Concentrations and spatial variations of cyclodienes and other organochlorines in herring and perch from the Baltic Sea

The Science of the Total Environment 215 Ž1998. 69]83

Concentrations and spatial variations of cyclodienes and other organochlorines in herring and perch from the Baltic Sea B. Strandberg a,U , L. Strandberg a,c , B. van Bavel a , P.-A. Bergqvist a , D. Bromanb , J. Falandysz c , C. Naf ¨ b , O. Papakostab , C. Rolff b , C. Rappea a Institute of En¨ironmental Chemistry, Umea ˚ Uni¨ersity, S-901 87 Umea, ˚ Sweden Aquatic Chemical Ecotoxicology, Department of Zoology, Stockholm Uni¨ersity, S-106 91 Stockholm, Sweden c Department of En¨ironmental Chemistry and Ecotoxicology, Uni¨ersity of Gdansk, PL 809 52 Gdansk, Poland b

Received 29 September 1997; accepted 19 December 1997

Abstract Herring Ž Clupea harengus. and perch Ž Perca flu¨iatilis. were collected in the northern and southern Baltic Sea and analyzed for the presence of the cyclodiene pesticides chlordane ŽCHL., heptachlor, aldrin, dieldrin, endrin, isodrin, endosulfan and mirex, as well as other organochlorine contaminants, hexachlorocyclohexanes ŽHCHs., DDTs, hexachlorobenzene ŽHCBz. and polychlorinated biphenyls ŽPCBs. in order to investigate concentrations, accumulation and differences in geographical distribution. In the northern part of the Baltic Sea, Gulf of Bothnia, herring were collected at two pelagic stations, one in the Bothnian Bay ŽBB. and the other in the Bothnian Sea ŽBS., respectively; perch were collected at four coastal locations along the Swedish coast. All these locations were selected to represent background areas except one in the vicinity of an industrialised and contaminated area. Both specimens were also caught in the southern part of the Baltic Sea, in the Gulf of Gdansk ŽGG., Poland, a potentially highly polluted area. From the eight cyclodiene pesticides analyzed, three were detected in herring and perch samples, including 12 different CHL-related compounds, dieldrin and mirex. To our knowledge, it is the first time that mirex has been detected in samples from the Baltic Sea. Neither heptachlor, aldrin, endrin, isodrin nor endosulfan were found. However, HCHs, DDTs, HCBz and PCBs were found in every sample investigated, and the concentrations ranged e.g. for the cyclodiene chemicals dieldrin and CHL-related compounds from 30 to 170 ngrg lipid and for PCBs from 360 to 5400 ngrg lipid, both fish species included. Differences in contamination burden between the sites

U

Corresponding author. Tel.: q46 90 7865672; fax: q46 90 186155; e-mail: [email protected]

0048-9697r98r$19.00 Q 1998 Elsevier Science B.V. All rights reserved. PII S0048-9697Ž98.00114-4

70

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

can be seen, e.g. for herring the BB and GG locals were similar, and generally lower than BS for all chemicals except that of DDT where GG was the highest. For the perch samples the industrialised location had markedly higher concentrations of HCBz and PCBs than the other locations. This species also identifies GG as the most DDT contaminated site among the three studied areas. Q 1998 Elsevier Science B.V. Keywords: Organochlorines; Baltic Sea; Fish; Cyclodienes; Contamination

1. Introduction The pollution situation of the Baltic Sea, the largest brackish-water area in the world, has been of great concern during recent years. Generally, biota in semi-enclosed seas are at risk because of slower dilution or removal rates of contaminants compared to the open ocean ŽLoganathan and Kannan, 1994.. This emphasizes even more the situation for life in the Baltic Sea with its special physical features e.g. the sea is shallow, has low salinity, partly covered with ice during winter and a vulnerable ecosystem very poor in species abundance. The organochlorine pesticides cyclodienes Žchlordane ŽCHL., heptachlor, dieldrin, aldrin, endrin, isodrin, endosulfan and mirex., hexachlorocyclohexanes ŽHCHs ., DDTs, hexachlorobenzene ŽHCBz. and technical products like polychlorinated biphenyls ŽPCBs. are being phased out or their use is strongly restricted in many countries e.g. Europe, USA and Japan. However, use is still continuing and even increasing in developing countries in tropical regions ŽBarrie et al., 1992; Iwata et al., 1993; Loganathan and Kannan, 1994.. These chemicals are ubiquitous contaminants in the environment and due to their physical chemical properties they can volatilise from temperate and tropical climate zones, be transported in air and redeposited in colder regions ŽRappe, 1971; Oehme, 1991; Wania and Mackay, 1996. and are therefore found in such remote areas like the Arctic and Antarctic ŽKawano et al., 1984; Norstrom et al., 1988; Muir et al., 1995.. The northern latitudes of the Baltic Sea suggest it to be in the way of such transport processes of these pollutants ŽOehme, 1991., and they can also be detected in all types of environmental compartments in Scandinavia such as air

ŽOehme, 1991; Buser et al., 1992., soil, compost ŽWagman et al., 1995., sediment ŽOden and Ekst˚ edt, 1976., biota ŽJansson et al., 1993. and humans ŽLindstrom ¨ et al., 1995, 1996.. The concern about these pollutants has also increased from the ecotoxicological point of view as to which of their residues poses a toxic threat to wildlife ŽOlsson et al., 1992., and also for humans ŽAhlborg et al., 1995.. The first report on organochlorine compounds found in the Baltic Sea seems to be in the late 1960s ŽJensen, 1966.. The Baltic Sea has been reported to be heavily polluted by a range of compounds. Many reports include DDT and PCB, sometimes also HCHs and HCBz, but the first one to detect CHL was Jansson et al. Ž1979.. Among the other cyclodiene compounds, there are few reports on aldrin and dieldrin ŽJansson et al., 1979; Westernhagen et al., 1981; Wikstrom ¨ and Pyssalo, 1981; Kannan et al., 1992; Falandysz et al., 1994. and one on endrin ŽWikstrom ¨ and Pyssalo, 1981.. To our knowledge there are no reports of either endosulfan, isodrin or mirex in the Baltic Sea. Generally, in the literature only the most abundant CHL compounds are included and data of the minor ones are mostly lacking. In this paper we include a total of 14 CHL components. The aim of this paper was to study the contamination status of a long range of organochlorine contaminants including cyclodiene chemicals like CHL-related compounds in different locations in the northern part of the Baltic Sea, viz. the Gulf of Bothnia, and also compare to the potentially heavily contaminated Gulf of Gdansk. This was done by investigating the body burdens of two different fish species, herring Ž Clupea harengus. and perch Ž Perca flu¨iatilis..

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

71

2. Materials and methods 2.1. Characteristics of the sampling areas The Gulf of Bothnia is the northernmost part of the Baltic Sea, surrounded by Sweden and Finland, and it consists of the two sub-basins, the Bothnian Bay ŽBB. Žmean depth of 41 m. and the Bothnian Sea ŽBS. Žmean depth of 62 m. which are separated by a more shallow area, Kvarken ŽFig. 1.. Along the coast, several industries are situated and there is shipping activity all year round. The drainage area is relatively large Ž270 000 km2 for BB and 230 000 km2 for BS. and includes large and high mountains on the Swedish side to the Norwegian border. The region is only sparsely populated Ž; 1 million.. A number of rivers with high feeder streams run into the Gulf of Bothnia, with approximately the same water volume to each of the BS and the BB, respectively. This high input of fresh water, together with low water exchange with the North Sea results in brackish water with a gradient of salinity in the Gulf of Bothnia from north to south of ; 2]7‰. The low salinity along with other factors such as cold temperatures has created a special, but also vulnerable ecosystem with a small number of species. The Gulf of Gdansk ŽGG. is a rather shallow basin in the south Baltic Sea on the Polish coast. Three large, heavily industrialised harbour cities are situated in this area and one large river ŽWisla River. flows out into the GG and is one of the longest in the Baltic drainage area, it covers most of the agricultural land of Poland and drains to the Gulf of Gdansk. It has a drainage area inhabited by ; 30 million people including the large cities of Warsaw, Lodz, Krakow and Gdansk and Poland’s most industrialised area, viz. Silesia. There are several industries located directly at the coast of the Gulf of Gdansk and the biggest include an oil refinery, shipyards, phosphate fertilizer factories, naval and merchantile ports. All fish from the Gulf of Bothnia were captured in autumn 1991 and those from the GG in summer 1992 and the sampling sites are shown in Fig. 1. The samples of perch Ž Perca flu¨iatilis. were collected at four coastal locations in the

Fig. 1. Sampling locations in the Baltic Sea

ŽHF. Harufjarden ¨ Umea ˚ ŽUM. Hornslandet ŽHL. ŽGB. Gavlebukten ¨ Gulf of Gdansk ŽGG. F9 Bothnian Bay ŽBB. SR5 Bothnian Sea ŽBS.

Latitude

Longitude

Depth

658269900 638369900 618469460 608469100 548329300 648429500 618059000

228559800 208319500 178289300 178329790 188309000 228049000 198359000

82 m 50]55 m 56]65 m 56 m 8]10 m 125]130 m 125]130 m

Gulf of Bothnia. From the north to south the ., sampling stations are denoted: HF ŽHarufjarden ¨ UM ŽUmea ˚., HL ŽHornslandet. and GB ŽGavlebukten .. The HF station is in the BB, UM ¨ in the northernmost BS just south of Kvarken, HL and GB both in the BS. All stations except that of GB are selected to represent background areas. The GB station is supposed to be affected by the city of Gavle, an industrialized area, as ¨ well as two rivers, Gavlean passing ¨ ˚ and Dalalven, ¨ through industrial and agricultural areas. Samples of herring Ž Clupea harengus. were collected at two pelagic stations, one in the BB and one in the BS. These places of collection are denoted F9 and

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

72

SR5, respectively. Samples of both fish species were also collected in the GG. 2.2. Characteristics of the analyzed species Herring Ž Clupea harengus. was chosen because they are a pelagic non-migratory species staying within their respective seas all their life. It is also a species well defined in a pelagic food web ŽBignert et al., 1993.. Perch Ž Perca flu¨iatilis. is a littoral stationary species. Thus both these fish species have relatively low migratory behaviour and can therefore be used as indicators for a contamination load in a certain area. In order to find differences in contamination load we tried to limit biological variations for the fish species by selecting fish of similar sizes and weights. Table 1 gives the number of fish used in each sample with ranges of weight and size, and sexes. 2.3. Sample clean up and analysis The sample enrichment process was a multi-re-

sidue method aimed to target a multitude of analytes in different analytical endpoints ŽBergqvist et al., 1992.. The analytical steps for those contaminants discussed in this paper, quality control materials as well as parameters used for the analytical instrumentation are thoroughly described by van Bavel et al. Ž1996., and will thus be given just briefly herein. Before extraction an internal standard mixture containing 13 C-labelled lindane Žg-HCH., p, p9DDT, dieldrin, PCB 80 and PCB 153, was added to the samples. Homogenisation of the whole fish including skin, intestines and bones ; 30]100 g was either extracted wet using a Soxhlet device with toluene as solvent during 24 h followed by hexaneracetone during 24 h or homogenised with sodium sulfate and extracted using mixtures of acetonerhexanerdiethylether. After solvent evaporation the extracted organic phase Žlipid. was measured by weighing and the bulk lipid removal was effected by means of a polyethylene film dialysis method ŽHuckins et al., 1990; Strandberg et al., 1997.. Following the above extraction

Table 1 Characteristics of the analyzed fish samples Species

Sampling station

Number of fish in each sample

Weight range Žg.

Length range Žcm.

Sex

Herring

F9 F9 F9 SR5 SR5 SR5 SR5 GG

3 3 3 3 3 3 3 3

27.1]28.4 27.9]28.3 27.9]28.4 26.0]27.0 27.0]28.0 28.0]30.0 30.0]31.0 35.0]40.2

13.2]13.8 13.5]13.7 13.4]13.6 13.2]14.2 13.4]14.0 13.4]14.0 13.6]13.8 16.3]21.3

Male Male Male Male Mixed Mixed Mixed Mixed

Perch

HF HF HF HF UM UM HL GB GB GG GG

5 5 5 5 5 5 5 5 5 8 8

23.6]24.7 23.1]23.8 23.1]23.8 23.9]24.8 23.1]24.8 22.8]24.9 23.0]25.0 23.0]25.0 23.0]25.0 22.3]30.2 22.9]31.2

11.2]11.7 11.1]11.5 11.0]11.5 11.0]11.5 11.0]11.5 11.0]12.7 11.0]11.5 11.0]11.5 11.0]11.5 10.5]15.7 10.3]14.9

Male Mixed Mixed Mixed Male Male Male Male Male Mixed Mixed

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

and lipid removal steps, the organochlorine compounds analyzed here were finally enriched on a column containing 8 g Florisil Žmodified according to Norstrom et al., 1988.. Additional procedure blanks were also performed. All the studied compounds in the blanks were well under 10% of the herein-quantified results. Analysis and detection was accomplished by means of high-resolution gas chromatographyr low resolution mass spectrometry ŽHRGCr LRMS.. The MS instrument, a Fisons MD 800 coupled to a Fisons GC 8000, was working in the electron impact ŽEI. ionisation mode using selected ion recording ŽSIR.. The PCB data given are the summation of 68 tri through deca chlorinated PCB congeners Žvan Bavel et al., 1996.. Identification of these congeners was made according to Schwartz and Stalling Ž1991.. Every PCB congener was quantified using relative response factors ŽRRF. to a native PCB congener in each homologue group, respectively. Of the 14 CHL-related compounds analyzed, 12 are included in technical chlordane and these were heptachlor ŽHCHL., eight octachlordanes Ž trans- and cis-chlordane Ž t-CHL and c-CHL., MC4, MC5, MC7, U81, U82 and U83., and three nonachlor components Ž transand cis-nonachlor Ž t-NCHL and c-NCHL. and MC6.. The two metabolites, not present in the technical chlordane mixture, cis-heptachlorepoxide ŽHCHLep. and oxychlordane ŽOxyCHL. were also analyzed. The MC4, MC5, MC6 and MC7, and U81, U82 and U83 are abbreviations of compounds identified by Miyazaki et al. Ž1985. and Dearth and Hites Ž1991a., respectively. We identified all CHL in a technical chlordane mixture except that of U81 and U83. These two compounds were then tentatively identified according to Dearth and Hites Ž1991b.. The identification of the MC4, 5, 6 and 7, and U82 compounds was further accomplished by comparing analysis of the same reference material ŽBaltic herring oil. as published by Buser et al. Ž1992.. The other CHLrelated native compounds in the external quantification standard represented components, ŽHCHL, t-CHL, c-CHL, t-NCHL, c-NCHL, HCHLep and OxyCHL.. The octa- and nonachlordane components not included in this

73

standard were quantified using RRF to c-CHL and t-NCHL, respectively. Native compounds in the external quantification standard represented all other chemicals in this study. 3. Results and discussion 3.1. Organochlorine concentrations Organochlorine concentrations with mean values and standard deviations Ž%RSD. for the replicates Ž n s 3 or 4. of all samples from the Gulf of Bothnia ŽBB and BS. and the Gulf of Gdansk ŽGG. are presented in Table 2. PCBs and DDTs showed the highest concentrations of all contaminants in all fish samples of the study. The concentrations of PCBs exceeded those of DDTs in all samples. HCHs followed mostly as the third compound group in concentration order except for the remarkably high concentration of HCBz in perch from station GB and the somewhat elevated levels of chlordane-related compounds ŽCHLs. in perch from station UM. Dieldrin and CHLs were mostly similar in concentration followed by HCBz and just minor residues of mirex in most of the samples. All other contaminants analyzed in this study remained below the detection limit. The analytical repeatability of different compounds expressed as %RSD for the replicates was generally acceptable, 30% or better on lipid weight basis. The herring samples from the BB showed the highest repeatability and the variation among the BS-herring samples was somewhat larger for most of the compoundsrcompound groups. The variation for the perch samples from HF was mostly between 20 and 30%, however to some extent larger for the HCHs. The samples that were analyzed in duplicates deviated rarely more than 30% from their mean values for each compound. The lipid content of the herring samples calculated on wet weight basis was lower for the BS Ž4.0%. compared to BB Ž6.6%. ŽTable 2.. There was a larger variation in lipid content ŽRSDs 30%. for the BB replicates compared to those from the BS station ŽRSDs 5%.. In spite of this variation of lipid content, the variation of the

8%

15%

24

Polychlorinated biphenyls PCBs 688

19% 50% 13%

Hexachlorobenzene HCBz

22% 8% 23% 3% ND

21% 26% 3% 16%

p, p9-DDE p, p9-DDMU DDTs

97 13 25 134

0.94 32 1.2 22 ND Ž- 0.31. 100 3.7 159

DDTs o, p9-DDT p, p9-DDT o, p9-DDD p, p9-DDD o, p9-DDE

Hexachlorocyclohexanes a-HCH b-HCH g-HCH HCHs

F9 Au 1991 Mean Ž n s 3.

Sampling time Compounds

%RSD

Herring Bothnian Bay ŽBB.

Species Sampling sites

1555

82

1.7 148 3.9 213 ND Ž- 0.45. 571 79 1017

90 54 59 202

SR5 Au 1991 Mean Ž n s 3.

25%

24%

28% 37% 10%

26% 19% 15% 20% ND

18% 27% 35% 24%

%RSD

Herring Bothnian Sea ŽBS.

1333

41

15 367 14 225 ND Ž- 0.30. 591 45 1257

19 21 35 75

GG Su 1992 Ž n s 1.

Herring Gulf of Gdansk

1171

24

158 3.6 275

8 58 11 36 0.77

69 25 47 141

HF Au 1991 Mean Ž n s 4.

23%

24%

22% 30% 28%

19% 21% 22% 37% 14%

45% 39% 40% 40%

%RSD

Perch Bothnian Bay ŽBB.

Table 2 Mean concentrations Žngrg lipid. for the analyzed compounds in herring and perch from the Baltic Sea

2156

33

289 36 582

35 119 36 64 2.2

61 24 27 112

UM Au 1991 Mean Ž n s 2.

1034

18

135 2.6 207

5 38 7 19 0.53

90 21 31 142

HL Au 1991 Ž n s 1.

Perch Bothnian Sea ŽBS.

5418

270

545 55 876

19 133 29 89 6.7

62 44 39 145

GB Au 1991 Mean Ž n s 2.

2930

19

767 51 1407

35 450 24 74 6.1

97 32 46 175

GG Su 1992 Mean Ž n s 2.

Perch Gulf of Gdansk

74 B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

Table 2 Ž Continued. Species Sampling sites

Herring Bothnian Bay ŽBB.

Sampling time Compounds

F9 Au 1991 Mean Ž n s 3.

Herring Bothnian Sea ŽBS.

%RSD SR5 Au 1991 Mean Ž n s 3.

Herring Perch Gulf of Gdansk Bothnian Bay ŽBB.

%RSD GG Su 1992 Ž n s 1.

HF Au 1991 Mean Ž n s 4.

Perch Bothnian Sea ŽBS.

%RSD UM Au 1991 Mean Ž n s 2.

HL Au 1991 Ž n s 1.

Perch Gulf of Gdansk GB Au 1991 Mean Ž n s 2.

GG Su 1992 Mean Ž n s 2.

Chlordanes Heptachlor ŽHCHL. ND Ž- 0.32. cis-heptachlorepoxide 4.9 ŽHCHLep. U81 ND Ž- 0.22. U82 0.62 U83 ND Ž- 0.22. MC4 0.60 trans-chlordane Ž t-CHL. 1.2 MC5 4.5 cis-chlordane Ž c-CHL. 16 MC7 0.79 Oxychlordane ŽoxyCHL. 6.4 MC6 4.3 trans-nonachlor 22 Ž t-NCHL. cis-nonachlor Ž c-NCHL. 8.0 CHLs 68 Endosulfan 1 Endosulfan 2 Endrin Isodrin Aldrin Dieldrin Mirex Wet weight % Lipid phase

ND Ž- 5.0. ND Ž- 3.2. ND Ž- 0.81. ND Ž- 1.5. ND Ž- 0.64. 36 0.48 35 6.6

ND 5%

ND Ž- 0.25. ND 12 5%

ND Ž- 0.29. 3.9

ND Ž- 0.62. ND 11 22%

ND Ž- 0.58. ND Ž- 0.41. ND Ž- 0.28. ND Ž- 0.15. 7.1 9.3 4.5 3.0

ND 27% ND 33% 19% 7% 8% 9% 9% 4% 4%

ND Ž- 0.20. 1.1 0.54 1.2 2.7 9.1 30 1.2 15 8.6 48

ND Ž- 0.15. 0.43 ND Ž- 0.15. 0.52 2.3 3.7 14 0.73 6 1.7 12

ND Ž- 0.45. 0.35 ND Ž- 0.45. 0.73 2.2 2.9 10 ND Ž- 0.45. 6.2 3.1 22

ND Ž- 0.69. ND Ž- 0.69. ND Ž- 0.69. 3.0 5.5 7.5 39 ND Ž- 0.69. 9.6 7.9 69

3% 1% ND ND ND ND ND 4% 7% 10% 30%

18 148 ND Ž- 3.5. ND Ž- 4.5. ND Ž- 0.74. ND Ž- 1.2. ND Ž- 0.45. 121 1.4 39 4.0

ND 19% 28% 30% 51% 16% 14% 14% 27% 36% 29% 38% 21% ND ND ND ND ND 4% 50%

5%

3.6 49 ND Ž- 4.8. ND Ž- 3.9. ND Ž- 0.65. ND Ž- 1.3. ND Ž- 0.38. 70 ND Ž- 0.22. 76 9.1

10 69 ND Ž- 3.2. ND Ž- 2.9. ND Ž- 0.45. ND Ž- 1.8. ND Ž- 0.25. 97 0.80 52 1.0

ND ND ND 32% 21% 28% 27% ND 36% 26% 23% 24% 20% ND ND ND ND ND 28% 25% 22% 21%

27 175 ND Ž- 5.2. ND Ž- 3.9. ND Ž- 0.66. ND Ž- 2.2. ND Ž- 0.48. 117 1.4 51 1.1

ND Ž- 0.32. ND Ž- 0.32. ND Ž- 0.32. ND Ž- 0.32. 1.5 1.8 6.1 ND Ž- 0.32. 5.8 3.1 15 5.7 48 ND Ž- 3.1. ND Ž- 2.6. ND Ž- 0.48. ND Ž- 1.3. ND Ž- 0.3. 79 0.56 65 1.3

ND Ž- 0.18. ND Ž- 0.15. ND Ž- 0.18. 0.18 ND Ž- 0.18. ND Ž- 0.15. 0.55 0.16 2.3 1.1 2.5 1.3 9.2 4.5 ND Ž- 0.18. 0.20 6.7 3.3 2.6 1.3 24 6.0 8.9 61 ND Ž- 3.9. ND Ž- 2.5. ND Ž- 0.38. ND Ž- 1.2. ND Ž- 0.32. 30 ND Ž- 0.15. 71 1.2

2.3 23 ND Ž- 2.8. ND Ž- 1.9. ND Ž- 0.29. ND Ž- 1.5. ND Ž- 0.28. 35 ND Ž- 0.16.

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

Cyclodiene compounds

110 5.0

The %RSD are given for replicates n ) 2. Detection limits for ND compounds are given in brackets. The lipid content in percent of wet weight of the fish samples from the sampling locations are given at the end of the table. ND, not detected; n, number of samples; %RSD, percent standard deviation; Au, autumn; Su, summer. 75

76

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

normalised contaminant levels in the BB fish samples is somewhat lower than those from the BS. The perch samples from the Gulf of Bothnia showed, however, greater uniformity in lipid content when compared between the various stations ŽTable 2.. Both herring and perch from the GG were larger in size than those from the Gulf of Bothnia ŽTable 1. and had also higher lipid content ŽTable 2.. Differences in sampling seasons, summer for the GG fish and autumn for the Gulf of Bothnia fish, might well influence the lipid content. Time trend studies ŽMoilanen et al., 1982; Andersson et al., 1988; Haahti and Perttila, ¨ 1988; Kannan et al., 1992; Bignert et al., 1993. performed in the Baltic Sea area using herring or other species show a rapid decline in DDT concentrations early after the phase out and regulative uses in many European countries during the 1970s. Levels of other chemicals e.g. PCBs, HCBz and dieldrin have decreased at a very slow rate indicating a continuous input of these chemicals or a slow clearance rate in the Baltic ecosystem. The HCHs, on the other hand, showed almost a steady state in temporal trends ŽKannan et al., 1992.. Moilanen et al. Ž1982. showed a decrease in DDT and PCB levels although an increase for CHLs between the years 1978 to 1982. Table 3 shows a summation of earlier studies and our study using herring as an organism for determining the levels of organochlorine compounds in the Baltic Sea and at some reference locations. Among the organochlorines listed in Table 3, DDTs and PCBs are the most commonly analyzed compounds. As can be seen in Table 3, the levels of DDTs have decreased since the 1970s although the levels during the last 10 years seem to be rather stable. The levels of CHLs have decreased by a factor of ; 5]10 since the earliest available data in spite of the tendency of increase in the early 1980s ŽMoilanen et al., 1982.. The information for the rest of the chemicals is sparse, but it displays a rather uniform picture of the levels since the middle of the 1980s. Notable also are the surprisingly high levels of PCBs, DDTs and CHLs in the Gulf of Bothnia compared to the Baltic proper in 1979 ŽAndersson et al., 1988.. Samples from the Baltic Sea have been re-

ported to contain high levels of organochlorines despite the fact that there have been regulations and prohibitions in the adjacent territories ŽKannan et al., 1992.. The concentrations of most of the organochlorines in the Baltic Sea herring of this study are however comparable to those found in herring from the Great Lakes USA an area reported as heavily polluted with these contaminants ŽNewsome and Andrews, 1993; Suns et al., 1993.. 3.2. Cyclodiene compound occurrences Originating from the eight cyclodiene pesticides analyzed, 12 chlordane-related compounds ŽCHLs., dieldrin and mirex were detected in the herring and perch samples ŽTable 2.. To our knowledge, this is the first time mirex has been detected in samples from the Baltic Sea. Neither heptachlor, aldrin, endrin, isodrin nor endosulfan were found above the detection limit. Among the minor CHLs only U81 was found to be undetected in all samples. Technical chlordane is a product consisting of up to 120 components, mainly hepta-, octa- and nona-chlorinated dicyclopentadienes ŽDearth and Hites, 1991a.. Many of its uses have been similar to those of DDT. The use of chlordane has stopped in the USA, Japan and most European countries ŽBarrie et al., 1992., but the consumption in other, mainly southern latitude countries, continues ŽLoganathan and Kannan, 1994.. We are not aware of any use of chlordane in the Scandinavian ŽAndersson et al., 1988. or the northern European countries but uncertain concerning the use in the former USSR ŽBarrie et al., 1992.. Heptachlor ŽHCHL. is one component in technical chlordane but also a pesticide manufactured on its own with similar applications as chlordane. It has had a limited use in the Finnish plywood industry, but according to Pyysalo et al. Ž1983. this has reportedly not caused any contamination. CHLs, also including the two toxic and persistent metabolites oxychlordane ŽOxyCHL. and cis-heptachlorepoxide ŽHCHLep., are among the most ubiquitous and important toxic environmental contaminants ŽBuser et al., 1992.. The major

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

77

Table 3 Literature data and data from this study Žngrg lipid weight. using herring as a monitoring organism of organochlorine contamination in different areas in the Baltic Sea, and at reference locations Sampling years

Location

CHLs

DDTs

PCBs

Dieldrin

Lindane

HCBz

Reference

NA

7700

3900

NA

NA

NA

Jensen et al., 1969

NA NA NA NA NA NA NA NA 520 410

6200 17 000 7000 14 000 3600 17 000 17 000 37 000 57 000 6762

3500 6800 6100 12 000 20 000 20 000 19 000 20 000 24 000 16 000

NA NA NA NA NA NA NA NA NA NA

NA NA NA NA NA NA NA NA NA NA

NA NA NA NA NA NA NA NA NA NA

Jensen et al., 1969 Jensen et al., 1969 Jensen et al., 1972 Jensen et al., 1972 Jensen et al., 1972 Jensen et al., 1972 Jensen et al., 1972 Jensen et al., 1972 Andersson et al., 1988 Moilanen et al., 1982

600 200 780 563

6800 1600 16 000 2140

7900 2500 24 000 4543

NA NA NA NA

NA NA NA NA

NA NA NA NA

Jansson et al., 1979 Andersson et al., 1988 Andersson et al., 1988 Moilanen et al., 1982

44

754

1030

NA

71

57

Paasivirta and Rantio, 1991 Haahti and Perttila, ¨ 1988 Haahti and Perttila, ¨ 1988

1985]89

Baltic Sea Baltic proper, Stockholm Archipelago Gulf of Bothnia, Bothnian Sea Baltic proper, south Sweden Gulf of Bothnia, Bothnian Bay Gulf of Bothnia, Bothnian Sea Baltic proper, Gulf of Gdansk Gulf of Bothnia, Bothnian Bay Gulf of Bothnia, Bothnian Sea Baltic proper, South Gotland Baltic proper, east of Gotland Baltic proper, Turku Archipelago Baltic proper, Gotland Baltic proper, east of Gotland Gulf of Bothnia, Bothnian Bay Baltic proper, Turku Archipelago Baltic proper, Gulf of Finland

1986

Gulf of Bothnia, Bothnian Bay

NA

330

780

NA

14

18

1986

Gulf of Bothnia, Bothnian Sea

NA

340

840

NA

11

16

1986

˚ Baltic proper, south Aland

NA

440

900

NA

19

14

1986

Baltic proper, Gulf of Finland

NA

510

1100

NA

13

16

1986 1987 1991 1991 1992

Gulf of Bothnia, Bothnian Sea Baltic proper, south Sweden Gulf of Bothnia, Bothnian Bay Gulf of Bothnia, Bothnian Sea Baltic proper, Gulf of Gdansk

189 183 46 96 34

1900 4300 154 932 1138

3600 1300 688 2030 1333

81 68 36 121 70

55 58 25 59 35

120 140 24 82 41

1971 1978 1987 1993

Reference locations Atlantic, east coast of Canada Atlantic, east coast of Canada Skagerak, Swedish west coast Great Lakes US

279 284 38 805

NA NA 570 403

3550 3670 2100 1570

NA NA 36 195

NA NA 35 140

NA NA 41 49

1965 1965]1968 1966]1968 1969 1969 1969 1970 1970 1970 1970 1978 1978 1979 1979 1982

Haahti and Perttila, ¨ 1988 Haahti and Perttila, ¨ 1988 Jansson et al., 1993 Jansson et al., 1993 This study This study This study

Zitko, 1978 Zitko, 1978 Jansson et al., 1993 Newsome and Andrews, 1993

CHLs is the summation of the c-CHL, t-CHL, t-NCHL and OxyCHL components and DDTs is the summation of the p, p9-DDTrDrE compounds. The PCBs, however, vary between references, our data is the summation of 68 congeners. NA, not analysed.

components in technical chlordane are transchlordane Ž t-CHL. Ž24%., cis-chlordane Ž c-CHL. Ž19%., trans-nonachlor Ž t-NCHL. Ž7%., cis-nonachlor Ž c-NCHL. Ž4%. and heptachlor ŽHCHL.

Ž13%. ŽEisler, 1990.. HCHL is however seldom found in biological samples since it is rapidly converted to HCHLep in the environment. The percentage composition of CHLs is shown

78

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

Fig. 2. Percentage composition of individual chlordane-related compounds ŽCHLs. as calculated to total CHLs in herring from the Bothnian Bay ŽBB., Bothnian Sea ŽBS. and Gulf of Gdansk ŽGG.. For abbreviations of CHL compounds, see Table 2.

Fig. 3. Percentage composition of individual chlordane-related compounds ŽCHLs. as calculated to total CHLs in perch from the Bothnian Bay ŽBB. station HL, Bothnian Sea ŽBS. stations UM, HL and GB, and Gulf of Gdansk ŽGG.. For abbreviations of CHL compounds, see Table 2.

in Figs. 2 and 3 for herring and perch, respectively. Composition of CHL patterns including the metabolites is considered as a useful tool for understanding and tracing origin and transport pathways, etc. In areas of recent use of technical chlordane, c-CHL is the most abundant component in fish ŽSchmitt et al., 1990.. Iwata et al. Ž1995. suggested elevated ratios of t-NCHL compared to those of c-CHL and t-CHL in remote areas to be explained by the higher transportability, due to a higher Henry’s law constant Ž H LC ., for t-NCHL. On the other hand this compound has also a higher persistency in the environment as compared to the other two. Equal or higher concentrations of t-NCHL may thus indicate an aged mixture of CHLs in the aquatic system or be suggestive of the input through air transport from low latitude areas. The ratios between the more persistent and air transportable t-NCHL to c-CHL were for the herring samples 1.4]1.6 in the Gulf of Bothnia and 0.8 in the GG, and for perch 1.8]2.6 in the Gulf of Bothnia and 1.3 in the GG. Both fish species indicate thus a higher ratio of t-NCHL in the northern part of the Baltic Sea compared to the GG location. Since we believe long range air transport as a main source for CHLs in the Baltic Sea region, this small, however evident, difference between north and south seems to confirm this assumption.

Fish species as compared to marine mammals or seabirds have been reported to have low potential to metabolise CHLs ŽKawano et al., 1988.. The persistent metabolites OxyCHL and HCHLep have been proven to be much more toxic than the parent chemicals ŽEisler, 1990., thus the abundance and concentration of these should be emphasised when evaluating the presence of CHLs in the environment. In the herring and perch samples investigated here, both OxyCHL and HCHLep are relatively abundant compounds among the CHLs. A slight difference can however be observed for the perch samples from two of the sites, i.e. the HL perch has the highest and the UM perch the lowest abundance of both these compounds despite the fact that these sites are only 400 km apart. Dieldrin is considered to be the strongest carcinogen and most persistent of the major organochlorine pesticides ŽBarrie et al., 1992.. It has been used for many reasons, e.g. as a soil insecticide for agriculture purposes ŽNewton and Asher, 1993. or as a mothproofing agent in the textile industry ŽBoer de, 1989.. Dieldrin was used heavily in Great Britain until the mid 1980s and it is still used with restriction ŽBarrie et al., 1992; Newton and Asher, 1993.. It was banned in Sweden in 1969, and in the US in 1974, the use in former USSR is unknown, but it is still registered for use in developing countries such as India

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

ŽBarrie et al., 1992.. The number of reports on this toxic compound is sparse since it is destroyed during the analytical procedure for biological samples where the lipid removal step is performed by a destructive method, e.g. sulphuricacid-containing reagents. The non-destructive analytical method used in this study enables us to determine this compound among all the other compounds. Mirex was the cyclodiene compound found in the third highest concentration in the fish samples from the Baltic Sea in this study. It is a fully chlorinated compound, very resistant to biodegradation since most organisms cannot metabolise it. It is also very lipophilic and has a high potential to accumulate in food chains ŽMudambi et al., 1992.. It has been used for control of pests e.g. ants, although more widely used as a flame retardant and additive agent in plastics ŽComba et al., 1993.. There are several reports from Lake Ontario which is a heavily polluted lake in America ŽSchmitt et al., 1990; Comba et al., 1993; Newsome and Andrews, 1993; Suns et al., 1993.. Even though it has been reported to have low mobility in the environment ŽWania and Mackay, 1996. it has been found in a marine food web in such a remote area like the Arctic ŽHargrave et al., 1992.. We have found no reports of use of mirex in Europe. Possibly, the use of mirex as a fire retardant in imported products can be the reason for the distribution of this chemical in the Baltic Sea environment. Mirex was found only in the Gulf of Bothnia fish samples and not in the southern Baltic proper i.e. GG. Aldrin was below the detection limit of 0.25]0.65 ngrg lipid weight for both herring and perch in this study ŽTable 2.. Aldrin was mainly used in agriculture for the control of, e.g. soil-inhabiting insects and was previously used intensively as such in Britain ŽNewsome and Andrews, 1993.. Aldrin is quickly oxidised to dieldrin under environmental conditions e.g. by soil bacteria, plants and organisms. It is thus difficult to differentiate the contribution of aldrin to measured values of dieldrin in the environment. Endrin and isodrin were also below the detection limit of 0.29]0.65 and 1.2]2.2 ngrg lipid, respectively in all samples of this study ŽTable 2..

79

To our knowledge these compounds have never been found in samples from the Baltic Sea. Endrin has been used on comparatively few crops in the USA, but is still used in many areas to protect orchards from rodent damage. Schmitt et al. Ž1990. reported data from the US contaminant biomonitoring program, 1976]1984, in freshwater fish. Endrin was found in just a few stations, and the mean concentrations had been declining since the earliest studies. Recalculating on a lipid weight basis, the concentrations should approximately correspond to 30]500 ngrg lipid. The detection limit in our study is thus well below those concentrations found in fish from that study. In, for example, the former USSR the use of endrin was not permitted and we have found no reports of any use in the rest of Europe. It exceeds many other pesticides in regard to toxicity, but it is relatively short-lived in the environment. Isodrin has practically never been used in agriculture in spite of its strong effects as an insecticide and has thus very small distribution in the environment. It is however manufactured for use in the production of endrin. Endosulfan is the final cyclodiene chemical discussed in this paper. It exists in two isomeric forms Žendosulfan 1 and 2. and is widely used as an insecticide for e.g. orchards ŽBarrie et al., 1992.. It is registered and used with restrictions also in Sweden. Endosulfan has high acute toxicity to fish and mammals, and has been shown to cause fish death at small concentrations ŽBarrie et al., 1992.. Although it is bioconcentrated, it is eliminated or metabolised rapidly in organisms and is therefore not accumulated. Both isomers were below the detection limit of 1.9]5.2 ngrg lipid in the analyzed samples of this study ŽTable 2.. Endosulfan has been proved to be widely distributed world wide since it was one of the most abundant cyclodiene compounds found in snowpacks in such remote areas as the Canadian Arctic ŽBarrie et al., 1992.. The presence of these compounds also in the Baltic Sea environment cannot be excluded. 3.3. Spatial ¨ariations Fig. 4 shows the spatial variation of detected

80

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

Fig. 4. Spatial variation of organochlorines in herring Ž Clupea harengus. from the Bothnian Bay ŽBB., Bothnian Sea ŽBS. and Gulf of Gdansk ŽGG..

organochlorines Žmirex not shown. in herring on lipid weight basis, for the pelagic stations BB and BS in the Gulf of Bothnia and GG in the southern Baltic proper. A comparison between the Gulf of Bothnia stations shows that the BS station Žin the southern Gulf of Bothnia. has an evidently higher pollutant burden than the BB station. Station BS shows two to three times higher concentrations for all contaminants compared to station BB, and for DDT the difference is even larger. The contamination load with respect to organochlorine levels in herring in the GG was surprisingly low since this area is known to be highly industrialised and under impact of untreated municipal effluents as well as of large volumes of agricultural runoff via the Wisla River. The levels were quite similar to those of station BB except that of the somewhat elevated levels of PCBs and high levels of DDTs where GG was the highest of the three stations studied. Fig. 5 displays in the same way the spatial variation for the four coastal stations in the Gulf of Bothnia and GG for the perch samples. At a first comparison of the HF, UM and HL stations, all representing background areas in the Gulf of Bothnia, station UM shows a factor of 2]3 times higher concentrations for most of the compounds. A carbon content analysis on a dry weight basis in surface sediment shows much lower TOC for the UM station Ž; 0.8%. compared to the other two

Fig. 5. Spatial variation of organochlorines in perch Ž Perca flu¨iatilis. from station HF in the Bothnian Bay ŽBB., UM, HL and GB in the Bothnian Sea ŽBS. and Gulf of Gdansk ŽGG..

background stations Ž; 2.5%.. The lower carbon content might be explained by lower biomass content in the water in this area. A lower biomass may result in less dilution and thus higher bioavailability of contaminants in the water column, giving a higher concentration in biota. Station GB that is influenced by industrial and agricultural activity shows markedly higher levels of only PCBs and HCBz compared to the other coastal stations. The somewhat higher DDT levels at station GB can be due to agricultural use of this pesticide in the past in this area, which is now slowly leaking out from the rivers or coastal water systems affecting the Gavlebukten. The domi¨ nance of the metabolite DDE in relation to DDT indicates however an aged mixture of this pesticide. The elevated levels of PCBs Žfive times. and HCBz Ž10 times. can be caused by a specific industrial point emission source. Pure PCB has been used in this region as cutting oil in steel-mills in the 1970s. In agreement with the picture for herring from GG ŽFig. 4., perch also had higher DDT levels than any station from the Gulf of Bothnia and also elevated levels of PCBs. For the rest of the contaminants however quite similar levels compared to a background station or open sea situation in the Gulf of Bothnia. A lower DDErDDT ratio in the fish from GG compared to those in the Gulf of Bothnia indicates a somewhat more recent use of DDT in this area.

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

81

4. Conclusion

Acknowledgements

The Baltic Sea is polluted with a notable amount of organochlorine compounds, including cyclodiene chemicals like chlordane, dieldrin and mirex. The pollution load in the studied areas shows a rather uniform picture and is comparable for many compounds to the Great Lakes, America, an area reported as heavily polluted with these contaminants. Some other cyclodiene pesticides like heptachlor, aldrin, endrin, isodrin and endosulfan were not detected in this study. Variation in pollution load can be seen in analyzed fish from different sampling locations in the Gulf of Bothnia and the Gulf of Gdansk ŽGG.. There can be several possible explanations e.g. differences in temperature, food chains, biomass content in the water, lipid weight for analyzed species, differences in washout from the atmosphere as well as point sources. In the Gulf of Bothnia there is higher pollution burden in the Bothnian Sea ŽBS. than the BothŽGB. region nian Bay ŽBB., and the Gavlebukten ¨ in BS is a point source of PCBs, HCBz and DDTs. The Gulf of Gdansk ŽGG. has a similar pollution level of organochlorine compounds compared to a background station or open sea situation in the Gulf of Bothnia except for the elevated levels of DDTs and PCBs. The lower DDE to DDT ratio that was found in both fish species indicate a point source in this area. It seems like the rest of the pesticides discussed in this paper have not been used in Poland. Other chemicals like HCHs, chlordanes, dieldrin and mirex seem to be more evenly distributed between the sampling stations indicating long range transport as the most important source. There are chlordane pattern similarities between the different sampling stations for both fish species. The somewhat elevated levels of the more easily air transportable Žhigher Henry’s law constant. trans-nonachlor, than trans- and cischlordane, in the north ŽGulf of Bothnia. compared to the south ŽGulf of Gdansk. seem to indicate long range air transport as the source for chlordanes in the Baltic Sea environment.

We gratefully acknowledge the financial support of the Kempe Foundation, the Swedish Forest Industries, the Finnish National Board of Water and the Environment, the Swedish Environmental Protection Agency, the Boliden Mineral AB and the Center for Environmental Research in Umea, ˚ Sweden. References Ahlborg U, Lipworth L, Titus-Ernstoff L, Hsieh C, Hanberg A, Baron J, Trichopoulos D, Adami H-O. Organochlorine compounds in relation to breast cancer, and endometriosis: an assessment of the biological and epidemiological evidence. Crit Rev Toxicol 1995;25:463]531. ¨ Linder C, Olsson M, Reuteraardh L, Uvemo U, Andersson O, Wideqvist U. Spatial differences and temporal trends of organochlorine compounds in biota from the Northwestern Hemisphere. Arch Environ Contam Toxicol 1988;17: 755]765. Barrie LA, Gregor D, Hargrave B, Lake R, Muir D, Shearer R, Tracey B, Bidleman T. Arctic contaminants: sources, occurrence and pathways. Sci Total Environ 1992;122:1]74. van Bavel B, Naf ¨ C, Bergqvist P-A, Broman D, Lungren K, Papakosta O, Rolff C, Strandberg B, Zebuhr ¨ Y, Zook DR. Levels of PCBs in the aquatic environment of the Gulf of Bothnia: bentic species and sediments. Mar Pollut Bull 1996;32:210]218. Bignert A, Gothberg A, Jensen S, Litzen K, Odsjo ¨ ¨ T, Olsson M, Reutergardh L. The need for adequate biological sam˚ pling in ecotoxicological investigations: a retrospective study of twenty years pollution monitoring. Sci Total Environ 1993;128:121]139. Bergqvist P-A, Bandh C, Broman D, Ishaq R, Lundgren K, Naf ¨ C, Pettersen H, Rappe C, Rolff C, Strandberg B, Zebuhr ¨ Y, Zook DR. Multi-residue analytical method including planar PCB, dioxins and other organic contaminants for marine samples. Organohalogen Compd 1992;9:17]20. Boer de J. Organochlorine compounds and bromodiphenylethers in livers of Atlantic cod Ž Gadus morhua. from the North Sea, 1977]1987. Chemosphere 1989;18: 2131]2140. Buser HR, Muller M, Rappe C. Enantioselective determina¨ tion of chlordane components using chiral high-resolution gas chromatography-mass spectrometry with application to environmental samples. Environ Sci Technol 1992;26: 1533]1540. Comba ME, Norstrom RJ, McDonald CR, Kaiser KLE. A Lake Ontario Gulf of St. Lawrence dynamic mass budget for mirex. Environ Sci Technol 1993;27:2198]2206.

82

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83

Dearth MA, Hites RA. Chlordane accumulation in people. Environ Sci Technol 1991a;25:1279]1285. Dearth MA, Hites RA. Complete analysis of technical chlordane using negative mass spectrometry. Environ Sci Technol 1991b;25:245]254. Eisler R. Chlordane hazards to fish, wildlife and invertebrates: a synoptic review. Biological report 85, Contaminant Hazard Reviews Report 21, Fish and Wildlife Service. US: Department of the Interior, 1990:1]21. Falandysz J, Kannan K, Tanabe S, Tatsukawa R. Organochlorine pesticides and polychlorinated biphenyls in cod liver oils: North Atlantic, Norwegian Sea, North Sea and Baltic Sea. Ambio 1994;23:288]293. Haahti H, Perttila ¨ M. Levels and trends of organochlorines in cod and herring in the northern Baltic. Mar Pollut Bull 1988;19:29]32. Hargrave BT, Harding GC, Vass WP, Erickson PE, Fowler BR, Scott V. Organochlorine pesticides and polychlorinated biphenyls in the Arctic ocean food web. Arch Environ Contam Toxicol 1992;22:41]54. Huckins JN, Tubergen MW, Lebo JA, Gale RW, Schwartz TR. Polymeric film dialysis in organic solvent media for cleanup of organic contaminants. J Assoc Off Anal Chem 1990;73:290]293. Iwata H, Tanabe S, Sakai N, Tatsukawa R. Distribution of persistent organochlorines in the oceanic air and surface seawater and the role of ocean on their global transport and fate. Environ Sci Technol 1993;27:1080]1098. Iwata H, Tanabe S, Ueda K, Tatsukawa R. Persistent organochlorine residues in air, water, sediments and soils from the Lake Baikal region, Russia. Environ Sci Technol 1995;29:792]801. Jansson B, Vaz R, Blomkvist G, Jensen S, Olsson M. Chlorinated terpenes and chlordane components found in fish, guillemot and seal from Swedish waters. Chemosphere 1979;5:181]190. Jansson B, Andersson R, Asplund L, Litzen K, Nylund K, Sellstrom ¨ U, Uvemo U-B, Nahlberg K, Wideqvist U, Odsjo ¨ T, Olson M. Chlorinated and brominated persistent organic compounds in biological samples from the environment. Environ Toxicol Chem 1993;12:1163]1174. Jensen S. Report of a new chemical hazard. New Sci 1966;32:612. Jensen S, Johnels AG, Olsson M, Otterlind G. DDT and PCB in marine animals from Swedish waters. Nature 1969;224:247]250. Jensen S, Johnels AG, Olsson M, Otterlind G. DDT and PCB in herring and cod from the Baltic, the Kattegat and the Skagerak. Ambio Special Rep 1972;1:71]85. Kannan K, Falandysz J, Yamashita N, Tanabe S, Tatsukawa R. Temporal trends of organochlorines in cod-liver oil from the southern Baltic Proper, 1971]1989. Mar Pollut Bull 1992;24:358]363. Kawano M, Inowa T, Hidaka H, Tatsukawa R. Chlordane compounds residues in weddell seals from Antarctic. Chemosphere 1984;13:95]100. Kawano M, Inoue T, Hidaka H, Tatsukawa R. Bioconcentra-

tion and residue patterns of chlordane compounds in marine animals: invertebrates, fish, mammals and seabirds. Environ Sci Technol 1988;22:792]797. Lindstrom M, Karlsson L, Rappe C, ¨ G, van Bavel B, Jaremo ¨ Hardell L. The use of supercritical fluid extraction ŽSFE. as a sample preparation method in analysis of PCDD, PCDF and PCB in human tissue. Organohalogen Compd 1995;23:27]30. Lindstrom ¨ G, van Bavel B, Broman C. Determination of chlordane in human adipose tissue by SFE-LC off-line MS. Organohalogen Compd 1996;27:253]256. Loganathan B, Kannan K. Global organochlorine contamination trends: an overview. Ambio 1994;23:187]191. Miyazaki T, Yamagishi T, Matsumoto M. Isolation and structure elucidation of some components in technical grade chlordane. Arch Environ Contam Toxicol 1985;14:475]483. Moilanen R, Pyysalo H, Wikstrom ¨ K, Linko R. Time trends of chlordane, DDT and PCB concentrations in pike Ž Esox lucius. and Baltic herring Ž Clupea harengus. in the Turku Archipelago, northern Baltic Sea for the period 1971]1982. Bull Environ Contam Toxicol 1982;29:334]340. Mudambi AR, Hassett JP, McDowell WH, Scrudato RJ. Mirex-photomirex relationships in Lake Ontario. J Great Lakes Res 1992;18:405]414. Muir D, Grift N, Lockhart WL, Wilkinson P, Billeck B, Bruskill G. Spatial trends and historical profiles of organochlorine pesticides in Arctic lake sediments. Sci Total Environ 1995;160r161:447]457. Newsome WH, Andrews P. Organochlorine pesticides and polychlorinated biphenyl congeners in commercial fish from the Great Lakes. J AOAC Int 1993;76:707]710. Newton I, Asher A. Long-term trends in organochlorine and mercury residues in some predatory birds in Britain. Environ Pollut 1993;79:143]151. Norstrom RJ, Simon M, Muir D. Organochlorine contaminants in arctic marine food chains: identification, geographical distribution, and temporal trends in polar bears. Environ Sci Technol 1988;22:1063]1071. Oden S, Ekstedt J. PCB and DDT in Baltic sediments. Ambio Special Rep 1976;4:125]129. Oehme M. Dispersion and transport paths of toxic persistent organochlorines to the Arctic } levels and consequences. Sci Total Environ 1991;106:43]53. ¨ Bergman A, ˚ Blomkvist G, Frank A, Olsson M, Andersson O, Rappe C. Contaminants and diseases in seals from Swedish waters. Ambio 1992;21:561]562. Paasivirta J, Rantio T. Chloroterpenes and other organochlorines in Baltic, Finnish and Arctic wildlife. Chemosphere 1991;22:47]55. Pyysalo H, Wikstrom ¨ K, Litmanen R. A baseline study on the concentrations of chlordane-, PCB-, and DDT compounds in Finnish fish samples. Chemosphere 1983;12:837]842. Rappe C. In ecological problems of the circumpolar area. International Symposium, June 1971, Lulea, ˚ Sweden, 1971. Schmitt CJ, Zajicek JL, Peterman PH. National contaminant biomonitoring program: residues of organochlorine chemi-

B. Strandberg et al. r The Science of the Total En¨ironment 215 (1998) 69]83 cals in US freshwater fish, 1976]1984. Arch Environ Contam Toxicol 1990;19:748]781. Schwartz TR, Stalling DL. Chemometric comparison of polychlorinated biphenyl residues and toxicologically active polychlorinated biphenyl congeners in the eggs of Forester’s terns Ž Sterna fisteri .. Arch Environ Contam Toxicol 1991;20:183]199. Strandberg B, Bergqvist P-A, Rappe C. Dialysis with semipermeable membranes as an efficient lipid removal method in the analysis of bioaccumulative chemicals. Anal Chem 1997:in press. Suns KR, Hitchin GG, Toner D. Spatial and temporal trends of organochlorine contaminants in spottail shiners from selected sites in the Great Lakes Ž1975]1990.. J Great Lakes Res 1993;19:703]714.

83

Wania F, Mackay D. Tracking the distribution of persistent organic pollutants. Environ Sci Technol 1996;30:390]396. Westernhagen.v H, Rosenthal H, Dethlefsen V, Ernst W, Harms U, Hansen PD. Bioaccumulating substances and reproductive success in Baltic flounder. Aquat Toxicol 1981;1:85]99. Wikstrom ¨ K, Pyssalo H. Organochlorine compounds in the liver of cod Ž Gadus morhua. in the northern part of Baltic. Chemosphere 1981;10:999]1004. ¨ Wagman N, Strandberg B, Bavel v B, Oberg LG, Rappe C. ˚ PCBs in earthworms Ž Eisenia foetida. from household composts. Organohalogen Compd 1995;26:401]404. Zitko V. Nonachlor and chlordane in aquatic fauna. Chemosphere 1978;7:3]7.