Fate of Chlorinated Fatty Acids in Migrating Sockeye Salmon and

Sep 18, 2004 - spawning lake of the salmon, contained higher concentrations of chlorinated fatty acids than grayling in a lake without migratory salmo...
0 downloads 0 Views 125KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 5548-5554

Fate of Chlorinated Fatty Acids in Migrating Sockeye Salmon and Their Transfer to Arctic Grayling HUILING MU,† GO ¨ R A N E W A L D , * ,‡ EINAR NILSSON,‡ PETER SUNDIN,§ AND C L A S W E S EÄ N ‡ BioCentrum-DTU, Center for Advanced Food Studies, Technical University of Denmark, Building 224, DK-2800 Lyngby, Denmark, Center for Chemistry and Chemical Engineering, Lund University, P.O. Box 124, SE-221 00 Lund, Sweden, and Department of Environmental Assessment, Swedish University of Agricultural Sciences, P.O. Box 7050, SE-750 07 Uppsala, Sweden

To investigate whether biotransport constitutes an entry route into pristine ecosystems for nonpersistent, nonvolatile xenobiotic compounds, extractable organically bound halogen in sockeye salmon (Oncorhynchus nerka) from Alaska was determined before and after spawning migration. The major organohalogen compounds in the salmon were halogenated fatty acids, predominantly chlorinated species that accounted for up to 35% of the extractable, organically bound chlorine (EOCl) in the fish tissues. The amount of chlorinated fatty acids in the salmon muscle decreased as a result of spawning migration. The decrease was correlated with that of triacylglycerols in the salmon muscle, indicating the chlorinated fatty acids to be mobilized and metabolized to approximately the same extent as the other fatty acids. Chlorinated fatty acids were also transferred to the maturing roe in a manner similar to that of the unchlorinated fatty acids. Lipids of the Arctic grayling (Thymallus arcticus), a fish resident to the spawning lake of the salmon, contained higher concentrations of chlorinated fatty acids than grayling in a lake without migratory salmon. This may reflect a food-chain transfer of the chlorinated fatty acids originating from the salmon, demonstrating a long-range transport route for this type of pollutants to pristine areas.

Introduction Organohalogen compounds are justifiably causing a great deal of public concern, due to their presence in our food, especially in fish and other seafood. The environmental distribution, transport, and bioaccumulation of persistent organic pollutants (POPs) as PCBs and DDT have been studied extensively, even though these well-known xenobiotics normally constitute less than 5% of the extractable, organically bound chlorine (EOCl) in fish (1-5). In studies of marine samples from Scandinavian waters, efforts were made to characterize the other organochlorine compounds, various chlorinated fatty acids (Cl-FAs) being found (6-12). As much as 90% of the EOCl present in fish and 25% of that present * Corresponding author phone: +46 46 2223878; fax: +46 46 149156; e-mail: [email protected]. † Technical University of Denmark. ‡ Lund University. § Swedish University of Agricultural Sciences. 5548

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004

in bivalves could be accounted for by Cl-FAs (8, 9). Since Cl-FAs have been found not only in fish from the receiving waters of chlorine-bleached-pulp mills (6, 7, 11) but also in fish from areas not influenced by industrial or other obvious sources, it can be asked whether Cl-FAs can also be of natural origin as well as whether these nonpersistent compounds can be subject to long-range transport. The toxicology of Cl-FAs is far from being adequately elucidated, and Cl-FAs have been shown to produce adverse effects, particularly in tests of toxicity involving reproductionrelated endpoints (4, 13-16). Fish probably fail to recognize Cl-FAs as xenobiotic compounds, as can be seen in the failure to activate the P-450 enzyme system of salmonid fishes (4, 17), and in 14C from labeled Cl-FAs being incorporated into the phospholipids and triacylglycerols (TAGs) of perch (Perca fluviatilis) to approximately the same extent as 14C from labeled stearic and oleic acids (18). Cl-FAs and other fatty acids are not resistant to concentrated sulfuric acid (4, 8), which normally is considered to be a requirement for a xenobiotic compound to be classified as persistent. Thus current theories of the bioaccumulation and excretion of POPs (19) are not applicable to chlorinated lipids. During migration, the feeding of anadromous fish (which spend most of their life-cycle in the sea but spawn in freshwater) ceases, their muscle lipids being converted to energy or being transferred to the maturing gonads. The latter process is especially pronounced in females, although no information concerning the possible concurrent mobilization and metabolism of halogenated lipids is available. After spawning, sockeye salmon die, their carcasses constituting an energy resource for the lake ecosystem involved. Juvenile salmon as well as other organisms in the spawning lakes are promoted both by the general fertilization of the lake this produces and by direct feeding on the roe and carcasses of the salmon (20, 21). The ”movement of biomass” which the salmon migration constitutes also brings POPs, having been accumulated at the ocean feeding-grounds, to their spawning lake ecosystem. The POPs are efficiently accumulated in the food chain (22) as well as being accumulated in the sediments (23). As a significant part of the salmon biomass is directly consumed by various organisms e.g. insects and fish, we hypothesize that also nonpersistent compounds such as Cl-FAs may be efficiently accumulated by the local food-chain. This possible transfer of chlorinated lipids from the salmon carcasses to other fish was studied by determining the content of chlorinated lipids found in a particular resident fish there, the Arctic grayling, Thymallus arcticus, both in the salmon-spawning lake and in a nearby lake without migratory fish.

Experimental Section Fish Sampling and Sample Preparation. The sockeye salmon (females) and the grayling were obtained in Alaska (U.S.A.) in 1994 (Table 1) and analyses were performed in 19941995. The salmon were collected both in the Gulf of Alaska, before entering Copper River, and in Summit Lake after they had completed the 400 km ascent to the lake (Figure 1), by which time they had reached maturity and were ready to spawn. The grayling were collected in Lower Fish Lake (LFL, situated 2 km upstream from Summit Lake), where the salmon spawn and die. Grayling were also collected in Round Tangle Lake (RTL), which had no connection with LFL and did not host any anadromous fish. Muscle and roe of the salmon were excised. The bodies of the grayling were homogenized individually, subsamples being collected. These samples were stored for up to six weeks at -30 °C. Following 10.1021/es048744q CCC: $27.50

 2004 American Chemical Society Published on Web 09/18/2004

TABLE 1. Description of the Fish Samples Studied, Obtained from Alaska sample

size (kg)

length (mm)

salmon (S1) salmon (S2) salmon (S3) average ( SD salmon (S4) salmon (S5) salmon (S6) average ( SD grayling (G1) grayling (G2) grayling (G3) grayling (G4) average ( SD grayling (G5) grayling (G6) grayling (G7) grayling (G8) average ( SD

2.9 2.7 2.4 2.7 ( 0.3 2.4 2.7 1.8 2.3 ( 0.5 0.37 0.19 0.39 0.45 0.35 ( 0.11 0.32 0.25 0.26 0.16 0.25 ( 0.07

641 642 622 635 ( 11 632 662 595 630 ( 34 342 260 352 370 331 ( 49 318 305 315 285 306 ( 15

homogenization, lipids were extracted by chloroform and methanol (24) where the lipid extraction efficiency was >95%. The solvents were evaporated under nitrogen. Prior to lipid weight determination the extracts were freeze-dried. Neutral lipids and polar lipids were separated on a silica gel column (silica gel 60, extra pure, 70-230 mesh, Merck, Germany) by using acetone and methanol, respectively (25). The solvents were then removed in a rotary evaporator until about 10 mL remained. The remaining solvent was removed under nitrogen gas, the lipid sample being freeze-dried for dry mass determination. A blank was prepared as above, but without lipids being added. The neutral and polar lipid fractions were examined by thin-layer chromatography (TLC). About 10 µg dry mass of each sample was applied to the TLC plate (silica gel 60 HPTLC aluminum plates, Merck, Germany) using a Camag TLC Spotter. Two different elution media were employed for the separations. The neutral lipids were eluted with n-hexane/diethyl ether/acetic acid (80:20:1, v/v/v), and the polar lipids were eluted with chloroform/methanol/acetic acid (80:20:1, v/v/v). After elution, the TLC plates were sprayed with 10% sulfuric acid and heated for visualization of the organic material. TLC confirmed that no phospholipids were found in the neutral lipids and that no TAGs were found in the polar lipids (data not shown). A minor portion of free fatty acids was detected in the neutral muscle lipids of the mature salmon (that were ready to spawn), which may have been formed by the mobilization of the lipid supplies during the spawning migration. Gas Chromatography (GC). The lipids were transesterified to fatty acid methyl esters (FAMEs), using sulfuric acid as a catalyst (8), and the halogenated FAMEs were enriched by the removal of polyunsaturated FAMEs and of straight-chain FAMEs, representing their silver ion complexes and urea complexes, respectively, the standard deviation of the method is less than 3% (11). The FAME composition was studied by GC, using halogen selective detection, ELCD (8, 11). Part of the fish FAMEs were transesterified to fatty acid propyl esters (FAPEs) and were studied by GC to confirm the fatty acid character of unidentified compounds in the ELCD gas chromatograms (11). Fatty acid-bound chlorine (FACl) was quantified in relation to the 1-chlorooctadecane which was added to the FAMEs that remained after the enrichment procedure. A Varian GC model 3700 was equipped with a Varian 1075 split/splitless injector and a fused-silica capillary column coated with a polar stationary phase (DB23, 30 m × 0.53 mm i.d., film thickness 1.5 µm, J & W Scientific, U.S.A.). The samples were injected at 260 °C, 2 min of splitless conditions

sample site Mouth of Copper River Mouth of Copper River Mouth of Copper River Summit Lake (2 km downstream from Lower Fish Lake) Summit Lake (2 km downstream from Lower Fish Lake Summit Lake (2 km downstream from Lower Fish Lake Lower Fish Lake (spawning lake) Lower Fish Lake (spawning lake) Lower Fish Lake (spawning lake) Lower Fish Lake (spawning lake) Round Tangle Lake (without migratory salmon) Round Tangle Lake (without migratory salmon) Round Tangle Lake (without migratory salmon) Round Tangle Lake (without migratory salmon)

being employed. Helium at a flow rate of 10 mL/min served as the carrier gas. A flame ionization detector (FID) in parallel with an ELCD (Tracor/Varian, model 1000), used in the halogen mode, was operated at base temperatures of 280 °C. The ELCD reactor was at a temperature of 850 °C, hydrogen at a flow rate of 50 mL/min serving as reaction gas, the flow rate of the n-propanol being 0.5 mL/min. The oven temperature of the column was programmed for 90 °C (3 min) followed by a 90-230 °C rise (4 °C/min) (8, 11). Neutron Activation Analysis. The total lipids, as extracted from the salmon muscle, the salmon roe, and the grayling, respectively, were dissolved in cyclohexane and were washed twice with 3 mL of water (pH ) 2, sulfuric acid). The organic solvent was evaporated then after drying by sodium sulfate. The residues were redissolved in cyclohexane to a concentration of 10 mg/mL. The concentrations of EOCl, extractable organically bound bromine (EOBr), and iodine (EOI) were determined by NAA at the Institute for Energy Technology (Kjeller, Norway). The relative standard deviation of the method is about 10%. Extractable organically bound halogen (EOX) was defined as the sum of EOCl, EOBr, and EOI. Statistics. Log-transformed concentration data was used in the statistical comparisons, because of the skewed distributions of the values (26). Descriptive statistics, unpaired t-tests, and simple regressions were performed using the software StatView for Macintosh.

Results and Discussion Composition of Salmon Lipids. Before the salmon entered the river, the total lipid content in the muscle tissue was in the range of 5-10% on a fresh-weight basis. Since the salmon cease feeding once they enter the river (27), lipids are consumed during ascent of the river, resulting in a lipid content of only 2-3% in the sexually mature salmon (Table 2). TAGs are mainly utilized here, being depleted for energy and the gonad growth, resulting in a 72% reduction in the amount of muscle TAGs during migration (Table 3). The phospholipids in the muscle tissue are consumed to a lesser extent, leading to an increase in the proportion of phospholipids, relative to TAGs from 10 to 18%. The content of total lipids in the salmon roe was about 10% of the fresh-weight and did not change as the gonads grew (Table 2). As the roe biomass increased, however, the amount of roe lipids likewise increased, more than doubling by the end of migration (Table 3). Phospholipids constituted 25-30% of total lipids in each of the roe samples. Organohalogen Compounds in Salmon. The ratios of the salmon EOCl to EOBr concentrations (Table 2) were VOL. 38, NO. 21, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5549

FIGURE 1. Map showing the 400 km migration route of the Copper River sockeye salmon (Oncorhynchus nerka) studied, from the Gulf of Alaska to one of its spawning lakes, Lower Fish Lake. The map also shows the reference lake, Round Tangle Lake, and the oceanic currents. similar to those found previously in fish lipids (2, 4). This indicates lipid extraction by chloroform/methanol, when chloroform and chloride ions have been thoroughly removed from the lipid residue, to be compatible with the determination of EOX by NAA. On a fresh weight basis, the salmon roe had a higher concentration of EOX than the salmon muscle (Table 2). The concentration of EOX in the salmon muscle, decreased during migration. The concentration of EOX in the muscles and the roe was positively related to the tissue lipid content (linear regression, r2 ) 0.61, p < 0.01, n ) 12). Following the esterification of fish lipids to FAMEs and the selective removal of polyunsaturated FAMEs and satu5550

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004

rated straight-chain FAMEs, the ELCD chromatograms (selective detection of halogens) of almost all the salmon samples were similar in profile, although the peak levels of the individual compounds varied. After the transesterification of FAMEs to the corresponding propyl esters, the fatty acid character of the compounds detected could be confirmed, the GC/ELCD traces obtained being similar but the retention times being longer. In each of the salmon samples, EOCl dominated over EOBr and EOI by a factor of at least 4.4 and 7.5, respectively (the average factors being 14 and 47) (Table 2). This suggests the ELCD response, which was quantified in terms of organically bound chlorine, to mainly involve the chlorinated

TABLE 2. Tissue Lipid Content (% of Fresh Weight) for the Individual Salmon and for Subsamples of Individually Homogenized Arctic Grayling and Concentrations of Extractable Organically Bound Chlorine, Bromine, and Iodine (EOCl, EOBr, and EOI), Determined by NAA in Different Fish Samples, and of the Chlorine (Determined by GC/ELCD) Associated with Chlorinated Fatty Acids (Cl in Cl-FAs), in µg/g Fresh Weighta sample

tissue

lipid content (% of fresh weight)

Cl in Cl-FAs (µg Cl/g fresh weight) 0.4 1.2 0.7 0.4 ( 0.3 0.6 1.1 1.2 1.0 ( 0.3

Salmon after Spawning Migration (Summit Lake) 2.4 2.4 0.1 2.9 3.0 0.2 2.5 2.0 0.1 2.6 ( 0.3 2.5 ( 0.5 0.1 ( 0.1 9.9 9.1 1.0 11.4 7.1 1.6 9.2 4.8 0.9 10.2 ( 1.1 7.0 ( 2.2 1.2 ( 0.4

0.2 0.1 0.3 0.2 ( 0.1 0.1 0.1 0.2 0.1 ( 0.1

0.2 0.3 0.3 0.3 ( 0.1 0.8 1.1 0.8 0.6 ( 0.4

Arctic Grayling from Lower Fish Lake 6.3 nd 8.0 nd na nd 3.5 nd 4.5 ( 3.5

nd nd nd nd

1.0 2.7 0.8 0.8 1.3 ( 0.9

Arctic Grayling from Round Tangle Lake 2.4 nd 5.5 nd 2.7 nd 1.9 nd 3.1 ( 1.6

nd nd nd nd

0.2 0.3 0.1 0.1 0.2 ( 0.1

S4 S5 S6 average ( SD S4 S5 S6 average ( SD

muscle muscle muscle

G1 G2 G3 G4 average ( SD

whole body whole body whole body whole body

5.6 8.7 5.4 3.8 5.9 ( 2.0

G5 G6 G7 G8 average ( SD

whole body whole body whole body whole body

3.8 6.0 3.8 2.3 4.0 ( 1.5

a

EOI (µg/g fresh weight) 0.1 0.4 0.2 0.2 ( 0.2 0.1 0.1 0.1 0.1 ( 0.0

muscle muscle muscle

roe roe roe

EOBr (µg/g fresh weight)

Salmon Prior to Spawning Migration (Copper River Outlet) 4.6 3.8 0.1 9.7 5.0 0.5 6.1 5.6 0.7 6.8 ( 2.6 4.8 ( 0.9 0.4 ( 0.3 9.2 7.5 1.3 9.7 11.8 1.6 9.5 11.5 2.2 9.5 ( 0.3 10.3 ( 2.4 1.7 ( 0.5

S1 S2 S3 average ( SD S1 S2 S3 average ( SD

roe roe roe

EOCl (µg/g fresh weight)

nd ) not detected; na ) not analyzed.

TABLE 3. Proportions of Neutral Lipids and Polar Lipids, the Total Amount of Lipids, and the Chlorine in the Chlorinated Fatty Acids (Cl-FAs) in the Muscle and Roe of Sockeye Salmon (( SD, n ) 3)a total tissue Cl in lipid total tissue amount of Cl Cl-FAs proportion amount of in Cl-FAs (µgCl/g (%) lipids (g) (µgCl) total lipid)

sample/ tissue

lipid class

muscle muscle roe roe

neutral polar neutral polar

Prior to Spawning Migration 92 ( 3 101 ( 40 1131 ( 645 10 ( 4 10 ( 1 48 ( 15 75 ( 3 12 ( 3 122 ( 14 25 ( 4 4(2 29 ( 12

11 ( 2 5(1 11 ( 4 7(1

muscle muscle roe roe

neutral polar neutral polar

After Spawning Migration 77 ( 2 28 ( 8 253 ( 87 18 ( 2 6(2 77 ( 14 70 ( 2 25 ( 7 232 ( 117 31 ( 1 11 ( 4 96 ( 46

9(2 13 ( 6 9(2 8(3

a The figures are calculated from the total individual tissue weight. The concentration of chlorine associated with Cl-FAs (determined by GC/ELCD) is expressed in relation to the lipid class in µg/g total lipid.

compounds. Thus, a considerable portion (8-24%) of the EOCl in the salmon could be accounted for by the Cl-FAs (Table 2). By comparison, the concentration of ∑PCBs and ∑DDT/DDE (22) accounted for 0.1-2.5% of the EOCl altogether. In accordance with the results for NAA, a positive correlation was found between the concentration of chlorine associated with Cl-FAs (on a fresh-weight basis) and the fat content of the muscle and roe of the salmon, an r2 value of

0.84 (p < 0.0001, n ) 12) being obtained. Omitting the roe samples resulted in an even higher r2-value of 0.98 (p < 0.0005, n ) 6) strongly indicative of the chlorinated lipids in the muscle being mobilized along with the unchlorinated ones. Metabolism of Cl-FAs in Salmon. The major part of the Cl-FAs (calculated as chlorine) in the ascending salmon was found in the muscle TAGs (Table 3). After migration, at maturity, this amount had been reduced significantly, by 78% (p < 0.01), a reduction in parity with the total consumption of muscle neutral lipids. During spawning migration, however, the total amount of Cl-FAs in the phospholipids of the salmon muscles and roe increased by a factor of 2 to 3 (Table 3). Thus, the Cl-FAs and the normal, unchlorinated fatty acids in the salmon muscle TAGs were metabolized to about the same extent, although part of the Cl-FAs were incorporated into the phospholipids, where the Cl-FAs tend to accumulate. The relatively constant concentration of chlorine associated with the fatty acids could also be seen in the pattern of GC/ELCD peaks, which tentatively represented the three chlorohydroxy fatty acids C17ClOH, C19ClOH, and C21ClOH (Figure 2). The concentration of C17ClOH was highest in the TAGs, in particular the roe TAGs, whereas the concentration of C21ClOH was highest in the roe phospholipids. The increase in the amounts of Cl-FAs in the roe lipids was similar to the increments of TAGs and phospholipids in the roe, suggesting the chlorinated and the unchlorinated fatty acids to be transferred to a similar extent from the muscles to the roe. Despite the amounts of Cl-FAs in the roe TAGs and the roe phospholipids being more than doubled, however, this increase did not fully account for the amounts mobilized from the muscle lipids. It has been reported that VOL. 38, NO. 21, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5551

FIGURE 2. Concentrations of monochlorohydroxyheptadecanoic acid (C17ClOH), monochlorohydroxynonadecanoic acid (C19ClOH), and monochlorohydroxyheneicosanoic acid (C21ClOH) in triacylglycerols (TAG) and phospholipids (PL) of salmon from the inlet of Copper River and after spawning migration to Summit Lake. Three individuals from each site were examined. chlorinated or brominated fatty acids can be metabolized in rat and that water-soluble metabolites such as chloride or bromide ions can be excreted via the urine (28-30). Our data indicate the fish to have catabolized and excreted some Cl-FAs. Our study shows that, during spawning migration of the salmon, Cl-FAs can be redistributed, without apparent discrimination, between different organs and lipid classes and that despite the chemically unstable character of these compounds, they resist being degraded over the 400 km spawning migration. Transfer of Organohalogen Compounds to Grayling. Higher concentrations of EOCl were found in the Arctic grayling from the salmon spawning lake (LFL) (Figure 1) than in the grayling from a neighboring lake (RTL) (Table 2). Since the concentrations of EOBr and EOI were below the NAA detection limits, the compounds detected by GC/ELCD are likely to have been chlorinated ones. Organochlorine compounds were detected by GC/ELCD in the grayling from both lakes (Figure 3A,B). The transesterification of FAMEs to FAPEs confirmed the GC detectable organochlorine compounds being of fatty acid character (Figure 3D,E). Quantification of the ELCD peaks showed the Cl-FAs to account for 15-35% of the EOCl in the grayling from LFL but for only 2-8% of the EOCl in the grayling from RTL. In both grayling populations, however, the Cl-FAs were prominent organochlorine components as compared with the concentrations of ∑PCBs and ∑DDT/DDE (22). GC/ELCD showed the concentration of Cl-FAs in the LFL grayling to be significantly (p < 0.01) higher than in the RTL grayling. Since both lakes are situated in the same area (Figure 1) and are exposed to a similar load from the surrounding land and atmosphere (22), the relatively high concentration of Cl-FAs in the grayling from LFL is most likely due to the transport of chlorinated lipids to LFL by the salmon. This accords with results concerning the transport by the migrating salmon of persistent pollutants to LFL (22). Chlorinated lipids may accumulate in the grayling either directly through their feeding on the salmon roe (20) or indirectly from the decaying salmon carcasses by way of the food chain. The latter explanation appears the most plausible, 5552

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004

since a stomach content analysis of twelve grayling from LFL showed only one of them to have fed on roe and their principal diet to consist of insects, insect larvae, and zooplankton (31). The GC/ELCD pattern of the FAMEs from the RTL grayling displayed only one major chlorinated component (Figure 3A), whereas a more complex pattern of chlorinated species was found among the FAMEs of the LFL grayling (Figure 3B). Since the salmon carcasses were presumably the source of Cl-FAs, it could have been expected that a similar pattern of chlorinated species would be found in both the salmon and in the LFL grayling. However, only traces of the Cl-FAs that were major constituents of the salmon lipids (Figure 3C) were found in LFL grayling lipids. Although an adequate explanation of this difference can only be obtained when the Cl-FAs in the grayling are identified, the probable uptake route from the salmon carcasses to the grayling is via the lake food-chain, as was discussed above, a major part of the Cl-FAs brought to the ecosystem by the salmon being likely to have been degraded or altered. Possible Origin of the Chlorinated Fatty Acids. Since during their time spent in the ocean sockeye salmon migrate over large areas of the Pacific to the south and east of Alaska (32), the area in which the chlorinated compounds accumulate is difficult to specify. The oceanic current that dominates the feeding grounds of the Copper River salmon is the Alaska current (33), which diverts to the north from the large Subarctic current and passes the Canadian coast before it reaches Alaska (Figure 1). Since several pulp mills are located along the west coast of the United States and Canada and along several of the rivers that empty into the Pacific there (34-36), it is possible that the Cl-FAs in the Copper River salmon are anthropogenic, particularly the dichlorinated fatty acids, which have been found in fish in the receiving waters of pulp mills using chlorine gas as a bleaching agent (8, 10-12). In view of chlorohydroxy fatty acids, which were the major Cl-FAs in the salmon, having also been found in jellyfish in the open sea (37), it is possible that such compounds are naturally produced. The Cl-FAs in the grayling from RTL have no distinct source, since no anadromous fish spawn in the lake and no industrial sources are

ing degradation by e.g. oxidation in the abiotic compartment of the ecosystem, to be efficiently transferred to biota. In a wider perspective, biotransport may be an important process for long-range distribution of compounds today emerging as environmental threats e.g. pharmaceuticals (40) and perfluorochemicals (41) not possessing the classical POP characteristics.

Acknowledgments We thank Nicole Szarzi of The Alaska Department of Fish and Game in Glennallen (presently in Homer) and Thomas Kline, of the University of Alaska, in Fairbanks (presently at the Prince William Sound Science Center in Cordova), for their invaluable help in the planning and conducting of the sampling in Copper River and Anders Karlsson of the Department of Ecology of Lund University, presently at Cureon A/S, Denmark, for providing reference compounds for TLC. Bruce McKague, University of Toronto, kindly commented on the mass spectrometric characterization of chlorohydrins. Financial support from the Swedish Environmental Protection Agency, the Swedish Council for Forestry and Agricultural Research and the Carl Trygger Foundation is gratefully acknowledged.

Supporting Information Available Mass spectrometric characterization of organohalogen compounds in salmon. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited

FIGURE 3. ELCD gas chromatograms of chlorinated FAMEs released from (A) total lipids of grayling from Round Tangle Lake, (B) total lipids of grayling from Lower Fish Lake, and (C) phospholipids of mature salmon muscle. Traces (D) and (E) represent chlorinated FAPEs corresponding to traces “A” and “B”, respectively. i.s. ) internal standard, 1-chlorooctadecane. (*) ) peaks not representing fatty acids. found in the vicinity of it. It has been reported, however, that dichlorostearic acid can be formed as a UV-catalyzed product following mixing of oleic acid and either DDT or methoxychlor (38). Since PCBs and DDTs are ubiqutous pollutants and unsaturated fatty acids are components common to living organisms, the Cl-FAs in the grayling from RTL may be rearrangement products of unsaturated fatty acids and chlorinated pollutants, both of which are present. Natural chlorination by chloroperoxidases (39) could possibly also lead to the formation of Cl-FAs followed by its uptake in fish. Environmental Significance. By ways of biotransport migrating salmon significantly influences the POP levels in the sediments and in the food-chain in their spawning lakes. One crucial aspect of biotransport is that the pollutants may be transferred into the local food-chain by direct consumption of salmon biomass without residence in any abiotic compartment prior potential bioaccumulation. This makes biotransported pollutants being highly bioavailable and thus being bioaccumulated to a much higher degree than pollutants reaching the ecosystem e.g. via atmospheric deposition. Our study demonstrates that biotransport is a process allowing pollutants neither having the persistence nor the volatility to be transported in the atmosphere to reach remote ecosystems. Biotransport also allows compounds, not resist-

(1) Grimvall, A. In Naturally-Produced Organohalogens; Grimvall, A., de Leer, E. W. B., Eds.; Kluwer Academic Publishers: Dordrecht, 1995; pp 3-20. (2) Lunde, G.; Gether, J.; Steinnes, E. AMBIO 1976, 5, 180-182. (3) So¨dergren, A.; Bengtsson, B. E.; Jonsson, P.; Lagergren, S.; Larsson, A° .; Olsson, M.; Renberg, L. Water Sci. Technol. 1988, 20, 49-60. (4) Håkansson, H.; Sundin, P.; Andersson, T.; Brunstro¨m, B.; Dencker, L.; Engwall, M.; Ewald, G.; Gilek, M.; Holm, G.; Honkasalo, S.; Idestam-Almquist, J.; Jonsson, P.; Kautsky, N.; Lundberg, G.; Lund Kvernheim, A.; Martinsen, K.; Norrgren, L.; Pesonen, M.; Rundgren, M.; Stålberg, M.; Tarkpea, M.; Wese´n, C. Pharmacol. Toxicol. 1991, 69, 459-471. (5) Newsome, W. H.; Andrews, P.; Conacher, H. B. S.; Rao, R. R.; Chatt, A. J. AOAC Int. 1993, 76, 703-706. (6) Wese´n, C.; Carlberg, G. E.; Martinsen, K. AMBIO 1990, 19, 3638. (7) Wese´n, C.; Martinsen, K.; Carlberg, G.; Mu, H. In Environmental Fate and Effects of Bleached Pulp Mill Effluents; So¨dergren, A., Ed.; Swed. Environ. Prot. Agency Report 4031; 1992; pp 207219. (8) Wese´n, C.; Mu, H.; Lund, K. A.; Larsson, P. J. Chromatogr. 1992, 625, 257-269. (9) Wese´n, C.; Mu, H.; Sundin, P.; Ringstad, O.; Odham, G. In Naturally-Produced Organohalogens; Grimvall, A., de Leer, E. W. B., Eds.; Kluwer Academic Publishers: Dordrecht, 1995; pp 307-316. (10) Wese´n, C.; Mu, H.; Sundin, P.; Frøyen, P.; Skramstad, J.; Odham, G. J. Mass Spectrom. 1995, 30, 959-968. (11) Mu, H.; Wese´n, C.; Nova´k, T.; Sundin, P.; Skramstad, J.; Odham, G. J. Chromatogr. 1996, 731, 225-236. (12) Mu, H.; Wese´n, C.; Sundin, P.; Nilsson, E. J. Mass Spectrom. 1996, 31, 517-526. (13) Cunningham, H. M. In Water Chlorination: Environmental Impact and Health Effects; Jolley, R. L., Brungs, W. A., Cumming, R. B., Jacobs, V. A. Eds.; Ann Arbor Science: Ann Arbor, 1980; Vol. 3, pp 995-1005. (14) Jones, B. A.; Tinsley, I. J.; Lowry, R. R. Lipids 1983, 18, 319-326. (15) Ewald, G.; Sundin, P. Pharmacol. Toxicol. 1993, 73, 159-162. (16) Cherr, G. N.; Shenker, J. M.; Lundmark, C.; Turner, K. O. Environ. Toxicol. Chem. 1987, 6, 561-569. (17) Goksøyr, A.; Larsen, H. E. Effekter av klorerte fettsyrer og fettekstrakter på cytokrom P450-systemet hos laks; Report, Marine molecular biology laboratory, Bergen, Norway, 1993. (In Norwegian) VOL. 38, NO. 21, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5553

(18) Ewald, G.; Sundin, P.; Skramstad, J.; Frøyen, P. Aquat. Toxicol. 1996, 36, 129-139. (19) Connell, D. W. Bioaccumulation of xenobiotic compounds; CRN Press: Boca Raton, Florida, 1990. (20) Skopets, M. B.; Prokopyev, N. M. Voprosy Ikhtiologii 1990, 30 (4), 564-576. (21) Kline, T. C., Jr.; Goering, J. J.; Mathisen, O. A.; Poe, P. H.; Papker, P. L.; Scalan, R. S. Can. J. Fish. Aquat Sci. 1993, 50 (11), 23502365. (22) Ewald, G.; Larsson, P.; Linge, H.; Okla, L.; Szarzi, N. Arctic 1998, 51, 40-47. (23) Kru ¨ mmel, E. M.; Macdonald, R. W.; Kimpe, L. E.; Gregory-Eaves, I.; Demers, M. J.; Smol, J. P.; Finney, B.; Blais, J. M. Nature 2003, 425, 255-256 (24) Bligh, E. G.; Dyer, W. J. Can. J. Biochem. Physiol. 1959, 37, 911917. (25) Bjo¨rn, H. Do phospholipids from fish contain chlorinated fatty acids? Honour’s Thesis, Dept of Ecology, Lund University, Sweden, 1994. (26) Newton, I. Environ. Pollut. 1988, 55, 29-40. (27) Foerster, R. E. In Bulletin 162; Stevenson, J. C., Ed.; Fisheries Research Board of Canada: Ottawa, 1968. (28) Cunningham, H. M.; Lawrence, G. A. Food. Cosmet. Toxicol. 1976, 14, 283-288. (29) Cunningham, H. M.; Lawrence, G. A. Food. Cosmet. Toxicol. 1977, 15, 101-103. (30) Lawrence, J. F.; Chadha, R. K.; Iverson, F.; McGuire, P.; Conacher, H. B. S. Lipids 1984, 19, 704-707.

5554

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004

(31) Ewald, G. Dept of Ecology, Lund University, unpublished data. (32) Burgner, R. L. In Pacific salmon life histories; Groot, C.; Margolis, L., Eds.; UBC Press: Vancouver, 1991. (33) Thomson, K. A.; Ingraham, W. J., Jr.; Healey, M. C.; LeBlond, P. H.; Groot, C.; Healey, C. G. Can. J. Fish. Aquat. Sci. 1994, 51, 441-449. (34) Macdonald, R. W.; Cretney, W. J.; Crewe, N.; Paton, D. Environ. Sci. Technol. 1992, 26, 1544-1550. (35) Thomas Register, Products & Services Section; Thomas Publ.: New York, U.S.A., 1993; Vol. 11, p 21837. (36) Anonymous. Pulp and paper from Canada: A trade directory; Canadian Pulp and Paper Association: Montreal, Que´bec, 1995. (37) White, R. H. Analyses of organic halogen compounds in marine organisms. Thesis, Dept of Biochemistry, University of Illinois, 1974. (38) Schwack, W. J. Agric. Food Chem. 1988, 36, 645-648. (39) Winterbourn, C. C.; van den Berg, J. J. M.; Roitman, E.; Kuypers, F. A. Arch. Biochem. Biophys. 1992, 296, 547-555. (40) Erickson, B. E. Environ. Sci. Technol. 2002, 141, A-145 A. (41) Giesy, J. P.; Kannan, K. Environ. Sci. Technol. 2002, 146, A-152 A.

Received for review August 11, 2004. Accepted August 13, 2004. ES048744Q