(Received 2 July 2003; Accepted 12 November 2003)

Transcription

1 Environmental Toxicology and Chemistry, Vol. 23, No. 7, pp , SETAC Printed in the USA /04 $ ACCUMULATION AND DISTRIBUTION OF POLYCHLORINATED DIBENZO-p-DIOXIN, DIBENZOFURAN, AND POLYCHLORINATED BIPHENYL CONGENERS IN ATLANTIC SALMON (SALMO SALAR) PIRJO ISOSAARI,* HANNU KIVIRANTA, ØYVIND LIE, ANNE-KATRINE LUNDEBYE, GORDON RITCHIE, and TERTTU VARTIAINEN National Public Health Institute, Department of Environmental Health, P.O. Box 95, FIN Kuopio, Finland National Institute of Nutrition and Seafood Research, P.O. Box 176, Sentrum, N-5804 Bergen, Norway Nutreco, Aquaculture Research Centre, P.O. Box 48, Sjøhagen 3, 4016 Stavanger, Norway University of Kuopio, Department of Environmental Sciences, P.O. Box 1627, FIN Kuopio, Finland (Received 2 July 2003; Accepted 12 November 2003) AbstractAdult Atlantic salmon (Salmo salar) were fed on four diets containing polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and polychlorinated biphenyls (PCBs) for 30 weeks. Lipid-normalized concentrations showed that all congeners were equally partitioned between whole-fish and fillet samples. Skinned fillet accumulated approximately 30% of the total PCDD/F and PCB content in fish. Accumulation efficiencies in whole fish were 43% for 2,3,7,8-chlorinated dibenzo-p-dioxins and dibenzofurans, 83% for dioxin-like PCBs, and 78% for other PCB congeners. Among PCDD/Fs, tetra- and pentachlorinated congeners were preferentially accumulated in salmon, whereas hepta- and octachlorinated dibenzo-p-dioxins were excreted in the feces. Substitution patterns that were associated with a preferential accumulation of PCBs in salmon included nonortho substitution and tetrachlorination. Accumulation efficiencies and lipid-normalized biomagnification factors (BMFs) were not influenced by the PCDD/F and PCB concentrations of the diets. Biomagnification (BMF 1) of tetra- and pentachlorinated dibenzop-dioxins and dibenzofurans and of all the PCBs was observed. Differences in the behavior of PCDD/F and PCB congeners resulted in a selective enrichment of the most toxic congeners in salmon. KeywordsDietary accumulation Farmed fish Salmon Dioxins Polychlorinated biphenyls INTRODUCTION Polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and polychlorinated biphenyls (PCBs) have some toxic properties in common, although their toxic potency is variable. According to the potency to induce dioxin-like toxicity, the World Health Organization (WHO) has worked out a revised list of toxic equivalence factors (TEFs) for human and mammalian species concerning the 17 PCDD/F congeners with the 2,3,7,8-chlorination pattern and the 12 dioxin-like PCB congeners with non- or mono-ortho chlorination pattern. Separate lists of WHO-TEFs have been established for fish and birds [1]. Tolerable daily and weekly intakes recommended by the WHO and the European Commission (EC) take into account the combined exposure to both PCDD/Fs and dioxin-like PCBs [2,3]. As yet, the EC s maximum permissible limits for various food and feed items have only been set for PCDD/Fs [4,5]. The existing regulations and risk assessments clearly point out that investigations into the sources, transportation routes, and accumulation properties of both PCDD/Fs and PCBs are needed to limit exposure in the first place. One of the critical sources of human dietary exposure to PCDD/Fs and PCBs is fish. It has been observed that some fish, especially fatty fish, may contain high levels of PCDD/ Fs and PCBs as a result of food-chain biomagnification from contaminated ecosystems or fish feed [6 9]. Lots of work has already been done to investigate accumulation, distribution, * To whom correspondence may be addressed (pirjo.isosaari@ktl.fi). elimination, and toxicity of these compounds in different fish species. In some of the most recent studies, researchers have paid special attention to the practicality of experiments by using environmentally relevant concentrations, adult fish, and long-term exposure [10 12]. In most studies, only 2,3,7,8- tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) or a number of PCDD/F or PCB congeners have been investigated, despite the coexistence of these compounds in the environment. In addition, the comparability of different studies is often poor, because the behavior of PCDD/Fs and PCBs can be influenced by a large number of factors, including the source, duration, and magnitude of exposure as well as the fish species and age. In addition, variation between individuals implies that several fish (or pooled samples) and parallel treatments need to be examined to obtain statistically significant results. The individual PCDD/F and PCB congeners also differ from each other in their behavior and fate, because they comprise a variety of structural and chlorination patterns that, in turn, affect their physicochemical and metabolic properties. The aim of the present study was to assess the properties that influence the accumulation of dioxin-like compounds, and for this purpose, comparable data regarding the accumulation of all dioxin-like PCDD/Fs and PCBs were required. Such data were obtained by exposing parallel groups of adult Atlantic salmon (Salmo salar) to graded levels of dioxin-like PCDD/ Fs and PCBs and other environmentally relevant PCBs in fish feed. The present report presents valuable new information concerning long-term accumulation and distribution of PCDD/ Fs and PCBs in a widely consumed fatty fish species using natural sources and levels of dietary PCDD/Fs and PCBs that 1672

2 Dioxin accumulation in salmon Environ. Toxicol. Chem. 23, are representative of the commercial feed used in aquaculture. Congener-specific accumulation efficiencies and biomagnification factors serve as a basis for environmental and human health risk assessments as well as for regulatory work. Fish, diets, and sampling MATERIALS AND METHODS Atlantic salmon were exposed to dietary PCDD/Fs and PCBs in 7,000-L flow-through tanks that were kept outdoors. Saltwater flow was maintained at 125 L/min, temperature at 7to10 C, and oxygen content at approximately 9 to 10 mg/ L. The tanks were kept clean with ultraviolet filters and by brushing the surfaces. A black net covered the tanks, giving some shadow; otherwise, the lighting conditions were natural. Before the exposure, the salmon were fed with commercial salmon feed. Salmon with an average weight of 1.8 kg were randomly allocated into four exposure groups (A, B, C, and D), each of which comprised three parallel tanks with 65 salmon in each. The diets for these four exposure groups were prepared by mixing a basal fish feed with a constant amount of fish meal and varying proportions of two fish oils to create a distinct concentration gradient in the diets. One of the fish oils was of Pacific origin and contained low levels of PCDD/ Fs and PCBs, whereas the selected lot of fish oil of Baltic origin contained high levels of these pollutants. The lipid content of the feeds was 33 to 36%, and the protein content was 39 to 40%. Each day during the 30-week trial, the salmon were fed to satiation, and the feed that remained uneaten was collected from the tanks. The average feed consumption rates during the first and last 15-week periods were and kg kg 1 d 1, respectively. At the end of the trial, the salmon had reached an average weight of 4.8 kg. Salmon were sampled at the beginning of the feeding trial (a composite sample comprising one fish from each tank) and after 15 and 30 weeks from the beginning (a composite sample of 12 fish from each tank at both sampling times). A starvation period of 3 to 4 d was used before the samplings to ensure that the guts were empty. Before pooling the fish for a composite sample, the weights and lengths of the fish were recorded. Feces were carefully stripped from each tank at 15 and 30 weeks. Four composite water samples (1 L) representing each exposure group were collected at 30 weeks. Sample pretreatment and analysis Concentrations of PCDD/Fs and PCBs were analyzed in whole fish, skinned fillet, feces, water from the tanks, all the diets, fish meal, and fish oil. The samples were freeze-dried and homogenized. The entire fillet and whole fish were included in the homogenates. A subsample of approximately 40 g was Soxhlet extracted for 20 h with toluene, and the lipid content was determined gravimetrically. The extract was spiked with 100 pg of PCDD and PCDF standards (16 PCDD/ F congeners), non ortho-pcb standards (PCBs 77, 81, 126, and 169), and other PCB standards (PCBs 30, 80, 101, 105, 138, 153, 156, 180, and 194). All the standards, except for PCB 30, were 13 C-labeled and purchased from the Cambridge Isotope Laboratories (Andover, MA, USA). The extracted lipids were decomposed with sulfuric acid and removed from the sample in a silica gel column. The sample was then eluted through columns filled with activated carbon (Carbopack C, 60/80 mesh; Supelco, Bellefonte, PA, USA) and Celite (catalog no. 2693; Merck, Darmstadt, Germany) to separate PCDD/Fs from PCBs. Further clean-up of the PCDD/F and PCB fractions was achieved with an activated aluminum column (catalog no. 1097, standardized, activity level II III; Merck). Non-ortho PCBs (coplanar PCBs) were separated from the other PCBs in another activated carbon column (without Celite). Individual PCDD/F, PCB, and non-ortho PCB congeners in these three sample fractions were separated with a HP 6890 gas chromatograph equipped with a fused silica capillary column (DB- Dioxin, 60-m column length, 0.25-mm inner diameter, m phase thickness; J&W Scientific, Folsom, CA, USA). Two microliters were injected into a split-splitless injector at 270 C. The temperature program for PCDD/Fs was started at 140 C and held for 4 min, then raised by 20 C/min to 180 C and held for 0 min, then raised by 2 C/min to 270 C and held for 36 min. For PCBs, the temperature program was started at 60 C and held for 3 min, then raised by 20 C/min to 200 C and held for 0 min, then raised by 4 C/min to 270 C and held for 14 min. For non-ortho PCBs, the temperature program was started at 140 C and held for 4 min, then raised by 20 C/min to 200 C and held for 0 min, then raised by 10 C/min to 270 C and held for 12 min. Congeners were quantified with a high-resolution mass-spectrometer (VG SE; VG Analytical, Manchester, UK) that was coupled to the gas chromatograph and operated on a selective-ion recording mode at a resolution of 10,000. The concentrations of individual congeners below the limit of quantification were considered to be zero. The limit of quantification for an individual congener was the concentration of an analyte that produced an instrumental response at two different ions to be monitored with a signal to noise ratio of 3: 1. On a lipid-weight basis, the limits of quantification were 0.5 to 5 pg/g for PCDD/Fs, 0.1 to 0.5 ng/g for PCBs, and 1 pg/g for non-ortho PCBs. The analytical methods have been accredited according to the International Organization for Standardization/International Electrotechnical Commission standard 17025, and the high precision of the analyses has been frequently proven during intercalibration studies [13 16]. Seven percent of the samples of the present study were analyzed in duplicate. The coefficients of variation of PCDD/F and PCB sum concentrations and WHO toxic equivalents (TEQs) in replicate analyses were less than 10%. Blank samples covering the whole analytical procedure did not indicate problems with laboratory or crosssample contamination. Statistical analyses Statistical analyses of the data were performed with SPSS for Windows 9.01 (SPSS, Chicago, IL, USA) statistical software. The analytical tests used were one-way analysis of variance with Scheffe s or Tamhane s T2 post-hoc tests for pairwise multiple comparisons in cases of equal or unequal variances, respectively. Combined effects of multiple factors were examined with univariate analysis of variance. Differences with p 0.05 were considered to be statistically significant. RESULTS AND DISCUSSION PCDD/Fs and PCBs in fish feeds Analyses of feeds A through D proved that they had the expected concentration gradient. The PCDD/F and PCB contents of the feeds reflected the contents in the fish meal and

3 1674 Environ. Toxicol. Chem. 23, 2004 P. Isosaari et al. Table 1. Sum and World Health Organization toxic equivalent (WHO-TEQ) concentrations of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) and polychlorinated biphenyls (PCBs) PCDD/F sum (pg/g) WHO PCDD/F -TEQ (pg/g) PCB sum (ng/g) WHO PCB -TEQ (pg/g) WHO SUM -TEQ (pg/g) PCB contribution to WHO SUM -TEQ (%) Baltic fish oil (lipid wt) Pacific fish oil (lipid wt) Fish meal (dry wt) Feed A (dry wt) Feed B (dry wt) Feed C (dry wt) Feed D (dry wt) Whole fish, start (fresh wt) Whole fish A D, 15 weeks (fresh wt) Whole fish A D, 30 weeks (fresh wt) Fillet, start (fresh wt) Fillet A D, 15 weeks (fresh wt) Fillet A D, 30 weeks (fresh wt) Feces, A D (dry wt) Water, A D (L 1 ) a a a a a 8 53 a Decreasing percentage from A to D. fish oils. The PCDD/F sum concentrations in the feeds ranged from 3.9 to 22 pg/g dry weight, or 0.2 to 4.9 pg WHO PCDD/F - TEQ/g dry weight (Table 1). Congener profiles were nearly the same in feeds B, C, and D (Fig. 1). Feed A only contained fish oil of Pacific origin, and its profile had the biggest deviation from the average profile, with a larger proportion of octachlorinated dibenzo-p-dioxin and a smaller proportion of 2,3,4,7,8-pentachlorodibenzofuran (2,3,4,7,8-PeCDF) than the other feeds. The 2,3,4,7,8-PeCDF was the main contributor to the WHO PCDD/F -TEQ of the feeds, and concentrations of 2,3,7,8-TCDD were 0.47 pg/g dry weight or less. The PCB sum concentrations in feeds A through D were 28 to 80 ng/g dry weight, or 2.7 to 5.4 pg WHO PCB -TEQ/g dry weight. The predominating congeners were PCBs 118, 138, and 153, whereas non-ortho PCB 126 had a highest share of WHO PCB -TEQ (Fig. 2). The percentage contribution of dioxin-like PCBs to the WHO SUM -TEQ (including PCDD/Fs and PCBs) decreased from 93% in feed A to 53% in feed D, because the PCB to PCDD/F ratio was lower in fish oil of Baltic origin than in fish meal and fish oil of Pacific origin. The concentrations in the pregrowth feed was not measured, but they were presumably at the same level or even higher than that in feed A, which led to a greater elimination than intake of contaminants in exposure group A at the beginning of the experiment. Therefore, the results from group A were not taken into account when calculating accumulation efficiencies and biomagnification factors. PCDD/Fs and PCBs in fish, feces, and water Concentrations of PCDD/Fs and PCBs in whole salmon and fillet increased in a dose-dependent manner during the trial, with the exception of group A. At its highest, the PCDD/ F concentration in group D salmon was 6.2 pg/g fresh weight in fillet and 8.2 pg/g fresh weight in whole fish. The WHO PCDD/F -TEQ content of the fillet was less than half the EC s maximum limit of 4 pg WHO PCDD/F -TEQ/g fresh weight [4], despite the fact that the levels in feeds C and D exceeded the maximum limit regarding foodstuffs for fish (2.25 pg WHO PCDD/F -TEQ/g [5]). The PCB concentrations were 47 ng/ g fresh weight in fillet and 59 ng/g fresh weight in whole fish. The lipid content was 16% in fillet and 20% in whole fish, and it remained unchanged in all exposure groups during the study. The lipid-normalized concentration ratios of fillet to whole fish showed an equal partitioning of PCDD/F ( ) (SD throughout) and PCB congeners ( ) during the last 15 weeks of the exposure, and no statistically significant differences were observed between the partitioning of different congeners in feeding groups B, C, and D. Comparison of PCDD/F congener profiles in feeds, fish, and feces showed that lower-chlorinated furans were preferably accumulated in fish tissues, whereas hepta- and octachlorinated dioxins and furans were depleted in fish and enriched in feces (Fig. 1). Congener profiles of PCBs in fish and feces did not seem to differ markedly from the profile in feed (Fig. 2). The amount of feces increased as the salmon grew bigger; however, the PCDD/F and PCB concentrations remained at the same level. The average amount of feces excreted in a tank during the trial was 390 g fresh weight. At its highest, the PCDD/F content of the feces collected from group D was 123 pg WHO PCDD/F -TEQ, and the amount of PCBs was 35 pg WHO PCB -TEQ. Compared to the total intake from feed, the amounts distributed into the feces were negligible. Of the WHO PCDD/F -TEQ and WHO PCB -TEQ intake, only 0.01 to 0.02% and to 0.005%, respectively, ended up in the feces. The water samples collected from tanks A through D contained 0.01 to 0.09 pg/l of WHO PCDD/F -TEQ and to 0.02 pg/l of WHO PCB -TEQ. Considering the high volumes of water, it is possible that water comprised a larger PCDD/F and PCB sink than feces. However, the low concentrations in the water samples would make exact budgets highly uncertain. Accumulation efficiencies Accumulation efficiency ( ), or the net effect of dietary absorption and elimination, was calculated as C fish (1) FtC feed where C fish is the concentration in fish, C feed is the concentration in feed, F is the feeding rate (kg kg 1 d 1 ), and t is time (210 d). All concentrations were fresh-weight concentrations, and concentrations in fish were corrected for growth dilution and

4 Dioxin accumulation in salmon Environ. Toxicol. Chem. 23, Fig. 2. Congener profiles of polychlorinated biphenyls (PCBs) in the feeds, whole fish, fillet, and feces illustrated as average profiles standard deviation in exposure groups B, C, and D. Number system for PCBs is according to the International Union of Pure and Applied Chemistry. Fig. 1. Congener profiles of 2,3,7,8-chlorinated dibenzo-p-dioxins and dibenzofurans (2,3,7,8-PCDD/Fs) in the feeds, whole fish, fillet, and feces illustrated as average profiles standard deviation in exposure groups B, C, and D. TCDF/TCDD tetrachlorinated dibenzofuran/ dibenzo-p-dioxin; PeCDF/PeCDD pentachlorinated dibenzofuran/ dibenzo-p-dioxin; HxCDF/HxCDD hexachlorinated dibenzofuran/ dibenzo-p-dioxin; HpCDF/HpCDD heptachlorinated dibenzofuran/ dibenzo-p-dioxin; OCDF/OCDD octachlorinated dibenzofuran/dibenzo-p-dioxin. initial contaminant concentration [7]. Accumulation efficiencies were calculated separately for each of the 12 fish tanks. Accumulation efficiencies showed that 43% of PCDD/Fs, 83% of dioxin-like PCBs, and 78% of other PCBs eaten by the salmon ended up in the whole fish (Table 2). When WHO- TEQs were applied for individual congeners, the resulting accumulation efficiencies of WHO PCDD/F -TEQ and WHO PCB -TEQ were higher. Approximately 30% of both pollutants (sum and WHO-TEQ) found in fish were located in skinned fillet. The concentrations in the feeds had no influence on the accumulation efficiencies. Factors other than the exposure group that could possibly affect the accumulation of different PCDD/F congeners were not tested statistically because of the small number of congeners available for examination. However, accumulation efficiency decreased with increasing chlorine substitution of the PCDD/F congener. The lowest accumulation efficiency (24%) was measured for 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin. The octachlorinated dibenzo-p-dioxin must have had even lower accumulation efficiency, because despite its high proportion in the feeds, it was not found at detectable levels in salmon. In previous studies, the effect of chlorination degree on the accumulation of PCDD/ Fs has also remained unclear because of the lack of comparable data regarding all the 2,3,7,8-chlorinated congeners. Investigations carried out by Loonen et al. [17] indicated very low dietary absorption efficiencies for PCDD/Fs in general, and the absorption of penta- and hexachlorinated dibenzo-p-dioxins and dibenzofurans was more efficient than the absorption of tetra-, hepta-, and octachlorinated dibenzo-p-dioxins and dibenzofurans. On the other hand, absorption of hepta- and octachlorinated dibenzo-p-dioxins and dibenzofurans in fish is

5 1676 Environ. Toxicol. Chem. 23, 2004 P. Isosaari et al. Table 2. Intake rates, accumulation efficiencies, and biomagnification factors (BMF) in exposure groups B, C, and D ab Intake rate (pg/kg/d, 10 3 ) Accumulation (%) BMF PCDD/Fs 2,3,7,8-TCDF 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDF 2,3,7,8-TCDD 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8-HpCDD OCDD Sum WHO PCDD/F -TEQ Dioxin-like PCBs PCB 77 d PCB 81 d PCB 126 d PCB 169 d PCB 105 PCB 114 PCB 118 PCB 123 PCB 156 PCB 157 PCB 167 PCB 189 Sum WHO PCB -TEQ Other PCBs PCB 18 PCB 28/31 PCB 33 PCB 47 PCB 49 PCB 51 PCB 52 PCB 60 PCB 66 PCB 74 PCB 99 PCB 101 PCB 110 PCB 122 PCB 128 PCB 138 PCB 141 PCB 153 PCB 170 PCB 180 PCB 183 PCB 187 PCB 194 PCB 206 PCB 209 Sum (61 67) 58 (54 62) 56 (54 59) 32 (26 38) c 72 (70 75) 64 (59 69) 34 (31 38) 24 (20 28) (41 45) 60 (57 64) 87 (84 91) 97 (90 105) 92 (87 96) 84 (80 88) 76 (73 79) 78 (70 87) 74 (67 80) 79 (75 82) c 83 (80 87) 90 (86 94) 80 (71 90) 52 (45 58) 37 (32 41) 126 ( ) c 153 ( ) c 94 (85 103) 93 (86 100) 80 (69 90) 92 (84 99) 86 (75 98) 96 (86 106) 77 (73 80) 68 (63 74) 83 (80 86) c 84 (82 86) 75 (71 78) 76 (71 81) 84 (81 87) 78 (75 81) 1.10 ( ) 1.05 ( ) 1.03 ( ) 0.73 ( ) c 1.38 ( ) 1.17 ( ) 0.65 ( ) 0.52 ( ) ( ) 1.89 ( ) 1.68 ( ) 1.57 ( ) 1.38 ( ) 1.42 ( ) 1.35 ( ) 1.42 ( ) c 1.57 ( ) c c 2.24 ( ) c 2.83 ( ) c 1.75 ( ) 1.71 ( ) 1.47 ( ) 1.69 ( ) 1.66 ( ) 1.79 ( ) 1.41 ( ) 1.36 ( ) 1.52 ( ) c 1.54 ( ) 1.17 ( ) 1.30 ( ) 1.56 ( ) a For definitions, see Table 1 and Figure 1. Number system for PCBs is according to the International Union of Pure and Applied Chemistry. b Values in parentheses represent 95% confidence intervals. c Coefficient of variation 25%. d Non-ortho PCB.

6 Dioxin accumulation in salmon Environ. Toxicol. Chem. 23, not fully inhibited because of their extreme structural and physicochemical properties, because they were actually among the main congeners in the salmonid fishes of the Great Lakes [18]. The lower accumulation efficiencies of PCDD/Fs as compared with PCBs were expected based on the results of previous studies [19]. It appears that dietary accumulation efficiencies of PCDD/Fs with more than four chlorines rarely exceed 50% [7,17,20]. One reason for this could be low absorption efficiency from the feed; however, only very small amounts of PCDD/Fs were excreted into the feces during the present study. Other reasons could be high elimination rates via the gills into the water or metabolism. Fish are known to metabolize non-2,3,7,8-chlorinated dibenzo-p-dioxins and dibenzofurans and, probably, also some lower-chlorinated dibenzo-p-dioxins and dibenzofurans with the 2,3,7,8-substitution pattern, including 2,3,7,8-TCDD/F, whereas evidence for the metabolism of more highly chlorinated dibenzo-p-dioxins and dibenzofurans is weak [19]. On the other hand, the elimination rates of PCDD/Fs are relatively high as compared with those of other organic compounds with similar log K ow values [21]. One explanation for this could be that a greater portion of PCDD/Fs was excreted into the feces than was actually recovered, possibly because of some redistribution of feces (and PCDD/Fs) into the water. Statistical comparisons between PCB congeners showed that non-ortho PCBs had higher accumulation efficiencies than PCBs with either one (p 0.001) or more (p 0.04) chlorine substituents in the ortho position. The difference in accumulation efficiencies could be a result of the slower elimination rates, as Coristine et al. [22] have reported for non-ortho PCBs 126 and 169. Based on their accumulation, the dioxin-like PCBs can be divided into two groups, one of which has a higher accumulation efficiency than the other. Interestingly, non-ortho PCBs have the most dioxin-like structures and the highest dioxin-like toxicity; however, they differed significantly from PCDD/Fs, which had lower accumulation efficiencies than PCBs. The high tendency of, especially, the toxic PCB congeners to be retained in fish tissues might explain why PCBs have a high contribution to the WHO SUM -TEQ in fish tissues. In aquatic species, non-ortho PCBs typically contribute to the WHO SUM -TEQ by an average of 69% 20%, which is higher than their contribution in terrestrial species [23]. Compared with other PCBs, higher accumulation efficiencies (or biomagnification rates) have been reported for non-ortho PCBs in the Baltic Sea fish and other wildlife [24] and for dioxin-like PCBs in the Lake Michigan (USA) food web [25]. Chlorination degree also had some impact on the accumulation of PCBs. Accumulation efficiencies of tetrachlorinated biphenyls were the highest and those of trichlorinated biphenyls the lowest, whereas PCBs with five to eight chlorines did not differ significantly from each other. Within the different chlorination groups, non-ortho PCBs have the highest log K ow values [26], which suggests that the chlorination degree and the number of ortho substituents could have combined effects on accumulation. Combined effects did appear in univariate tests, but the trends were inconsistent because mono-ortho PCBs only represented penta- and hexachlorinated biphenyls. The preferential accumulation of tetrachlorinated biphenyls is not supported by previous studies. Correlation between log K ow and bioaccumulation of several hydrophobic compounds shows that the highest bioconcentration factors of nonionic compounds and the longest half-lives of several hydrophobic organochlorine compounds occur at log K ow values of approximately seven [27,28]. The log K ow values closest to seven are found among hexa- and heptachlorinated biphenyls [26]. In practice, some studies have reported preferential food-chain accumulation of penta- to heptachlorinated biphenyls in fish [29,30], whereas the longest half-lives measured in other studies have been attributed to hepta- and octachlorination patterns [8,27]. Some statistically insignificant trends in bioaccumulation of PCB congeners that are related to chlorination degree, log K ow, toxicity, or non-ortho structure have also been presented in other studies [19,31]. In addition to the previously discussed factors that might influence the absorption and accumulation of PCBs in fish, one important and improperly known factor is the mechanism that transports hydrophobic organic compounds from the diet into the fish tissues. The mechanisms proposed are diffusive (fugacity based) [32] and mediated/active uptake [33]. Mediated transport of hydrophobic compounds could take place via lipid or protein carrier molecules, and this process favors compounds with a molecular weight of approximately 450, such as hepta- and octachlorinated dibenzo-p-dioxins and dibenzofurans and nona- and decachlorinated biphenyls. However, our investigation did not lend support to the hypothesis that mediated transport would be a significant transport mechanism for PCDD/Fs and PCBs. Finally, the congener-specific differences in accumulation were investigated with a reference to their metabolism, which is favored in the presence of vicinal H-atoms [34,35]. According to the structures of the PCBs measured in the present study, fish should have the best metabolic capacity toward PCBs 18, 33, 110, and 123 and no metabolic capacity toward PCBs 153, 169, 180, 183, 187, and 194. Metabolism seemed to diminish the accumulation efficiency, but the statistical significance of this finding is questionable because of the coinfluence of chlorination degree on the accumulation of these few congeners. Judged by the high accumulation efficiencies, the metabolic capacity of salmon in general must have been low. Biomagnification factors Congener-specific differences in biomagnification factor were similar to the trends observed in accumulation efficiencies (Table 2). This was evident because both parameters were based on the same growth-corrected, whole-fish concentrations, and no change in the lipid content of the fish occurred during the trial. Minor biomagnification ( 1) was shown for tetra- and pentachlorinated dibenzofurans and dibenzo-p-dioxins, but not for the higher-chlorinated congeners. Biomagnification factors were higher for PCBs than for PCDD/Fs, and all the measured PCB congeners biomagnified. The lipid-based concentrations of individual PCDD/F and PCB congeners seemed to stabilize between 15 and 30 weeks. Decreasing accumulation rate of the pollutants was not caused by an increasing growth rate, as evidenced by the growthcorrected concentrations and an unchanging ratio of growth and feeding rates. Hence, a steady state equilibrium was assumed, as required for the calculation of biomagnification factors. CONCLUSION Significant differences were found between the dietary accumulation efficiencies of 2,3,7,8-substituted PCDD/Fs, nonortho PCBs, and other PCBs in farmed Atlantic salmon. Ac-

7 1678 Environ. Toxicol. Chem. 23, 2004 P. Isosaari et al. cumulation efficiencies of tetra- and pentachlorinated dibenzop-dioxins and dibenzofurans were much higher than those of hepta- and octachlorinated dibenzo-p-dioxins and dibenzofurans. Among PCBs, non-ortho substitution pattern and tetrachlorination indicated preferential accumulation as compared with the other PCBs. Considering that PCBs, especially the most toxic congeners, accumulated more efficiently than PCDD/Fs in fish, contamination of fish feed or aquatic environments with PCBs could pose a risk of enhanced human exposure to dioxin-like compounds. AcknowledgementWe acknowledge the valuable technical assistance of the personnel at the Laboratory of Chemistry at the National Public Health Institute, the National Institute of Nutrition and Seafood Research, EWOS Innovation, and Nutreco Aquaculture Research Centre. This paper is a contribution to the European Union (EU) CRAFT Project Q5CR , jointly sponsored by the EU Commission and the International Fishmeal and Fish Oil Organisation and some of its members. REFERENCES 1. Van den Berg M, Birnbaum L, Bosveld ATC, Brunström B, Cook P, Feeley M, Giesy JP, Hanberg A, Hasegawa, R, Kennedy SW, Kubiak T, Larsen JC, van Leeuwen FXR, Liem AKD, Nolt C, Peterson RE, Poellinger L, Safe S, Schrenk D, Tilitt D, Tysklind M, Younes M, Wærn F, Zacharewski T Toxic equivalency factors (TEFs) for PCBs, PCDDs, and PCDFs for humans and wildlife. Environ Health Perspect 106: Van Leeuwen FXR, Younes M WHO revises the tolerable daily intake (TDI) for dioxins. Organohalogen Compounds 38: European Commission, Health & Consumer Protection Directorate-General, Scientific Committee on Food Opinion of the Scientific Committee on Food on the risk assessment of dioxins and dioxin-like PCBs in food, update based on new scientific information available since the adoption of the SCF opinion of 22nd November CS/CNTM/DIOXIN/20 final. Brussels, Belgium. 4. The Council of the European Union Council Regulation (EC) No 2375/2001 of 29 November 2001 amending Commission Regulation (EC) No 466/2001 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Communities L 321: The Council of the European Union Council Directive 2001/102/EC of 27 November 2001 amending Directive 1999/ 29/EC on the undesirable substances and products in animal nutrition. Official Journal of the European Communities L 6: Easton MDL, Luszniak D, Von der Geest E Preliminary examination of contaminant loadings in farmed salmon, wild salmon, and commercial salmon feed. Chemosphere 46: Isosaari P, Vartiainen T, Hallikainen A, Ruohonen K Feeding trial on rainbow trout: Comparison of dry fish feed and Baltic herring as a source of PCDD/Fs and PCBs. Chemosphere 48: Jacobs MN, Covaci A, Schepens P Investigation of selected persistent organic pollutants in farmed Atlantic salmon (Salmo salar), salmon aquaculture feed, and fish oil components of the feed. Environ Sci Technol 36: Kiviranta H, Vartiainen T, Tuomisto J Polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls in fishermen in Finland. Environ Health Perspect 110: Jones PD, Kannan K, Newsted JL, Tillitt DE, Williams LL, Giesy JP Accumulation of 2,3,7,8-tetrachlorodibenzo-p-dioxin by rainbow trout (Onchorhynchus mykiss) at environmentally relevant dietary concentrations. Environ Toxicol Chem 20: Wu WZ, Schramm KW, Xu Y, Kettrup A Accumulation and partition of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/F) in the muscle and liver of fish. Chemosphere 43: Giesy JP, Jones PD, Kannan K, Newsted JL, Tillitt DE, Williams LL Effects of chronic dietary exposure to environmentally relevant concentrations of 2,3,7,8-tetrachlorodibenzo-p-dioxin on survival, growth, reproduction and biochemical responses of female rainbow trout (Oncorhynchus mykiss). Aquat Toxicol 59: International Union of Pure and Applied Chemistry Collaborative study on novel and conventional analytical techniques for the determination of toxicologically relevant PCB congeners in fish and human adipose tissuepreliminary Report of the Evaluation Meeting, April Project 650/90/97: IUPAC/CFC/ WG-HHEC. Barcelona, Spain. 14. Lindström G, Småstuen Haug L, Nicolaysen T International intercalibration on dioxin in food Final Report 2000:9. Folkehelsa, Oslo, Norway. 15. Becher G, Nicolaysen T, Thomsen C Interlaboratory comparison on dioxins in food Final Report 2001:4. Folkehelsa, Oslo, Norway. 16. Småstuen Haug L, Nicolaysen T, Thomsen C, Frøshaug M, Becher G Interlaboratory comparison on dioxins in food Report 2002:4. Norwegian Institute of Public Health, Oslo, Norway. 17. Loonen H, Parsons JR, Govers HAJ Dietary accumulation of PCDDs and PCDFs in guppies. Chemosphere 23: Giesy JP, Kannan K, Kubitz JA, Williams LL, Zabik MJ Polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) in muscle and eggs of salmonid fishes from the Great Lakes. Arch Environ Contam Toxicol 36: Niimi AJ Evaluation of PCBs and PCDD/Fs retention by aquatic organisms. Sci Total Environ 192: Muir DCG, Yarechewski AL Dietary accumulation of four chlorinated dioxin congeners by rainbow trout and fathead minnows. Environ Toxicol Chem 7: Opperhuizen A, Sijm DTHM Bioaccumulation and biotransformation of polychlorinated dibenzo-p-dioxins and dibenzofurans in fish. Environ Toxicol Chem 9: Coristine S, Haffner GD, Ciborowski JJH, Lazar R, Nanni ME, Metcalfe CD Elimination rates of selected di-ortho-, monoortho-, and non-ortho-substituted polychlorinated biphenyls in rainbow trout (Oncorhynchus mykiss). Environ Toxicol Chem 15: Focant J-F, Eppe G, Pirard C, Massart A-C, André J-E, De Pauw E Levels and congener distributions of PCDDs, PCDFs, and non-ortho PCBs in Belgian foodstuffs. Assessment of dietary intake. Chemosphere 48: Paasivirta J Bioaccumulation of POPs in Baltic Sea wildlife as a possible source of ecological damages. Toxicol Environ Chem 66: Trowbridge AG, Swackhamer DL Preferential biomagnification of aryl hydrocarbon hydroxylase-inducing polychlorinated biphenyl congeners in the Lake Michigan, USA, lower food web. Environ Toxicol Chem 21: Hawker DW, Connell DW Octanol-water partition coefficients of polychlorinated biphenyl congeners. Environ Sci Technol 22: Fisk AT, Norstrom RJ, Cymbalisty CD, Muir DCG Dietary accumulation and depuration of hydrophobic organochlorines: Bioaccumulation parameters and their relationship with the octanol/water partition coefficient. Environ Toxicol Chem 17: Meylan WM, Howard PH, Boethling RS, Aronson D, Printup H, Gouchie S Improved method for estimating bioconcentration/bioaccumulation factor from octanol/water partition coefficient. Environ Toxicol Chem 18: Willman EJ, Manchester-Neesvig JB, Armstrong DE Influence of ortho-substitution on patterns of PCB accumulation in sediment, plankton, and fish in a freshwater estuary. Environ Sci Technol 31: Kiviranta H, Vartiainen T, Parmanne R, Hallikainen A, Koistinen J PCDD/Fs and PCBs in Baltic herring during the 1990s. Chemosphere 50: Jackson L, Carpenter SR, Manchester-Neeswig J, Stow CA PCB congeners in Lake Michigan coho (Oncorhynchus kisutch) and chinook (Oncorhynchus tshawytscha) salmon. Environ Sci Technol 35: Gobas FAPC, Muir DCG, Mackay D Dynamics of dietary bioaccumulation and fecal elimination of hydrophobic organic chemicals in fish. Chemosphere 17: Burreau S, Axelman J, Broman D, Jakobsson E Dietary

8 Dioxin accumulation in salmon Environ. Toxicol. Chem. 23, uptake in pike (Esox lucius) of some polychlorinated biphenyls, polychlorinated naphthalenes, and polybrominated diphenyl ethers administered in natural diet. Environ Toxicol Chem 16: Kannan N, Reusch THB, Schulz-Bull DE, Petrick G, Duinker JC ChlorobiphenylsModel compounds for metabolism in food-chain organisms and their potential use as ecotoxicological stress indicators by application of the metabolic slope concept. Environ Sci Technol 29: Looser R, Ballschmiter K Biomagnification of polychlorinated biphenyls (PCBs) in freshwater fish. Fresenius J Anal Chem 360: