis length or observation of reproductive organs would have a potentially to a total allowable error of 0.25 units.On this basis, the data from all laboratories are considered satisfactory (-2<Z<2) greater impact on the final reported values of the summary indices. other than Q156P (questionable, Z=2.36) and Q158P (unsatisfactory, Z=4.76). It is the intention of QUASIMEME to oVer further exercises as part of its overall programme of laboratory performance studies to support marine monitoring activity.Future exercises Littorina—prostate length. The reference laboratory found only one female Littorina with a measurable prostate gland in will take account of the diYculties that were encountered in the current exercise and attempts will be made to obtain more the 133 females that were examined. It is therefore concluded that the population was insuYciently aVected by TBT to allow strongly aVected Littorina. Investigations are required of the possible eVect of transport and storage on the length of the reliable comparisons to be made between laboratories of average female prostate length .All participants, except labora- penis in Nucella. It is also necessary to develop parallel study procedures in relation to the determination of imposex in tory Q154P, agreed with the reference laboratory in that they found 0 or 1 female with a measurable prostate gland.Buccinum undatum.18,19 A further training workshop is planned under QUASIMEME and the EU Biological EVects Quality Assurance in Monitoring Programmes (BEQUALM) pro- Discussion gramme. Particular attention will be paid to accurately sexing specimens, the measurement of penis length and the calculation This exercise has demonstrated that it is possible to conduct international laboratory performance studies for imposex and of summary indices.Discussions are in progress to integrate the laboratory intersex using live marine snails, provided that suYcient care is taken in packaging the animals and that transport time is performance studies of imposex/intersex with the chemical measurements of organotin compounds in shellfish tissue.This kept short. Participants were based in seven European countries and Canada. Where customs authorities did not intervene, combination would reflect the integrated biological–chemical approach to monitoring currently being developed by inter- samples were received in good condition. The success of the study is confirmed by the return of data from 87% of the national monitoring agencies such as OSPAR and the Helsinki Commission (HELCOM).participating laboratories. The data emphasise the care required in preparing sub-samples from a bulk collection of snails. As described Acknowledgements above, most of the samples were prepared by coning and quartering the bulk collection, and these samples showed The authors thank the participating laboratories (Table 5) for submitting their data to the study, the QUASIMEME project relatively small diVerences in mean shell height.Samples distributed on other occasions, for example to replace lost oYce staV (Caroline Warwick, Alison Allen, Kieren Smith, Judith Gourlay) for assistance with the distribution of samples, samples, showed greater diVerences.J. Environ. Monit., 1999, 1, 233–238 2378 B. Bauer, P. Fiorini, I. Ide, S. Liebe, J. Oehlmann, E. Stroben and collation of data and other administrative support, and the B. Watermann, Hydrobiologia, 1995, 309, 15. QUASIMEME scientific group for their support and advice 9 OSPAR, Joint Assessment and Monitoring Programme, Oslo and regarding the exercise.Paris Commission, Guidelines for Contaminant-specific Biological EVects Monitoring, 1997. 10 I. M. Davies, A. Minchin and M. Harding, Marine Laboratory Aberdeen Report 9/97, 1997, p. 33. References 11 D. E. Wells, A. Aminot, J. de Boer, D. Kirkwood and W. P. Cofino, Mar. Pollut. Bull., 1997, 35, 3. 1 G. W. Bryan, P. E. Gibbs and G. R. Burt, J. Mar. Biol. Assoc. 12 W. P. Cofino and D. E. Wells, Mar. Pollut. Bull., 1994, 29, 149. UK, 1988, 68, 733. 13 I. M. Davies, QUASIMEME Bulletin, 1998, Issue 5, pp. 19–20. 2 G. W. Bryan and P. E. Gibbs, ed. M. C. Newman and 14 P. E. Gibbs, P. L. Pascoe and G. W. Bryan, Comp. Biochem. A. W. McIntosh, in Metal Ecotoxicology: Concepts and Physiol., 1991, 100C, 231. Applications, Lewis Publishers, Michigan, 1991, pp. 323–361. 15 P. Fiorini, J. Oehlmann and E. Stroben, Zool. Anz., 1991, 226, 1. 3 P. E. Gibbs, G. W. Bryan, P. L. Pascoe and G. R. Burt, J. Mar. 16 M. Thompson and R. Wood, J. Pure Appl. Chem., 1993, 65, 2123. Biol. Ass. UK, 1987, 67, 507. 17 D. E. Wells and W. P. Cofino, Mar. Pollut. Bull., 1997, 35, 18. 4 S. K. Bailey and I. M. Davies, J. Mar. Biol. Assoc. UK, 1989, 18 C. C. Ten Hallers-Tjabbes, J. F. Kemp and J. P. Boon, Mar. 69, 335. Pollut. Bull., 1994, 28, 311. 5 D. Minchin, J. Oehlmann, C. B. Duggan, E. Stroben and 19 B. P. Mensinck, B. Van Hattum, C. C. Ten Hallers-Tjabbes, J. M. Everaarts, H. Kralt, A. D. Vethaak and J. P. Boon, M. Keatinge, Mar. Pollut. Bull., 1995, 30(10), 633. Nederlands Institut voor Onderzoek der Zee, Rapport, 1997, 6 D. Minchin, E. Stroben, J. Oehlmann, B. Bauer, C. B. Duggan 1997–6. and M. Keatinge, Mar. Pollut. Bull., 1996, 32(2), 188. 7 J. Oehlmann, E. Stroben and P. Fiorini, J. Molluscan Stud., 1991, 57, 375. Paper 9/02797A 238 J. Environ. Monit., 1999, 1, 233–238