Chlorobiphenyl contaminants at Pladda and Garroch Head in the

Firth of Clyde in relation to sewage sludge input

L. Webster,* D. E. Wells and L. A. Campbell

FRS Marine Laboratory, Aberdeen, PO Box 101, Aberdeen, UK AB11 9DB

Received 29th August 2000, Accepted 26th October 2000First published as an Advance Article on the web 13th November 2000

Sewage sludge dumped at Garroch Head in the Firth of Clyde contains signi®cant quantities of chlorinated

hydrocarbons, such as polychlorinated biphenyls (CBs). These compounds are lipophilic and resistant to

degradation. They accumulate in the biota either from the water column or through the food chain,

particularly in tissue with a high lipid content. Bottom dwelling ®sh, such as plaice, in the vicinity of the dump

site will accumulate CBs from their environment. Eighteen of the 209 CBs were measured in plaice livers from

the Garroch Head dump site and from Pladda, a site reasonably remote from the dump site but also in the

Clyde, over a 7 year period prior to the cessation of dumping in December 1998. Concentrations of the

congeners in the liver of ®sh caught at the dump site were, in general, higher than those in the liver of ®sh

caught at Pladda. Concentrations in the plaice livers for the sum of 18 CBs ranged from 1611 to 8471 mg kg21

lipid for Garroch Head samples and from 336.9 to 2635 mg kg21 lipid for samples from Pladda. The data were

evaluated using principal component analysis (PCA). Pattern analysis was undertaken by normalising to the

recalcitrant CB 153. Livers from the dump site were found to have a higher proportion of the lower chlorinated

CBs. CB patterns were similar at the Garroch Head dump site from year to year, but multivariate techniques

showed that there were differences in pattern when normalised to CB 153.

Introduction

The Firth of Clyde, on the west coast of Scotland, is a partiallyenclosed sea loch with a substantial input of industrial anddomestic waste. Ef¯uent and accidental discharges fromengineering works, military base operations and textile andpaper industries have resulted in environmental contaminationby polychlorinated biphenyls (CBs). However, one of the mostconcentrated inputs of contaminants into the area is a result ofsewage sludge and dredge spoil dumping. Sewage sludge hasbeen dumped at Garroch Head, south of the Isle of Bute, since1904 (Fig. 1). In 1974, with the introduction of the Dumping atSea Act, the designated disposal area was moved 4 km south,with dumping con®ned to a circular area of 1 mile. More than1 500 000 tonnes of sewage sludge were disposed of annually atGarroch Head up to 1998. The water in this area is, on average,80 m deep and currents are relatively slow at 0.5 m s21.However, Kelly et al.1,2 demonstrated that sewage particulateswere widespread over the entire Firth. Elevated CB concentra-tions have been found in most of the Firth, with 20±25% of theFirth being classed as contaminated (w20 mg g21 Aroclorequivalents). An area of 13 km2 around the dump site wasclassed as heavily contaminated (w100 mg g21 Aroclor equiva-lents). These concentration guidelines were proposed in 1993by the UK Ministry of Agriculture, Fisheries and Food(MAFF) for categorising CBs in sediment.3 The chemical andbiological effects of sludge disposal have been investigated; oneof the effects of dumping has been to cause organic enrichmentof the sediment, with a maximum organic carbon content of11.9% at the centre of the dump site, resulting in changes in thebenthic infauna within a 2 km radius of the centre of the dumpsite.4±6 The disposal of sewage sludge at sea ceased on 31December 1998, to meet the UK's obligations under the ECUrban Waste Water Treatment Directive 91/271/EEC. Thedisposal of dredged material at sea, however, is still authorisedunder the Food and Environment Protection Act 1985 (FEPA).The main disposal point within the Clyde estuary is at ClochPoint. The disused dredge spoil site at Hunterston is also a

source of CB contamination with the sediment still beingmoderately contaminated.

Fig. 1 Map showing the dump sites at Garroch Head, pre- and post-1974, and the Pladda site in the Firth of Clyde. Plaice were collectedfrom the post-1974 dump site at Garroch Head and from Pladdabetween 1992 and 1998, and the livers analysed for CBs.

DOI: 10.1039/b007010f J. Environ. Monit., 2000, 2, 591±596 591

This journal is # The Royal Society of Chemistry 2001

Publ

ishe

d on

13

Nov

embe

r 20

00. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

23/1

0/20

14 2

1:13

:37.

View Article Online / Journal Homepage / Table of Contents for this issue

In addition to the organic enrichment of the sediment,elevated levels of heavy metals and other organic compoundshave also been measured at the Garroch Head dump site.1,2,5±8

The sludge contained organic contaminants, such as CBs, andpesticides, such as dichlorodiphenyltrichloroethane (DDT) anddieldrin. Commercial formulations of CBs, such as Aroclors,have been widely used in transformers, capacitors and anti-fouling paints. Since 1970, they have only been authorised foruse in closed systems; however, they can still enter theenvironment following the destruction and disposal ofindustrial plants and equipment. Adverse environmentalimpacts of CBs include reproductive and immune dysfunction,vitamin A de®ciency, immunosuppression and weight loss.9±12

The major source of CBs comes from seepage of ground waterinto the sewage disposal system. Since CBs are lipophilic, theycan accumulate in the organic-rich sediment around the dumpsite. High CB levels, ranging from 1 to 185 mg g21 of Aroclorequivalents, were found in Glasgow sewage sludge in the late1960s as a result of a transmission ¯uid spill.13 Elevated levelsof CBs were also found in the organically enriched sedimentcollected in 1989 from Garroch Head, with a maximumconcentration of 203 mg kg21 of Aroclor equivalents.1 CBconcentrations in sediments collected in 1997 from GarrochHead showed a decrease relative to concentrations found in thisarea in 1989, with the highest concentration found in theGarroch Head area being 54.8 mg kg21(sum of 21 congeners),re¯ecting a reduction in the disposal of CBs.14

CBs are persistent and lipophilic, and accumulate in ®shtissues, especially those with a high lipid content such as theliver. The congeners transfer directly from the water column tothe biota via the gills or body surfaces into the circulatory ¯uids(bioconcentration) or can be transferred along the food chain(biomagni®cation).8,15±17 Bioaccumulation is the sum of theseprocesses and is of concern due to the biological impact of CBson marine life and the possible risk to consumers. The uptakeof CBs via food is likely to absorb all congeners similarly, andhence the pattern will be determined by the diet. Sedentary,bottom dwelling ®sh, such as plaice, should accumulate highercontaminant levels compared to pelagic species since their preyand habitat may be more highly contaminated. The concentra-tions of CBs in plaice should also correlate with the octanol/water partition coef®cient (log Kow). Fish, in general, do notmetabolise organochlorines extensively. CBs are metabolisedby the cytochrome P450 group of enzymes and the extent ofthis is affected by the age, sex and species. Most of the lesschlorinated CB congeners (tri- and tetrachloro) and some ofthe more highly chlorinated CBs can be metabolised by ®sh.18±21

This study reports the concentrations of 18 CBs in livers ofplaice caught around the Garroch Head dump site in the Firthof Clyde over a 7 year period. Kelly et al.1,2 have shown that thecontents of the dump site at Garroch Head could have awidespread effect on ®sh; therefore, concentrations at GarrochHead were compared with those in samples from Pladda, alsoin the Clyde, south of Arran.

Methods

Sample collection

Fish were taken by the research vessel FRV Clupea or Aorafrom the sludge disposal area at Garroch Head and fromPladda, south of Arran. Sample collection took place inOctober/November each year to correspond with the pre-spawning period. Samples were dissected on board ship andlivers and muscle ®llets removed, wrapped in aluminium foiland frozen. From 1992 to 1995, male and female samples wereanalysed individually. From 1996 to 1998 samples were pooled,either at sea or on arrival at the laboratory. Five males werepooled to make one sample.

Isolation of chlorobiphenyls

Frozen liver samples (0.5±3.0 g) were defrosted and groundwith anhydrous sodium sulfate (y106weight of sample). Thedried tissue was placed in a cellulose thimble, previouslycleaned for 4 h with tert-butyl methyl ether (MTBE), andplaced into a Soxhlet apparatus. A recovery standard, CB 209(Promochem, Welwyn Garden City, Herts, UK), was addedprior to extraction. The plaice livers were extracted with 180 mlof MTBE for 12±16 h. An aliquot of the extract was taken forthe determination of extractable lipid content. A furtheraliquot containing 50±200 mg lipid was removed and trans-ferred into hexane before passing through alumina and silicacolumns.22 The internal standards (2,4-dichlorobenzyl alkylhexyl ether with C6 and C16 alkyl chains)22 were added to bothextracts before concentrating using a turbovap system.

Determination of CBs by gas chromatography with electroncapture detection (GC-ECD)

The concentration and composition of 18 CB congeners (CB31, 28, 52, 44, 70, 101, 149, 118, 114, 153, 105, 138, 158, 128,156, 190, 170, 194) were determined by GC-ECD using either aVarian 3500 GC (Varian, Walton-on-Thames, Surrey, UK) ora PE GC autosystem (Perkin-Elmer, Beacons®eld, UK) ®ttedwith a cool on-column injector. A medium polarity column wasused for the analyses (HP5 column 60 m60.25 mm, 0.25 mm®lm thickness: Hewlett-Packard, Stockport, UK) along with anuncoated pre-column (2.5 m60.53 mm id). The carrier gas washydrogen (1±3 ml min21) and the make-up gas nitrogen(30¡5 ml min21). The initial oven temperature was 80 ³Cwhich was held for 1 min. The temperature was raised at15 ³C min21 up to 180 ³C and held at this temperature for12 min. Thereafter, the temperature was raised at 3 ³C min21

up to a ®nal temperature of 290 ³C and held at this temperaturefor 20 min. The chromatograph was calibrated using a series ofexternal standards and the two 2,4-dichlorobenzyl alkyl ethers.From 1992 to 1995, the data were quanti®ed using a VGMinichrom data system (VG Data Systems, Chester, UK).From 1996 to 1998, a Client Server Turbochrom data system(Perkin-Elmer) was used.

Data analysis

As CBs are known to be lipophilic, the concentrations werenormalised to lipid to give CB concentrations in mg kg21 lipid.CB patterns and temporal trends were investigated bynormalising to the recalcitrant CB 153 and evaluating thedata using principal component analysis (PCA). CB 153 wasselected as it was present at the highest concentrations of all theCBs in all plaice liver samples analysed.

Quality control

Procedural blanks and laboratory reference materials (LRM)were analysed with each batch of samples. The LRM resultswere monitored using Shewhart charts. Further quality controlwas assured through successful analysis of CBs in biologicaltissue in the QUASIMEME (Quality Assurance of Informationfor Marine Environmental Monitoring in Europe) Laboratoryperformance studies. The limit of detection was less than0.03 mg kg21 for all of the 18 CBs, with values of 0.024 mg kg21

for CB 31 and 0.015 mg kg21 for CB 194.

Results

The ®sh length range for all years was similar and is shown inTable 1, along with the percentage lipid of the plaice liver. The®sh length range was similar from year to year. However, therewas a wide range of lipid contents for the plaice livers. Ingeneral, the lipid contents for the livers of plaice from Pladda

592 J. Environ. Monit., 2000, 2, 591±596

Publ

ishe

d on

13

Nov

embe

r 20

00. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

23/1

0/20

14 2

1:13

:37.

View Article Online

were lower than those for the livers of plaice from the dump siteat Garroch Head. However, analysis of variance demonstratedthat only the percentage lipid for Pladda livers in 1992 wassigni®cantly different from that of the Garroch Head livers(pv0.05). In 1992 and 1993, both male and female plaice liverswere analysed. In 1994 and 1995, only female samples wereanalysed from Garroch Head, due to dif®culties in obtainingmales at this site, although the control samples were a mixtureof male and female plaice livers. The percentage lipid waspositively correlated with the total CB concentration withpv0.001. From 1996 to 1998, only pools of ®ve males wereanalysed.

Eighteen CBs were investigated in plaice liver caught in thevicinity of the Garroch Head dump site and from Pladda. Themean concentration for each year is shown in Table 2 andindividual results for samples from 1992 to 1995 are showngraphically as a box and whisker plot in Fig. 2. The total CBconcentrations for plaice livers, for all 7 years, from Pladdawere signi®cantly lower than those from the dump site(pv0.001). The mean total CB concentration in livers fromGarroch Head plaice ranged from 1611 mg kg21 lipid in 1995 to8471 mg kg21 lipid in 1997, and in livers from Pladda plaice, theconcentrations ranged from 336.9 mg kg21 lipid in 1996 to2635 mg kg21 lipid in 1997 (Table 2). There was one exception,with the mean total CB concentration being higher in plaicelivers collected from Pladda than from Garroch Head in 1995,

with total CB concentrations of 2024 mg kg21 lipid and1610 mg kg21 lipid, respectively. Fig. 3 shows the differencein the concentrations of the seven ICES (International Councilfor the Exploration of the Sea) CBs between the livers of plaicefrom Garroch Head and Pladda. The seven ICES CBs (28, 52,101, 118, 153, 138, 180; ICES 7) were recommended by theEuropean Union Community Bureau of Reference; these CBswere selected as indicators due to their relatively highconcentrations in technical mixtures, their wide chlorinationrange and their persistence. In 1992, 1994, 1996 and 1997, theconcentrations of all seven ICES CBs were higher in plaice liverfrom Garroch Head. In 1993 and 1998, the concentration washigher for CB 138 in plaice livers from Pladda than in thosefrom Garroch Head. In 1995, the concentrations of CB 101,118 and 153 were highest in the Pladda plaice livers. Aroclor1254 equivalents were calculated by multiplying the sum of theICES 7 CBs (mg kg21 wet weight) by 2.5 to give an approximatetotal concentration for a direct comparison with the JointMonitoring Programme (JMP) guidelines of the Oslo and ParisCommission (OSPAR) for CBs (Table 3).1 The mean Aroclor

Table 1 Summary of ®sh length and percentage lipid range for plaicelivers from the dump site at Garroch Head (GH) and the control site atPladda (PL) in samples collected between 1992 and 1998. Samples from1996 to 1998 were pooled to give two pools of ®ve, and the percentagelipid was determined for the pooled samples

Year Site Length/cm Lipid rangea (%)

1992 GH 17.5±32.1 7.3±34.4 (20.0¡8.8)PL 18.4±28.7 4.6±18.4 (8.5¡4.1)

1993 GH 19.0±32.2 9.7±34.9 (24.0¡7.2)PL 19.2±23.7 12.3±29.8 (17.8¡6.3)

1994 GH 15.8±34.3 4.4±40.7 (19.4¡8.4)PL 23.0±29.0 9.7±24.0 (15.8¡4.4)

1995 GH 19.5±31.5 8.4±43.3 (20.5¡10.7)PL 20.5±24.3 12.4±26.5 (18.5¡4.5)

1996 GH 21.5±26.0 22.8, 24.6 (23.7)PL 18.5±29.5 12.8, 9.8 (11.3)

1997 GH 19.0±25.0 27.6, 32.8 (30.2)PL 16.5±29.5 9.8, 13.1 (11.4)

1998 GH 18.5±25.0 24.9, 22.0 (23.4)PL 17.0±28.5 5.1, 13.8 (9.4)

aFigures in parentheses represent the mean¡s.

Table 2 Mean CB concentrations, sum of 7 ICES congeners (CB 28, 52, 101, 118, 153, 138, 180) and sum of all 18 congeners studied (mg kg21 lipid)in plaice livers collected between 1992 and 1998 from Garroch Head (GH) and Pladda (PL)

Year CB31

CB28

CB52

CB44

CB70

CB101

CB149

CB118

CB114

CB153

CB105

CB138

CB158

CB128

CB156

CB180

CB170

CB194

SumICES 7

Sumall CBs

GHÐ1992 42.0 75.5 223.5 73.0 138.5 246.0 152.0 175.0 3.5 377.5 63.5 251.0 21.5 30.0 17.5 195.0 70.0 68.0 1544 22231993 34.2 58.3 227.1 70.8 109.2 203.8 109.6 140.8 0.4 338.8 49.6 203.8 13.8 24.2 9.6 142.9 55.4 54.2 1315 18471994 57.2 97.4 341.8 427.3 184.0 319.6 191.2 214.9 1.0 443.3 63.9 276.3 22.7 26.8 17.5 196.9 72.2 82.5 1890 30371995 25.9 48.3 128.8 30.2 84.9 157.1 108.3 123.4 9.3 320.0 39.5 199.5 17.1 20 16.6 177.1 53.7 51.2 1154 16111996 42.3 89.3 334.6 91.0 177.8 350.0 173.5 182.9 6.8 429.9 68.8 238.5 24.8 32.1 19.7 188.9 71.8 39.7 1814 2562

1997 242.4 521.9 1181.5 275.5 666.6 981.8 647.4 614.2 23.8 1336.1 243.4 665.6 64.9 79.5 57.3 550.3 227.2 91.1 5851 84711998 28.2 104.3 254.3 73.9 156.8 213.2 166.2 165.4 4.3 534.6 61.5 189.3 23.5 32.1 21.8 247.0 110.3 38.9 1708 2426

PLÐ1992 4.7 9.4 102.4 31.8 54.1 130.6 91.8 107.1 ND 288.2 35.3 224.7 3.5 21.2 12.9 147.1 54.1 81.2 1009 14001993 1.1 7.9 96.6 34.8 47.2 137.1 100.0 119.1 ND 344.4 37.6 242.1 10.1 27.5 10.1 144.4 60.7 78.1 1092 14991994 10.1 10.1 76.6 14.6 39.9 117.1 85.4 98.7 ND 355.1 21.5 228.5 5.7 14.6 6.3 160.1 61.4 96.8 1046 14031995 23.3 29.8 151.0 49.6 85.8 210.1 149.5 156.0 3.5 408.7 51.5 282.6 24.3 34.4 19.4 201.4 73.0 70.0 1440 20241996 3.5 5.3 31.0 9.7 17.7 37.2 26.5 29.2 ND 65.5 10.6 56.6 4.4 8.8 3.5 15.0 7.1 5.3 239.8 336.91997 41.2 47.4 354.4 107.0 138.6 327.2 91.2 244.7 7.0 528.9 86.0 388.6 21.1 56.1 12.3 101.8 52.6 28.9 1993 26351998 31.6 37.9 221.1 69.5 56.8 215.8 101.1 72.6 3.2 431.6 58.9 293.7 21.1 41.1 10.5 161.1 56.8 43.2 1434 1928

Fig. 2 Box and whisker plot, produced using Systat version 6.0,showing the total CB concentration (mg kg21) for plaice livers collectedbetween 1992 and 1995. The central horizontal line inside the boxmarks the median and the edges of the box (hinges) mark the ®rst andthird quartiles. The plot also shows outliers. An asterisk indicates anoutside value, a point that falls more than 1.5 times the hspread (thedistance between the hinges) outside the closest hinge. The graph showsthat, in all years except 1995, the total CB concentrations were higher inplaice livers from Garroch Head.

J. Environ. Monit., 2000, 2, 591±596 593

Publ

ishe

d on

13

Nov

embe

r 20

00. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

23/1

0/20

14 2

1:13

:37.

View Article Online

equivalents ranged from 591.5 mg kg21 wet weight in 1995 to4418 mg kg21 wet weight in 1997 for Garroch Head plaicelivers. The mean Aroclor equivalents in plaice livers fromPladda ranged from 67.8 mg kg21 wet weight in 1996 to665.7 mg kg21 wet weight in 1995. CB levels were previouslyreported in plaice livers from the Clyde and ranged from 61.9 to1172 mg kg21 wet weight of Aroclor equivalents in samplescollected in 1988±1991, showing little indication of any declinein the CB concentrations in plaice.7 By contrast, CBconcentrations of 128.9, 1.7, 52.9 and 7.3 mg kg21 wet weightof Aroclor equivalents were found in plaice liver from MorayFirth, Shetland, Minches and Atlantic, respectively.7 Usingdiscriminant analysis, McKenzie21 found CB concentrations tobe species speci®c, separating the pelagic species (mackerel,hake and whiting) from the two benthic species (monk®sh andplaice). McKenzie21 found total CB concentrations in ®sh(whiting, cod, hake, mackerel, monk®sh and plaice) caught atthe Garroch Head site in November 1995 ranging from 71.1 to1160 mg kg21 lipid in liver, with the demersal species (plaice andmonk®sh) having the highest CB concentrations. Total CB (22congeners) concentrations in plaice liver ranged from 1043 to1160 mg kg21 lipid.

Only two of the mean CB concentrations in plaice liverfrom Garroch Head fell within the upper contaminationlevel category (w1.0 mg kg21wet weight Aroclor equiva-lents) set by the JMP for CBs in ¯ounder liver. The meanconcentrations in 1996 and 1997 were 1061 and 4418 mg kg21

wet weight Aroclor equivalents, respectively (Table 3). Allother concentrations fell within the medium contaminationcategory (0.50±1.00 mg kg21 wet weight). All but twoPladda plaice livers were in the lowest contaminationcategory (v0.5 mg kg21 wet weight). The mean totalAroclor equivalents for the 1995 and 1997 Pladda plaicelivers were 665.7 and 568.0 mg kg21 wet weight, respectively.Ecotoxicological assessment criteria (EAC) and backgroundreference concentrations have previously been used byOSPAR to assess monitoring data and are used to identifypotential areas of concern. The estimated EAC for plaicelivers is 1100 mg kg21 lipid weight for the sum of the 7 ICESCBs. The mean values for the sum of the 7 ICES CBs forplaice livers collected between 1992 and 1998 were all abovethis value in Garroch Head samples (Table 2), indicatingthat there is a potential for environmental effects in thisarea. For the Pladda plaice livers, only plaice livers from1995, 1997 and 1998 were above this value (1440, 1993,1434 mg kg21 lipid weight; Table 2).

In all cases, at both Pladda and Garroch Head, CB 153had the highest concentration of any congener. CB 153occurs in all technical formulations and tends to lie in themiddle of the range of physical parameters, such as the logKow. The log Kow value is directly related to the watersolubility of a compound and gives an indication of thelipophilicity. CB 153 is not metabolised by marine mammalsand has not been found to be metabolised by plaice.Cytochrome P450 1A iso-enzyme metabolises CBs with avicinal hydrogen atom in the ortho and meta positions of atleast one ring in combination with one ortho chlorine. CB 28and CB 118 ®t these criteria, and also have the lowestconcentration in nearly all plaice livers analysed from thedump site and at Pladda. In addition, CB 28 is the leastlipophilic, and therefore the least likely to bioaccumulate inplaice livers from both sites. The hexa- and hepta-CBs madethe largest contribution to the total CB concentration,ranging from 42.3% to 53.3% in Garroch Head samples andfrom 42.9% to 61.0% in samples from Pladda. However, adifference was observed between control and dump sites.The proportion of tri- and tetra-CBs was less in the Pladdasamples (10.1±25.2%) than in samples from Garroch Head(18.6±35.3%). Hope et al.23 have shown a high percentage ofthe hexa- and hepta-CBs in ®sh, with 67±80% of the total CBconcentration being accounted for by these compounds. Ingeneral, the proportion of higher chlorinated CBs (¢5chlorines) increases through the food web as they arerelatively less volatile, more lipophilic and more resistant tometabolic and microbial degradation.23 The congeners witha greater af®nity for particles (higher log Kow) will betransported to locations remote from the input sourcemainly as suspended material. The lower chlorinated, morewater-soluble CBs are less likely to be associated withsediment and biota remote from the input source. Thisindicates that, at Pladda, the source of CBs was mainly fromGarroch Head, and they were not introduced by atmo-spheric deposition, which would result in a higher propor-tion of the lower chlorinated, more volatile CBs. Bycomparison, Garroch Head showed a higher proportion ofthe lower chlorinated CBs, as these are being continuallyreplenished along with the sewage at this site. A similarchange in pattern was found in sediments by Kelly andCampbell1 in the vicinity of the dump site. A lowerproportion of the less chlorinated CBs was found withincreasing distance from the source. More recently, Hess8

found a higher proportion of tri- and tetra-CBs in sedimentsat Garroch Head in comparison with other areas in theClyde. Both plaice livers and sediment from Pladda showeddifferent CB patterns from the Garroch Head samples,having a higher proportion of more highly chlorinated CBs.

Fig. 3 Graph showing the difference in concentrations betweenGarroch Head plaice livers and Pladda plaice livers (mg kg21 lipid)for each of the seven ICES CBs. In most years, the CB concentrationswere higher at Garroch Head.

Table 3 Mean and standard deviation of the sum of the concentrationsof the 7 ICES CBs (CB 28, 52, 101, 118, 138, 153, 180) in mg kg21 wetweight and the Aroclor equivalents (2.56ICES 7) for plaice livers fromboth Pladda and Garroch Head between 1992 and 1995. Samples from1996 to 1998 were pooled to give two pools of ®ve. The CBconcentrations for each of the pooled samples is shown, along withthe Aroclor equivalents of the mean ICES concentration

Year

7 ICES CBs/mg kg21 wet weight Aroclor equivalents

Garroch Head Pladda Garroch Head Pladda

1992 308.7¡184.0 85.8¡38.6 771.8 214.51993 315.7¡108.5 194.3¡102.5 789.2 485.81994 366.7¡161.4 165.3¡61.4 916.8 413.21995 236.6¡122.9 266.3¡136.0 591.5 665.71996 467.2, 381.8 29.5, 24.6 1061 67.8

(Mean~424.5) (Mean~27.1)1997 1725, 1808 250.1, 204.2 4418 568.0

(Mean~1767) (Mean~227.2)1998 590.3, 209.1 138.6, 133.8 999.2 340.5

(Mean~399.7) (Mean~136.2)

594 J. Environ. Monit., 2000, 2, 591±596

Publ

ishe

d on

13

Nov

embe

r 20

00. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

23/1

0/20

14 2

1:13

:37.

View Article Online

Statistical analysis

PCA was used to analyse the CB data. PCA is a data reductiontechnique that allows the comparison of a large number ofvariables, by forming a new number of variables from linearcombinations of the original variables. Pattern analysis of CBsis often carried out by normalising to CB 153.8,19,21±25 Thisapproach was adopted for this investigation, and the mean CBconcentrations for each year were normalised to CB 153 forboth the Garroch Head and Pladda sites to remove the varianceand to produce relative contaminant patterns. The factorloading plot for normalised CB data in plaice livers fromPladda and Garroch Head is shown in Fig. 4(a). This gives agraphical representation of the extent to which each factoraccounts for the variance in the data and shows the relationshipbetween the different CB congeners. The outer circle of thefactor loading plot represents the total variance of the data set.Determinants the same distance from the centre and withsimilar positions are positively correlated. Those in theopposite direction are negatively correlated. The factor loadingplot showed three main groups: (1) CB 28, 31, 52, 44, 70, 101,118, 149; (2) CB 105, 138,158, 114, 128,156; (3) CB 170, 180,194.

The principal components that accounted for the greatestvariance were plotted against each other [Fig. 4(b)] for both theGarroch Head and Pladda sites. The ®rst and secondcomponents accounted for 40% and 17% of the variance,respectively. The two circles in Fig. 4(b) represent the 95%con®dence limits of the two sites. Pladda, in general, scorednegatively on Factor 1, which is essentially a contrast betweenhigh chlorinated, hepta- and octa-CBs and lower chlorinatedCBs. The data in Fig. 4(b) indicate that there are higher levelsof lower chlorinated CBs at the Garroch Head dump site. The

higher chlorinated congeners are more persistent in theenvironment. They will therefore have higher relative concen-trations at sites remote from the source, whereas at the GarrochHead dump site there will be a regular input of CBs into theenvironment.

PCA was also applied to the data obtained from the GarrochHead dump site samples to investigate any chronologicaldifference resulting from dumping. The factor loading plots forthese data for each year from 1992 to 1998 are shown inFig. 5(a). There are again three main groups: (1) CB 31, 52, 28,70, 101, 118, 149, 105, 44; (2) CB 138, 128, 158, 114; (3) CB 156,180, 170, 194. The ®rst and second components accounted for40% and 21% of the variance, respectively, and were plottedagainst each other in Fig. 5(b). There was little evidence ofclustering, indicating that the CB patterns changed littlebetween years. However, the 1995±1998 samples were mainlysituated in the top right-hand corner of the biplot and samplesfrom 1992±1994 in the lower part of the plot. The 1992±1994group is associated with higher levels of CBs from group (3),mainly the more highly chlorinated, hexa-, hepta- and octa-CBs (CB 156, 180, 170 and 194), and the 1995±1998 group isassociated with the lower chlorinated CBs. This suggests thatthere has been a gradual change in the CB content of thesewage dumped at Garroch Head over the 7 year periodinvestigated.

Conclusions

Levels of CBs in plaice livers at the Garroch Head dump sitewere found to be higher than those in ®sh livers from otherScottish waters. Concentrations in plaice have not decreasedover the time scale investigated, although previous work hasshown that the CB concentrations in sediment have decreased

Fig. 4 (a) Factor loading plots for CBs normalised to CB 153 for plaicelivers from the Garroch Head dump site and Pladda from 1992 to 1998.(b) Principal component analysis of plaice livers from Pladda andGarroch Head using the CB concentrations normalised to CB 153. Thecircles represent the 95% con®dence limits. A separation is observedbetween the two sites indicating that there is a difference in CB pattern.

Fig. 5 (a) Factor loading plots for CBs normalised to CB 153 for plaicelivers from the Garroch Head dump site between 1992 and 1998. (b)Principal component analysis for the Garroch Head dump site usingthe CB concentrations normalised to CB 153 for plaice livers from 1992to 1998.

J. Environ. Monit., 2000, 2, 591±596 595

Publ

ishe

d on

13

Nov

embe

r 20

00. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

23/1

0/20

14 2

1:13

:37.

View Article Online

at Garroch Head. Unusually high CB levels were found in1997, with the samples being classed as highly contaminatedusing JMP guidelines or EAC. The total CB concentrationswere higher at the Garroch Head dump site than at Pladda inall years except 1995. The ®sh at the dump site showed a higherproportion of the lower chlorinated tri- and tetra-CBs incomparison with ®sh from the Pladda site. PCA also showedthat there were differences in CB patterns, when normalised toCB 153, between Garroch Head and Pladda ®sh. The lowermolecular weight and more water-soluble CBs may desorbfrom the sediment more readily and be released into the watercolumn. The more particle reactive, more highly chlorinatedCBs will be transported far from the source, resulting in a lowerproportion of the lower chlorinated CBs in the ®sh at Pladda.

There is no evidence that the CB concentrations in plaicelivers over the 7 year period investigated have decreased. Thecessation of dumping in December 1998 should result in achange in CB pattern, with the Garroch Head site becomingmore like Pladda with a lower proportion of the lesschlorinated CBs. The decrease in total CB concentration willdepend on the dispersive nature of the dump site, but asGarroch Head is thought to be an accumulative site this maytake a number of years. It is therefore necessary to monitorchanges in CB concentrations following the cessation ofdumping.

References

1 A. G. Kelly and L. A. Campbell, Mar. Environ. Res., 1995, 41, 99.2 A. G. Kelly and P. W. Balls, Coast. Zon. Top., 1997, 3, 88.3 MAFF Aquatic Environment Monitoring Report, No. 36, MAFF

Fisheries Laboratory, Lowestoft, Suffolk, 1993.4 D. W. Mackay, Mar. Pollut. Bull., 1986, 17, 91.5 A. G. Kelly and L. A. Campbell, Mar. Pollut. Bull., 1994, 28, 103.

6 A. G. Kelly, Environ. Pollut., 1995, 88, 207.7 A. G. Kelly and L. A. Campbell, Scottish Fisheries Research

Report, The Scottish Of®ce Agriculture and Fisheries Department,Aberdeen, 1994, No. 54.

8 P. Hess, PhD Thesis, Robert Gordons University, Aberdeen, 1998.9 D. A. Clark, G. Sweeney, S. Safe, E. Hancock, D. G. Kilburn and

J. Gualdie, Immunopharmacology, 1983, 6, 143.10 B. Leece, M. A. Denomme, R. Towner and S. M. A. Li, J. Toxicol.

Environ. Health, 1985, 16, 379.11 H. A. Tilson, J. L. Jacobson and W. J. Rogan, Neurotoxicol.

Tetratol., 1990, 12, 239.12 B. Brunstrom, H. Hakansson and K. Lundberg, Pharmacol.

Toxicol., 1991, 69, 421.13 A. V. Holden, Nature, 1970, 228, 1220.14 P. J. Edgar, I. M. Davies, A. S. Hursthouse and J. E. Mathews,

Mar. Pollut. Bull., 1999, 38, 486.15 K. A. Kidd, R. H. Hesslein, B. J. Ross, K. Koczanski,

G. R. Stephens and D. C. G. Muir, Environ. Pollut., 1998, 102, 91.16 A. Ruus, K. I. Ugland, O. Espeland and J. U. Skarre, Mar.

Environ. Res., 1999, 48, 131.17 A. Goksoyr and L. Forin, Aquatic Toxicol., 1992, 22, 287.18 J. P. Boon, F. Eijgenraam, J. M. Everaarts and J. C. Duinker,

Mar. Environ. Res., 1989, 27, 159.19 D. G. Randall, D. W. Connell, R. Yang and S. S. Wu, Chemo-

sphere, 1998, 37, 1263.20 J. P. Boon, J. Van der Meer, C. Allchin, R. J. Law, J. Klungsoyr,

P. E. G. Leonards, H. Spliid, E. Stor Hansen, C. McKenzie andD. E. Wells, Arch. Environ. Contam. Toxicol., 1997, 27, 419.

21 C. McKenzie, PhD Thesis, Robert Gordons University, Aberdeen,1999.

22 D. E. Wells, A. Cowan and E. G. Christie, J. Chromatogr., 1985,328, 172.

23 B. Hope, S. Scatolini, E. Titus and J. Cotter, Mar. Pollut. Bull,1997, 34, 548.

24 C. McKenzie, E. Rogan, R. J. Reid and D. E. Wells, Environ.Pollut., 1997, 98, 15.

25 I. Echarri, C. Nerin, D. E. Wells, C. Domeno and J. Cacho,Analyst, 1998, 123, 421.

596 J. Environ. Monit., 2000, 2, 591±596

Publ

ishe

d on

13

Nov

embe

r 20

00. D

ownl

oade

d by

Nor

thea

ster

n U

nive

rsity

on

23/1

0/20

14 2

1:13

:37.

View Article Online