Spatial and Temporal Trends of POPs in Norwegian and UK

One can imagine two different mechanisms by which global fractionation can occur. .... PCB 180, 1000, 1100, 1500, 340, 410, 820, 2400, 2800, 290, 180,...
0 downloads 0 Views 172KB Size
Download PDF
Environ. Sci. Technol. 2003, 37, 454-461

Spatial and Temporal Trends of POPs in Norwegian and UK Background Air: Implications for Global Cycling S. N. MEIJER,† W. A. OCKENDEN,† E. STEINNES,‡ B. P. CORRIGAN,† AND K . C . J O N E S * ,† Environmental Science Department, Institute of Environmental and Natural Sciences, Lancaster University, Lancaster LA1 4YQ, UK, and Department of Chemistry, Norwegian University of Science and Technology, N-7034 Trondheim, Norway

Data are presented for PCBs and HCB measured by passive air samplers (SPMDs) along a latitudinal transect from the south of the UK to the north of Norway during 19982000. This work is part of an ongoing air sampling campaign in which data were previously gathered for 1994-1996. Comparisons of the masses of chemicals sequestered by the SPMDs during these different time intervals are used to investigate spatial and temporal trends. Results are discussed in the context of sources, long-range atmospheric transport, fractionation/cold condensation, and global clearance processes controlling ambient levels of POPs. Spatial trends show a decrease in absolute sequestered amounts of PCBs with increasing latitude i.e., with increasing distance from the source area. However, relative sequestered amounts of the homologue groups (expressed as a ratio to penta-PCB) show a clear latitudinal trend, with the relative contribution of the lighter congeners increasing with increasing latitude, providing evidence of latitudinal fractionation. Absolute amounts of HCB increase with latitude, suggesting this compound is undergoing cold condensation. Sequestered amounts of PCBs generally decreased between the two sampling periods by a factor 2-5 over 4 years, suggesting half-lives on the order of 1.7-4 years. The relative rates of decline (19982000 data as a percentage of the 1994-1996 data) were compared for different congeners and latitudes. No clear latitudinal trends were found, with all sites/congeners showing a similar marked decline over time to ca. 30% of the former value. We discuss the interpretation of these observations and conclude they imply that the underlying trends of current ambient levels of PCBs in European background air are still largely controlled by primary emissions, rather than recycling/secondary emissions from the major environmental repositories such as soils or water bodies.

Introduction Concerns over persistent organic pollutants (POPs) in the environment arise because of their susceptibility to long* Corresponding author phone: +44-1524-593972; fax: +44-1524593985; e-mail: [email protected]. † Lancaster University. ‡ Norwegian University of Science and Technology. 454

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 3, 2003

range atmospheric transport (LRAT), persistence, bioaccumulation tendency, and potential toxicity. This has led to international measures to control their release into the environment (1). Environmental monitoring data are required to assess the effectiveness of these source reduction measures. These data need to provide information on the spatial and temporal trends of POPs in different media (2), to help understand the processes that influence their fate and transport on a regional and global scale. The atmosphere, being the main transport mechanism for these compounds, is a particularly important medium for investigation. However, there is often a lack of reliable trend data for POPs in air, and the data that are available cover very few areas, primarily the Great Lakes (3, 4), parts of Europe (5-7), and the Arctic (810). This provides rather incomplete information on the regional and global picture and makes it difficult to assess spatial and temporal changes in air concentrations, which is essential to understanding global POPs dynamics. Many factors control air concentrations of POPs, namely primary and secondary emissions, advection, and various loss processes (see Figure 1). Primary emissions of POPs to the atmosphere include leakage of PCBs from buildings, waste disposals, etc. or direct application of pesticides, while secondary emissions include revolatilization from environmental reservoirs such as soils and water bodies (11, 12). Environmental loss processes include OH-radical degradation in the atmosphere and loss processes in soil or water bodies, such as biodegradation, occlusion into organic matter (e.g. the formation of bound residues), and physical removal processes (e.g. burial in deep ocean waters or soils), which deplete the “pool” of recyclable POPs. These factors will all vary spatially and temporally, for example due to differences in temperature, and with the physical/chemical properties of the chemical, leading to a highly complex situation regionally/globally. It has been suggested that semivolatile chemicals such as POPs could volatilize from warm source/usage areas and migrate to colder remote regions via LRAT (13), where they would be deposited onto surfaces such as soil, plants, or water bodies. This idea, which has also been referred to as “cold condensation”, was taken a step further by Wania and Mackay (14) who suggested a “global fractionation” might occur, whereby a POPs mixture would become fractionated during LRAT, based on the ambient temperature and the physical-chemical properties of the individual compounds. For example, the relative concentrations of high volatility congeners in the PCB mixture would increase with increasing latitude. One can imagine two different mechanisms by which global fractionation can occur. First, in a “primary source” scenario, air concentrations would be dominated by emissions from primary sources. After release from a primary source, the chemical is deposited and subsequently prevented from volatilizing, e.g. through permanent retention in the soil or burial in deep ocean waters. In this scenario environmental reservoirs would act as sinks. Different chemicals have different LRAT potential, dependent on their modes of deposition and atmospheric degradation rates. This would result in fractionation of a chemical mixture away from the source. In this primary source scenario, absolute amounts would therefore be expected to decrease with latitude/distance from the source area. A second scenario is a world dominated by secondary sources. Here the emission of chemicals from environmental reservoirs would control levels in the atmosphere. Repeated air-surface exchange would see the chemical move in a series of “hops” (the “grasshopper effect”). This would also result in fractionation, becoming more pronounced over time. In this case, absolute concen10.1021/es025620+ CCC: $25.00

 2003 American Chemical Society Published on Web 12/17/2002

FIGURE 1. Conceptual diagram of input and loss mechanisms controlling atmospheric concentrations of POPs (adapted from ref 6). In the primary source scenario there is more physical removal of the chemical (as indicated by the larger arrow) and therefore less revolatilization to the atmosphere (as indicated by the smaller arrow). trations of some chemicals may become higher at higher latitudes (“cold condensation”). Temporal trends in atmospheric concentrations would be expected to show a shift in relative concentrations over time at one site, e.g. more volatile congeners becoming more abundant over time in higher latitudes. There are still major uncertainties as to whether environmental reservoirs act as sources or as sinks, i.e., whether primary (scenario 1) or secondary (scenario 2) sources are controlling ambient levels of POPs at the present time (see Figure 1). To resolve these issues, simultaneous measurements should be taken at multiple locations over a long period of time. Since active sampling methodologies are prohibitively expensive, considerable incentive exists to develop passive air sampling procedures. Several approaches are being developed. They have recently been reviewed, and the potential problems and research needs discussed (15). An approach we first used for regional scale monitoring was to deploy semipermeable membrane devices (SPMDs) as passive air samplers (16). SPMDs consist of a polyethylene dialysis bag filled with 1 mL of triolein lipid and have been gaining acceptance as an approach to short- and long-term, inte-

grated passive air sampling for POPs (16-22). There is no need for a power supply, which is required with conventional active sampling using high-volume air samplers (HiVols), hence they are easy to deploy at remote sites. SPMDs are therefore useful for large-scale monitoring studies with high spatial resolution. Recently, a spatial study was carried out in the northwest of England to help pinpoint local sources; SPMDs gave good reproducibility between replicate samplers and reflected spatial differences in air quality and composition (22). During 1994-1996 we deployed SPMDs as passive air samplers at remote sites on a latitudinal transect from southern England to northern Norway, to look for evidence of the LRAT of PCBs and their potential to undergo global fractionation (14, 23). These sites were carefully selected, away from local sources. Full details of this campaign are given elsewhere (16). Our intention is to continue to use the same sampling network over the long-term, to gain insights into how concentrations of PCBs and other POPs change in background air spatially and temporally. In this paper, we present new data from samplers deployed on the network during 1998-2000 in an identical manner to the earlier study and compare the results with the 19941996 deployment. Comparison of the masses of chemical sequestered by the SPMDs during these different time intervals is used to discuss spatial and temporal trends in the context of sources, LRAT, fractionation, and global clearance processes controlling ambient levels of POPs.

Methods SPMDs. Standard U.S. Geological Survey (USGS) SPMDs (8090 cm × 2.5 cm, 75 µm membrane thickness, 1 mL triolein) were purchased from Environmental Sampling Technologies, St. Joseph, MO. Sampling Sites and Deployment. Two SPMDs were deployed at each site in Stevenson screen boxes. Care was taken not to contaminate the SPMDs during deployment. To this end, the SPMDs were transported in airtight containers and handled with solvent-rinsed tweezers at all times. The sampling sites where the SPMDs were deployed are the same as the ones used by Ockenden et al. (16) in the 1994-1996 campaign, except for two sites. An additional site was added (site 10) and the Svalbard site (79 °N, previously site K) was

FIGURE 2. SPMD sampling sites. VOL. 37, NO. 3, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

455

TABLE 1. Sample Information deployment deployment temp range temp range mean site ref 1994-1996 (16) 1998-2000 1994-1996 latitude annual a site (°N) temp (°C) min max min max (16) 1 2 3 4 5 6 7 8 9 10 11 12

50.75 52.45 54.02 56.10 58.05 58.53 61.25 61.33 64.97 67.38 69.83 75.00 78.92

9.9 9.8 9.1 9.2 7.1 6.3 2.0 6.5 1.0 4.4 0.0 -2.5

-7 -11 -6 -4 -7 -14 -31 -12 -28

+32 +33 +31 +32 +30 +31 +26 +28 +28

-31

+26

-33

+16

-6 -6 -6 -1 -6 -13 -34 -9 -29 -15 -36

+29 +34 +30 +29 +29 +32 +33 +31 +30 +32 +30

A B C D E F G H I J K

a

Averaged for the period 1961-1990 (data from UK Met Office and Norwegian Met Office).

replaced with a site on Bear Island (75 °N, site 12). The sampling sites are shown in Figure 2. The SPMDs were deployed in the summer of 1998 (between 9 August and 1 September) and were collected after 2 years in the summer of 2000 (between August 1 and August 29), i.e., for an average of 720 days. The SPMDs at Bear Island (site 12) were deployed for a shorter period, between July 21, 1999 and September 17, 2000, i.e., 423 days. A possible implication of the shorter exposure time in site 12 would be the underestimation of sequestered amounts for the less chlorinated PCBs at this site. Previous studies have shown that equilibrium between the air and the SPMDs is established for these compounds after 200 (tri-PCBs) to 400 (tetra-PCBs) days (24). Table 1 shows temperature information and the location of the sampling sites, including a comparison with the sampling sites from the 1994-1996 study (16) for the sake of clarity. Each sampling site is within a few kilometers of a meteorological station (UK Meteorological Office and the Norwegian Meteorological Institute). However, due to incompleteness of meteorological datasets, it was not possible to obtain mean annual temperatures for the sampling period for certain sites. Therefore, mean annual temperatures were compiled using daily air temperature data averaged over several decades, depending on availability of temperature data (1961-1990 for Norway and 1960-1996 for the UK). As mean annual temperatures tend to be very similar year to year, we assumed that this was a good approximation. To further check that temperatures were similar in the two sampling periods, minimum and maximum temperatures as measured inside

the Stevenson screen boxes are shown for each sampling period (Table 1). Extraction and Cleanup. To be able to compare SPMD concentrations between this study and the previous one (16), care was taken to deploy, store, and extract the SPMDs in exactly the same way as reported previously (16, 21). Briefly, after collection the samples were stored in a freezer until they were extracted. The SPMDs were shaken for 20 s with 100 mL of hexane to remove particulates and exterior lipids (hereafter referred to as “exterior fraction”). This exterior fraction was spiked with a recovery spike containing 13C12labeled PCBs 28, 52, 101, 138, 153, and 180. The SPMDs were then cut open and spiked internally with the 13C12-PCB recovery spike and, after resealing, were dialyzed in hexane in the dark for 2 periods of 24 h (with fresh hexane replacing the decanted dialysate between dialization periods). The two dialysates were combined and will be referred to as the “interior fraction”. Both fractions went through the same cleanup method, but only the interior fraction was used in the interpretation of the results and the exterior fraction was merely used as a check in case of anomalous results, in line with the previous study (16, 21). The cleanup procedure used here differs somewhat from the previous study, but both methods gave similar and good recoveries; therefore the results are considered comparable (see below and ref 16). SPMD extracts were cleaned on a mixed silica gel/alumina column, followed by size exclusion chromatography as described elsewhere (22). Finally the samples were fractionated on a mixed silica gel/alumina column. A 10 mm i.d. column was packed with (from bottom) 3 g of silica gel (3% deactivated), 2 g of alumina (6% deactivated), and 1 cm Na2SO4. The column was precleaned with 50 mL of DCM followed by 50 mL of hexane under N2 pressure. The sample was quantitatively transferred to the column in a small volume of hexane and eluted with 22 mL of hexane (F1) followed by 20 mL of DCM (F2). Both fractions were reduced and solvent exchanged to 50 µL of dodecane containing PCB 30 and 13C12 PCB 141 as internal standards. All fractions (interior and exterior, F1 and F2) were analyzed by GC-MS with an EI+ source in selected ion monitoring mode (SIM). Details of the instruments, GC temperature program, and monitored ions are given elsewhere (21, 25). The following compounds were routinely detected in all SPMDs: tri-PCB 18, 31, 28, 22; tetra-PCB 52, 49, 44, 74, 70; penta-PCB 95, 90/101, 99, 87, 110, 123, 118, 105; hexa-PCB 151, 149, 153, 132, 141, 138, 158; hepta-PCB 187, 183, 174, 180, 170; octa-PCB 199, 203, 194 and hexachlorobenzene (HCB). QA/QC. Field blanks (SPMDs stored in sealed cans during the deployment period), laboratory blanks (newly purchased SPMDs), and solvent blanks (extraction and clean up

TABLE 2. Sequestered Amounts of HCB, PCB Homologue Groups, and ICES Congeners (pg/SPMD) site

1

2

3

4

5

6

7

8

9

10

11

12

HCB sum 3-Cl sum 4-Cl sum 5-Cl sum 6-Cl sum 7-Cl sum 8-Cl total PCB PCB 28 PCB 52 PCB 90/101 PCB 118 PCB 138 PCB 153 PCB 180

9600 1700 6000 16000 16000 4000 330 45000 550 1800 4000 2100 4000 4300 1000

9100 2200 7000 20000 17000 4500 400 51000 590 2000 4800 2700 4400 4300 1100

10000 6400 17000 39000 28000 6100 560 97000 2700 4300 9500 5400 7000 7000 1500

14000 1700 4100 8400 7100 1500 71 23000 600 1200 2300 980 1600 1900 340

16000 1700 3800 8500 8000 1700 120 24000 500 1100 2400 750 1600 2000 410

10000 2100 6300 17000 15000 3200 220 44000 820 1800 4500 1700 3500 3800 820

13000 4000 6900 15000 29000 8900 620 64000 1700 1800 4900 1000 5500 7500 2400

14000 2000 5300 15000 30000 9900 740 63000 560 1500 4800 1300 6100 7400 2800

19000 2200 4500 8100 6100 1100 48 22000 990 1300 2300 560 1400 1400 290

15000 1700 4000 7500 4800 780 48 19000 540 1300 2000 690 1000 1200 180

16000 1900 3700 6500 3800 570 48 17000 850 1300 1700 590 830 930 130

34000 4900 6000 10000 8600 1500 48 31000 2500 1700 3100 780 1500 1900 330

456

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 3, 2003

FIGURE 3. Sequestered amounts (pg) per SPMD for the ICES congeners. Comparison between this study (1998-2000) and the 1994-1996 study. Note that no data were available for PCB 28 and sites 10 and 12 in the 1994-1996 study. procedure carried out without the SPMD) were included at a rate of 50%, 15%, and 8%, respectively. The field blanks were used to correct the results for blank levels and to calculate a limit of detection (LOD ) 3 × SD of the mean blank). The other blanks were merely included as a method check. In cases where the concentration was found to be below detection limit, the value of 1/2 LOD was inserted. Recoveries were routinely monitored in all samples using the 13C12 PCB spike described above. Recoveries were very good for all compounds, averaging 99% (range 77-121%) for the interior fraction and 101% (range 79-132%) for the exterior fraction. In addition, a recovery test of the method was carried out by extracting and cleaning 3 SPMDs spiked with 2.5 ng of all compounds of interest. Again, excellent recoveries were obtained, averaging 95% (range 81-115%) for the interior fraction and 94% (range 65-123%) for the

exterior fraction. Therefore, even though the cleanup method differs from the earlier study, results are comparable between the deployment periods. A table with mean recoveries for the interior and exterior fractions (F1 and F2 results combined in each case)sfor the routine recovery monitoring and the method recovery testsis given in the Supporting Information (Table SI-1).

Results After blank correction, the results for F1 and F2 of the interior fraction were added together, and for each site the mean sequestered amount of the two SPMDs was calculated for each congener. For the sake of brevity, only homologue and selected congener results are reported in this paper. A table with detailed results at all sites for all congeners is available in the Supporting Information (Table SI-2). The comparison VOL. 37, NO. 3, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

457

FIGURE 4. Relative concentration of the different homologue groups (expressed as ratio to penta-PCB) versus latitude. Site 7 and 8 are indicated by crosses, the other sites by diamonds. of the two deployment periods will focus on specific PCB congeners analyzed on both occasions (1998-2000 and 1994-1996). Initial Discussion of the Results. Figure 3 shows a comparison of sequestered amounts of several major congeners (PCB 52, 90/101, 118, 153, 138, and 180) between the two sampling periods at all sites. Sequestered amounts for PCB 28 over the 1998-2000 period are also shown, although a comparison with the earlier period is not possible because of missing data (16). In the two sampling periods, absolute sequestered amounts of heavier PCBs generally decreased with increasing latitude, whereas absolute sequestered amounts of the lighter PCBs were well mixed latitudinally (Figure 3). This decline has been observed in other environmental compartments and is often related to distance from source areas (26-28). Sequestered amounts decreased between the 1994-1996 and 1998-2000 deployment periods. The only exceptions to this were the hexa-octa congeners including PCB 153, 138, and 180 at sites 7 and 8, where sequestered amounts apparently increased (see also Supporting Information, Table SI-2); we do not have an explanation for this. It is unlikely that this anomaly is caused by a laboratory contamination artifact, as laboratory contamination is usually more apparent for the more volatile congeners (29). Likewise, local contamination is unlikely to be the cause, because a. the two sites showing the anomaly are on opposite sides of Norway and experience different meteorology and b. the congener patterns do not match the typical Aroclor-like patterns normally associated with a local contamination source (30). The results were confirmed by reanalysis on GC-MS, and replicates at each site were in 458

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 3, 2003

agreement. The exterior fractions of all sites were investigated, and no abnormalities were found. The data for these congeners at sites 7 and 8 are therefore treated as outliers. We hope that the findings from the next sampling period (2000-2002) will help to clarify this issue. Spatial Trends. To investigate global fractionation, it is useful to look at relative amounts of the different PCB congeners and how they change with latitude and temperature. Figures 4 and 5 show plots of the relative concentrations of the different homologue groups (expressed as ratio to the penta-PCB homologue group) plotted against latitude and mean annual temperature, respectively. It is clear from Figures 4 and 5 that the hexa, hepta, and octa congeners at sites 7 and 8 represent outliers, and they have therefore been omitted from the regression lines. The contribution of triand tetra-PCBs increases with latitude and decreases with increasing temperature, while the contribution of hexathrough octa-PCBs (excluding sites 7 and 8) shows the opposite effect. These trends are consistent with the global fractionation theory (23) and confirm the results obtained in the 1994-1996 study (16). Similar results were found by Agrell et al. (7) who measured air concentrations of PCBs in the Baltic Sea region. In that study, the relative proportion of the high-volatility congeners was positively correlated with latitude. HCB distribution is different to that of PCBs on the transect, reflecting its higher volatility and LRAT potential (23, 31). Figure 6 shows a clear latitudinal trend for HCB, with absolute sequestered amounts increasing with increasing latitude (Figure 6a) and decreasing with increasing temperature (Figure 6b). This is evidence that HCBsin

FIGURE 5. Relative concentration of the different homologue groups (expressed as ratio to penta-PCB) versus mean annual temperature. Sites 7 and 8 are indicated by crosses, the other sites are indicated by diamonds. actually accumulating in the condensed phasesssoils/water bodies etc. of polar latitudes). Other studies have found similar trends for HCB in plants (32) and for thessimilarly volatileshexachlorocyclohexanes (HCHs) in seawater (23, 33).

FIGURE 6. Sequestered amounts of HCB (pg per SPMD) versus a) latitude and b) mean annual temperature. contrast to PCBs where absolute amounts decrease with increasing latitudesis undergoing cold condensation (i.e.

Temporal Trends. The declines between the 1994-1996 and the 1998-2000 periods were very similar between sites and for different congeners (excluding the hexas-octas at sites 7 and 8), with the sequestered amounts decreasing by factors of 2-5 over the 4-year period (see also Figure 3). Declines in the ambient concentrations of POPs in recent years have generally been assumed to be first order (3, 4, 6). If it is assumed that first-order declines apply to the atmospheric PCB concentrations between the 1994-1996 and the 1998-2000 studies, these reductions translate to atmospheric half-lives on the order of 1.7-4 years. This is very comparable to other studies. For example, PCB trends in Great Lakes and UK air over the past decade or so show declines with half-lives in the ranges of ∼2-∼6 years, depending on congener/study/location/time frame (3, 4, 6). A Swedish study on temporal trends of PCBs in various marine and freshwater biota over a 28 year period does not report half-lives (34). However, visual examination of the results suggests half-lives of ca. 10 years (34), perhaps somewhat longer than has been observed in air. Air would be expected to respond more quickly to declines in sources/emissions than these biotic matrices. In summary, the European background air measured at these sites shows a decline in air concentrations which is broadly consistent with that measured for other locations in the Northern Hemisphere. VOL. 37, NO. 3, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

459

FIGURE 7. Relative change in concentration between the two sampling periods at different latitudes. The 1998-2000 data are expressed as a percentage of the 1994-1996 data and plotted against latitude (excluding site 7 and 8). Note that a smaller percentage signifies a greater decline. Spatial Trends in Relative Rates of Decline. Important information on the relative importance of primary/secondary sources and loss processes can be gained by comparing the relative rates of decline in air concentrations of different congeners in different places. As discussed in the Introduction, in a world dominated by secondary sources of PCBs (scenario 2), a change in the PCB congener pattern would occur over time and would vary between the source and sink areas. For example, the relative contribution of lighter congeners to the total PCB concentration would be increased at higher latitudes (i.e. fractionation), and this effect would become more important over time. On the other hand, if advection from primary sources still controls air concentrations, then latitudinal fractionation would also be expected, but this would not change over time (assuming the source signature remained constant). The relative changes in sequestered amounts were calculated by expressing the 1998-2000 data as a percentage of the 1994-1996 data. This is plotted against latitude in Figure 7 for several congeners (excluding sites 7 and 8). The most striking feature of the data is that all the congeners plotted have declined to ∼30% of their 1994-1996 values. There are no discernible differences in percentage decline between the congeners. The second clear observation is that there are no extremely marked variations with latitude (ANOVA carried out on the regressions between percentage decline and latitude shows that regression lines are not significant except for PCB 90/101 and PCB 118 (p < 0.05)). Do Primary Emissions Still Dominate Ambient PCB Levels? The data presented in this study implies that at the present time PCB levels in background European air are still 460

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 3, 2003

mostly controlled by primary emissions, rather than recycling/ secondary emissions from the major environmental repositories such as soils or water bodies. A similar conclusion was reached by Bignert et al. (34), who did not see any difference in half-lives of POPs in biota between the north and the south of Sweden. At first, this conclusion is surprising and requires a comment on earlier studies and their interpretation. Results from an earlier study (35) were interpreted as showing extensive revolatilization of previously deposited PCBs from an annually ploughed arable soil. Observations of diurnal cycling in air were originally interpreted as evidence that the high environmental burden of PCBs in soils was subject to re-emission/recycling (36, 37). However, a later study favored the explanation that air-vegetation exchange could better account for the dynamic nature of air concentration changes (38). Indeed, POPs could cycle many times between the air and vegetation, on their journey from primary sources to environmental repositories and sinks (38, 39). Volatilization of PCBs from soils clearly can occur and has been experimentally measured under appropriate conditions (40, 41). It would be enhanced by ploughing (35, 42), the presence of high levels at the soil surface layer, which is in intimate contact with the air, and by elevated temperatures. However, a recent regional background soil survey (28) showed a strong source regionsremote region gradient, indicating that PCBs are generally effectively retained by soils. Indeed, recent studies indicate that only a small proportion of the soil PCB burden is volatilized to air from undisturbed soils (28, 43). An important observation from the present study is that different PCB congeners are declining at similar rates in air at different places, suggesting a loss mechanism

which is largely independent of physical/chemical properties. Burial to deeper soil layers provides an effective long-term/ permanent sink for POPs. This process appears largely noncongener specific in woodland and pasture grassland soils (28, 44) and has been identified as a major factor involved in the transport of POPs to deeper soil layers (44). Burial effectively excludes the majority of the soil burden from the “recyclable pool”. A contrasting situation is agricultural soils, which have high soil burdens of pesticides, where regular ploughing makes them an important source to air (42). Largescale advection/dilution processes would also be consistent with the data presented here (e.g. a net supply/dilution of Northern Hemisphere air to the Southern Hemisphere). The conclusions from the present study regarding the importance of primary emissions as a contemporary source of PCBs in airsif confirmed by other studiessclearly have important implications for policy makers and regulators. If primary emissions still dominate for these compounds, it implies further efforts at source reduction may be worthwhile. Passive air monitoring networks measuring spatial and temporal trends, such as the one described in the present study, are clearly invaluable to gaining insights into this issue.

Acknowledgments We would like to thank the site owners for their kind cooperation. Also many thanks to Gareth, Marian, and Bas for help with the deployment/collection. Finally, we are grateful to Gareth Thomas, Andy Sweetman, and Tom Harner for the many useful discussions.

(12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31)

Supporting Information Available Quality control information (Table SI-1) and congenersspecific sequestered amounts (Table SI-2). This material is available free of charge via the Internet at http://pubs.acs.org.

(32) (33) (34)

Literature Cited (1) UN-ECE. Draft Protocol on Persistent Organic Pollutants EB.AIR/ WG.5/R.94. 1998. (2) Evaluation of persistence and long-range transport of organic chemicals in the environment; Klecka, G. et al., Eds.; SETAC Special Publication Series, Pensacola, FL, 2000. (3) Hillery, B. R.; Basu, I.; Sweet, C. W.; Hites, R. A. Environ. Sci. Technol. 1997, 31, 1811-1816. (4) Simcik, M. F.; Basu, I.; Sweet, C. W.; Hites, R. A. Environ. Sci. Technol. 1999, 33, 1991-1995. (5) Coleman, P. J.; Lee, R. G. M.; Alcock, R. E.; Jones, K. C. Environ. Sci. Technol. 1997, 31, 2120-2124. (6) Sweetman, A. J.; Jones, K. C. Environ. Sci. Technol. 2000, 34, 863-869. (7) Agrell, C.; Okla, L.; Larsson, P.; Backe, C.; Wania, F. Environ. Sci. Technol. 1999, 33, 1149-1156. (8) Halsall, C. J.; Bailey, R.; Stern, G. A.; Barrie, L. A.; Fellin, P.; Muir, D. C. G.; Rosenberg, B.; Rovinsky, F. Y.; Kononov, E. Y.; Pastukhov, B. Environ. Pollut. 1998, 102, 51-62. (9) Harner, T.; Kylin, H.; Bidleman, T. F.; Halsall, C.; Strachan, W. M. J.; Barrie, L. A.; Fellin, P. Environ. Sci. Technol. 1998, 32, 3257-3265. (10) Hung, H.; Halsall, C. J.; Blanchard, P.; Li, H. H.; Fellin, P.; Stern, G.; Rosenberg, B. Environ. Sci. Technol. 2001, 35, 1303-0311. (11) Breivik, K.; Sweetman, A.; Pacyna, J. M.; Jones, K. C. Sci. Total Environ. 2002, 290, 181-198

(35) (36) (37) (38) (39) (40) (41) (42) (43) (44)

Jones, K. C. Environ. Sci., Pollut. Res. 1994, 1, 172-177. Goldberg, E. D. Proc. R. Soc. London, Ser. B 1975, 189, 277-289. Wania, F.; Mackay, D. Ambio 1993, 22, 10-18. Ockenden, W. A.; Jaward, F. M.; Jones, K. C. Sci. World 2001, 1, 557-575. Ockenden, W. A.; Sweetman, A. J.; Prest, H. F.; Steinnes, E.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 2795-2803. Petty, J. D.; Huckins, J. N.; Zajicek, J. L. Chemosphere 1993, 27, 1609-1624. Prest, H. F.; Huckins, J. N.; Petty, J. D.; Herve, S.; Paasivirta, J.; Heinonen, P. Mar. Pollut. Bull. 1995, 31, 306-312. Prest, H. F.; Jacobson, L. A. Chemosphere 1995, 30, 1351-1361. Booij, K.; van Drooge, B. L. Chemosphere 2001, 44, 91-98. Ockenden, W. A.; Prest, H. F.; Thomas, G. O.; Sweetman, A.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 1538-1543. Lohmann, R.; Corrigan, B. P.; Howsam, M.; Jones, K. C.; Ockenden, W. A. Environ. Sci. Technol. 2001, 35, 2576-2582. Wania, F.; Mackay, D. Environ. Sci. Technol. 1996, 30, 390-396. Ockenden, W. A.; Corrigan, B. P.; Howsam, M.; Jones, K. C. Environ. Sci. Technol. 2001, 35, 4536-4543. Thomas, G. O.; Sweetman, A. J.; Parker, C. A.; Kreibich, H.; Jones, K. C. Chemosphere 1998, 36, 2447-2459. Muir, D. G.; Omelchenko, A.; Grift, N. P.; Savoie, D. A.; Lockhart, W. L.; Wilkinson, P.; Brunskill, G. J. Environ. Sci. Technol. 1996, 30, 3609-3617. Simonich, S. L.; Hites, R. A. Environ. Sci. Technol. 1997, 31, 999-1003. Meijer, S. N.; Steinnes, E.; Ockenden, W. A.; Jones, K. C. Environ. Sci. Technol. 2002, 36, 2146-2153. Alcock, R. E.; Halsall, C. J.; Harris, C. A.; Johnston, A. E.; Lead, W. A.; Sanders, G.; Jones, K. C. Environ. Sci. Technol. 1994, 28, 1838-1842. Schulz, D. E.; Petrick, G.; Duinker, J. C. Environ. Sci. Technol. 1989, 23, 852-859. Beyer, A.; Mackay, D.; Matthies, M.; Wania, F.; Webster, E. Environ. Sci. Technol. 2000, 34, 699-703. Calamari, D.; Bacci, E.; Focardi, S.; Gaggi, C.; Morosini, M.; Vighi, M. Environ. Sci. Technol. 1991, 25, 1489-1495. Iwata, H.; Tanabe, S.; Sakai, N.; Tatsukawa, R. Environ. Sci. Technol. 1993, 27, 1080-1098. Bignert, A.; Olsson, M.; Persson, W.; Jensen, S.; Zakrisson, S.; Litzen, K.; Eriksson, U.; Haggberg, L.; Alsberg, T. Environ. Pollut. 1998, 99, 177-198. Alcock, R. E.; Johnston, A. E.; McGrath, S. P.; Berrow, M. L.; Jones, K. C. Environ. Sci. Technol. 1993, 27, 1918-1923. Lee, R. G. M.; Hung, H.; Mackay, D.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 2172-2179. Lee, R. G. M.; Jones, K. C. Environ. Sci. Technol. 1999, 33, 705712. Hung, H.; Thomas, G. O.; Jones, K. C.; Mackay, D. Environ. Sci. Technol. 2001, 35, 4066-4073. Gouin, T. Unpublished MSc Thesis, Trent University, Canada, 2002. Cousins, I. T.; Hartlieb, N.; Teichmann, C.; Jones, K. C. Environ. Pollut. 1997, 97, 229-238. Cousins, I. T.; Jones, K. C. Environ. Pollut. 1998, 102, 105-118. Harner, T.; Bidleman, T. F.; Jantunen, L. M. M.; Mackay, D. Environ. Toxicol. Chem. 2001, 20, 1612-1621. Sweetman, A. J.; Cousins, I. T.; Seth, R.; Jones, K. C.; Mackay, D. Environ. Toxicol. Chem. 2002, 21, 930-940 Cousins, I. T.; Gevao, B.; Jones, K. C. Chemosphere 1999, 39, 2507-2518.

Received for review March 5, 2002. Revised manuscript received November 7, 2002. Accepted November 7, 2002. ES025620+

VOL. 37, NO. 3, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

461