Vol. 43, No. 4, pp. 34>358, 1997 Copyright 0 1996 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0141-I 136/97 $17.00+0.00
Polychlorinated Biphenyls in Fish and Shellfish from the Mersey Estuary and Liverpool Bay Richard T. Leah, Michael S. Johnson, Les Connor & Carolyn Levene Industrial Ecology Research Centre, Department of Environmental and Evolutionary University of Liverpool, PO Box 147, Liverpool L69 3BX, UK
(Received 3 March 1996; revised version received 23 June 1996; accepted 27 June 1996)
ABSTRACT Muscle tissue of$at$sh (dab, sole, flounder and plaice) from the Mersey Estuary and inner Liverpool Bay contain concentrations of CPCB(ICES 6) that range from 7.7-43.4 pg kg-‘. Using a derived conversion factor of 3.0 for CPCB(ICES 6) to CPCB(A1254), concentrations of CPCB(A12.54) in J4at$sh from the Mersey Estuary and Liverpool Bay range from 23-130 pg kg-‘, the latter of which classifies into the ‘upper’ category of contamination as defined by the Joint Monitoring Programme of the Oslo and Paris Commissions. Values of CPCB(A12.54) in Mersey Estuary flatjish are 1.2-4x higher than data reported for outer Liverpool Bay and the Irish Sea over the last twenty years. CPCBs in roundJish (cod, whiting) from the Mersey Estuary are low, though higher than in Liverpool Bay by a factor of up to 3x. Congeners No. 138 and No. 153 contribute much of the CPCB in all Jlatjish, with No. 180 and No. 101 present in significant but lesser amounts. The lighter, less chlorinated congeners No. 28 and No. 52, are absent from roundfish except those from the inner estuary. CPCBs in Mersey Estuary fish are 20-200 times higher than for the north-west Atlantic Ocean and the Solway Firth. For some Mersey Estuary fish CPCB values exceed national proscriptive limits set by the Environmental Protection Agency for the USA. Caution is advocated regarding the consumption ofjish from industrialised estuaries, including the Mersey, though there are no statutory limits for PCBs in fish or fishery products within the European Union. Copyright 0 1996 Elsevier Science Ltd
INTRODUCTION The Mersey Estuary is the outlet for one of the most extensive river catchment systems in Britain, covering an area of some 5000 km2 including the major conurbations of Liverpool and Manchester. Discharges of industrial effluent to the River Mersey and historic inputs of pollutants to Liverpool Bay from sewage disposal and dredge spoil have been the stimuli behind the long term monitoring programme for pollutants in fish populations of the Irish Sea (e.g. Franklin, 1987, 1990, 1991). 345
R. T. Leah et al.
Emphasis in monitoring has mainly been towards fish that dominate the commercial catch from offshore areas. However, coastal and estuarine sites are also important marine habitats where biota may have a greater exposure risk from pollutants than in the deep sea. This is especially true for toxic environmental contaminants with a long residence time. Polychlorinated biphenyls (PCBs) are prominent in this regard as their lipid solubility enhances amplification along the marine food chains that lead to wildlife and humans. Whilst acute effects of PCBs have never been viewed as a serious problem, their chronic effects are so similar to those of DDT [ l,l, 1 trichloro-2,2 bis (Cchlorophenyl) ethane] and DDE [ l,l, dichloro-2,2 bis (Cchlorophenyl) ethylene], that they may well act synergistically (WHO, 1979). Consequently, there has been speculation that PCBs contribute to the lower immune competency of seal populations that is thought to result in greater susceptibility to phocine distemper (Harwood, 1990). Organochlorines have also been implicated in the reproductive failure of seal populations in the Baltic and elsewhere (Harding & Addison, 1986; Reijnders, 1986; Morris et al., 1989). Exposure of fish to PCBs is potentially significant as regards the Mersey Estuary and Liverpool Bay since there has been speculation that elevated levels in certain species of Mersey fish compromises the health of the offshore seal population (Johnston et al., 1991). PCBs have also been linked to impaired reproductive success in fish of the Baltic and North Seas (von Westernhagen et al., 1981; Barnthouse et al., 1990), and also seabirds (Kubiak et al., 1989; Becker et al., 1993). The research reported here had two principal objectives. Firstly, to establish the contemporary levels of PCBs in biota of the Mersey Estuary and Liverpool Bay, and the PCB status of these ecosystems in the context of the British coastline and other European waters. The sampling and analysis programme thus focused on fish and shellfish from sites between the inner Mersey Estuary and Liverpool Bay, plus a reference site in the Solway Firth. The second main objective surrounded the increasing recreational use of the Mersey Estuary as an angling resource, and the need to establish a database on PCBs in fish in relation to possible health risks to humans from consumption of fish caught in the estuary. Subsidiary objectives were to examine inter-specific differences in contamination and small scale spatial variation in the PCB congener signature of fish from different sites within the Mersey Estuary ecosystem.
Field sampling Sampling was undertaken between 1990 and 1992, using local fishing boats contracted during seasons of the year, as appropriate to the six target species of fish. Two sites were sampled in the inner Mersey Estuary, Eastham and Garston, and two within the outer Mersey Estuary/Liverpool Bay, namely Rock Channel and Great Burbo Bank (Fig. 1). Targeted flatfish were plaice ( > 25 cm; Pleuronectes platessa), dab ( > 20 cm; Limanda limanda), flounder (> 25 cm; Platichthys jlesus) and Dover sole (Solea solea). Roundfish taken for analysis were whiting (> 27 cm; Merlangius merlangus) and cod (> 35 cm; Gadus morhua). Site-species combinations where fishing was conducted successfully are shown in Table 1. Comparison samples were to be collected from the Solway Firth where, in practice, only flounder were caught in sufficient numbers. Small scale sampling of the invertebrate community of the Mersey Estuary was also undertaken, including: shrimps
PCBs in jish and shell&h
Fig. 1. Sampling sites in the Mersey Estuary and Liverpool Bay.
TABLE 1 CPCB(ICES 6) in Fish from the Mersey Estuary and Solway Firth” Species Cod Whiting Whiting Whiting Whiting Dab Dab Dover Sole Dover Sole Dover Sole Flounder Flounder Flounder Flounder Plaice Plaice Plaice Plaice
Wet wt basis (pg kg-‘)
HEL basis (pg kg-‘)
Rock Channel Garston Eastham Rock Channel Gt. Burbo Bank Rock Channel Gt. Burbo Bank Garston Eastham Gt. Burbo Bank Garston Rock Channel Gt. Burbo Bank Solway Firth Eastham Garston Rock Channel Gt. Burbo Bank
1.8 l 0.6 (22) 6.2 ztO.7 (22) 7.8 f 0.6 (22) 2.2 f 0.5 (22) 1.7*0.2 (11) 43.4* 13.5 (11) 7.7 f 3.4 (44) 14.5 f 2.4 (22) 11.3h2.9 (11) 12.7*2.9 (22) 25.2zt2.1 (11) 14.9*2.1 (11) 17.9zt3.1 (33) 0.5*0.1 (22) 10.1* 1.2 (11) 10.1 i2.2 (22) 15.1 f 3.0 (33) 9.6f 1.6 (41)
765 f 613 (22) 2253 f 921 (22) 2691 f 501 (22) 892 f 660 (22) 637rt387 (11) 3283%2159 (11) 725 f 448 (44) 4545 l 7092 (22) 6051&9059 (11) 4045 f 4022 (22) 5653* 1824 (11) 5880*5345 (11) 3896 f 2308 (33) 408 f 75 (22) 1016rt412 (11) 2118f2131 (22) 2019% 1014 (33) 2311*1181 (41)
“CPCB(ICES 6): mean ;t SD (n). HEL: Hexane extractable lipids.
R. T. Leah et al.
(Crangon vulgaris); mussels (Mytilus edulis); starfish (Asteria rubens); whelks (Buccinium undatum) and hermit crabs (Pagurus bernhardus). Analytical
The lipid-rich liver is where organochlorines are known to accumulate in fish (Franklin, 1987; Franklin & Jones, 1993). However, in this study, muscle tissue was analysed since, unlike wildlife predators, humans usually eat only the muscle tissue of fish. Specimens were frozen upon landing, later defrosted, weighed, measured and dissected. Muscle samples were then refrozen to await analysis. Invertebrates were depurated passively in clean seawater for 24 h to eliminate sediment and residual food items from the gut, and were analysed on a composite soft tissue basis. Analytical methods for PCBs mainly followed proven analytical techniques developed by MAFF (Allchin et al., 1989) but were adapted for the discrimination and determination of individual congeners. 5 g of muscle tissue were ground with 30 g of anhydrous Na2S04 and stored for 48 h to ensure complete dehydration. Samples were then solvent-extracted for 4 h with 120 ml of n-hexane in a soxhlet assembly. After cooling, the extract was quantitatively transfered to a 100 ml flask, and 50 ml of the volume-adjusted extract were then removed for the gravimetric determination of hexane-extractable lipids (HEL). Clean-up of the extracts was achieved using 3 g of 5% deactivated A1203 topped with a little NazS04. To avoid exceeding the maximum lipid capacity of an Al203 column, typically 50 mg, the appropriate volume of extract was calculated, reduced in volume to 1 ml using a stream of nitrogen, and transferred quantitatively to the A1203 column. The column was then eluted with 20 ml n-hexane and the first 16 ml collected and again reduced to 1 ml volume. Separation of PCBs from the more polar pesticides was undertaken using 2 g of 5% deactivated SiOz topped with Na2S04. Extracts were eluted with 20 ml of 2% tetrahydrofuran in n-hexane, typically collecting successive fractions of 5 ml and 9.5 ml. Analysis was undertaken on a Varian 3500 GC fitted with an electron capture detector, a Varian 8035 autosampler and an automated DS651 data capture station. The gas chromatograph (GC) was configured with a cool, on-column injector and a 3 m long, 320 urn i.d. deactivated fused silica retention gap connected to a 60 mx0.25 mm (i.d.) DB5 column of film thickness, 0.25 urn. The GC was operated using H2 as the carrier gas (40 cm se-’ average linear velocity), an injection volume of 1 ul and an injector temperature of 60°C held for 6 set with a temperature programme of 180 “C min-’ to 280 “C maintained to the end of the run. The initial column temperature was 80°C for 1 min, operated with a twopart temperature programme: 7.5”C min-’ to 200°C and 2.5 “C min-’ to 280°C held for 5 min. The detector temperature was isothermal at 300°C and the make-up gas was N2 used at 30 ml min-‘. The internal standard was octachioronaphthalene. Quality assurance involved the use of rotating procedural blanks, concentrating pesticide grade solvents to detect the presence of interfering peaks. Accuracy and precision were checked periodically by the analysis of the BCR certified reference material, CRM No. 349 (Chlorobiphenyls in cod liver oil; Table 2). Thimbles were spiked with decachlorobiphenyl as a standard for which > 95% recovery was obtained consistently. Data handling
Environmental PCB contamination comprises varying amounts of the 209 possible congeners, and concentrations of ‘total’ PCB in environmental samples can be expressed
PCBs in jsh and shellJish
by any of several different methods, none of which is adequate in all respects. Included in the options for expressing XPCB in environmental samples are the expressions: E ICES congeners; Aroclor 1254; and mixtures of Aroclor (Balls&miter & fill, 1980). In this study the analytical programme comprised various intra-calibrated methods that enabled calculation of the results in different ways for the purposes of comparison. For example, XPCB expressed as Aroclor 1254 was calculated for historical reasons, and to interpret Mersey Estuary data against other published information. Data are expressed mostly on a wet tissue weight basis, though summary data are also given in terms of hexane-extractable lipid (HEL). The diversity of analytical methods and standards means that inter-comparisons of published PCB concentrations in biota are difficult. Early method of analysis relied upon separation and quantitation, by height or area, of a very small number of broad chromatographic peaks, standard&d against a technical Aroclor mixture. This approach usually identifies the particular technical mixture, the most appropriate of which for studies of marine biota is Aroclor 1254 because the pattern of congener composition in Al254 is similar to that found in fish. These methods give an estimate of the concentration of CPCB in Aroclor equivalents but little information on congener composition. Latterly, quantitation of individual congeners has been favoured, driven by the hypothesis that a meaningful environmental assessment of the PCB group can be achieved only by focusing on those congeners that bioaccumulate or are potentially toxic (McFarland & Clarke, 1989). High resolution capillary chromatography has enabled accurate quantitation of many single congeners in PCB mixtures; but still only a few such congeners are targeted by the analytical programmes used in large scale environmental sampling and analysis. Those targeted in this study are six of the congeners known as the ICES 7 (Table 3). TABLE 2 Analysis of Polychlorinated Congener No.
Biphenyls in Cod Liver Oil (BCR 349)
CertiJied value f 95% conjdence interval
Experimental value f 95% confidence interval
68zk7 149zk20 370+ 17 765+45 938k40 282 f 22
65zt8 143 f 12 413*15 850+24 907 * 25 284zt9
101 138 (f 163) 153
180 “All values in pg kg-‘.
TABLE 3 Numbering and Structure of Polychlorinated Biphenyl Congeners in CPCB(ICES 7) Congener No. 28 52
101 118” 138 153 180 “Omitted from the expression CPCB(ICES 6).
trichlorobiphenyl [ tetrachlorobiphenyl pentachlorobiphenyl pentachlorobiphenyl hexachlorobiphenyl hexachlorobiphenyl heptachlorobiphenyl
2,4,4’] [2,2’,5,5’] [2,2/,4,5,5/l [2,3’,4,4’,5] [2,2’,3,4,4’,5’] [2,2’,4,4’,5,5’] [2,2’,3,4,4’,5,5’]
R. T. Leah et al.
The ICES 7 congeners have increasingly become accepted as a ‘standard’ set for environmental monitoring (Lang, 1992), but with supplementary congeners added to the list wherever possible, in order to better describe the PCB profile. The ICES 7 list contributes a large percentage of the peak area generated by environmental mixtures, and so provides a valid basis for the calculation of XPCB. Six congeners from the ICES 7 list were used as prime, quantitative standards in this study, and to report CPCB values under the label ZPCB(ICES 6). The use of the ICES 7 group less congener No. 118 (i.e. No. 28, 52, 101, 138, 153, 180), as a basis for reporting CPCB(ICES 6), was because of the perceived historical conflict in the resolution of No. 118 from No. 149 (Schultz et al., 1989). Adequate resolution of these two congeners can, in fact, now be achieved routinely (Franklin & Jones, 1993). The analytical values reported here are expressed on the basis of the congener which is expected to make the dominant contribution to the chromatographic peak in question. In a number of cases more than one congener will co-elute (e.g. No. 138 and No. 163; Larsen 8z Riego, 1990). The early published information on PCB contamination is dominated by Aroclor quantitation using packed column chromatography. These data are extremely important for the purposes of determining trends through time; but the cross-comparison of databases acquired over timescales during which analytical methods have evolved requires the use of a conversion factor if they include a mixture of data expressed as CPCB(ICES) or XPCB(Aroclor 1254). Each PCB congener has a different ECD response factor which is curvilinear over the range of concentrations found in the environment. This produces a wide range of conversion factors depending on the relative and absolute concentrations of the congeners present. In this study a conversion factor of x3.0 for XPCB(ICES 6) to XPCB(Al254) was generated experimentally. Calibration of the gas chromatograph using a standard solution of the ICES 6 congeners was followed by quantitative analysis of these congeners in a standard Al254 solution. The conversion factor was derived by dividing the concentration of the Al254 solution by the XICES 6 congener concentrations in the Al254 solution.
RESULTS For sites at, or seaward of, the mouth of the Mersey Estuary, CPCB(ICES 6) data showed consistent but low values in muscle of roundfish, generally < 2 pg kg-’ in cod and < 3 pg kg-’ in whiting (Table 1). Values for flatfish varied with species and site. The highest XPCB(ICES 6) mean was for dab from Rock Channel (43.4& 13.5 pg kg-‘). On a wet weight basis, this was nearly double the mean for flounder from the inner estuary site at Garston (25.2 f 2.1 pg kg-‘) and more than double that for Gt. Burbo Bank (17.9* 3.1 pg kg-‘). It is notable that on an HEL basis, the reverse was true in that CPCB(ICES 6) values for the same flounder were higher than for dab by 1-2x . The wet weight values for the Mersey Estuary are 36--50x higher than the mean for the Solway Firth population of flounder (0.5 f 0.1 pg kg-‘). The lowest means for flatfish from the Mersey Estuary were for dab and plaice from Gt. Burbo Bank, at 7.7*3.4 and 9.6& 1.6 pg kg-‘, respectively (Table 1). Data for XPCB(ICES 6) in inshore invertebrates are summarised in Table 4. Concentrations of XPCB were mostly higher for invertebrates than for fish based on comparisons within-site, with the mean results for hermit crabs (105 f 12 pg kg-‘) and
PCBs infish and shelljsh
TABLE 4 CPCB(ICES 6) in Invertebrates Species
Shrimp Mussels Starfish Whelk Hermit crab
from the Mersey Estuary and Liverpool Bay
Wet wt basis (pg kg-‘)
HEL basis (Fg kg-‘)
New Ferry New Ferry Gt.Burbo Bank Gt.Burbo Bank Gt.Burbo Bank
8.2 f 6.2 20& 1.1 9.9 f 6.0 91 zt6.6 105 f 12
1230* 128 2180* 123 650 f 49 6710*622 1754* 123
“CPCB(ICES 6): mean f SD (n = 11).
whelks (91 f 6.6 ug kg-i) from Gt. Burbo Bank being 5-15x those for the corresponding flatfish populations. Recently, whelks, along with Dover sole and plaice, have started to contribute to the commercial fishery in Liverpool Bay. Moreover, in recent years commercial fishing activity has intensified within the Rock Channel and Great Burbo Bank areas, as a result of declining catches elsewhere. The contribution of different congeners to XPCB(ICES 6) varied between fish species and between sites. In the outer Mersey Estuary and Liverpool Bay, congeners No. 28 and No. 52 were not detected in cod and whiting in which No. 138 and No. 153 were co-dominant and accounted for > 70% of CPCB. Congener No. 180 was the next most abundant and was present in higher concentrations relative to congeners No. 138 and No. 153 than is usual for fish (Leah & Rogers, 1991). This contrasts with the pattern in the inner Mersey Estuary where whiting from Eastham and Garston showed all six congeners, with No. 28 and No. 52 only marginally less prominent than No. 138 and No. 153. Congener No. 52 was undetectable in most of the dab population at Gt. Burbo Bank (Fig. 2); but other species of flatfish from the same site showed a consistent presence of congeners No. 28 and No. 52 in trace to modest amounts (Figs 2-5). With the exception of Dover sole, congeners No. 153 and No. 138 were found in approximately equal (1:l) ratios. Together, these two congeners contributed 5&70% of XPCB(ICES 6). Congener No. 180 contributed significantly more (up to 33%) towards CPCB(ICES 6) in some flatfish, than in roundfish, especially in flounder and plaice. Only congeners No. 138, No. 153 and No. 180 contributed to CPCB in flounder at the reference site in the Solway Firth, with No. 153 marginally dominant. Invertebrates from Gt. Burbo Bank showed a different congener pattern to that for flatfish from the same site. Whilst all six ICES congeners were clearly present, the ratio of the more heavily chlorinated congeners (No. 153, No. 138 and No. 180) to lighter congeners (No. 28 and No. 52) in hermit crabs and whelk (combined ratio - 16.2 and 33: 1, respectively) was greater than in flounder (5.2:1), plaice (6.8:1) and Dover sole (12.7:1). Dab were an exception to this pattern, due to the virtual absence of congenem No. 28 and No. 52 in this species at Gt. Burbo Bank.
DISCUSSION For two of the most abundant congeners (No. 138 and No. 153), the patterns in fish may be linked to their environmental persistence. For a mixture of congeners emanating from a common source, a close correlation in contamination levels would be expected between
R. T. Led
101 153 138 180
PCB CongmerNumber Fig. 2. Congener contribution (No. 28-No. 180) to CPCB(ICES 6) in dab from Gt. Burbo
101 153 138 180
101 153 138 180
Fig. 3. Congener contribution (No. 28-No. 180) to CPCB(ICES 6) in Dover sole from Gt. Burbo Bank.
Fig. 4. Congener contribution (No. 28 -No. 180) to EPCB(ICES 6) in flounder from Gt. Burbo Bank.
101 153 138 180
PCB Cowncr Number Fig. 5. Congener
contribution (No. 28-No. 180) to EPCB(ICES 6) in plaice from Gt.
members of the group. Regressions of the (HEL) concentrations of congeners No. 138 and No. 153 in flounder showed a close correlation (r2 = 0.97; p < 0.001). With a regression slope close to unity and an intercept very near to the origin (Fig. 6) there must be a close similarity in the processes leading to bioaccumulation of these two congeners. All the ICES congeners measured were regressed against No. 180, the most chlorinated member of the group. A typical dataset of HEL values, for Dover sole from Garston in the inner Mersey Estuary, is shown in Fig. 7. Concentrations of the different congeners different
PCBs in fish and shellfish
( pg kg-l x1000, HEL)
Fig. 6. Congener ratios (No. 138: No. 153) in flounder from Gt. Burbo Bank (HEL basis).
y= 0.15487+1.7030x R-2 zO.996 U153 4
y= 0.19495+1.4671x R-2.0.993
=i E! 0‘ g
y=O.11937+0.75158x R-2 = 0.947 (101 y=O.l4732 +0.11624x R-2:0.414 3
PC828 PCB52 PCBlOl PCB153 PCB138
y=0.12138+5.7869c3x R-2 I 0268
K 2 iI?2 2 B s
PCB t180 (erg kg-l x1000, HEL) Fig. 7. Congener ratios (ICES: No. 180) in Dover Sole from Garston (HEL basis).
are mostly highly correlated but the slopes of the regression lines vary, and the coefficient of correlation decreases with the slope. This decrease appears to correspond with the chlorination level of the congener, and may reflect the greater susceptibility of less chlorinated congeners to degradation. The highly chlorinated and toxic congeners, such as No. 138, No. 153 and No. 180, are known to magnify in food chains to higher relative amounts, with implications for higher marine mammals and other higher organisms (Neilsen, 1994).
R. T. Leah et al.
In 1984, MAFF reported residues of CPCB(A1254) in muscle of fish from Liverpool Bay in the range 5569 pg kg-’ (wet wt; Franklin, 1987). More recently the CPCB(A1254) value has been shown to be stable, with data for 1987 (748 pg kg-‘) and 1988 (366 pg kg-‘) comparable with 1984 but significantly lower than some twenty years ago (l&660 pg kg-‘; Franklin, 1987). With the exception of dab, which usually has the highest values in MAFF surveys of the Irish Sea, the range of CPCB (A1254) for Mersey Estuary fish is 1.24x higher than recent MAFF data for Liverpool Bay (Franklin & Jones, 1993). MAFF results for CPCB in whiting (5-12 ltg kg-’ for 19841988) are significantly lower than the data acquired in this study for inner estuary sites (18.623.4 pg kg-‘) where sampling was conducted in 199 1-1992. The dab datasets for both the MAFF and present studies are surprising for they cover a very wide range of mean values and include both the highest and also the lowest mean CPCB values in the range that represent the extremes of mean CPCB data covering the four flatfish species common to both projects. The Oslo and Paris Commissions have set non-statutory guidelines for CPCB(A1254) in fish muscle (wet wt; ICES, 1992). Under the OSPARCOM Joint Monitoring Programme (JMP), data for CPCB(A1254) in fish classify into three categories: < 10 ng kg-’ (lower); 10-50 pg kg-’ (medium); and > 50 pg kg-’ (upper; Franklin, 1991). Using the 3x conversion factor, any EPCB(ICES 6) mean of > 16.7 ug kg-’ classifies into the ‘upper’ JMP category. This ‘upper’ classification applies to flounder from Garston and dab from Rock Channel, but also to flounder from Gt. Burbo Bank. Inner Mersey Estuary data for flatfish were all within the XPCB(A1254) range of 30.3-75.6 ug kg-‘. Of particular interest are the XPCB results for Rock Channel and Great Burbo Bank, both outer estuary sites. Whilst XPCB(A1254) data for whiting showed lower concentrations at these down river sites compared to Garston and Eastham, this did not apply to flatfish, the mean values for Rock Channel and Great Burbo Bank being 44.7 and 53.7 ng kg-’ in flounder, and 28.8 and 45.3 pg kg-’ in plaice, respectively. However, not all flatfish followed this trend, flounder from Garston, in the inner estuary, having a mean XPCB(A1254) value of 75.6 pg kg-‘. The considerable inter-specific variation in XPCB concentrations is not unexpected and may reflect the contrasting spatial or territorial histories of different species, and the consideration of XPCB levels in species with a largely open sea history alongside values for those that spend more time in coastal and estuarine areas. The clear distinction between the flatfish and roundfish in terms of CPCB concentrations probably reflects their contrasting habitat preferences and feeding strategy. Flounder are known to inhabit the Mersey and also the adjacent Dee Estuaries for feeding, and also breeding, and dab similarly spend a great deal of their life within estuaries, moving only a short distance offshore. In the Mersey Estuary these two species have generally higher levels of XPCB than do plaice. The latter show distinct migration patterns in Liverpool Bay, feeding inshore on shellfish in the Mersey Estuary but only in late spring and summer. Cod and whiting, which migrate to the coastal areas only in the late autumn and winter, have the lowest concentrations of XPCB. Data for flounder sampled between 1986 and 1988 along the French coastline from the Seine Estuary in the north, to the Gironde Estuary off the south-west coast, place the Mersey Estuary figures in an interesting perspective. Typical Mersey Estuary data at 44.7-75.6 pg kg-’ CPCB(A1254) are mostly higher than for the Loire Estuary (50.4 pg kg-‘; converted from dry to wet wt by dividing by 5) but very much lower than
PCBs injish and shelljish
the Seine Estuary and eastern English Channel (861-1361 c(g kg-‘; Cossa et al., 1992). Moreover, these latter data compare with those for the highly industrial&d Elbe Estuary in Germany (Kohler et al., 1986). Flounder from the Solway Firth contained much lower concentrations of PCBs than the Mersey Estuary, with a mean XPCB(ICES 6) of 0.6 ug kg-i (CPCB(A1254): 1.8 ug kg-‘). Moreover, XPCB(A1254) in Mersey flounder is higher than Liverpool Bay and most other industrialised estuaries around the British coastline by 2-4x (Franklin, 1987). The database for flatfish considered in totality is indicative of a significant historic or contemporary source of PCB contamination in the Liverpool Bay area, perhaps of estuarine origin or from waste disposal offshore. Clear differences in XPCB for any one site do exist between species in some cases so it is not just spatial variation that explains wide ranging PCB data across the study area. Probably, there is also influence from differential exposure relating to the feeding habits and strategies of different species, and also their migratory movements relating to breeding patterns. Assessment of the invertebrate data for the Mersey Estuary is hindered by the lack of published data for the same species in other EC waters. The closest comparison as regards mussels and coastal Britain, is based on MAFF data for 1990 from inshore waters of the eastern coastline from Berwick to Whitstable (Franklin & Jones, 1993). CPCB(A1254) reported by MAFF ranged from < 1 pg kg-’ to 95 l.tg kg-‘, with the higher values, > 30 ug kg-‘, restricted mainly to the south-east coast. Most values were less than one-half of the calculated peak YZPCB(Al254) for the Mersey Estuary of 60 ug kg-‘, in mussels from New Ferry. Comparisons of CPCB in Liverpool Bay hermit crabs with results for the same area published by Knickmeyer & Steinhart (1990) are hindered by differences in data expression and the congener composition of the standard mixtures. However, it is noticeable that the low clearance rates for the higher chlorinated biphenyls apparent in this earlier sampling were also evident in the present project. Common features include the significant co-dominance in hermit crabs of congeners No. 153 and No. 138 and their contribution of a minimum 35% towards the expression CPCB, however derived. This same co-dominance has been described in starfish from the German Bight (Knickmeyer et al., 1992). Consistency in the contribution of individual congeners to CPCB is thought to be independent of the lipid content of tissue, the value for L’PCB, and the location (Knickmeyer & Steinhart, 1990; Knickmeyer et al., 1992).
CONCLUSION Contamination by PCBs is of importance to the Mersey Estuary and the wider area of Liverpool Bay. Using a CPCB(ICES 6) to XPCB(A1254) conversion factor of 3.0, most flatfish from the inner Mersey Estuary gave mean CPCB(A1254) values that classify into the ‘upper’ or most polluted category of the OSPARCOM Joint Monitoring Programme (JMP) guidelines, the threshold for which is 50 pg kg-‘. Excluding dab, which consistently has the highest XPCB values in MAFF surveys of the Irish Sea, the range of ZPCB(A1254) in flatfish from the Mersey Estuary is 1.2-4x higher at its mid-point than MAFF data for Liverpool Bay in the last twenty years. An alternative benchmark for the Mersey Estuary is other industrialised estuaries and coastal waters under the influence of pollution flumes originating in estuarine discharges. In this respect flatfish from the Seine
R. T. Leah et al.
(France) and Elbe (Germany) estuaries, as well as the eastern English Channel, are higher than in the Mersey Estuary by factors of up to 23x (Seine and Elbe) and 4x (English Channel; Cossa et al., 1992; Kohler et al., 1986). Fish from the Mersey Estuary and Liverpool Bay showed a high degree of inter-correlation between the concentrations of the more chlorinated PCB congeners, specifically No. 153, No. 138 and No. 180. Congeners No. 138 and No. 153 were dominant in most Mersey Estuary fish, with No. 180 and No. 101 significant minority contributors to the expression CPCB(ICES 6). The lighter, less chlorinated, congeners No. 28 and No. 52 were present only in trace amounts. Congener ratios were broadly similar for the flatfish as a group, and appear not to vary between the inner estuary sites and those towards Liverpool Bay. However, there were notable differences in the congener composition of the roundfish and flatfish. Since this is known to reflect toxicity hazards and the biological significance of the PCB pool, this pattern very much supports the need for assessments and comparisons at the individual congener level in environmental analysis. The present database supports conclusions by MAFF concerning the absence of significant risk from the consumption of fish from the Irish Sea, despite it being recognised as the area of highest organochlorine contamination along the entire British coastline (MAFF, 1989). However, there is never room for complacency in the setting of standards that safeguard ecological and human health, and the Mersey Estuary sites used in this study were much further into the estuary than the fishing limits that equate with the statutory survey and analysis obligations of MAFF. Between 1988 and 1993, MAFF issued advisory notices to anglers concerning the risk to health of consuming fish caught in estuaries affected by industrial activity. ZPCB values for some Mersey Estuary fish exceed not only the non-statutory JMP guidelines, which refer to descriptive classifications rather than health risk, but also most of the national proscriptive limits set by the Environmental Protection Agency for the USA (US EPA, 1989). Reservations therefore exist as regards the acceptability of the CPCB levels in some fish from the inner and outer estuary, though there are no statutory limits for PCBs in fish or fish products within the European Union.
ACKNOWLEDGEMENTS This research was sponsored by a partnership entitled the Mersey Estuary and Liverpool Bay Research Consortium comprising ICI plc, National Rivers Authority, North West Water plc and Elf Atochem, in association with the Industrial Ecology Research Centre of the University of Liverpool. Special mention must be made of the time and effort of Sally Collings, Sandy McNeish, Sarah Robinson, Susanne Saunders, Julie Kent and Kathryn Porter.
REFERENCES Allchin, C. R., Kelly, C. A. & Portmann, J. E. (1989). Methods of analysis for chlorinated hydrocarbons in marine and other samples. Aquatic Environmental Protection: Analytical Methods Report No.6 MAFF, Lowestoft, UK. 25 pp.
PCBs in jish and shelljish
Ballschmiter, K. & Zell, M. (1980). Analysis of polychlorinated biphenyls (PCB) by glass capillary gas chromatography. Composition of technical Aroclor and Chlophen mixtures. Fresenius Z. Anal. Chem., 302, 20-31.
Barnthouse, L. W., Suter, G. W. & Rosen, A. E. (1990). Risks of toxic contaminants to exploited fish populations: influence of life history, data uncertainty snd exploitation intensity. Environ. Toxicol. Chem., 93,291-3
Becker, P. H., Schuhmann, S. & Koepff, C. (1993). Hatching failure in common terns (Sterna hirundo) in relation to environmental chemicals. Environ. Pollut., 79, 207-213. Cossa, D., Auger, D., Averty, B., Lucon, M., Masselin, P. & Noel, J. (1992). Flounder (Platichthys jfesus ) muscle as an indicator of metal and organochlorine contamination of French Atlantic coastal waters. Ambio, 212, 176182. US EPA (1989). Assessing Human Health Risks from Chemically Contaminated Fish and Shellfish: a Guidance Manual. US EPA Rept., EPA-503/S-89-002, 136 pp. Franklin, A. (1987). The Concentration of Metals, Organochlorine Pesticide and PCB Residues in Marine Fish and Shellfish: Results from MAFF Fish and Shellfish Monitoring Programmes, 1977-84. Aquatic Environmental Monitoring Report No. 16. MAFF, Lowestoft, UK. 38 pp. Franklin, A. (1990). Monitoring and Surveillance of Non-Radioactive Contaminants in the Aquatic Environment, 19841987. Aquatic Environmental Monitoring Report No. 22. MAFF, Lowestoft, UK. Franklin, A. (1991). Monitoring and Surveillance of Non-Radioactive Contaminants in the Aquatic Environment and Activities Regulating the Disposal of Wastes at Sea, 1988-89. Aquatic Environmental Monitoring Report No. 26. MAFF, Lowestoft, UK. 90 pp. Franklin, A. & Jones, J. (1993). Monitoring and Surveillance of Non-Radioactive Contaminants in the Aquatic Environment and Activities Regulating the Disposal of Wastes at Sea, 1991. Aquatic Environmental Monitoring Report No. 36. MAFF, Lowestoft, UK. 78 pp. Harding, G. C. & Addison, R. F. (1986). Accumulation and effects of PCBs in marine invertebrates and vertebrates. In Wade, J. S. (ed.) PCBs in the Environment, Vol 2. CRC Press, Fl, USA. Harwood, J. (1990). A short history of the North Sea seal epidemic. Ocean Challenge, 1,4-8. ICES (1992). International Council for the Exploration of the Seas. A Compilation of Standards and Guidance Values for Contaminants in Fish, Crustaceans and Molluscs for the Assessment of Possible Hazards to Human Health. Document JMG 17/3/10E. Johnston, P. A., Stringer, R. L. & French, M. C. (1991). Pollution of UK estuaries: historical and current problems. Sci. Total Environ., 106,55-70. Knickmeyer, R., Landgraff, 0. & Steinhart, H. (1992). Cyclic organochlorines in the sea star (Asterias rubens) from the German Bight, December 1988 to May 1989. Mar. Environ. Res., 33, 127-143.
Knickmeyer, R. & Steinhart, H. (1990). Concentrations of organochlorine compounds in the hermit crab, Pagarus bernhardus from the German Bight, December 1988 to May 1989. Neth. J. Sea Res., 253, 365-376.
Kohler, A., Harms, U. & Luckas, B. (1986). Accumulation of organochlorines and mercury in flounder - an approach to pollution assessments. Helgo. Meeres, 40, 431440. Kubiak, T. J., Harris, H. J., Smith, L. M., Schwartz, T. R. & Stalling, D. L. (1989). Microcontaminants and reproductive impairment of the Foster’s tern on Green Bay, Lake Michigan, 1983. Arch. Environ. Contam. Toxicol., 18, 706-727. Lang, V. (1992). Review: polychlorinated biphenyls in the environment. J. Chromatog., 595, 143. Larsen, B. & Riego, J. (1990). Interference from 2,3,5,6-3’,4’-hexachlorobiphenyl (CB163) in the determination of 2,3,4-2’,4’,5’-hexachlorobiphenyl (CB138) in environmental and technical samples. Int. J. Environ. Anal. Chem., 40, 59-68. Leah, R. T. & Rogers, H. (1991). Chlorinated Organic Pollutants in Fish from the North-east Irish Sea. NRA Rept. NR-2788. Water Research Centre. MAFF (1989). Report of the Working Party on Pesticide Residues: 1985-l 988. Surveillance Paper No. 25. London, Ministry of Agriculture, Fisheries and Food. 71 pp. McFarland, V. A. & Clarke, J. U. (1989). Environmental occurrence, abundance and potential toxicity of polychlorinated biphhenyl congeners: considerations for a congener-specific analysis. Environ. Health Perspec., 81, 225-239.
R. T. Lenh et al.
Morris, R. J., Law, R. J., Allchin, C. R., Kelly, C. A. & Fileman, C. F. (1989). Metals and organochlorines in dolphins and porpoises of Cardigan Bay, west Wales. Mar. Pollut. Bull, 20, 512-523.
Neilsen, A. H. (1994). Chemicals in the Aquatic Environment: Distribution, Persistence and Toxicity. Lewis Publishers, Boca Raton, Fl, USA, 438 pp. Reijnders, P. J. H. (1986). Reproductive failure in common seals feeding on fish from polluted coastal waters. Nature (Lond.), 324, 456457. Schultz, D. E., Petrick, G. & Duinker, J. C. (1989). Complete characterization of polychlorinated biphenyl congeners in commercial Aroclor and Chlophen mixtures by multi-dimensional gas chromatography-electron capture detection. Environ. Sci. Technol., 23, 852-859. von Westernhagen, H., Rosenthal, V., Dethlefsen, V., Ernst, W., Harms, U. & Hansen, P. D. (1981). Bioaccumulating substances and reproductive success in Baltic flounders, Platichthys jlesus. Aquat. Toxicol., 1, 85-99. WHO (1979). DDT and its Derivatives. Environmental Health Criteria 9. World Health Organisation, Geneva, Switzerland, 194 pp.