Time Trends of Arctic Contamination in Relation to Emission History

How long does it take for organic contaminant concentrations to decline in the Arctic after regulatory measures have succeeded in reducing emissions g...
0 downloads 0 Views 250KB Size
Environ. Sci. Technol. 2007, 41, 5986-5992

Time Trends of Arctic Contamination in Relation to Emission History and Chemical Persistence and Partitioning Properties TODD GOUIN AND FRANK WANIA* Department of Physical and Environmental Sciences, University of Toronto Scarborough, 1265 Military Trail, Toronto, Ontario, Canada, M1C 1A4

How long does it take for organic contaminant concentrations to decline in the Arctic after regulatory measures have succeeded in reducing emissions globally? This question is explored by using a zonally averaged global distribution model to estimate the lag-time between the period when emissions begin to decrease and when a decline in a chemical’s Arctic Contamination Potential is observed. A long lag is problematic, as contaminant concentrations can continue to increase well after a potential hazard is recognized. Using three different emission scenarios, the chemical property combinations that are most likely to experience a lag on the order of decades were identifed among 96 hypothetical chemicals with different partitioning and reactivity properties. The first such property combination comprises the persistent “swimmers” that reach the Arctic by slow long-range oceanic transport. They require a half-life (t1/2) in water of more than 10 years for a significant lag to occur. The second group of compounds experiencing a long lag includes semivolatile chemicals that are in dynamic exchange between atmosphere and ocean. These “multihoppers”, with air-water partition coefficients, KAW of approximately 0.01, need to be highly persistent in air (t1/2 > 3 years) and surface media (t1/2 > 10 years). Their lag depends both on the oceans’ large storage capacity and relatively low stickiness, i.e., a high likelihood of return to the atmosphere. Notably, no lag is predicted for less water soluble multihoppers (KAW > 1), which are more likely to distribute into soils and foliage, because the terrestrial environment is “stickier” than the oceans, greatly reducing the number of hops these chemical will experience. The oceans thus play a crucial role in facilitating delayed Arctic contamination, either by transporting dissolved contaminants slowly to higher latitudes, or by providing a relatively nonsticky temporary storage reservoir which is in constant exchange with the atmosphere. Precaution advises a swift regulatory response to increasing concentrations in remote marine organisms of substances that have property combinations that are predicted to result in a significant delay between emission reductions and concentration declines.

Introduction Regulating persistent organic pollutants (POPs) under international treaties, such as the Stockholm Convention, is * Corresponding author [email protected]. 5986

9

phone:

(416)287-7225;

e-mail:

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

motivated by the desire to protect remote ecosystems from contaminants used in areas of intensive industrial and agricultural activity (1, 2). The underlying concern is the potential detrimental effect such contaminants may have on sensitive ecosystems (3) and populations (4, 5). Regions with prevailing low temperatures, such as the poles, are believed to be particularly vulnerable to the accumulation of POPs (1-5). Consequently, chemicals combining properties of persistence, potential for bioaccumulation and longrange transport are scrutinized, and their use and manufacture is highly regulated. Several models have been developed to assist regulatory bodies in identifying chemicals that may be subject to accumulation in polar regions. Mu ¨ ller-Herold et al. (6), for instance, rely on a simple model to assess the exposure ratios of organic chemicals in the water and soil compartments of two boxes, representing polar and nonpolar regions. Taking a more complex approach, Scheringer et al. (7) and Wania (8, 9) use zonally averaged global multimedia models to assess an organic chemical’s Cold Condensation Potential (CCP) and Arctic Contamination Potential (ACP), respectively. In particular, Wania (8) defines an eACP indicator for organic chemicals as the fraction of the total globally emitted amount of a chemical that is in Arctic surface compartments after a defined period of time. When this indicator is calculated and plotted as a function of the “chemical space”, defined by the partitioning and degradation properties of hypothetical organic chemicals, it can aid in the identification of organic chemical property combinations that are likely to lead to contamination of the Arctic (9). The properties of a real chemical can be located within this space, and an assessment be made regarding its potential to accumulate in the Arctic. Such models can provide an effective hazard assessment tool for regulatory bodies to adopt when screening thousands of chemicals for their potential to impact remote ecosystems. Previously, the eACP of a substance was calculated after 10 years of its continuous release into the environment (8, 9). This was believed to be sufficient to account for the longterm or delayed contamination potential of a substance (8, 9). This seemed reasonable, considering that the movement of radionuclides from point sources in temperate regions to the Arctic via oceanic transport occurs on the time scale of 4 to 5 years, whereas the atmospheric transport pathway would likely require significantly less time, on the order of days (10-13). In their investigation of the CCP of POPs, however, Scheringer et al. (7) predicted the concentration of mirex in Arctic soil to increase for 18 years following its initial release into the environment, whereas other substances reached a maximum of Arctic contamination within 2 years of their release. This suggests that the time before maximum concentrations in polar regions are observed is chemicalspecific, i.e., is dependent on a compound’s partitioning and persistence attributes. The time for contamination in the Arctic to reach a maximum is also influenced by how a substance is released to the environment, i.e., the estimated ACP-values are strongly dependent on where, when, and for how long an emission took place. The emission profile for many POPs, such as the polychlorinated biphenyls (PCBs), followed a bell-shaped curve, where a maximum associated with peak manufacture and usage was followed by a decline due to the implementation of regulatory measures (14, 15). Most models that currently assess the environmental fate and distribution of chemicals, however, assume a steady-state release scenario, where chemical is constantly being emitted into the model environment. An assessment of the time response to emission 10.1021/es0709730 CCC: $37.00

 2007 American Chemical Society Published on Web 08/01/2007

changes is thus difficult. The use of time-variant emission trends should allow for a more realistic estimate of the time required for chemical concentrations to decline after emission reductions. Assessing the time when a maximum ACP for a substance is reached in relation to the emission profile is particularly important when evaluating the effectiveness of regulatory action. It is likely that some substances respond rapidly to reductions in usage and manufacture, while a lag time between peak usage and maximum ACP is expected for others. Here we investigate how temporal emission variability influences the time course of the ACP of hypothetical chemicals with different partitioning properties and reactivities. The aim is a better understanding of how chemical properties influence the lag time between the moment when emissions of a regulated substance begin to decline and the time when polar contamination with that substance peaks.

Method Global Distribution Model Globo-POP. Globo-POP, a zonally averaged, dynamic global distribution model, was used to calculate the contamination of the Arctic for 96 hypothetical chemicals with different partitioning and reactivity properties. The same model was used in previous studies on the ACP (8, 9) and consists of 10 interconnected multimedia box models, each with four air, two soil, one sediment, and two water compartments, that are representative of well-mixed latitudinal bands of homogeneous climatic conditions. Chemical transport is described two-dimensionally in the atmosphere and in one dimension in the surface ocean. This implies a simplistic macrodiffusive representation of ocean mixing that ignores advection. Model parameters are described in detail elsewhere (16, 17) and in the Supporting Information. Arctic Contamination Potential. Arctic contamination is quantified with the eACP (8, 9), defined as the ratio of the amount present in Arctic surface media to the cumulative globally emitted amount:

eACP ) (mT1 - mA1)/eTG × 100%

(1)

where mT1 and mA1 are the amount of chemical in all compartments and in the four atmospheric compartments of zone 1 (N-Polar) of the Globo-POP model, respectively, and eTG is the cumulative amount of the chemical emitted globally. The eACP provides a relative indicator of the hazard associated with a substance with respect to its potential to contaminate the Arctic. It is not an indicator of risk, since the value of eACP is normalized by the emission rate and thus not indicative of any exposure. Unlike previous studies (8, 9), where the eACP was calculated using a generic steady-state emission scenario, we investigate here the influence of the emission profile on the eACP, with an emphasis on estimating the lag-time associated with a chemical’s potential to contaminate the Arctic. We define the lag as the length of time between the peak in emissions and the time when a maximum eACPvalue is observed. For comparison, three emission scenarios are employed, whereby the chemical release is to the atmosphere and scaled according to the zonal distribution of the global human population (8, 9). This implies release primarily to the temperate, subtropical, and tropical zones of the Northern hemisphere. The first scenario, “pulse”, assumes a 1-year emission pulse followed by no emissions for the next 99 years. In the second scenario, “bell curve”, emissions increase gradually for the first 25 years, reach a maximum, and then decrease steadily to zero over the next 25 years. Model calculations are continued for an additional 50 years. Emissions remain constant for 100 years in the third scenario, “steady”.

FIGURE 1. Hypothetical chemical space, characterizing the major transport modes for organic chemicals, based on their partitioning properties log KAW and log KOA (adapted from ref 9). Six hypothetical property combinations have been selected to represent chemicals that will be transported as single hoppers (SH), swimmers (SW), multiple hoppers (MO, MM, and MT), and fliers (FL). Hypothetical Chemicals. Chemicals undergo four modes of global transport depending on their physical-chemical properties (9). On the basis of their most probable mode of transport, they can be classified as being either “fliers”, “multiple hoppers”, “swimmers”, or “single hoppers” (Figure 1). Swimmers and multiple hoppers are estimated to have the greatest ACP (9). The eACP was calculated for six hypothetical combinations of partitioning properties (Table 1). These include particle-bound chemicals (SH), watersoluble chemicals (SW), water soluble and relatively volatile chemicals (MO), multimedia chemicals (MM), very hydrophobic, semivolatile chemicals (MT), and volatile chemicals (FL). They are subject to different modes of global transport, as specified in Table 1. To render these six chemical types less abstract, we can further identify real persistent chemicals with similar combinations of partitioning properties (Table 1). For each combination of partitioning properties, 16 classes of environmental reactivity were considered by varying by increments of a factor of 10 both the degradation half-lives (t1/2) in surface media (ranging from 876 to 876 000 h, i.e. from 0.1 to 100 years), and the reaction rate constant with OH radicals in air (ranging from 10-15 to 10-12 cm3 molecule-1 h-1, corresponding to degradation half-lives between 1 day and 30 years). Note that the reference t1/2 refers to water and does not include processes other than degradation. The t1/2 for soil and sediment are estimated to be double and ten times that of the t1/2 in water, respectively. For the resulting 96 hypothetical chemicals, Globo-POP calculated the eACP as a function of time for each of the three emission scenarios.

Results and Discussion Variability of the Arctic Contamination Potential with Time. The time course of the eACP over 100 years for all 288 simulations (6 partitioning property combinations × 16 degradation property combinations × 3 emission scenarios) is presented in the SI. Selected results for the six hypothetical chemical partitioning property combinations, when assumed to be highly persistent in air and surface media, are displayed in Figure 2. The different partitioning property combinations influence both the absolute eACP values (8, 9) and the temporal variability of the eACP. This is consistent with earlier studies that also reported temporal differences in Arctic contamination (7). For all three emission scenarios, the chemical property combinations with the highest maximum VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5987

TABLE 1. Partitioning Property Combinations, Mode of Global Transport and Example Chemicals of Similar Partitioning Behavior for Each of the Six Types of Hypothetical Chemicals Used in the ACP Calculations ID

KAW

KOA

SH

10-4

10+12

single hopper

SW MO

10-4 10-2

10+5 10+3

MM

10-2

10+7

swimmer multiple hopper exchanging with the oceans multimedia multiple hopper

MT

10+1

10+7

FL

10+1

10+5

mode of global transport

multiple hopper exchanging with terrestrial surfaces flier

examples decabromodiphenyl ether (decaBDE), polychlorinated dioxins and furans (PCDD/Fs) perfluorocarboxylates (PFCAs), trifluoroacetic acid (TFA) short perfluorinated alcohol, e.g., perfluoropropanol, lighter polychlorinated biphenyls (PCBs) and hexachlorobenzene (HCB) long chain perfluoroalkanes fluorotelomer alcohols (FTOHs)

to the latter scenario, which is most realistic and most relevant within a regulatory context. The types of chemicals experiencing such a delay and the underlying mechanisms will be discussed in detail below. First, however, we would like to highlight that the lines in the plots of Figure 2 cross, implying that a set of chemicals could be ranked differently in terms of their eACP, depending on the chosen time interval. For instance, after 10 years of steady emissions (Figure 2C) the eACP10 of the swimmer (SW) is less than that of the multihoppers (MO and MM). After 50 years, however, a change in the ranking is observed, where the eACP50 for MO > SW > MM, and after 100 years the eACP100 ranks SW > MO > MM. Similar changes in ranking are observed for the other two emission scenarios (Figure 2A,B). This suggests that the temporal variability of the ACP may be important when ranking persistent chemicals in terms of their potential to contaminate the Arctic. It is of interest to note that the differences in ranking is largely influenced by differences in the rate of decline between the six chemicals, following their peak in eACP. Whereas their rate of increase is similar, the rate of decline for the multihoppers, MO and MM, is faster than for the swimmers. For instance, in the pulse and bell-curve emission scenarios the multihoppers approach an eACP of 0% after 100 years, while the eACP of the SW is still about 2.5% after 100 years. The differences in these rates of decline are likely related to differences pertaining to how these substances are transported to the Arctic, as discussed later.

FIGURE 2. Absolute arctic contamination potential eACP versus time for each of the three emission scenarios and all six chemical partitioning property combinations, assuming persistence in both surface media (t1/2 > 100 years) and atmosphere (kair ) 10-14 cm3 molecule-1 h-1, t1/2 ≈ 3 years).

eACP values (>2%, SW, MO, and MM) display the most significant delay in the occurrence of that maximum. This delay can be on the order of several decades (Figure 2A and 2B), which suggests that the levels of some contaminants in the Arctic could continue to increase, relative to the total amount of the chemical that has been emitted globally, for decades following the implementation of regulatory restrictions. Because similar lag-time behavior is calculated in the pulse and the bell-curve emission scenarios (Figure 2), we shall restrict the remaining discussion to results pertaining 5988

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

Relationship between Persistence, the ACP, and the Occurrence of a Lag. The reactivity of a substance strongly influences the maximum eACP that it can achieve (9). Chemicals that are highly degradable are unlikely to contaminate the Arctic to any significant extent, simply because their transport to the Arctic is limited. For instance, substances with t1/2 in surface media less than a year and atmospheric t1/2 less than approximately 10 days (kA > 10-12 cm3 molecule-1 s-1) have maximum eACP values that are below 0.75% (9). Not only do degradable chemicals have low absolute eACP values, but they also show no significant delays in the occurrence of maximum eACP: Their highest eACP values generally occur less than 2 years after a chemical’s initial release into the environment. The influence of the t1/2 in air and surface media on the occurrence of a lag-time is illustrated quantitatively in Figure 3. No lag is observed for any of the six chemical property combinations when the t1/2 in surface media is 1 year or less. Similarly, no lag is simulated when the t1/2 in air decreases below 100 days. The only exception is the swimmers, for which the degradation rate constant in air has an extremely limited influence on the length of the lag (Figure 3B). Whereas the multihoppers (Figure 3C, MT and MM) only show a significant lag when they are persistent in both air and surface media, a lag-time never occurs for SH (Figure 3A) and FL (Figure 3F), regardless of their persistence.

FIGURE 3. Dependence of the length of the lag time between the peak in emission, from the bell-curve emission scenario, and the maximum eACP, in years on the persistence in air and surface media for the six chemical property combinations (Figure 1 and Table 1) (d ) day; a ) year; h ) hour).

The results displayed in Figures 2 and 3 suggest that only chemical partitioning property combinations SW, MO, and MM experience significant delays in achieving maximum eACP. Chemicals with the partitioning property combinations SW, MO and MM need to be persistent in surface media (t1/2 of at least a few years) for a delay to occur. Chemicals with partitioning properties corresponding to MO and MM additionally need to be very persistent in air (t1/2 g 1 year) for the delay to manifest itself, whereas the SW show a delay even if the chemical is degradable in air. Mechanism of Delayed Arctic Contamination by Swimmers. Among the six hypothetical partitioning property combinations investigated in this study, swimmers are predicted to experience the greatest lag-time associated with their movement to the Arctic, if their persistence in surface media exceeds 1 year (Figures 2 and 3). Such chemicals are both soluble and persistent in water and therefore subject to long-range oceanic transport (LROT) to the Arctic. The delay in reaching the Arctic is related to the slow pace of oceanic mixing, as parametrized in the Globo-POP model. The increase in lag-time with increasing persistence in water can be understood by visualizing a pool of contaminated seawater slowly dispersing to higher latitudes, while at the same time being depleted by degradation. In order to confirm the role of LROT in the occurrence and magnitude of a lag for swimmers, eACP calculations for the most persistent chemical property combination were repeated with the oceanic mixing terms set to zero. By shutting off the oceanic transport pathway, we are able to differentiate between the relative importance of advection due to ocean currents, and the atmospheric advection pathway. Both the eACP value and the occurrence of a lagtime for SW are dramatically reduced when the oceanic advection term is set to zero (Figure S8A). This implies that the mechanism influencing the occurrence of a lag time for the swimmers in the Globo-POP model is related to the

mobility of these chemicals in the oceans and confirms that the atmospheric pathway for these substances is minor (see discussion in relation to perfluroalkyl compounds in the Supporting Information). Thus, for persistent chemicals with properties of swimmers, models, such as Globo-POP, will require careful parametrization of the movement of ocean currents to adequately predict their global environmental fate and their time trends in Arctic regions. Mechanism of Delayed Arctic Contamination by Multihoppers. High eACP values and lag times on the order of 10 to 15 years are observed for multihoppers MO and MM that are persistent in air (t1/2 > 3 years) and slow to degrade in surface media (t1/2 > 10 years) (Figure 3). The length of the lag for these substances depends strongly on their reactivity in air. Whereas chemicals with t1/2 in air of 100 days or less are predicted to experience a rapid response to emission reductions, (Figure 3), the predicted lag can increase to just over a decade when the t1/2 in air is above 3 years. The partitioning properties of MO and MM are at the limit of the chemicals classified as swimmers (Figure 1), and it is conceivable that their mechanism of delay is the same as that for the swimmers, i.e., the residence time of these multihoppers in the oceans may be sufficiently long to allow for LROT to the Arctic. Calculations without ocean mixing, however, reveal that for chemicals MO and MM, eliminating the possibility of LROT has only a minor effect on lag-time and eACP (Figure S8B). Multi-hoppers MO and MM therefore enter the Arctic primarily via atmospheric transport, and their delay in reaching maximum arctic contamination cannot be related to slow oceanic transport. Instead, it appears that the process of “grass-hopping”, that these substances are believed to undergo when persistent in air, is responsible for the slow build-up of arctic contamination even after emissions have peaked (1, 2). To investigate this, we calculated the average number of hops nhop that these chemicals will experience, VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5989

FIGURE 4. The average number of hops calculated with Globo-POP for different half-life combinations of the multihopper MM using the bell-shaped emission scenario (A), and the influence on the eACP of the most persistent MM of reducing the ocean depth to 2 m (B). adopting the definition for nhop by Gouin et al. (18):

nhop ) NSANAS/(NRSNRA + NRSNAS + NRANSA)

(2)

where NSA is the volatilization flux from all surfaces, NAS is the deposition flux to all surfaces, NRS is the degradation rate in all surface media, and NRA is the rate of degradation in air, for the entire global environment and averaged for the entire 100 year simulation period (all in units of kg‚h-1). Calculating fluxes and rates using Globo-POP, we estimated nhop for different half-life combinations of partitioning property combination MM. Figure 4A indicates that only chemicals that are persistent in both air and surface media experience a large number of hops. The most persistent combination of the multihopper MM will undergo an average of approximately 90 hops. As the reactivity of a substance with partitioning properties MM increases, both the lag and nhop is greatly reduced. The interpretation is that the repeated cycles of deposition and evaporation require considerable time (on the other order of many years and decades) and thus high persistence in both surface and atmosphere. If sufficiently persistent, however, this cycling process can continue to deliver contaminant to remote areas for years following emission reductions. This reduction in nhop with decreasing persistence is related to “stickiness” (19), a term which describes the fraction of a substance that is not available for exchange because it degrades in a compartment (18). Grass-hopping, conversely, is dependent on the fraction that is available for exchange (18). Highly “sticky” substances are either reactive or partition predominantly into a single compartment. Stickiness can be defined for surface media as well as air (18). Swimmers, for instance have high stickiness in water, fliers have high stickiness in air, and therefore these substances do not experience significant hopping. An interesting observation is that among the three multihopping partitioning property combinations, MT shows much less of a lag time than MO and MM (Figure 3). This can also be explained by hopping and stickiness, as chemical MT is more “sticky” and experiences much lower nhop than either MO or MM, for a given combination of half-lives. The reason is that the highly hydrophobic MT exchanges mostly between terrestrial surfaces and the atmosphere, whereas the lower log KAW of MO and MM allows them to exchange with the world’s oceans. Stickiness in the oceans tends to be lower than in terrestrial surfaces (20). We can use eq 2 to calculate nhop separately for terrestrial and aquatic surfaces 5990

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

in the total global environment, which indeed reveals that the average number of hops for the most persistent chemical MM is much higher for the oceans (nhop ) 88) than the terrestrial environment (nhop ) 12). This dominance of the oceans is partly due to their large surface area in the GloboPOP model environment, causing a significant fraction of contaminant to be deposited to the oceans, making them a large reservoir for exchange with the atmosphere (20, 21). We should caution, however, that a zonally averaged model such as Globo-POP will tend to overstate the importance of oceans, because it ignores that most contaminant sources are based on the continents. To visualize the influence that the oceans have on a lag for the multihoppers MO and MM, we calculated the eACP in a hypothetical model environment where the depth of the surface ocean is reduced from 200 to 2 m. Reducing the ocean volume, and thus the storage capacity of this compartment, dramatically decreases the magnitude of the eACP and eliminates the occurrence of a lag for the multihoppers (Figure 4B). At the same time, reducing the storage capacity of the oceans causes an increase in nhop, as it results in increased exchange between atmosphere and ocean. This, incidentally, highlights that there is no simple correlation between lag time and nhop, i.e. extensive hopping does not necessarily imply delayed Arctic contamination. The oceans thus play a crucial, though different, role in the delayed Arctic contamination of both swimmers and multihoppers (13). For the former it is the slow transport medium to the Arctic, and for the latter it is a relatively nonsticky storage reservoir which is in constant exchange with the atmosphere. Many of the Stockholm POPs have partitioning properties corresponding to the MM. The atmospheric degradation halflives of most POPs, for instance the PCBs and PCDD/Fs, are less than 100 days (22-24), with only HCB having an air t1/2 of almost 3 years (23). Even if we assume that the surface t1/2 for these POPs is greater than 10 years, Globo-POP predicts that their Arctic contamination would not continue to increase after their widespread ban, i.e., that these POPs will not experience a lag. As an illustrative example, measurements indicate that concentrations of DDT, PCBs, and HCB in Arctic biota declined sharply, shortly following the implementation of regulatory measures in the 1970s and 1980s (3, 25-27), as did levels of the HCHs in surface waters and Arctic air (13). Interestingly, this also suggests that these POPs will experience only a limited number of hops (Figure 4A), i.e., their persistence in air is too small to allow for extremely efficient grass-hopping. This is consistent with

previous modeling studies aimed at quantifying the grasshopper effect (18, 21). Only substances that are very persistent in air, such as HCB (23) or perfluoropolyethers (28), can be expected to truly express the global transport and distribution behavior associated with the grasshopper effect. Regulatory Implications. By suggesting that highly persistent organic contaminants, which move in the environment as either swimmers or multihoppers, have the potential for relatively long lag-times, the results from this study are particularly relevant within a regulatory context. The study results imply that, following the implementation of regulatory action to reduce emissions of such substances, their levels in remote marine regions could increase for several years before they begin to decline. For persistent and bioaccumulative “swimmers”, we believe that a strong argument can be made for implementing the precautionary principle as advocated in the Stockholm Convention on POPs. If such substances show increasing concentrations in remote marine organisms, swift action to limit their release may be even more urgent than for atmospherically transported contaminants, because the levels in the environment will respond slowly to regulatory initiatives (see discussion in Supporting Information). The situation is not unlike that observed for issues relating to ozone-depleting substances, greenhouse gases, and sulfur, where a significant time lag exists between emission reductions and maximum detrimental effects (29-32). The problem may be further exacerbated if there is an additional lag between restrictions on the manufacture and use of a substance and a decline in its emissions. Such lags would be expected if emissions are not only associated with manufacturing processes, but also with the use and disposal of products containing the substance, especially if that use occurs for extended periods of time (15). Clearly, precaution also advises increased efforts toward proactively identifying substances that may be persistent in water and have the potential to bioaccumulate.

Acknowledgments We are grateful to David Nguyen for performing some of the computer simulations and to the Natural Sciences and Engineering Research Council of Canada for funding.

Supporting Information Available Additional figures with model results, including a discussion on the relevance of the model results for the transport pathways and time trends of persistent perfluroalkyl compounds in the Arctic. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Wania, F.; Mackay, D. Global fractionation and cold condensation of low volatility organochlorine compounds in polarregions. Ambio 1993, 22, 10-18. (2) Wania, F.; Mackay, D. Tracking the distribution of persistent organic pollutants. Environ. Sci. Technol. 1996, 30, 390A-396A. (3) Muir, D.; Braune, B.; DeMarch, B.; Norstrom, R.; Wagemann, R.; Lockhart, L.; Hargrave, B.; Bright, D.; Addison, R.; Payne, J.; Reimer, K. Spatial and temporal trends and effects of contaminants in the Canadian Arctic marine ecosystem: A review. Sci. Total Environ. 1999, 230, 83-144. (4) Godduhn, A.; Duffy, L. K. Multi-generation health risks of persistent organic pollution in the far north: use of the precautionary approach in the Stockholm Convention. Environ. Sci. Policy 2003, 6, 341-353. (5) Odland, J. O.; Deutch, B.; Hansen, J. C.; Burkow, I. C. The importance of diet on exposure to and effects of persistent organic pollutants on human health in the Arctic. Acta Paediatr. 2003, 92, 1255-1266. (6) Mu ¨ ller-Herold, U.; Smieszek, T.; Peter, P.; Scheringer, M.; Morosini, M. A simple measure for precautionary assessment of organic chemicals with respect to global cold condensation. Ecol. Model. 2006, 194, 266-273.

(7) Scheringer, M.; Wegmann, F.; Fenner, K.; Hungerbu ¨ hler, K. Investigation of the cold condensation of persistent organic pollutants with a global multimedia fate model. Environ. Sci. Technol. 2000, 34, 1842-1850. (8) Wania, F. Assessing the potential of persistent organic chemicals for long-range transport and accumulation in polar regions. Environ. Sci. Technol. 2003, 37, 1344-1351. (9) Wania, F. Potential of degradable organic chemicals for absolute and relative enrichment in the Arctic. Environ. Sci. Technol. 2006, 40, 569-577. (10) Kershaw, P.; Baxter, A. The transfer of reprocessing wastes from north-west Europe to the arctic. Deep-Sea Res. Part II-Top. Stud. Oceanogr. 1995, 42, 1413-1448. (11) Gobeil, C.; Macdonald, R. W.; Smith, J. N.; Beaudin, L. Atlantic water flow pathways revealed by lead contamination in Arctic basin sediments. Science 2001, 293, 1301-1304. (12) Li, Y.-F.; Macdonald, R. W.; Jantunen, L. M. M.; Harner, T.; Bidleman, T. F.; Strachan, W. M. J. The transport of β-hexachlorocyclohexane to the western Arctic Ocean: A contrast to R-HCH. Sci. Total Environ. 2002, 291, 229-246. (13) Li, Y. F.; Macdonald, R. W.; Ma, J. M.; Hung, H.; Venkatesh, S. Historical R-HCH budget in the Arctic Ocean: The Arctic Mass Balance Box Model (AMBBM). Sci. Total Environ. 2004, 324, 115-139. (14) Li, Y.-F.; Scholtz, M. T.; Van Heyst, B. J. Global gridded emission inventories of R-hexachlorocyclohexane. Environ. Sci. Technol. 2003, 37, 3493-3498. (15) Breivik, K.; Sweetman, A.; Pacyna, J. M.; Jones, K. C. Towards a global historical emission inventory for selected PCB congeners - A mass balance approach: 2. Emissions. Sci. Total Environ. 2002, 290, 199-224. (16) Wania, F.; Mackay, D. A global distribution model for persistent organic chemicals. Sci. Total Environ. 1995, 161, 211-232. (17) Wania, F.; Mackay, D. The global distribution model. A nonsteady state muti-compartmental mass balance model of the fate of persistent organic pollutants in the global environment; Technical report and computer program on CD-ROM; www.utsc.utoronto.ca/∼wania (accessed July 26, 2007), 2000. (18) Gouin, T.; MacKay, D.; Jones, K. C.; Harner, T.; Meijer, S. N. Evidence for the “grasshopper” effect and fractionation during long-range atmospheric transport of organic contaminants. Environ. Pollut. 2004, 128, 139-148. (19) Beyer, A.; Mackay, D.; Matthies, M.; Wania, F.; Webster, E. Assessing long-range transport potential of persistent organic pollutants. Environ. Sci. Technol. 2000, 34, 699-703. (20) Beyer, A.; Matthies, M. Long-range transport potential of semivolatile organic chemicals in coupled air-water systems. Environ. Sci. Pollut. Res. 2001, 8, 173-179. (21) Semeena, V. S.; Lammel, G. The significance of the grasshopper effect on the atmospheric distribution of persistent organic substances. Geophys. Res. Lett. 2005, 32, 1-5. (22) Brubaker, W. W.; Hites, R. A. Jr. OH reaction kinetics of polycyclic aromatic hydrocarbons and polychlorinated dibenzo-p-dioxins and dibenzofurans. J. Phys. Chem. A 1998, 102, 915-921. (23) Brubaker, W. W., Jr.; Hites, R. A. OH reaction kinetics of gasphase R- and γ-hexachloro-cyclohexane and hexachlorobenzene. Environ. Sci. Technol. 1998, 32, 766-769. (24) Anderson, P. N.; Hites, R. A. OH radical reactions: The major removal pathway for polychlorinated biphenyls from the atmosphere. Environ. Sci. Technol. 1996, 30, 1756-1763. (25) Bignert, A.; Olsson, M.; Persson, W.; Jensen, S.; Zakrisson, S.; Litze´n, K.; Eriksson, U.; Ha¨ggberg, L.; Alsberg, T. Temporal trends of organochlorines in Northern Europe, 1967-1995. Relation to global fractionation, leakage from sediments and international measures. Environ. Pollut. 1998, 99, 177-198. (26) Kannan, K.; Falandysz, J.; Yamashita, N.; Tanabe, S.; Tatasukawa, R. Temporal trends of organochlorine concentrations in codliver oil from the southern Baltic proper, 1971-1989. Mar. Pollut. Bull. 1992, 24, 358-363. (27) Hickey, J. P.; Batterman, S. A.; Chernyak, S. M. Trends of chlorinated organic contaminants in Great Lakes trout and walleye from 1970 to 1998. Arch. Environ. Contam. Toxicol. 2006, 50, 97-110. (28) Young, C. J.; Hurley, M. D.; Wallington, T. J.; Mabury, S. A. Atmospheric lifetime and global warming potential of a perfluoropolyether. Environ. Sci. Technol. 2006, 40, 2242-2246. (29) Montzka, S. A.; Butler, J. H.; Myers, R. C.; Thompson, T. M.; Swanson, T. H.; Clarke, A. D.; Lock, L. T.; Elkins, J. W. Decline in the tropospheric abundance of halogen from halocarbons: Implications for stratospheric ozone depletion. Science 1996, 272, 1318-1322. VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5991

(30) Mahlman, J. D. The long timescales of human-caused climate warming: Further challenges for the global policy process. Adv. Econ. Environ. Resources 2006, 3-29. (31) Wetherald, R. T.; Stouffer, R. J.; Dixon, K. W. Committed warming and its implications for climate change. Geophys. Res. Lett. 2001, 28, 1535-1538. (32) Snucins, E.; Gunn, J.; Keller, B.; Dixit, S.; Hindar, A.; Henriksen, A. Effects of regional reductions in sulphur deposition on the

5992

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

chemical and biological recovery of lakes within Killarney Park, Ontario, Canada. Environ. Monit. Assess. 2001, 67, 179-194.

Received for review April 24, 2007. Revised manuscript received June 22, 2007. Accepted June 28, 2007. ES0709730