Article pubs.acs.org/est
Exploring the Role of Shelf Sediments in the Arctic Ocean in Determining the Arctic Contamination Potential of Neutral Organic Contaminants James M. Armitage,*,† Sung-Deuk Choi,†,§ Torsten Meyer,† Trevor N. Brown,† and Frank Wania† †
Department of Physical and Environmental Sciences, University of Toronto Scarborough, 1265 Military Trail, Toronto, Ontario, Canada, M1C 1A4 § School of Urban and Environmental Engineering, Ulsan National Institute of Science and Technology (UNIST), UNIST-gil 50, Eonyang-eup, Ulju-gun, Ulsan 689-798, Korea S Supporting Information *
ABSTRACT: The main objective of this study was to model the contribution of shelf sediments in the Arctic Ocean to the total mass of neutral organic contaminants accumulated in the Arctic environment using a standardized emission scenario for sets of hypothetical chemicals and realistic emission estimates (1930−2100) for polychlorinated biphenyl congener 153 (PCB-153). Shelf sediments in the Arctic Ocean are shown to be important reservoirs for neutral organic chemicals across a wide range of partitioning properties, increasing the total mass in the surface compartments of the Arctic environment by up to 3.5-fold compared to simulations excluding this compartment. The relative change in total mass for hydrophobic organic chemicals with log air−water partition coefficients ≥0 was greater than for chemicals with properties similar to typical POPs. The long-term simulation of PCB-153 generated modeled concentrations in shelf sediments in reasonable agreement with available monitoring data and illustrate that the relative importance of shelf sediments in the Arctic Ocean for influencing surface ocean concentrations (and therefore exposure via the pelagic food web) is most pronounced once primary emissions are exhausted and secondary sources dominate. Additional monitoring and modeling work to better characterize the role of shelf sediments for contaminant fate is recommended.
■
INTRODUCTION The global-scale fate and transport of organic chemicals and their long-range transport to and accumulation in remote environments remain important research and regulatory priorities.1−5 For example, the detection of organic chemicals in remote regions is a key consideration for assessments and decision-making under the Stockholm Convention on Persistent Organic Pollutants (http://www.pops.int), an international treaty that was ratified by a sufficient number of signatories to enter into force in 2004. Besides the development of monitoring networks (e.g., the Global Atmospheric Passive Sampling network, GAPS),6,7 a great deal of effort has also been devoted to the development, application, and evaluation of modeling tools that simulate the behavior of organic chemicals in the environment.5 These modeling tools and the various outputs representing long-range transport potential (LRTP) such as Characteristic Travel Distance (CTD), Spatial Range (SR), and Arctic Contamination Potential (eACP10)8−11 provide valuable information for chemical assessments (e.g., ranking exercises) and can also be used to formulate and explore different hypotheses regarding chemical fate, presence in the global environment, and bioaccumulation.12,13 © 2012 American Chemical Society
Modeling the fate and transport of organic chemicals in the Arctic is particularly challenging, given that this environment experiences wide swings in solar radiation and ambient air temperature, has both temporary and permanent cyrospheric compartments (e.g sea-ice, seasonal snow packs, permafrost soils), and exhibits complex circulation patterns in the surface ocean water (vertical and large-scale horizontal mixing).14,15 The Arctic environment is also unique in that its continental shelf sediments are extensive, underlying approximately half of the total surface area of the Arctic Ocean (compared to ≤20% for most other ocean bodies).16 However, many of the chemical fate and transport models applied to persistent organic pollutants (POPs) in the past17−24 omit shelf sediments (in all regions). This means that the mass of chemical associated with suspended particulate matter in the water column is not properly accounted for (e.g., 100% transported to the deep ocean instead of some fraction being retained in shelf sediments Received: Revised: Accepted: Published: 923
October 19, 2012 December 10, 2012 December 15, 2012 December 15, 2012 dx.doi.org/10.1021/es304276g | Environ. Sci. Technol. 2013, 47, 923−931
Environmental Science & Technology
Article
matrices1,2,29,30 and because this modification involves the addition of an entirely new compartment (and set of process descriptions) as opposed to improved parameter values or descriptions of processes already included in the model. Shelf Sediments in the Arctic Ocean. As noted above, the Arctic Ocean is unique among the world’s oceans in that its shelf sediments (i.e., seafloor from shore to shelf break, operationally defined to be the point where the slope of the seafloor begins to exhibit a rapid change) underlie a substantial proportion of the total surface area of the water column (∼53%).16 The International Bathymetric Chart of the Arctic Ocean (IBCAO Version 3) (http://www.ngdc.noaa.gov/mgg/ bathymetry/arctic/ibcaoversion3.html)31 was used to calculate the surface area of the Arctic Ocean with depth ≤200 m (i.e., to match the depth of the surface ocean compartment in the GloboPOP model) and, subsequently, the average depth of these regions. The results of this analysis are illustrated in Figure 1.
and available for resuspension). Irrespective of this omission, modeled concentrations of POPs (e.g., polychlorinated biphenyls, PCBs) typically agree well with available monitoring data, at least in the lower atmosphere of the Northern Hemisphere (e.g., ref 25.). Agreement between modeled and measured concentrations in the atmosphere implies that modeled deposition fluxes to surface compartments (i.e., soils, surface ocean water) may be reasonable and that shelf sediments, while potentially important as a mass reservoir, do not (currently) exert a key influence on global transport patterns. While shelf sediments (and organic matter in general) may not strongly influence the fate and transport of chemicals of concern such as hexachlorocyclohexanes (HCHs) and some perfluoroalkyl acids (PFAAs), which tend to distribute predominantly in the dissolved phase of the water column,19,24,26 this compartment is important for more hydrophobic chemicals. For example, shelf sediments are substantial reservoirs of PCBs.27 Furthermore, as human and ecological exposure to chemicals can occur via the marine benthic food web (e.g., sediment/pore water→benthic bivalve/infauna→ walrus→human) and hydrophobic chemicals typically exhibit relatively high bioaccumulation potential, there is a strong rationale to explore the potential role of shelf sediments in more detail. The main objective of this study is to introduce continental shelf sediments to the Arctic environment of a global-scale fate and transport model and then explore the role of this compartment in determining the Arctic Contamination Potential (eACP10) of neutral organic chemicals. Simulations are conducted for hypothetical chemicals covering a wide range of partitioning properties and susceptibility to degradation and also for PCB-153 using long-term realistic emission estimates. Other modifications recently made to the global-scale model applied here are also described.
■
MATERIALS AND METHODS Global-Scale Fate and Transport Model (GloboPOP). GloboPOP is a fugacity-based chemical fate and transport model 17 that divides the globe into 10 latitudinal bands. Each latitudinal band includes well-mixed compartments representing the atmosphere (four layers), surface ocean water (0−200m depth), soils (agricultural, uncultivated), and freshwater and underlying sediments (active layer only). The chemical is distributed between two phases in the atmosphere (gaseous and particulate), two phases in water and sediments (dissolved and particulate), and three phases in soils (gaseous, dissolved, and particulate) based on its partitioning properties. The original version of GloboPOP was updated to include forest canopies and underlying soils in 2005.28 Additional modifications implemented more recently include the (i) representation of intermittent precipitation, (ii) distinction between rain and snowfall as forms of precipitation, (iii) inclusion of dynamic snow packs, (iv) estimation of a global water mass balance (including large-scale riverine transport to Arctic), (v) revised parametrization of soils in the high Arctic, and (vi) introduction of shelf sediments in the Arctic Ocean (active sediment layer only). The parametrization of the shelf sediments in the Arctic Ocean is discussed below. Details of the other modifications are summarized in the Supporting Information (SI, Section S1). We have chosen to focus on the role of shelf sediments in determining the fate and transport of organic contaminants in the Arctic environment because of the general lack of monitoring data in this compartment compared to other
Figure 1. Map of the Arctic Ocean indicating areas where the depth of the water column is ≤200 m and hence included when estimating the surface area of shelf sediments interacting with the surface ocean water in the model (∼1/3 of total surface area of Arctic Ocean above 64 °N) and >200 m (∼2/3 of total surface area of Arctic Ocean above 64 °N). Note that the average depth of the water column across all areas ≤200 m is approximately 70 m.
There are extensive areas of relatively shallow water in the Barents Sea stretching continuously along the Siberian coast from the Kara Sea to the Chukchi Sea. A substantial portion of the water in the Canadian Archipelago is also ≤200 m deep. Overall, we estimate that the surface area of the Arctic Ocean above 64°N (matching the North Polar zone in the GloboPOP model, see SI, Figure S1) with depth ≤200 m is approximately one-third of the total surface area of this compartment and that the average depth of these regions is 70 m. For comparison, the average depths of the entire Arctic Ocean and deep basins of the central Arctic Ocean are approximately 1200 and 2400 m, respectively.16 As a first approximation, the surface area of shelf sediments was set equal to the surface area of water with depth 924
dx.doi.org/10.1021/es304276g | Environ. Sci. Technol. 2013, 47, 923−931
Environmental Science & Technology
Article
Figure 2. Arctic Contamination Potential (eACP10) excluding and including shelf sediments and the ratio between these two model scenarios (i.e., eACP10 including shelf sediments divided by eACP10 excluding shelf sediments) assuming (A) perfect persistence, (B) 10:20:100 year baseline degradation half-lives in water, soil, and sediments, respectively, and a baseline kOH of 1 × 10−13 cm3 molecule−1 s−1, and (C) 1:2:10 year baseline degradation half-lives in water, soil, and sediments, respectively, and a baseline kOH of 1 × 10−12 cm3 molecule−1 s−1. eACP10 values are normalized to the maximum value for each scenario. Log KOW values of 5 and 8 are indicated by the white diagonal lines.
≤200 m (∼5.5 million km2), and the volume of the overlying water column was calculated using the average depth and surface area. This approach can lead not only to underestimation of the shelf sediment surface area, as the slope of the sediment bed is not accounted for, but also to overestimation, as sediment accumulation bottoms are not distinguished from areas of the seafloor that tend to be scoured more effectively. The active sediment layer (i.e., depth of sediment considered well-mixed and in direct contact with overlying water column) was assumed to be 0.05 m deep with a porosity of 0.827,32,33 and an average organic carbon content of 2.0%.34−37 Based on these dimensions, the total volume of the active layer of shelf sediments in the Arctic environment is approximately 275 km3 (compared to approximately 0.4 million km3 of overlying
surface ocean water, 2.5 million km3 of surface ocean water in total and 400 km3 of well-mixed soil; see SI, Table S8 for soil data). Note that this volume of sediment reflects only that portion that is in direct contact with the surface ocean layer in the model (i.e., shelf sediments below 200 m are not included). The default mass transport coefficient (MTC, in m h−1) for particle sinking in the surface ocean compartment in the North Polar zone (MTCPS = 2.25 × 10−8 m h−1) corresponds to a settling velocity of approximately 0.5 m d−1. No adjustments were made to this parameter for the current set of calculations. Following from the surface area estimation above, one-third of the total particle deposition flux is directed to the shelf sediment compartment while two-thirds is directed to the deep ocean (i.e., transported out of the model domain). Due to the 925
dx.doi.org/10.1021/es304276g | Environ. Sci. Technol. 2013, 47, 923−931
Environmental Science & Technology
Article
degradation) in all compartments; (ii) baseline degradation half-lives at 25 °C of 10, 20, and 100 years in surface water, soils, and sediments, respectively, and a second-order rate constant for reactions with OH radicals (kOH) in the atmosphere equal to 1 × 10−13 cm3 molecule−1 s−1 (equivalent to a baseline degradation half-life of approximately 100 days); and (iii) baseline degradation half-lives at 25 °C of 1, 2, and 10 years in surface water, soils, and sediments, respectively, and a kOH equal to 1 × 10−12 cm3 molecule−1 s−1. Simulations with the different degradation half-live scenarios were conducted including and excluding the shelf sediment compartment. Emissions were directed 100% to air and distributed spatially according to the pattern shown in the SI (Section S2). Other model assumptions (e.g., temperature dependence of partitioning constants and degradation half-lives) are also summarized in the SI. The long-term simulation (1930−2100) of PCB-153 is included in order to illustrate the influence of shelf sediments on fate and transport up to and long after global emission rates peak (i.e., the primary emission phase vs the secondary emission phase).40 PCB-153 was selected because it is a representative and well-studied POP and because reported levels of these compounds are still relatively high in Arctic biota.41 Monitoring data are also available for the Arctic environment.27,42 Partitioning properties and environmental degradation half-lives assumed for these simulations are summarized in the SI (Section S3). Details on the emission scenario are also presented in the SI.
coarse spatial resolution of the model, it is not possible to distinguish between areas (or depths) that exhibit different sedimentation dynamics (e.g., near-shore areas with high rates of resuspension vs accumulation bottoms). The MTCs for shelf sediment resuspension and sediment burial were set to 0.7 and 0.3 MTCPS, respectively, and are meant to be broadly representative of average conditions across different depths and levels of turbulence at the sediment−water interface (physically and/or biologically induced). Arctic Contamination Potential (eACP10). Arctic Contamination Potential (eACP10) is a metric characterizing longrange transport potential (LRTP) and accumulation in the surface compartments of the Arctic environment normalized to emission rate.10,11 It is calculated from non-steady state model output as follows: eACP10 =
(MNP,T − MNP,A ) E T,G
(1)
where MNP,T is the total mass of chemical in the North Polar zone after 10 years of constant emissions, MNP,A is the total mass of chemical in the atmosphere of the North Polar zone, and ET,G is the total mass of chemical emitted to the global environment during the simulation. Note that the total mass of chemical in the North Polar zone excludes the mass of chemical transported out of the surface ocean water to the deep ocean and transported out of the active sediment layer via burial. The mass of chemical in the atmosphere is excluded as substances with partitioning properties favoring extensive distribution to air (i.e., high air−water partition constant, low octanol-air partition constant) are not expected to exhibit substantial exposure potential.12 Equal weighting is given to the mass of chemical in all other compartments. In other words, only the presence of the chemical in the remaining exposure media matters, and there is no further consideration of relative exposure potential. This approach follows from our interpretation of the relevant information requirements for screening and risk profiling defined in the Stockholm Convention on Persistent Organic Pollutants (Annex D, E) (http://www.pops.int). Model Application. Two model applications are presented in this study. The first application is a generic assessment of the role of continental shelf sediments in the Arctic Ocean in determining eACP10 of neutral organic chemicals as a function of partitioning properties and degradation half-lives in the environment. The second application is a long-term simulation (1930−2100) of PCB-153 using realistic emission estimates38 including and excluding the shelf sediment compartment in the North Polar zone of the GloboPOP model. The generic assessment utilizes chemical space plots10,11 to summarize model output across a wide range of air−water (KAW), octanol−air (KOA), and octanol−water (KOW) partition constants. In this case, simulations were conducted for a set of hypothetical chemicals with log KAW ranging from −5 to 3, log KOA ranging from 4 to 15 and log KOW ranging from 1 to 12 (at 25 °C). These partition constants are adjusted to the corresponding values at ambient temperature in the various compartments over the year. Note that log KOW is used to calculate organic carbon−water partitioning (KOC), which is estimated here using a proportionality constant equal to 0.35 (i.e., KOC = 0.35KOW).39 To be consistent with a previous model application,11 the following scenarios for degradation half-lives were simulated: (i) perfect persistence (i.e., negligible
■
RESULTS AND DISCUSSION Generic Assessment. The Arctic Contamination Potential (eACP10) of the different sets of hypothetical chemicals excluding and including shelf sediments in the Arctic Ocean are presented and compared in Figure 2. The general features and basic underlying processes driving the model results shown in Figure 2 for the scenarios excluding shelf sediments in the Arctic Ocean (i.e., Figure 2Ai, Bi, and Ci) have been discussed in detail in previous publications.11,43 Important features of these outputs are as follows. First, the highest eACP10 values assuming perfect persistence (Figure 2Ai) are exhibited by hypothetical neutral organic chemicals with log KAW ranging from −1.5 to −3 and log KOA ≤ 7. These chemicals are present in soils and freshwater sediments only to a limited extent because they exhibit a relatively low affinity for organic matter and tend to be distributed predominantly in aqueous phases (i.e., fresh and marine surface waters). Both long-range atmospheric and oceanic transport are important for determining the eACP10, though. As the affinity for organic matter increases (i.e., log KOA and log KOW increases), longrange transport potential is reduced for several reasons including (i) increased deposition and accumulation in terrestrial reservoirs at lower latitudes and (ii) increased ‘permanent’ losses (i.e., chemical mass transported out of model domain) due to particle sinking from surface to deep ocean waters and sediment burial (freshwater). Relatively volatile chemicals (log KAW > 0, log KOA < 6) also have reduced eACP10 even though long-range atmospheric transport is highly efficient. This model output simply reflects the fact that eACP10 only considers the mass of chemical in surface compartments and net deposition of volatile chemicals to the surface compartments of the North Polar zone is not favorable (e.g., precipitation scavenging is not efficient, even considering the influence of temperature). 926
dx.doi.org/10.1021/es304276g | Environ. Sci. Technol. 2013, 47, 923−931
Environmental Science & Technology
Article
hexachlorocyclohexanes (HCHs).19,24 The final feature of the model results worth noting is that the relative importance of shelf sediments in the Arctic Ocean increases as the susceptibility to degradation increases (see Figure 2Aii vs Cii). These results largely reflect the assumptions regarding the relative stability in water, soil, and sediments (1:2:10). Illustrative Case Study (PCB-153). Selected model outputs from the long-term simulation of PCB-153 excluding and including shelf sediments in the Arctic Ocean are presented and compared in Figure 3. The simulated total mass of PCB153 in the surface compartments (and atmosphere) of the North Polar zone including and excluding shelf sediments in the Arctic Ocean from 1930−2100 is compared in Figure 3A. The total mass of PCB-153 in the shelf sediments (if included) and soil compartments is also displayed. The ratios of total mass, total concentration in the lower atmosphere (CA), and
The role of degradation in reducing eACP10 can be seen in Figure 2Bi and Ci, which presents model output assuming 10:20:100 year baseline degradation half-lives in water, soil, and sediments, respectively, and a baseline kOH of 1 × 10−13 cm3 molecule−1 s−1 and 1:2:10 year baseline degradation half-lives in water, soil, and sediments and a baseline kOH of 1 × 10−12 cm3 molecule−1 s−1, respectively. In addition to the maximum value of eACP10 being reduced (5.5% to 1%), the center of maximum values shifts its position in the chemical space such that the highest values are observed for hypothetical chemicals, which are much more water-soluble (i.e., log KAW < 4). This model output largely reflects the fact that long-range atmospheric transport is increasingly hindered as kOH increases to the values assumed here. The long-range oceanic transport potential is also reduced but to a lesser extent. Another important feature of the model results presented in Figure 2Bi and Ci is that the eACP10 contour plots tend to flatten out as susceptibility to degradation increases (i.e., reduced absolute differences in eACP10 across chemical space). This model behavior is driven by the shift in the relative importance of partitioning properties versus degradation half-lives in determining eACP10. The role of shelf sediments in determining the eACP10 of neutral organic chemicals can be ascertained from Figure 2Aii− Cii and the ratios presented alongside this model output in Figure 2. The largest relative increases in the total mass in the surface compartments of the Arctic environment are observed for volatile hydrophobic organic chemicals (VHOCs, log KAW > 0, log KOA from 6 to 9, log KOW ≥ 6) that exhibit eACP10 including shelf sediments two to greater than 3-fold higher than the model output excluding shelf sediments in the Arctic Ocean. Although these chemicals tend to be predominantly distributed in the atmospheric compartments of the model and are not efficiently transported to the surface compartments via wet deposition (esp. rain events) and dry deposition, the mass of chemicals that is deposited via gaseous deposition becomes strongly associated with suspended particulates (driven by hydrophobicity). In this context, it is more appropriate to consider the air-bulk water partition coefficient (i.e., account for the sorption capacity related to the presence of organic matter in the water column) as opposed to only the air−water partition coefficient (KAW). In the scenario excluding shelf sediments in the Arctic Ocean, this mass of chemical is lost from the surface compartments through particle sinking to the deep ocean (i.e., out of the surface ocean water and hence, model domain). A substantial proportion of this mass is preserved in the scenario including shelf sediments in the Arctic Ocean (i.e., present in the active sediment layer), accounting for the relative increase. “Nonvolatile” chemicals (log KOA > 10, log KOW > 7), which are predominantly distributed to the soil compartment, also exhibit elevated relative increases in eACP10 (i.e., a factor of 2 or more) for the same reason. The eACP10 of neutral organic chemicals with partitioning properties similar to hexachlorobenzene (HCB) and PCBs (see Figure 2) are also elevated in the scenario including shelf sediments by up to a factor of 2. There is also a substantial portion of chemical space where the eACP10 is not influenced by the inclusion of shelf sediments in the Arctic Ocean. These chemicals are less hydrophobic (i.e., log KOW ≤ 5) and hence tend not to associate with organic matter in the water column to a great extent (i.e., are predominantly in the dissolved phase). Therefore, particle sinking is not an efficient transport process, and the shelf sediments are not as important. This type of behavior in the global environment is exemplified by
Figure 3. Model output for the long-term (1930−2100) simulation of PCB-153 illustrating (A) the total mass in the North Polar zone (i.e., Arctic environment) including and excluding shelf sediments as well as the mass in shelf sediments and soils over time; (B) the concentrations in the lower atmosphere of the Arctic (CA) and in surface ocean water (CSO), and the total mass in the scenario including shelf sediments relative to the scenario excluding shelf sediments (i.e., ratios) over time; and (C) the fraction of total mass in soils, shelf sediments and the surface ocean compartment of the North Polar zone over time for the scenario including shelf sediments. 927
dx.doi.org/10.1021/es304276g | Environ. Sci. Technol. 2013, 47, 923−931
Environmental Science & Technology
Article
are only slightly elevated (20−25%) during the period when primary emissions (e.g., those that occur during manufacturing, use, and disposal) strongly dominate (1930−2000). The shelf sediments are still accumulating PCB-153, and resuspension back into the water column is not a significant flux compared to ongoing atmospheric (and oceanic) inputs. However, resuspension of PCB-153 into the water column from shelf sediments is increasingly significant as secondary emission sources (e.g., volatilization from environmental reservoirs such as boreal soils) begin to dominate (2000−2100) resulting in modeled concentrations approaching 3-fold higher. This model behavior translates into a slower response time to emission decreases than would be expected based on the simulations conducted with the version of the model excluding shelf sediments. Given that reported concentrations of PCB-153 (ng g−1 lipid) in walrus (benthic feeder) are similar to or higher than those in ringed seal,46 the potential to use modeled concentrations in Arctic shelf sediments to assess human and ecological exposure via marine benthic food webs over time should be explored further. Implications. The generic assessment (Figure 2) and PCB153 case study (Figure 3) demonstrate that shelf sediments in the Arctic Ocean can play an important role in determining the Arctic Contamination Potential (eACP10) and long-term fate and transport of neutral organic chemicals across a wide range of partitioning properties. The public availability of the IBCAO 3 map means that it is possible to add shelf sediments to globalscale and regional-scale models regardless of the spatial resolution/geographical segmentation and that this modification should be a priority for future versions of such tools. Improved estimates of shelf sediment surface area and inclusion of the sediments out to the shelf break (as opposed to limited to depths ≤200 m) in all oceans are required for modeling exercises seeking to approach a more complete mass inventory.27 With respect to the Arctic Ocean, fate and transport models with greater spatial resolution than the GloboPOP model are better suited for interpreting monitoring data obtained at different depths (e.g., near-shore vs off-shore in one region, shallow parts of the Barents Sea vs deeper sections of the Beaufort Sea) (Figure 1) where sediment-water interactions exert more or less influence on concentration and bioavailability of chemicals in the surface ocean. Incorporating a better representation of the horizontal and vertical mixing regime of the upper water column in different regions of the Arctic Ocean14,15 in addition to shelf sediments should also be considered (if not accounted for already). With respect to identifying new potential Arctic contaminants of concern, the most novel finding of this model exercise is that the eACP10 of VHOCs are higher in relative terms than previously recognized (Figure 2).11−13 Neutral organic chemicals exhibiting log KOW from 6 to 8 and log KOA greater than 6 have relatively high inherent bioaccumulation potential in aquatic piscivorous food chains and medium to high inherent bioaccumulation potential (as log KOA ↑, bioaccumulation potential ↑) in marine and terrestrial mammalian and human food chains, as quantified using the degree of modeled food web magnification.46,47 Note that inherent bioaccumulation potential does not address the role of biotransformation in organisms, a process that can substantially reduce exposure to parent compounds.48,49 While the results for the sets of hypothetical chemicals are useful, a more directly relevant question is whether any “real world” examples of VHOCs are present at levels in the Arctic
total concentration in surface ocean water (CSO) (i.e., model results from scenario including shelf sediments divided by model results from scenario excluding shelf sediments) over time are presented in Figure 3B. The modeled concentrations in the lower atmosphere (CA) are essentially identical throughout the simulation period regardless of the inclusion or exclusion of shelf sediments in the Arctic Ocean. The mass fraction (% of total mass inventory) in different surface compartments over time for the scenario including shelf sediments is presented in Figure 3C. Note that a comparison of model output to monitoring data in the Northern hemisphere (atmosphere, surface ocean water) using the same emission scenario for PCB-153 is presented in the Supporting Information of Armitage et al.44 As shown in Figure 3A and B, the simulated total mass of PCB-153 in the North Polar zone (i.e., Arctic environment) is up to 2-fold higher when shelf sediments are included. The largest differences occur in the late 1990s at which point global emissions have crested and are beginning to decline substantially from the peak values in the 1970s. Modeled concentrations of PCB-153 have reached 60 pg g−1 dry weight in the shelf sediments (3 ng g−1 OC) (SI, Section S4) and then begin to decline with an apparent dissipation rate of approximately 2.2% per year. The modeled concentrations are consistent with OC-normalized sediment levels reported for the upper 5 cm of the Thule and Schades sediment cores sampled in 1997 in northern Baffin Bay (1.3−3.8 ng g−1 OC).36 The model output is also in reasonable agreement with published estimates of the total mass inventory of PCB-153 in shelf sediments of the Arctic Ocean,27 given that assumptions used in those calculations (e.g., active sediment layer depth, surface area) are not identical to the assumptions in the model. Assuming that the fraction of ΣPCB accounted for by PCB-153 observed in the Thule and Schades sediment cores is typical (approximately 5−10%), the model output is also broadly consistent with monitoring data from the Canadian Archipelago and Chukchi Sea, where ΣPCB levels in marine sediment in the range 10−350 ng g−1 OC and 260 pg g−1 dry weight were reported, respectively.29,45 The total mass of PCB-153 in the active layer of the shelf sediments is greater than the total mass of PCB-153 in soils up until the year 2020 (Figure 3A and C). These results illustrate the importance of shelf sediments with respect to the mass inventory of hydrophobic POPs such as PCB-153 in the Arctic environment. Note that the total mass and fraction of total mass shown in Figure 3A and C do not include the mass of chemical present in deeper sediment layers (i.e., transported out of the active layer via sediment burial). Soils in the North Polar zone become relatively more important as a mass reservoir compared to shelf sediments as emissions are reduced and PCB-153 levels decline in the Arctic (and global) environment (Figure 3A and C). These model results reflect the lower residence time of PCB-153 in shelf sediments compared to soils. It is also worth noting that volatilization of PCB-153 from soils in the North Polar zone to the atmosphere is blocked throughout a substantial portion of the year due to the presence of the seasonal snowpack. The potential importance of Arctic shelf sediments when considering human and ecological exposure via marine (pelagic) food webs (e.g., phytoplankton→zooplankton→ Arctic cod→ringed seal) can be assessed by comparing the modeled concentrations in the surface ocean water over time for each scenario (CSO, Figure 3B). As shown in Figure 3B, the modeled concentrations of PCB-153 in the surface ocean water 928
dx.doi.org/10.1021/es304276g | Environ. Sci. Technol. 2013, 47, 923−931
Environmental Science & Technology
Article
Arctic Contamination Potential of neutral organic contaminants. The results indicate that shelf sediments are influential determinants of eACP10 and chemical fate in the Arctic environment for neutral organic chemicals spanning a wide range of partitioning properties and susceptibilities to degradation in the environment. Factors such as emission rate, mode of entry, and spatial distribution should be considered in addition to LRTP (eACP10), however. Collection of additional monitoring data from marine sediments in the Arctic environment may be warranted in certain cases as this compartment can potentially be an important reservoir of hydrophobic POPs and exposure source for benthic-feeding organisms (e.g., walrus, eider duck).46 Sampling campaigns should target shallow/near-shore locations (i.e., depth
