Latitudinal Fractionation of Polychlorinated Biphenyls in Surface

Mar 30, 2004 - ... Figure 3, but this satellite- derived data is unavailable for ice-covered regions. ... We also thank the captain and crew on the I/...
1 downloads 0 Views 119KB Size
Research Latitudinal Fractionation of Polychlorinated Biphenyls in Surface Seawater along a 62° N-89° N Transect from the Southern Norwegian Sea to the North Pole Area ANNA SOBEK AND O ¨ RJAN GUSTAFSSON* Stockholm University, Institute of Applied Environmental Research (ITM), SE-10691 Stockholm, Sweden

Surface seawater concentrations of PCBs, relative congener abundance, and possible effects of cold condensation were studied along a transect from the southern Norwegian Sea to the central Arctic Ocean (62° N-89° N). Large volume samples were collected from an ice breaker using a stainless steel surface seawater intake connected online to an ultra-clean laboratory. Concentrations of all studied PCB congeners, except for trichlorinated PCB 18, decreased with latitude. For instance, PCB 52 decreased from 470 fg L-1 at 62° N to 110 fg L-1 at 89°N and PCB 180 from 110 to 12 fg L-1. Concentrations in the central Arctic Ocean were on the order of 10-100 fg L-1 for the most abundant congeners. The relative contribution of trichlorinated PCBs to the total PCB concentration increased with latitude, the tetrachlorinated contribution to the total PCBs did not show any correlation to latitude, and the relative contribution of heavier congeners decreased with latitude. This study establishes the occurrence at very low abundances of PCBs in seawater in the central Arctic Ocean and demonstrates a northward concentration decrease. The latitudinal shift in congener pattern is reflecting the relative propensity of the PCB congeners to undergo long-range transport in the Arctic and is consistent with their relative vapor pressures.

Introduction Persistent organic pollutants (POPs) such as polychlorinated biphenyls (PCBs) are widely distributed and are today ubiquitous in the environment. Measurements have detected PCBs and many other POPs in remote areas, such as the Arctic, far away from any significant sources (e.g., refs 1-4). The occurrence of POPs in the Arctic has even become a concern for the health of the indigenous people due to the toxicity, persistence, and high bioaccumulation potential of many of these compounds (5). Transport of contaminants to the Arctic occur through the atmosphere, which is recognized as a significant transport route for POPs, river transport, and transport with ocean currents (6). The mobility of compounds in the environment is governed by compound-specific physicochemical properties, such as water solubility, vapor pressure, and partitioning behavior between air, water, sediment, and terrestrial compartments. It has been proposed that long-range at* Corresponding author e-mail: [email protected]; tel.: +46-8-6747317; fax: +46-8-6747638. 2746

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

mospheric transport of POPs occurs by the global fractionation or cold condensation concepts (7, 8). Temperature is here a key parameter as it affects the partitioning equilibrium between different phases. For instance, a decreasing temperature would lead to increased condensation out of the gaseous phase and thus less atmospheric transport. Temperature is generally decreasing with increasing latitude due to a lower angle of the incoming solar irradiation. The subcooled liquid vapor pressure (pL; at 25 °C) for PCBs span from 10-5 to 1 Pa, implying a possibility for a latitudinal fractionation, with increased relative contribution of more volatile congeners with decreasing temperature. Already in 1976, Harvey and Steinhauer (9) measured seawater concentrations of PCBs in the Atlantic Ocean (36° S-54° N) and concluded that codistillation with water and subsequent temperature-dependent condensation play an important role in determining the global distribution of PCBs. Iwata et al. (10) observed a latitudinal fractionation of PCBs in seawater (65° S-75° N), with a higher relative contribution of less chlorinated PCBs at higher latitudes. The same pattern was observed for DDT and its metabolites. Latitudinal congener fractionation of PCBs has also been observed in the atmosphere (50° N-70° N; refs 11 and 12), in soils (13), and in sediments (14). Grimalt et al. (15) observed the same phenomena of increasing relative contribution of more volatile PCB congeners with decreasing temperature along an altitude transect in mountain lakes in Europe. Some compounds, which are more volatile than the PCBs, increase in concentration toward the north, as an effect of cold condensation. This has been observed for the hexachlorocyclohexanes (HCHs) in seawater (10) and hexachlorobenzene (HCB) in air (11). Seawater concentrations of R-HCH in the Arctic are the highest in the world (e.g., refs 10 and 16). For PCBs, despite the extensive work conducted on biological matrixes within the Arctic Monitoring and Assessment Program (AMAP) (6), there is very limited knowledge of seawater concentrations throughout the Arctic Ocean. A global distribution model simulating the accumulation potential of POPs in the Arctic is, however, suggesting PCBs to be effectively transported to the Arctic (17). Therefore, the objectives of the present study were to (a) provide reliable information on the PCB seawater concentration along a transect extending into the high Arctic Ocean, (b) test the hypothesis of efficient transport of PCBs to the Arctic, and (c) investigate the mechanism of transport of PCBs to and within the high Arctic. To this end, data on the individual congener concentration and the latitudinal variation in congener fingerprint are evaluated along a transect from the southern Norwegian Sea to the North Pole area (62° N-89° N).

Materials and Methods Sampling Procedure. Surface seawater samples were collected onboard the ice-breaker ODEN during the Swedish Arctic Ocean expedition (SWEDARCTIC 2001) June to August 2001. Samples were taken along a transect through the Norwegian Sea to Barents Sea to Nansen Basin to Amundsen Basin to Makarov Basin toward the North Pole (62° N-89° N) (Table 1; Figure 1). A stainless steel seawater intake system, specially designed for this expedition, situated at the prow of the ship at a water depth of 8 m was used to take samples online into an ultra-clean laboratory. The ultra-clean laboratory used hepa-filtered incoming air and was set under high pressure to minimize contamination from the air in the surrounding wet laboratory. It was restricted to limited 10.1021/es0353816 CCC: $27.50

 2004 American Chemical Society Published on Web 03/30/2004

TABLE 1. Sampling Positions and General Characteristics of Each Water Sample position start

position end

sample ID

°N

°E

°N

°E

vol (L)

salinitya

tempa (°C)

HSb (µg QSE L-1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

62.10 66.43 71.38 77.49 78.20 78.36 80.27 80.31 81.20 82.30 84.16 85.11 88.57 87.59 87.55

4.56 10.3 1 18.2 4 29.5 3 27.1 7 33.1 1 15.5 4 11.5 7 24.2 4 25.5 8 33.3 9 38.3 2 1.13 °W 69.4 5 154. 18

63.53 68.59 71.57 77.53 78.21 78.38 80.25 80.59 c 81.58 84.20 87.29 88.46 88.23 87.52

6.56 14.0 18.5 9 29.6 8 27.3 0 33.8 15.2 5 20.2 1 c 26.2 0 34.9 0 31.5 3 2.48 °W 95.1 3 154. 44

1079 972 225 1137 701 1136 628 335 1476 2645 823 1270 5103 999 2814

33.20 33.84 34.89 33.48 33.24 33.02 34.18 34.56 32.90 34.27 34.29 34.47 31.43 34.01 31.01

10.8 10.3 9.2 -1.6 -1.5 0.1 3.4 3.6 -1.2 -1.5 -1.3 -1.7 -1.6 -1.7 -1.7

1.25 0.87 0.33 0.28 0.40 0.36 0.31 d 0.09 0.47 d 0.07 3.63-5.67 2.64 6.41

a Salinity and temperature were continuously measured and logged by a thermosalinograph coupled to the seawater intake system. b Fluorescence based determination following Wedborg et al. (34). Fluorescence intensities (350 nm extinction and 450 nm emission) were normalized to quinine sulfate equivalents (units of µg QSE L-1). c Position not available. d Not detected.

min-1, which was registered by a flow meter, was directed through the ultra-clean POP sampling line. The pressure over the filter was constantly monitored with a pressure indicator and never allowed to exceed 1 bar to minimize cell lyzing. During collection of sample 3, the flowmeter stopped logging, and the volume of this sample is therefore estimated from the time of sampling and the flow at start and stop. Particleassociated PCBs were collected on precombusted borosilicate filters (GF/F 293 mm, nominal pore size 0.7 µm; Whatman International Ltd., Maidstone, England), which were followed by polyurethane foam adsorbents (PUF; diameter 37 mm, length 160 mm; Sunde So¨m & Skumplast AS, Norway and punched out by Special-Plast AB, Vallentuna, Sweden) to collect the dissolved PCBs. The PUFs were extensively cleaned prior to the expedition to minimize the blank. The cleaning procedure included 90 °C washing with detergent (1 h), drying at 50 °C (24 h), Soxhlet extraction in toluene (24 h) and in acetone (24 h), and finally drying under vacuum in a desiccator with a PUF scrubber at the gas inlet. The prepared PUF adsorbents were placed in double layers of precombusted Al foil envelops in turn placed in double plastic bags and stored in a freezer (-18 °C) until sampling. Collected samples were placed in the same precombusted Al envelops and stored in double-sealed plastic bags in a freezer (-18 °C) until analysis.

FIGURE 1. Location of sampling sites. Samples 1, 2, 8, 12, and 14 were collected while the ship was moving (start and stop indicated by dashed lines). Maximum and minimum ice extents represent the multi-year average positions (6). personnel to enter the enclosed ultra-clean laboratory and the working procedure included changing to full-body lowlint nylon dresses (including caps) in the intermediate lock. Large volumes of water were collected for each sample (Table 1). The approximate total flow through the seawater intake system was 50 L min-1, whereas approximately 2 L

Extraction and Analysis. Concentrations of 15 PCB congeners were determined in filters and adsorbents (PCB IUPAC numbers 18, 28, 52, 70, 90/101, 110, 118, 105, 149, 153, 138, 180, 199, and 194) following previously described methods (18, 21, 22). Briefly, internal standards in the form of seven 13C-labeled PCB congeners were added to each sample prior to 24-h Soxhlet extraction with toluene (glassdistilled quality; Burdick & Jackson, Fluka Chemie AG, Buchs, Switzerland) using a Dean-Stark trap for collection of water. All extracts were eluted on an open silica column prior to further cleanup and HPLC separation on an amino column (µBondapak NH2, 7.8 × 300 mm; Waters Corporation, Milford, MA). The PCB fraction was thereafter eluted on an open column containing three layers of modified silica (SiO2/H2SO4 10 mm, SiO2/KOH 10 mm, and SiO2/H2O 10 mm). Finally, samples were quantified on a HP6890 (Hewlett-Packard, Avondale, PA) gas chromatograph (GC) equipped with a PTE-5 capillary column (30 m × 0.25 mm i.d., 0.25 µm film thickness; Supelco Inc, Bellefonte, PA) with a high-resolution mass spectrometer (HRMS) (Autospec Ultima; Micromass, Altrincham, UK) operated in the electron impact mode. A 13 C -labeled recovery standard (PCB 153) was added to all samples before injection on the GC-HRMS. VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2747

Quality Assurance. In addition to employing the ultraclean sampling procedure onboard I/B ODEN described previously, the sampling technique has been previously evaluated in field campaigns in the open Baltic Sea (18, 19). Those studies demonstrated that an identical filtration and extraction system to the one used in this Arctic study, operated under the same range of flow rates (1-3 L min-1), and also for sampled seawater volumes above 1000 L, was able to quantitatively collect the PCBs with good efficiency. Specifically, any breakthrough of the less hydrophobic congeners through the PUF adsorbent appears to be negligible with this system, and operating conditions, as both the dissolved concentrations of PCB 18 and PCB 180 in the open Baltic Sea stayed invariant around 0.7 and 0.08 pg L-1, respectively, irrespective of the tested seawater sample volume (range 300-1400 L; ref 19). The Baltic Sea may in this respect be considered a worse case scenario than the Arctic as its higher load of POPs and dissolved organic matter, if anything, would putatively lead to more fouling and thus lower column capacity of the PUF adsorbents. The surface seawater intake system was blank tested, on a pilot expedition, before the SWEDARCTIC 2001 expedition. Water samples were taken with the intake system in parallel with an in situ pump connected to silicone tubing, which was lowered into the water at a depth of 8-10 m on the side of the ship. The in situ pump and silicone tubing system has previously been thoroughly tested and evaluated (18, 19). There was no difference in PCB concentrations between the two sampling methods, supportive of the seawater intake system not contaminating the samples. Parallel samples were also taken during the Arctic expedition with the seawater intake system and Kiel in situ pumps (KISP; ref 20); the latter deployed to approximately the depth of the seawater intake system by another research group. For several stations of parallel sampling in the northern Barents Sea marginal ice zone, good agreement between the two sampling systems was achieved. For instance, the seawater intake system recorded concentrations of PCB 52, which was only 8 ( 26% (n ) 5) lower than for the KISP (personal communication, Peter Ko¨mp and Michael McLachlan, Institute of Baltic Sea Research, Warnemu ¨ nde, Germany, September 25, 2003). In parallel with field samples, field blanks and laboratory blanks were analyzed to control any contamination originating from the described sampling and analytical steps. The amount of PCB 52 in blanks varied between 3 and 11 pg (n ) 15) (field blanks 3-8 pg (n ) 6) and laboratory blanks 3-11 pg (n ) 9)). Blanks were subtracted from the raw data. Obtained PCB concentrations that did not exceed the blank levels three times were excluded. The orthochlorinated PCB 199 and PCB 194 had concentrations close to the blank in all samples and have therefore been excluded from evaluation. Average recoveries of the seven 13C -labeled internal standards for both GF/F filters and PUF adsorbents at all 15 stations were 71 ( 19%. Specifically, the recoveries for the GF/Fs were 74 ( 7% and for the PUFs 68 ( 3%, whereas for the most volatile internal standard (13C-PCB 28) the recoveries were for the GF/Fs 71 ( 15% and for the PUFs 65 ( 15%. The reported concentrations are corrected for the individual sample recoveries.

Results and Discussion Sampled Water Masses. Basic information about the sampling sites and sampled volumes are given in Table 1. Seawater sampling temperature varied from 10.8 to -1.7 °C. The salinity generally varied between 33.2 and 34.9 throughout the transect but decreased to 31.0 and 31.4 for samples 13 and 15 in the central Arctic Ocean. These two samples and sample 14 also had the highest content of humic substances (Table 1), indicating a freshwater influence, probably from the Lena River, which is known to affect the surface 2748

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

water signal at these locations amidst the well-known Transpolar Drift (23). Water sample 15 was the only sample collected west of the Lomonosov Ridge in the Makarov Basin. Freshwater from the Kara Sea rivers Ob and Yenisey is also transported into the Arctic Ocean and may have influenced the water at stations 9-12 (6). At the polar front (in the northern Barents Sea; ref 24), warm Atlantic water meets colder and fresher Arctic water from the north and therefore subducts, resulting in a clear decrease in surface sea temperature north of this boundary (Table 1). Samples south and north of the polar front therefore represent two different water masses; Atlantic and Arctic water, respectively. Samples 4-6, east of Spitsbergen, were sampled north of the polar front in Arctic waters, whereas samples 7 and 8 appear to have been collected in the northernmost stretches of the West Spitsbergen Current (Atlantic water mass). The long-term average maximum and minimum ice extent in the Barents Sea, based on many years of observations, is marked in Figure 1. The relative degree of ice coverage and contribution of first-year ice (FYI) and multi-year ice (MYI) among the stations was assessed by a combination of visual observations and intermittently updated ice maps (http.// www.natice.noaa.gov/pub/East-Arctic/ Barents-Sea/Barents_Sea_North/2001/Northwest). Naturally, stations 1-3 in the Norwegian Current were free of ice. Station 6 was at an open water location but had been ice covered two weeks before the time of sampling. Similarly, stations 7 and 8 were in open water in the warm West Spitsbergen Current but close to the edge of the FYI and had been ice covered intermittently in June to July. Stations 4 and 5 were located in the marginal ice zone (MIZ) in dense pack ice (>90% coverage of FYI). Stations 9 and 10 were in the northern Barents MIZ within dense pack ice (90-100% coverage) of which 80% had survived a melt season (i.e., MYI) and about 10% FYI. Finally, the high Arctic stations 11-15 were all in dense MYI (90-100% coverage). Surface Seawater Concentrations. Total seawater concentrations (i.e., dissolved + particulate) of analyzed PCB congeners are given in Table 2, and concentrations of five congeners (representing different homologues) along the latitudinal transect are illustrated in Figure 2 A-E. One sample from 89° N consistently had higher concentrations than other samples in this area (data not shown). This sample was collected in the same water but 4 days after sample 13, while the ship was attached to a large ice flow and not actively steaming. We strongly suspect that this sample was contaminated by PCB from material on the ship that had entered into the stagnant surface water around the ship. This sample, in contrast to all others, had been collected from the surface water next to the ship after the ship had been stationary in the ice pack for a long period (4 days). This sample was therefore excluded from the data set. There was a general trend of a concentration decrease with increasing latitude, except for the lightest congener PCB 18. For instance, PCB 52 decreased from 470 fg L-1 at 62° N to 110 fg L-1 at 89° N (Figure 2B) and PCB 180 from 110 fg to 12 fg L-1 (Figure 2E). Generally, sampling sites north of the minimum ice extent tended to have rather similar but low PCB concentrations. Ice cover prevents air-water exchange; therefore, not only latitude but also ice cover presumably had an effect of lowering PCB surface seawater concentrations. Note that we do not infer that latitude, per se, is the governing parameter but rather represents a proxy reference, incorporating both distance to sources and decreasing average temperatures. Station 8 had similar concentrations of PCB 18, 52, and 110 as the other stations north of the ice extent but higher concentrations of PCB 153 and 180. This station had a large influence of Atlantic water, as illustrated by the high temperature and had also by far the highest particulate organic carbon concentration (POC) (300

TABLE 2. Surface Seawater (Particulate + Dissolved) Concentrations (fg L-1) of 12 PCB Congenersa sample ID

PCB 18

PCB 28

PCB 52

PCB 70

PCB 90/ 101

PCB 110

PCB 118

PCB 105

PCB 149

PCB 153

PCB 138

PCB 180

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

279 171 * 252 347 68 175 * 140 281 * * 129 244 167

376 265 * 254 302 50 108 * 88 142 * 114 62 140 82

471 346 276 313 322 * 93 123 127 201 * * 110 260 69

541 480 270 278 239 55 68 122 111 106 * 21 60 107 31

565 289 363 293 269 * 70 61 84 113 80 29 78 119 25

460 234 238 202 152 * 26 95 38 50 * * 53 52 16

238 190 * 247 131 68 40 95 * 69 * 22 37 58 10

84 62 52 84 53 26 18 70 18 26 * 10 14 22 11

300 278 219 165 133 * 34 127 33 47 26 * 39 38 10

357 366 250 323 162 * 47 228 56 74 * 30 47 47 12

142 144 222 120 185 85 50 224 20 74 25 * 21 45 13

110 70 99 93 63 * 19 88 * 20 * * 13 14 *

a

An asterisk (*) indicates PCB amount in sample 100 higher than what was found in the current study for the same region. There are several other unpublished data sets referred to in CACAR (28), from other parts of the Arctic Ocean, which also report much higher PCB seawater concentrations. As pointed out in the Canadian Arctic Contaminants Assessment Report (28), the much lower PCB seawater concentrations reported in Schulz-Bull et al. (25) and this study are probably reflecting that our two studies, in contrast to the other studies, were employing strict ultraclean protocols and thereby avoided contamination of the ultra-trace sample levels.

Concentrations that we measured north of 85° N (stations 12-15) varied between 70 and 260 fg L-1 for PCB 52 and 12-15 fg L-1 for PCB 180. Sample 15, from the Makarov Basin, had for most congeners the lowest concentrations (Figure 2). This could be an effect of relatively pristine freshwater from the Lena River (29), which during transport with the Transpolar Drift may have been further cleaned by particle scavenging and settling. Alternatively, atmospheric circulation patterns may have limited the impact of dirty European continent air at this distant location. Utschakowski has reported sum PCB concentrations in the surface water at 75° N outside the Lena delta (29). On the basis of the same assumptions as stated previously (27), we estimated that PCB 52 constitutes around 5% of their sum PCBs, corresponding to 140 fg PCB 52 L-1, which is consistent with our observations in this same water mass but further to the north. Observations of Latitudinal Fractionation. In contrast to R-HCH, which is increasing in concentration with latitude, PCB concentrations are decreasing with latitude over the Arctic Ocean scale (Figure 2). These data are thus consistent with Jo¨nsson et al. (30), who concluded that while the Arctic Ocean shelf makes up 14% of the global shelf area, its inventory of PCB 52 and PCB 180 was estimated to be only 1 and 0.3% of the global shelf inventory of these PCBs. Thus, PCB concentrations in the major abiotic matrixes of seawater and sediment are lower in the Arctic than elsewhere. To diagnose any change in PCB fingerprint with latitude, the relative contribution of each chlorination degree to the total quantified PCBs was plotted as a function of latitude(Figure 3A-D). The relative contribution of trichlorinated PCBs increased with latitude, while the tetrachlorinated fraction did not show any trend, and the relative importance of heavier PCBs decreased with latitude (Figure 3). Station 8 was excluded from the regression of the hexachlorinated PCBs since the high contribution of hydrophobic PCB congeners at this station was clearly different from other stations and therefore dominated the regression. Samples 6, 11, and 12 were excluded from Figure 3 since they had a large fraction of congeners below three times the blank. Some of the other samples lacked detection of one or two congeners, leading to total PCB concentrations varying some between the samples. However, these observed patterns clearly indicate a congener fractionation of PCBs during northward transport, also at very high latitudes, confirming previously presented results of multimedia box models (e.g., refs 17, 31, and 32). This latitudinal PCB fingerprint fractionation is further consistent with reports of fractionation of PCBs at lower latitudes (11-13) as well as observations of different PCB congener profiles in sediments from the European shelves versus in the northern Baffin Bay (14). Earth’s mean annual surface skin temperature (SST) may have been a more accurate parameter than latitude in Figure 3, but this satelliteVOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2749

FIGURE 2. Concentration (fg L-1) in surface seawater for PCB 18 (A), PCB 52 (B), PCB 110 (C), PCB 153 (D), and PCB 180 (E) as a function of latitude. derived data is unavailable for ice-covered regions. The relative contribution of PCBs at certain latitudes is also affected by other processes than atmospheric and ocean transport, such as HO• degradation, retention in soils and vegetation, fractionation in terms of sedimentation processes 2750

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

FIGURE 3. Relative contribution of tri (A), tetra (B), penta (C), and hexa (D) chlorinated congeners to the total quantified PCBs as a function of latitude. Station 8 was excluded from the regression in panel D. from the surface water, and possible releases from local sources. It has been suggested that local Russian sources may contribute significantly to the POP burden in the Arctic

Literature Cited

FIGURE 4. Ratio of the concentration at 88° N (average of samples 13-15) and 62° N (sample 1) for the nine PCB congeners that were both detected and for which there are consistent vapor pressure data available, as a function of log subcooled liquid vapor pressure at 25 °C (33). Ocean (e.g., ref 24). However, the low PCB concentrations in samples affected by the Russian Arctic Rivers, documented in this work (e.g., Table 2), support that there was no significant contamination from local Russian sources. To investigate the influence of volatility on the propensity for long-range Arctic transport of the individual PCB congeners, the ratio of the mean concentration at 88° N (stations 13-15) divided by the concentration at 62° N (station 1) as a function of their log subcooled liquid vapor pressure (pL) is shown (Figure 4; only nine congeners included for which pL from the same source was known; ref 33). The correlation between increasing North/South ratio and volatility (i.e., pL) further emphasizes the importance of this physicochemical property in determining the extent of long-range atmospheric transport for semivolatile POPs. This relationship provides a valuable field-established tool in estimating chemicals’ potential of being transported to the Arctic. As espoused previously, other physicochemical properties and processes may also influence the chemical’s potential for long-range transport. For instance, chemicals with high atmospheric reactivity (e.g., light polycyclic aromatic hydrocarbons) may not be transported as efficiently as would be predicted just from their PL. This study has provided the first data on PCB concentrations in surface seawater in the central Arctic Ocean. Concentrations of PCBs decreased significantly along the 62° N-89° N transect and were in the high Arctic on the order of 10-100 fg L-1 for the most abundant congeners. The change in relative importance of PCB homologue groups along the latitudinal transect supports that latitudinal fractionation occurs during atmospheric transport of PCBs to the high Arctic.

Acknowledgments We thank Zofia Kukulska for help with HS measurements and Ralf Dahlqvist for providing the map and for running the seawater intake thermosalinograph. Peter Ko¨mp and Michael McLachlan are acknowledged for collaboration with the onboard sampling intercomparison. We also thank the captain and crew on the I/B Oden and the Swedish Polar Research Secretariat (SWEDARCTIC 2001) for cruise logistical support. This work was financially supported by funding from the Swedish Research Council (VR MIZ Flux G-5103-11651999). Stockholm Marine Research Center (SMF) is thanked for a Ph.D. stipend to A.S., and the Swedish Research Council provided a senior research fellowship to O ¨ .G. (VR Grant 6292002-2309).

(1) Hargrave, B. T.; Vass, W. P.; Erickson, P. E.; Fowler, B. R. Tellus 1988, 40B, 480-493. (2) Chernyak, S. M.; McConnell, L. L.; Rice, C. P. Sci. Total Environ. 1995, 160/161, 75-85. (3) Hop, H.; Borga˚, K.; Gabrielsen, G. W.; Kleivane, L.; Skaare, J. U. Environ. Sci. Technol. 2002, 36, 2589-2597. (4) Muir, D. C. G.; Wagemann, R.; Hargrave, B. T.; Thomas, D. J.; Peakall, D. B.; Norstrom, R. J. Sci. Total Environ. 1992, 122, 75-134. (5) Gilman, A.; Dewailly, E.; Feeley, M.; Jerome, V.; Kuhnlein, H.; Kwavnick, B.; Neve, S.; Tracy, B.; Usher, P.; Van Oostdam, J.; Walker, J.; Wheatley, B. In Canadian Arctic Contaminants Assessment Report; Jensen, J., Adare, K., Shearer, R., Eds.; Indian and Northern Affairs Canada: Ottawa, 1997. (6) AMAP Assessment Report: Arctic Pollution Issues. Arctic Monitoring and Assessment Programme (AMAP); Oslo, Norway, 1998. (7) Wania, F.; Mackay, D. Ambio 1993, 22, 10-18. (8) Wania, F.; Mackay, D. Environ. Sci. Technol. 1996, 30, 390-396. (9) Harvey, G. R.; Steinhauer, W. G. J. Mar. Res. 1976, 34, 561-575. (10) Iwata, H.; Tanabe, S.; Sakai, N.; Tatsukawa, R. Environ. Sci. Technol. 1993, 27, 1080-1098. (11) Ockenden, W. A.; Sweetman, A. J.; Prest, H. F.; Steinnes, E.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 2795-2803. (12) Agrell, C.; Okla, L.; Larsson, P.; Backe, C.; Wania, F. Environ. Sci. Technol. 1999, 33, 1149-1156. (13) Meijer, S. N.; Steinnes, E.; Ockenden, W. A.; Jones, K. C. Environ. Sci. Technol. 2002, 36, 2146-2153. (14) Gustafsson, O ¨ .; Axelman, J.; Broman, D.; Eriksson, M.; Dahlgaard, H. Chemosphere 2001, 45, 759-766. (15) Grimalt, J. O.; Ferna´ndez, P.; Berdie, L.; Vilanova, R. M.; Catalan, J.; Psenner, R.; Hofer, R.; Appleby, P. G.; Rosseland, B. O.; Lien, L.; Massabuau, J. C.; Battarbee, R. W. Environ. Sci. Technol. 2001, 35, 2690-2697. (16) Lakaschus, S.; Weber, K.; Wania, F.; Bruhn, R.; Schrems, O. Environ. Sci. Technol. 2002, 36, 138-145. (17) Wania, F. Environ. Sci. Technol. 2003, 37, 1344-1351. (18) Sobek, A.; Gustafsson, O ¨ .; Axelman, J. Int. J. Environ. Anal. Chem. 2003, 83, 177-187. (19) Sobek, A.; Gustafsson, O ¨ .; Hajdu, S.; Larsson, U. Environ. Sci. Technol. 2004, 38, 1375-1382. (20) Petrick, G.; Schulz-Bull, D. E.; Martens, V.; Scholz, K.; Duinker, J. C. Mar. Chem. 1996, 54, 97-105. (21) Zebu ¨ hr, Y.; Na¨f, C.; Bandh, C.; Broman, D.; Ishaq, R.; Pettersen, H. Chemosphere 1993, 27, 1211-1219. (22) Bandh, C.; Ishaq, R.; Broman, D.; Na¨f, C.; Ro¨nquist-Nii, Y.; Zebu ¨ hr, Y. Environ. Sci. Technol. 1996, 30, 214-219. (23) Carmack, E. C. In The Geophysics of Sea Ice; Untersteiner, N., Ed.; Plenum: New York, 1986; pp 171-222. (24) Pfirman, S. L.; Bauch, D.; Gammelsrod, T. In The Polar Oceans and Their Role in Shaping the Global Environment; Johanessen, O. M., Muench, R. D., Overland, J. E., Eds.; Geophysical Monograph 85; American Geophysical Union: Washington, DC, 1994. (25) Schulz-Bull, D. E.; Petrick, G.; Bruhn, R.; Duinker, J. C. Mar. Chem. 1998, 64, 101-114. (26) Hargrave, B. T.; Phillips, G. A.; Vass, W. P.; Bruecker, P.; Welch, H. E.; Siferd, T. D. Environ. Sci. Technol. 2000, 34, 980-987. (27) Schulz, D. E.; Petrick, G.; Duinker, J. C. Environ. Sci. Technol. 1989, 23, 852-859. (28) Muir, D.; Strachan, W. In Sources, Occurrence, Trends, and Pathways in the Physical Environment. The Canadian Arctic Contaminants Assessment Report II; Bidleman, T., Macdonald, R., Stow, J., Eds.; Indian and Northern Affairs: Ottawa, Canada, 2003; pp 92-99. (29) Utschakowski, S. Anthropogenic Organic Trace Compounds in the Arctic Ocean. Ph.D. Thesis, University of Kiel, 1998. (30) Jo¨nsson, A.; Gustafsson, O ¨ .; Axelman, J.; Sundberg, H. Environ. Sci. Technol. 2003, 37, 245-255. (31) Mackay, D.; Wania, F. Sci. Total Environ. 1995, 160/161, 25-38. (32) Beyer, A.; Wania, F.; Gouin, T.; Mackay, D.; Matthies, M. Environ. Sci. Technol. 2003, 37, 766-771. (33) Li, N.; Wania, F.; Lei, D. L.; Daly, G. L. J. Phys. Chem. Ref. Data 2003, 32, 1545-1590. (34) Wedborg, M.; Skoog, A.; Fogelqvist, E. In Humic Substances in the Global Environment and Implications on Human Health; Senesi, N., Miano, T. M., Eds.; Elsevier Science: Amsterdam, The Netherlands, 1994.

Received for review December 11, 2003. Revised manuscript received February 20, 2004. Accepted February 25, 2004. ES0353816 VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2751