Article pubs.acs.org/est
Observation-Based Assessment of PBDE Loads in Arctic Ocean Waters Joan A. Salvadó,† Anna Sobek,† Daniel Carrizo,‡ and Ö rjan Gustafsson*,† †
Department of Environmental Science and Analytical Chemistry (ACES), Stockholm University, Stockholm, 10691 Sweden Institute for Global Food Security, Queen’s University, Belfast, BT9 5BN United Kingdom
‡
S Supporting Information *
ABSTRACT: Little is known about the distribution of polybrominated diphenyl ethers (PBDE) -also known as flame retardants- in major ocean compartments, with no reports yet for the large deep-water masses of the Arctic Ocean. Here, PBDE concentrations, congener patterns and inventories are presented for the different water masses of the pan-Arctic shelf seas and the interior basin. Seawater samples were collected onboard three cross-basin oceanographic campaigns in 2001, 2005, and 2008 following strict trace-clean protocols. ∑14PBDE concentrations in the Polar Mixed Layer (PML; a surface water mass) range from 0.3 to 11.2 pg· L−1, with higher concentrations in the pan-Arctic shelf seas and lower levels in the interior basin. BDE-209 is the dominant congener in most of the pan-Arctic areas except for the ones close to North America, where pentaBDE and tetra-BDE congeners predominate. In deep-water masses, ∑14PBDE concentrations are up to 1 order of magnitude higher than in the PML. Whereas BDE-209 decreases with depth, the less-brominated congeners, particularly BDE-47 and BDE-99, increase down through the water column. Likewise, concentrations of BDE-71 -a congener not present in any PBDE commercial mixtureincrease with depth, which potentially is the result of debromination of BDE-209. The inventories in the three water masses of the Central Arctic Basin (PML, intermediate Atlantic Water Layer, and the Arctic Deep Water Layer) are 158 ± 77 kg, 6320 ± 235 kg and 30800 ± 3100 kg, respectively. The total load of PBDEs in the entire Arctic Ocean shows that only a minor fraction of PBDEs emissions are transported to the Arctic Ocean. These findings represent the first PBDE data in the deep-water compartments of an ocean.
■
INTRODUCTION Polybrominated diphenyl ethers (PBDEs) have been used for several decades as flame retardants in various consumer products, including polyurethane foam cushions, high-use textiles, electronic appliances, and printed circuit boards.1 They were commercialized as congener mixtures of three technical products; the deca-product is mainly composed of BDE-209 (>97%), the octa-product consists mainly of BDE183 followed by BDE-153 and BDE-154, and the penta-product consists of a mixture of tetra- to hexa-BDEs.2 The production and use of penta- and octa-BDE mixtures was banned in Europe in 2004, and only deca-BDE was still permitted until 2008. However, there are still stocks of all PBDEs in products, both in service and waste. PBDEs are similar to polychlorinated biphenyls (PCBs) in molecular structure and environmental behavior and are considered as persistent organic pollutants. PBDEs are known to be toxic, persistent, bioaccumulative, and prone to long-range atmospheric transport (LRT).3−5 It has been reported that BDE-209 can undergo both photolytic debromination and metabolic degradation to lower-brominated BDE congeners that are more toxic and bioaccumulative (e.g., BDE-71, BDE-47, BDE-99, and BDE-100).6,7 Low- and highbrominated congeners are ubiquitous in the environment and © XXXX American Chemical Society
have been detected in environmental and biological samples collected from all over the world, including high mountain lakes and deep-sea sediments.4,8−10 In the past, the Arctic was considered a clean area regarding anthropogenic pollution, but during the last decades, it has become an area of concern due to the impact of persistent organic pollutants on its ecosystems and top consumers, including humans.11−13 Furthermore, the Arctic is an important indicator region to evaluate LRT potential, persistence, and bioaccumulation of persistent organic pollutants.14,15 There is little knowledge on the distribution and transport of PBDEs to the Arctic Ocean, particularly for BDE-209 due to its low vapor pressure. In Oceans, the biological pump, the process by which phytoplankton fix carbon dioxide, and transport from continental shelves to the deep sea both cause the major downward export of organic pollutants from surface waters and thereby also from active cycling and exposure in the atmosphere−biosphere system.16−21 Received: November 18, 2015 Revised: January 26, 2016 Accepted: February 3, 2016
A
DOI: 10.1021/acs.est.5b05687 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 1. Spatial distribution of Σ14PBDEs concentrations in the Arctic Polar Mixed Layer sampled during the expeditions AO-01, Beringia-05, and ISSS-08. SNCAA: shelf of the Northern Canadian Arctic Archipelago.
It is well-known that the marine Arctic food web is exposed to PBDEs.11,22,23 In 1998 Jacob de Boer et al. reported the occurrence of PBDEs in sperm whales, which normally stay and feed in deep waters.22 These authors suggested the transfer of PBDEs to the deeper ocean as a source of concern and a threat to deep-sea ecosystems. Nevertheless, PBDE data in the deepwater masses of oceans is still completely lacking. There are only three PBDE studies performed on seawater of the Arctic Ocean, and all of them in surface waters of low-latitude Arctic regions.24−26 The general lack of data on persistent organic pollutants concentrations in high Arctic Ocean waters reflect the inaccessibility of the ice-covered interior basin and the technical challenges to sample the large volumes of deep seawater required to quantify trace levels of PBDEs and overcome shipboard and laboratory contamination. The present study therefore aims to (i) provide spatial and vertical PBDE concentrations in the water masses of the pan-Arctic shelf seas and the deep interior basins during the past decade, (ii) explore BDE congener profiles and their spatial distributions to explore potential sources and transport modes, and (iii) constrain the overall inventories of PBDEs in the different water layers (Polar Mixed Layer, intermediate Atlantic Water Layer, and the Arctic Deep Water Layer) of the entire Arctic Ocean. We present the first PBDE data in seawater
covering a vertical profile, here extending down to 2500 m depth.
■
MATERIALS AND METHODS Oceanographic Setting. The Arctic Ocean is composed of the Central Arctic Ocean Basin (CAOB) and the continental shelf system, the largest yet least-explored shelf seas in the world ocean, covering over 50% of the Arctic Ocean area27 (Figure 1). The water masses of the Arctic Ocean consist of a mixture of Atlantic water, Pacific water, river runoff, and sea ice melt and brine. An important characteristic of the Arctic Ocean is its strong stratification due to the approximately 4000 km3· y−1 of freshwater inflow and the seasonal ice melt, causing the formation of a low-density surface layer, which is, in the Arctic, named the Polar Mixed Layer (PML).28 Detailed geochemical characteristics of surface water regimes of the pan-Arctic shelf areas and interior basins are described elsewhere.29,30 The deep water, including the intermediate Atlantic Water Layer (AtWL) and the Arctic Deep Water Layer (ArDWL), is supplied by dense saline water from the Atlantic Ocean.21,31 Salinitytemperature-depth (CTD) profiles performed during the three cross-basin expeditions that were carried out to collect seawater samples for this study suggest that the PML is about 15 m deep in the shelf seas and 30 m in the central basin (the halocline B
DOI: 10.1021/acs.est.5b05687 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology layer was ∼40−50 m deep); the AtWL extends to ∼900 m water depth and below expands the ArDWL. Moreover, those CTD measurements demonstrated the homogeneity of deep water masses (Figure S1). The residence times of the water masses differ between these layers. The transport or replacement of the PML takes about 10 years, whereas the residence times of the AtWL are about 30 years, and the deep basin water (ArDWL) time scale is measured in centuries.31,32 The oceanic surface water circulation in the Arctic is well understood, in part due to ice-drifting data collected with drifting buoys. The surface layer is characterized by the large clockwise Beaufort Gyre, centered near 80°N 150°W over the Canadian Basin, and the Transpolar Drift originating in the East Siberian Sea and the Laptev Sea, which crosses the CAOB and the North Pole area along the Lomonosov Ridge (Figure 1).33 North Atlantic water enters the Arctic Ocean through both the Fram Strait (northwest Spitsbergen current) and Barents Sea (St. Anna Trough). Pacific Ocean water enters the Bering Strait and is exported across the Eurasian Basin of the CAOB following the Transpolar Drift. Large quantities of water, particularly freshwater, can be trapped by the Beaufort Gyre.34 The AtWL also moves along the continental margins and the ridges with a cyclonic circulation around the Arctic Ocean; nevertheless, the ArDWL has almost no direct contact with the surrounding oceans, and its transport or replacement could take up to several hundred years.32 Sample Collection. Seawater sampling was conducted during three extensive expeditions in the high Arctic. The AO01-2001 expedition was performed onboard the icebreaker Oden during the Swedish Arctic Ocean expedition (SWEDARCTIC 2001) from June to August 2001. During this campaign, we collected water samples from the PML using a stainless-steel seawater intake system and from the subsurface water masses using a stainless-steel filter-absorbent system mounted on a submersible pump (Kiel in-situ pump; KISP).35 Both systems used for sampling were previously described and tested.36,37 The sampling stations follow a transect from the Norwegian Sea to Barents Sea and both Nansen and Amundsen Basins, passing the North Pole area and extending into the Makarov Basin of the Canadian Basin toward the North Pole (62° N−89° N) (Figure 1 and Tables S1 and S2). The Beringia 2005 expedition was also accomplished onboard the icebreaker Oden from July to August 2005 (SWEDARCTIC 2005). This campaign focused on the Beringia region and the eastern Arctic Ocean. We collected water samples in the PML along a transect from the North Sea across the northern North Atlantic Ocean, rounding south of Greenland and extending through Baffin Bay. Samples were also collected through the Canadian Arctic Archipelago and Beaufort Sea and passing out and then back in through the Bering Strait. Furthermore, we followed a northward track through the American sector of the Chukchi Sea and into the Beaufort Gyre of the deep Canada Basin (Figure 1 and Table S1). The ISSS-08 expedition was performed onboard the H/V Yacob Smirnitsky (Archangelsk) from August to September 2008 as part of the International Polar Year (IPY) activities. Samples were collected from the PML in the Kara Sea, Laptev Sea, East Siberian Sea, and Russian sector of the Chukchi Sea (Figure 1 and Table S1). Water-sampling protocols for both I/B Oden and H/V Yacob Smirnitskyi were detailed previously.21,29,30 Briefly, the stainless steel seawater intake systems, placed under the prow of the
ships at a water depth of 8 m, and the stainless steel KISP allowed us to take samples into an ultraclean laboratory, where we changed and handled filters and absorbents. The pressure over the filter for the seawater intake sampling was constantly monitored with a pressure indicator and never allowed to exceed 1 bar to minimize cell lysing. Seawater sample volumes were in the order of 1000 L to meet the detection limits (e.g., Tables S1 and S2). The atmosphere in the ultraclean laboratory was set under high pressure, and incoming air was filtered through double-activated carbon and high-efficiency particulate air filters. Particle-associated PBDEs were collected on precombusted borosilicate filters (GF/F, 293 mm, nominal pore size 0.7 μm; Whatman International Ltd., Maidstone, England), with no mesh before the filter, which were followed by polyurethane foam absorbents (PUF; diameter of 37 mm, length of 160 mm, Sunde Söm & Skumplast) to collect the dissolved PBDEs. The PUFs were extensively cleaned prior to the expedition to minimize any contamination. The prepared PUF absorbents were placed in precombusted Al foil envelopes, which were in turn placed in plastic bags and stored in a freezer (−18 °C) until sampling. We used two PUFs per sample. Collected samples were placed in the same precombusted Al envelopes and stored in double-sealed plastic bags in a freezer (−18 °C) until analysis. Analysis. In the laboratory, samples were spiked with a surrogate standard (13C-labeled PCB-180) and extracted by Soxhlet for 24 h with toluene (glass-distilled quality; Burdick and Jackson, Fluka Chemie AG, Buchs, Switzerland) using a Dean−Stark trap for the collection of water. The use of PCBs as internal standard of PBDEs was previously tested and used in other studies.9,38,39 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 PBDE fraction was thereafter eluted on an open column containing three layers of modified silica (SiO2/H2−SO4 (10 mm), SiO2/KOH (10 mm), and SiO2/H2O (10 mm)). The instrumental analytical conditions for PBDE analyses are described elsewhere.9,39 Briefly, sample aliquots (1 μL) were injected into a gas chromatograph coupled to a mass spectrometer (Trace DSQ instrument from Thermo Scientific, TX) operating in negative-ion chemical ionization mode (NICI). The instrument was equipped with a low-bleed SGEBPX5MS fused silica capillary column (15 m long, 0.25 mm internal diameter, and 0.10 μm film thickness). Internal standards (13C-labeled BDE-209 and BDE-118) were added to all samples before injection on the GC-NICI-MS. The use of those internal standards is of special importance, particularly [C13]BDE-209, to correct the degradation of BDE 209 in the liner. Quality Assurance. All results were subjected to a strict quality assurance and control procedure. We used ultraclean protocols to avoid contamination during sampling, cleanup, and analysis.21,29,30 The sampling technique was previously evaluated in field campaigns in the open Baltic Sea and along a surface water transect from the North Sea into the Arctic Ocean.30,35,36 The same type of filtration and extraction system, operated under the same range of flow rates (1−3 L·min−1) and with water volumes up to 1000 L, was used in this study. The previous method evaluations in the contaminated Baltic Sea demonstrated that there was no breakthrough of any of the PCBs, irrespective of sample volume (range 300 to 1400 L). Field blanks and laboratory blanks (n = 10) were analyzed in C
DOI: 10.1021/acs.est.5b05687 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Table 1. Comparison of PBDE Concentrations (pg·L−1) in Surface Seawater (Dissolved and Particulate) Measured in This Study with Concentrations Reported from Other Locations in the World
a
sampling areas
na
BDE-209
∑PBDEs
sampling year
references
Chukchi Sea East Siberian Sea Laptev Sea Kara Sea Barents Sea Norwegian Sea Beaufort Sea SNCAAd Central Arctic Ocean Basin East of Greenland East of Greenland Chukchi Sea and East of Asia Kara Sea Tropical Atlantic Ocean Atlantic and Southern Ocean Hong Kong (China)
14 14 14 14 14 14 14 14 14 14 10 10 43 10 9 8
0.1−1.5 (0.8)b 0.6−0.9 (0.6) 0.6−5.5 (2.9) 0.7 0.6−7.8 (2.6) 1.5−10.4 (6.0) 2.1−4.1 (3.1) 0.4 0.2−3.4 (0.9) 0.3 (0.3) ndc−0.48 nd−0.2 − nd−30 (7) − −
1.1−2.4 (1.8) 0.9−1.0 (1.1) 1.0−6.2 (4.3) 1.3 0.9−8.3 (3.0) 1.9−11.2 (6.6) 3.9−5.9 (4.9) 5 0.2−4.2 (1.2) 0.9−1.5(1.2) 0.005−0.64 nd−0.8 1.8−10.8 nd−32 0.09−2.19 nd−297.3
2005 2008 2008 2008 2001 2001 2005 2005 2001 2005 2009 2010 2003, 05 2009 2008 2005
this study this study this study this study this study this study this study this study this study this study Möller et al., 2011 Möller et al., 2011 Carroll et al., 2008 Lohmann et al., 2013 Xie et al., 2011 Wurl et al., 2006
Number of congeners analyzed. bAverage values between brackets. cnd = not detected. dShelf of Northern Canadian Arctic Archipelago.
detectable BDE congener concentrations, indicating that these pollutants are widespread in the Arctic Ocean. Concentrations of ∑14PBDEs (sum of 14 BDE congeners defined above) ranged between 0.3 and 11.2 pg·L−1 (Figure 1 and Table S4). The highest ∑14PBDE concentrations were observed at station AO-01-1 (11.2 pg·L−1). This location suggests a strong source of PBDEs as it is one of the southernmost samples and the site closest to civilization, near the southern coast of Norway, and thus not part of the Arctic. The lowest PBDE concentrations were observed in the northernmost stations of the CAOB, particularly in the sample from AO-01-12. A previous study suggested that the water at AO-01-12 is influenced by freshwater inputs from the Lena River, and it could have been cleaned during the Transpolar Drift through particle scavenging and settling.29 However, we do not observe any influence of Arctic rivers, as there were no correlations of PBDE concentrations with salinity. There was a general decreasing concentration pattern with increasing latitude in the seawater samples from the Barents Sea−Nansen− Amundsen transect (AO-01 samples; Figure 1) and a slightly negative relationship between ∑14PBDEs and latitude (r2 = 0.53, p value
