Recent Declines in PAH, PCB, and Toxaphene Levels in the Northern

Aug 22, 2001 - BMW taps Solid Power for battery program. BMW says it will partner with Colorado-based Solid Power, a 2017 C&EN Start-up to Watch. The ...
2 downloads 12 Views 80KB Size
Research Recent Declines in PAH, PCB, and Toxaphene Levels in the Northern Great Lakes As Determined from High Resolution Sediment Cores ABBY R. SCHNEIDER,† HEATHER M. STAPLETON,† J E F F C O R N W E L L , ‡ A N D J O E L E . B A K E R * ,† Chesapeake Biological Laboratory, University of Maryland Center for Environmental Science, Solomons, Maryland 20688, and Horn Point Environmental Laboratory, Cambridge, Maryland

Sediment cores were collected from two sites in Grand Traverse Bay, Lake Michigan in May 1998, dated using 210Pb geochronology, and analyzed for polychlorinated biphenyl (PCB) congeners, polycyclic aromatic hydrocarbons (PAHs), and toxaphene. The extraordinarily high sediment focusing and accumulation rates in these cores relative to other Great Lakes sediments allowed quantification of highresolution temporal trends in the burial of hydrophobic organic contaminants. The focus-corrected accumulation rate of total PCBs (sum of 105 congeners) in 1998 was 0.50 ng/cm2-year at both sites. Toxaphene and total PAH (t-PAH; sum of 33 compounds) surficial accumulations varied at each site and ranged from 0.08 to 0.41 ng/cm2year for toxaphene and 25 to 52 ng/cm2-yr for t-PAHs at the two sites. The maximum t-PAH accumulation rate was in sediment dated from 1942, and PAH accumulation decreased from 1942 to 1980 with a first-order rate of decline 0.017 yr-1. Both toxaphene and t-PCB accumulations peaked in sediment deposited in 1972, after which their accumulations decreased with nearly identical rates of decline (0.027 yr-1 and 0.028 yr-1, respectively).

Introduction Hydrophobic organic contaminants (HOCs), such as polycyclic aromatic hydrocarbons (PAHs), and polychlorinated biphenyls (PCBs), are relatively persistent in the Great Lakes and actively cycle among the atmosphere, waters, and sediments (1-3). The United States government banned the production of PCBs and the organochlorine insecticide toxaphene in 1977 and 1981, respectively. Since then their levels have declined in the atmosphere (4), water column (2,5), sediment (2), and biota (6, 7) of the Great Lakes. Calculated first-order rate loss constants for PCBs within these compartments range from 0.078 yr-1 in the water column to 0.12 yr-1 in lake trout from Lake Michigan (3, 6-9). Despite the abundance of data describing the declining levels of HOCs in the Great Lakes, it is not clear whether levels are continuing to decline or have reached an inventory * Corresponding author phone: (410)326-7205; fax: (410)326-7341; e-mail: [email protected]. † University of Maryland Center for Environmental Science. ‡ Horn Point Environmental Laboratory. 10.1021/es002044d CCC: $20.00 Published on Web 08/22/2001

 2001 American Chemical Society

supported by continual “new” loadings from external sources (e.g., dynamic exchange of HOCs between the atmosphere and surface waters (10)) and recycling of “in place” contaminants within the lakes (e.g., release of HOCs from contaminated sediments (11)). Sediments are a large repository of HOCs and burial is an important removal process from the Great Lakes water column. Therefore, accurate estimates of net HOC burial fluxes are required to accurately model HOC dynamics. Accumulation of HOCs in undisturbed sediment columns may record information about historical inventories of HOCs during the past several decades. Here we explore the idea that the rate of ecosystem-level response to HOC loadings reductions can be estimated by analyzing how HOC burial rates have changed during the past 20 years, as recorded in sediment cores collected from high sedimentation rate areas. We focus on three classes of HOCs with different production and release histories: the industrial chemicals PCBs, the insecticide mixture toxaphene, and combustion-derived PAHs. In the Great Lakes, many studies have investigated the rate at which HOCs are buried in the sediment by measuring contaminant profiles in sediment cores ((3, 4, 11, 12) Table 1). While valuable, many of these cores were collected in areas with relatively low mass sedimentation rates, which limits the inherent temporal resolution of the geochronology. In addition, many of these cores were collected within 1015 years of the PCB and toxaphene phase-outs in the Great Lakes region and have only a few data points reflecting recent HOC declines. For example, cores collected in 1991 from Lake Michigan and analyzed for PCBs had only three sediment layers above the layer of peak PCB accumulation (12). Similarly, Pearson et al. (3) quantified toxaphene in the same Lake Ontario cores and found the maximum toxaphene accumulation occurred within the first 3-6 sediment layers. While many of these cores hint at a record of ecosystemlevel HOC declines, new sediment cores collected from Great Lakes basins with high sedimentation rates (and, therefore, high temporal resolution) are required to quantify these declines. Grand Traverse Bay is a very narrow, deep, fjord-like embayment in northern Lake Michigan. The two deep, steepwalled basins of Grand Traverse Bay are ideal locations for large sediment focusing (the transport of sediments from nearshore to depositional basins in deep water), leading to very high sedimentation rates that facilitate investigating historical trends. Using these cores, the historical accumulation of HOCs can be examined more thoroughly. The results presented in this paper are part of a larger project conducted within Grand Traverse Bay examining the seasonal cycling of HOCs within the water column to determine the factors influencing the bioaccumulation of HOCs in the food web (14, 15). Grand Traverse Bay has similar water column characteristics as the rest of northern Lake Michigan. HOC levels in the bay are typical of those in the open northern Great Lakes, suggesting that HOC cycling is driven by atmospheric deposition and volatilization.

Experimental Section Sampling. Sediment cores were collected using the U.S. Environmental Protection Agency’s R/V Lake Guardian on May 17, 1998 from two depositional sites in the western arm of Grand Traverse Bay (Figure 1). HMS1 was collected at VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3809

TABLE 1. Summary of Focus-Corrected Accumulations and Inventories for PCBs, PAHs, and Toxaphene in the Great Lakes PCBs

lake

year core collected

sed accum rate (g/cm2-yr)

focus factor

1990a 1991, 1994

0.01 0.01-0.06

1.14 1.14-1.63

0.07 N/A

1984 1991 (LM-18) 1992 (LM-47s) 1992 (LM-68k) 1998 (HMS1) 1998 (HMS2)

0.03-0.04 0.04 0.036 0.028 0.11 0.10

0.71-1.41 2.43 1.50 1.27 8.5 5.9

1990 (LO-E30) 1990 (LO-19) 1993

0.066 0.033 0.02-0.1

2.04 1.06 0.62-1.67

suficial accum (ng/cm2-yr)

PAHs

toxaphene

suficial accum (ng/cm2-yr)

inventory (ng/cm2)

suficial accum (ng/cm2-yr)

inventory (ng/cm2)

4.0 N/A

N/A N/A

N/A N/A

N/A 0.097-0.14

N/A 3.1-7.9

12 3

0.44-2.9 2.1 2.9 2.9 0.5 0.5

24-144 90 80 205 33 44

N/A 55 65 70 25 52

N/A 5500 6200 7800 2100 7800

N/A 0.24 0.52 0.70 0.07 0.41

N/A 7.5 15 43 4 28

13 12, 4, 3 12, 4, 3 12, 4, 3 this study this study

8.2 7.4 N/A

555 605 N/A

N/A N/A N/A

N/A N/A N/A

0.50 0.39 0.3-0.6

37 22 22-84

12, 3 12, 3 9

inventory (ng/cm2)

reference

Superior Michigan

Ontario

FIGURE 1. Location of sediment cores collected in Grand Traverse Bay. 45°01.221′N and 85°33.706′ W and HMS2 was collected at 44°49.500′ N and 85°36.933′ W. The two sites are separated by a sill, and the more southern HMS2 site is closer to the Traverse City, MI urban area. A 0.5 m × 0.5 m box core was used to collect sediment, which was then subsampled using 7.5 cm diameter plexiglass tubes. Two subcores from each site were vertically sectioned, and the corresponding layers from the two cores were combined into precleaned glass jars in the following increments: every 1/2 cm from the surface to 2 cm, every 1 cm from 2 to 10 cm, every 2 cm from 10 to 38 cm, and every 4 cm from 38 to 46 cm. Sediment samples were frozen aboard ship, transported, and stored frozen in the glass jars until analyzed. Chemical Analysis. In the laboratory, samples were thawed, homogenized, and subsampled to determine the percent water content, the 210Pb content, and concentrations of PCBs, PAHs, and toxaphene. The analysis of 210Pb followed Sugai (1990) and consisted of a HCl-HNO3 acid digestion and plating of 210Po on silver plates for alpha counting. Po209 was used as a yield tracer. Secular equilibrium between 210Pb and 210Po was assumed. Wet sediment was ground with anhydrous sodium sulfate (1:8 ratio of wet sediment to sodium sulfate) for HOC 3810

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 19, 2001

extraction. The dried sample was transferred to a Soxhlet flask, spiked with PAH and PCB surrogate standards, d8naphthalene, d10-fluorene, d10-fluoranthene, d12-perylene, 3,5-dichlorobiphenyl (IUPAC #14), 2,3,5,6-tetrachlorobiphenyl (IUPAC #65), and 3,3′,4,4′,5,6-hexachlorbiphenyl (IUPAC #166), and extracted for 24 h in dichloromethane (16). The samples were then concentrated, transferred to hexane, and purified by elution through deactivated alumina with petroleum ether. Afterward, the samples were again concentrated and transferred to hexane and perdeuterated PAH internal standards (d10-acenaphthene, d10-phenanthrene, d12-benz[a]anthracene, d12-benzo[a]pyrene, and d12benzo[g,h,i]perylene) were added. A suite of 33 PAHs were quantified in all sediment samples using a Hewlett-Packard model 5890 gas chromatograph connected to a HewlettPackard 5972 mass spectrometer using electron ionization in selected ion monitoring (SIM) mode (16). Laboratory blanks and replicate samples were run simultaneously to detect and monitor any interfering compounds and to document the precision of the method. After PAH analysis, the extract was further fractionated by liquid-solid chromatography using deactivated Florisil (16). PCB congeners and some organochlorine pesticides were eluted from the column with petroleum ether, while the remaining pesticides were eluted into a second fraction with 50:50 petroleum ether:dichloromethane. The first fraction was concentrated, transferred to hexane, further concentrated under nitrogen, and placed into autosampler vials with PCB internal standards, 2,4,6-trichlorobiphenyl (IUPAC # 30) and 2,2′,3,4,4′,5,6,6′ octachlorobiphenyl (IUPAC #204). Total PCBs were quantified as the sum of 105 congeners (47 resolved and 58 unresolved) using a Hewlett-Packard model 5890 gas chromatograph equipped with a 63Ni electron capture detector (16-18). To quantify toxaphene, the two Florisil fractions were recombined and concentrated under a stream of nitrogen. PCB congener 204 was used as an internal standard for toxaphene, and the samples were analyzed on a HewlettPackard 5890 gas chromatograph connected to a 5989A mass spectrometer utilizing negative chemical ionization in SIM mode. Quantification of total toxaphene and the toxaphene homologue groups was performed by running an automated macro and Qbasic program (19). Analytical Quality Assuarance. The high 210Pb activity within the cores resulted in small counting errors that ranged from 0.19 to 2.5 dpm/g, with a mean of 1.2 dpm/g, or only 2-4% of the total activity. As a measure of analytical precision and accuracy, laboratory matrix blanks spiked with PCB and deuterated PAH surrogate compounds, NIST Standard Reference Material 1941a (Organics in Marine Sediment), and replicates were analyzed. Laboratory matrix blanks

2), presumably due to changes in sediment loads to the southern basin of Grand Traverse Bay. This change in the 210 Pb profile produces variable sedimentation rates. From the surface down to the first break at 15 cm the average sedimentation rate is 0.10 ( 0.02 g/cm2-year. After this first break to a depth of 34 cm the sedimentation rate is 0.08 ( 0.03 g/cm2-year. From the second break to the bottom of the core the average sedimentation rate is 0.05 ( 0.01 g/cm2year. These sedimentation rates uncorrected for focusing are high relative to other cores collected throughout the Great Lakes, but this is to be expected because of the very narrow dimensions of Grand Traverse Bay. For comparison, the sedimentation rate measured from 16 cores collected throughout Lake Michigan ranged from 0.0010 g/cm2-year to 0.074 g/cm2-year (see Table 1). FIGURE 2. Excess cores.

210

Pb profiles in HMS 1 and HMS 2 sediment

consisted of approximately 30 g of Na2SO4 spiked with surrogate PCB and PAHs and extracted adjacent to sediment samples. Average percent recoveries for the surrogates were 82 ( 18%, 66 ( 11%, and 75 ( 12%, for PCB congeners 14, 65, and 166, respectively, and 67 ( 11%, 92.8 ( 12%, and 80 ( 11% for d10-fluorene, d10-fluoranthene, and d12-perylene, respectively. Analytical precision determined from the percent difference of replicate samples ranged from 6 to 45% for the various PAHs and 2-60% on a congener by congener basis for PCBs. Analysis of NIST SRM 1941a resulted in PCB and PAH levels ranging from 60 to 105% of the certified levels. 210Pb Data Analysis. The constant rate of supply model (CRS) was used to determine the rate of sediment accumulation in each core. The CRS model assumes the net supply of unsupported 210Pb is constant despite variations in the mass sedimentation rate. The date of each section of the core was calculated according to methodology described by Appleby and Oldfield (20). To apply the CRS model, the density of the sediment particles was assumed to be 2.5 g/cm3, and the unsupported 210Pb activity of each section was calculated by subtracting the background level of 2 dpm/g of 210Pb, determined by analyzing deep sediments from the cores, from the measured value. Focusing refers to the movement of sediment to deeper parts of a lake due to turbulence or overturning (21). If focusing occurs, the 210Pb inventory measured in the sediments is greater than that supported by the input of 210Pb from the water column. Plug flow models and field measurements indicate that the background 210Pb inventory should be 34.4 dpm/cm2 (21-23). The focusing factor (FF) is the ratio of the unsupported 210Pb inventory integrated over the depth of the core to the expected 210Pb inventory resulting from the regional atmospheric flux. PCB, toxaphene, and PAH accumulations for each section of the core were divided by the dimensionless focusing factor to account for their accumulation and build up due to lateral sediment transport. HOC inventories were also divided by the focusing factor to normalize the total areal burden to non depositional areas.

Results and Discussion 210Pb

Dating. At HMS1, the unsupported 210Pb activity decreases exponentially with depth after the first 3 cm (cumulative dry mass ) 0.4 g/cm2) indicating a constant sedimentation rate with minimal mixing in the upper layers (Figure 2). The average sedimentation rate for HMS1 is 0.11 ( 0.03 g/cm2-year. At HMS2, the mixed depth also appears to be around 3 cm (cumulative dry mass ) 3 g/cm2). However, there is a noticeable break in the 210Pb profile at a depth of 15 cm (cumulative dry mass ) 2.8 g/cm2) and 34 cm (cumulative dry mass ) 7.2 g/cm2; Figure

Focusing factors for these two cores were 8.5 and 5.9 at HMS1 and HMS2, respectively, significantly higher than other reported values for open Lake Michigan, which range from 0.32 to 2.9 ((12, 13) Table 1). These high focusing factors produce high sedimentation rates which provide greater temporal resolution in the sediment record when combined with fine incremental sectioning of the sediment cores. To estimate the temporal resolution within the core, the dry mass of each section (g/cm2) was divided by the average sedimentation rate (g/cm2-yr). According to this equation, sediment cores previously collected in Lake Michigan had temporal resolutions ranging between 8 and 12 years (dry mass values ranging between 0.3 and 0.5 g/cm2 and sedimentation rates of 0.028-0.040 g/cm2-yr (12)). Performing the same calculation using data obtained from the surface sediment at HMS1 and HMS2 yields temporal resolutions of 1.9 and 1.7 years, respectively, at the two sites. The high sedimentation rates in the Grand Traverse Bay cores provide excellent temporal resolution and insight into historical trends within the Great Lakes basin. Postdepositional Mobility. To investigate the possibility of postdepositional mobility and degradation of contaminants within the sediment, the percent composition of individual compounds was examined throughout the core. Lighter and branched PAH molecules are more labile than their unbranched parent compounds, and, therefore, the percentage of labile vs nonlabile PAHs might shift in a core experiencing any degradation reactions. For example, phenanthrene and anthracene degrade microbially (24, 25). In HMS1 and HMS2, the proportions of phenanthrene and anthracene relative to their more labile mono- and dimethylated derivatives are consistent throughout the core indicating there was no significant degradation (16). Experiments show that lightweight PCB congeners undergo microbrial degradation in oxic and anoxic conditions in the Hudson River estuary (26-28). To investigate the possibility of microbrial degradation of PCBs within the sediment of GTB, congeners previously shown to degrade under aerobic conditions were examined, and the percentage of lightweight congeners remains invariant throughout the core (16). For example, coeluting congeners 8,5 (2,4′dichlorobiphenyl and 2,3 dichlorobiphenyl) and 16,32 (2,2′,3trichlorobiphenyl and 2,4′,6-trichlorobiphenyl) have been shown to experience reductive dechlorination within the Hudson River (29), while congeners 49, 44, 41, 64, and 71 (2,2′,4,5-tetrachlorobiphenyl, 2,2′,3,5′-tetrachlorobiphenyl, 2,2′,3,4-tetrachlorobiphenyl, and 2,3,4′,6-tetrachlorobiphenyl) are more resistant to dechlorination. The consistency of these congener compositions within HMS1 and HMS2 suggests dechlorination is not occurring and is comparable to results from sediment cores collected in a suite of midlatitude and arctic lakes (30). From this evidence, it seems unlikely that PCBs are being degraded or remobilized within the core. VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3811

Toxaphene homologue groups have varying numbers of chlorine atoms associated with their central structure and have the potential to bioaccumulate depending on their paritioning coefficients. Toxaphene displays no evident change in homologue composition over time in these cores (16). The average toxaphene homologue distribution in HMS1 and HMS2 was dominated by the heptachlorinated toxaphene compounds. This distribution is similar to homologue patterns found in Lake Ontario sediment (9) and lake trout from Lake Michigan (31). However, Pearson et al. (3) observed a pattern dominated by octachlorinated compounds in Lake Michigan sediment. The commercial mixture of toxaphene that was initially released into the environment was dominated principally by the octachlorinated compounds with significant contributions from the hexa-, hepta-, and nonachlorinated compounds. It is possible that there has been some degradation of toxaphene in the sediment cores that preferentially degrades the heavier chlorinated compounds. Howedeshell and Hites (9) noticed a decrease in the ratio of the 6 chlorinated:7 chlorinated homologue groups that they attributed to in situ degradation. However, the homologue compositions in Grand Traverse Bay sediment did not appear to change significantly throughout the core. The lack of significant changes in composition of the PAH, PCB, and toxaphene compounds suggests that insignificant degradation or remobilization has occurred within these cores. Therefore the trends apparent throughout the core most likely reflective historical deposition in this area. Contaminant Inventories. Inventories of HOCs represent the total integrated mass of the compound of interest per unit area

inventory )

∑(C - C )F d i

B

B

where Ci is the concentration in each depth increment (ng/g dry), CB is the background concentration for the HOC of interest (in this case CB ) 0 for PAHs, PCBs, and toxaphene), FB is the dry mass of each increment (g/cm3), and d is the thickness of each increment. Inventories were divided through by the focusing factor to account for the lateral movement of sediment (Table 1). The focus-corrected PCB inventories in Grand Traverse Bay were 30 and 44 ng/cm2 at HMS1 and HMS2, respectively. Compared to previous studies, the sediment cores collected in this study had around 10 additional years of PCB accumulation, yet the PCB inventories were quite low. For comparison, Hermanson et al. (13) found PCB inventories ranging from 24 to 144 ng/cm2 in five cores collected throughout Lake Michigan in 1984 (Table 1). Golden et al. (12) collected three cores in Lake Michigan in 1990-1992 and calculated PCB inventories ranging from 80 to 205 ng/ cm2 (Table 1). For comparison in the early 1990s, Lake Ontario had PCB inventories as high as 788 ng/cm2 due to proximity to more urbanized locations, and the inventory in more isolated Lake Superior was 4 ng/cm2 ((12) Table 1). The PCB inventories in Grand Traverse Bay are similar to those of northern Lake Michigan. The toxaphene inventory at HMS1 was only 4 ng/cm2, while at HMS2 it was 28 ng/cm2. HMS2 is closer to Traverse City, and these results suggest there may have been an input of toxaphene from the local watershed. Pearson et al. (3) reported higher toxaphene concentrations in northern Lake Michigan than in the southern portion of the lake suggesting that the northern region was also influenced by local input of toxaphene (Table 1). The toxaphene inventory at HMS1 is similar to the inventory in the southern Lake Michigan core collected by Pearson et al. (3), while the inventory at HMS2 is similar to their northern Lake Michigan results. The inventory at HMS2 is also comparable to values found within 3812

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 19, 2001

FIGURE 3. Observed concentrations of (a) PCBs, (b) toxaphene, and (c) PAHs in sediment core HMS 1 and the focus-corrected accumulation rates of (d) PCBs, (e) toxaphene, and (f) PAHs. Lake Ontario, while the inventory at HMS1 is comparable to that found in Siskiwit Lake, WI (3). The focus-corrected t-PAH inventories were 2100 ng/cm2 and 7800 ng/cm2 at HMS1 and HMS2, respectively. The higher inventory at HMS2 could be due to urban input from Traverse City; the core was collected 5 miles north of the city. HMS1, the more northern site, was isolated from urban input and had a lower PAH inventory. The inventories at both sites are consistent with results from other cores collected in Lake Michigan. Simcik et al. (4) collected sediment cores throughout Lake Michigan and found focus corrected PAH inventories ranging from 1500 to 7800 ng/cm2. The lowest PAH inventory documented in that study was attributed to sediment mixing; the actual PAH inventory throughout the depositional zone in Lake Michigan was between 5000 and 7000 ng/cm2 (Table 1). PCB Profiles. The PCB concentration in sediment at HMS1 has been steadily decreasing since its peak in 1972 ( 4 years of 60 ng/g to 30 ng/g in 1998 (Figure 3a). Concentrations at HMS2 are similar to those at HMS1; however, due to the break in the sedimentation rate the profiles are more difficult to interpret and are not discussed further. PCB levels were first observed in sediment sections deposited during 1928 ( 7.5, corresponding to the time these chemicals were first commercially manufactured. In sediment deposited between 1928 and 1972, PCB levels steadily increased, reflecting increased use and production of PCBs, until production was stopped in early 1970s and the chemicals were banned in 1977. Correcting the contaminant accumulation rates for focusing events removes the influence of high sedimentation rates that are experienced in the bay, standardizing the sediment to other cores collected throughout Lake Michigan. In Grand Traverse Bay, focus-corrected accumulation rates of PCBs displayed maximum accumulation rates of 0.90 ng/ cm2-yr in 1972 and have steadily declined in recently deposited sediment to a surface flux of 0.50 ng/cm2-yr (Figure 3d). Hermanson et al. (13) measured PCB fluxes for various regions of Lake Michigan and found levels ranging from 0.44 to 2.9 ng/cm2-yr, which are comparable to our findings (Table 1). Golden et al. (12) also measured fluxes throughout Lake

Michigan and observed higher fluxes ranging from 2.1 to 2.9 ng/cm2-yr which may have resulted from proximity to more urbanized areas such as Green Bay and Chicago (Table 1). Strachan and Eisenreich (32) estimated an atmospheric contribution of PCBs to Lake Michigan was approximately 0.69 ng/cm2-yr, which is in good agreement with our estimated sedimentation fluxes, indicating that the atmosphere inputs and sedimentations losses of PCBs to the water column are in balance. Toxaphene Profiles. Though the toxaphene concentrations at both sites were different, the depth profiles were similar, and the toxaphene concentration peaked at the same depth interval as the PCB concentration (Figure 3b). At HMS1, the surficial toxaphene concentration was 5 ng/g dry sediment, while at HMS2 the surficial concentration was 28 ng/ g. The maximum concentration of toxaphene occurred in 1972 ( 4 years and was 9.8 ng/g at HMS1 and 54 ng/g at HMS2. The concentration profiles with depth are similar in maximum peak date to the Lake Michigan toxaphene profiles documented by Pearson et al. (3). In those cores, toxaphene values ranged from about 15-20 ng/g near the surface and had maximum levels of about 40-50 ng/g (Table 1). However, the northern Lake Michigan cores displayed relatively little or no change in toxaphene levels since their peak in early 1970s, and, in one northern core, the levels appeared to increase at the surface. In the Grand Traverse Bay cores, toxaphene concentrations decreased 50% over the past 30 years. The focus-corrected accumulation rates for toxaphene in HMS 1 range from 0.07 ng/cm 2-yr near the surface to a maximum value of 0.14 ng/cm2-yr in the same core section as the observed PCB maximum peak. The maximum concentration appears approximately 10 years prior to the banning of toxaphene in 1982, and this peak has been documented previously in Lake Ontario and Lake Michigan ((3, 9) Table 1). The similarities in contaminant profiles between PCBs and toxaphene shown here may not be coincidental. Other northern Lake Michigan cores, specifically core LM68k, analyzed by two different labs, separately, for PCBs and toxaphene, also displayed similar profiles between the two contaminants (3, 9). PAH Profiles. PAHs display a different profile compared to the chlorinated contaminants, with peak concentrations earlier in the core, around 1942 ( 6.5 years (Figure 3c). The PAH patterns throughout the cores are dominated by the heavier molecular weight molecules such as indeno[1,2,3c,d]pyrene and benzo[g,h,i]perylene. The PAH depth profiles in Grand Traverse Bay sediment are similar to those reported by Simcik et al. (4) for other Lake Michigan sediment cores. Loadings of PAHs to the atmosphere increased during the WWII era due to an increase in coal consumption in the United States. This increase was accentuated in the Great Lakes due to their proximity to the industrial centers. Coke signals are dominated by heavier PAHs such as benzo[b+k]fluoranthene and chrysene, which are in high proportions in Lake Michigan sediments (4) and the sediment collected in Grand Traverse Bay. The shift to alternative energy sources in industrial processes and particulate emission controls reduced large point-source emissions of PAHs from the 1940s till 1980. Since 1980, the concentration of PAHs in the sediment has remained relatively constant and is approximately 40% of its peak value. The focus-corrected surficial t-PAH accumulation rate at HMS1 is 25 ng/cm2-yr. At HMS2, the PAH depth profile is similar to HMS1 except the concentrations and accumulation of PAHs are greater perhaps due to input from Traverse City. The surficial accumulation of t-PAH at HMS2 is 52 ng/cm2yr and is similar to accumulation rates observed in southern Lake Michigan (4). Northern Lake Michigan has higher t-PAH accumulation rates because contaminated sediment is

FIGURE 4. Total PCB concentrations over the past 25 years and their associated rate of decline in the air, water, and sediment. Data replotted from ref 5, ref 12, and this study. transported out of the southern basin, along the lake bottom, and deposited in the northern region. (4, 12). Grand Traverse Bay is not affected by this horizontal transport of sediment, and its PAH accumulation rates are similar to southern Lake Michigan (see Table 1).

Recent Decreases in HOCs Due to their low sedimentation rates and collection dates, previously reported sediment core geochronologies do not have the temporal resolution needed to quantify decreases in HOC accumulation rates that have occurred over the past 30 years. These cores collected from Grand Traverse Bay display similar contaminant profiles as open Lake Michigan but possess more data points in recently deposited sediment, with greater resolution than most other cores. The first-order rate of decline in t-PCB accumulation in the sediment core was 0.027 years-1. The rates of t-PCB decline in Lake Michigan air (33) and water (5) are four to six times faster than that in the sediment (Figure 5). This differences is probably due in part to the hydrophobic nature of PCBs, which causes them to be retained in the sediment as well as to the efficient recycling of PCBs in the benthic region. Only 1-35% of the PCB flux associated with the settling particles is actually buried in the sediment (1). The slow settling of the PCB enriched particles in the benthic nepheloid layer could be causing the slower rate of decline in the sediment. PCBs are primarily lost from the water column through sedimentation and volatilization. To assess the influence each of these loss processes on water column concentrations in Lake Superior, Jeremiason et al. (2) calculated values for ksed, or the proportion of PCBs lost from the water column that are actually buried in the sediment. The calculated ksed values are similar to the rates of decline of PCBs in Grand Traverse Bay sediment and were only a small component (2-10%) of the overall rate loss of PCBs from the water column of Lake Superior (k ) 0.20 years-1). The low values calculated in both studies reflect the inefficient burial of PCBs in the sediment and imply that air-water exchange is the dominant processes controlling PCB levels in the water column. VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3813

Coke emissions were enriched in the HMW PAH compounds, whereas lower temperature combustion is dominated by the LMW PAH compounds. This trend in the declination rates probably reflects the decreased production of HMW PAHs rather than selective degradation of these compounds.

Acknowledgments We thank Jeff McDonald of Indiana University for his guidance and assistance in the toxaphene analysis; the captain and crew of the RV Lake Guardian for the use of their vessel and aid in extracting the cores; Holly Bamford for her assistance in field sampling; and Dr. Jeff Jeremiason for his continued support in this project.

Literature Cited

FIGURE 5. First-order rate of decline calculated in HMS 1 for individual (a) PCB congeners and (b) PAH compounds plotted against their log Kow values. For PCBs the rate constants were calculated using the upper 18 cm of sediment. For PAHs, the rate constants were calculated using sediment sections collected from 12 to 30 cm. Rate constants for individual congeners were calculated in these cores to examine the congener-specific temporal trends. Figure 4 displays the relationship between rate constants of individual PCB congeners and their log Kow values. The lighter congeners are decreasing more rapidly in the sediment than the heavier congeners up to a log Kow value of around 7, after which the rate constants decrease. This trend in declination rates probably occurs because the lighter weight congeners desorb off the sediment particles more rapidly than the higher molecular weight congeners (11). These lower molecular weight congeners are found primarily in the dissolved phase in the water column (4). The slowly declining higher molecular weight PCB congeners such as 2,2′,4,4′,5,5′-hexachlorobiphenyl (IUPAC #153), 2,2′,3,3′4,6′hexachlorobiphenyl (IUPAC #132), and 2,3,3′,4,4′-pentachlorobiphenyl (IUPAC #105) comprise a large portion of the total PCB burden in the sediment. The overall rate of decline of toxaphene was 0.028 years-1 and is similar to the PCB rate of decline. This similarity suggests that the processes controlling the rate of toxaphene decline in the sediment are similar to the ones controlling PCB decline. It is possible that toxaphene is also efficiently recycled in the benthic region. The rate constant for toxaphene homologue groups was not calculated due to the higher uncertainties in our estimates provided by the low concentrations. The maximum peak in t-PAH accumulation occurred earlier than for PCBs and toxaphene, and the rate of decline for t-PAHs was calculated from 1942 to 1980. After 1980 the t-PAH concentration in the sediment cores appears to have leveled off reflecting relatively constant current inputs of t-PAHs to the sediment. From 1942 to 1980 t-PAHs declined at a rate of 0.017 years-1 with the HMW PAHs declining more rapidly than the LMW PAHs (Figure 5). This trend probably reflects changes in PAH production rather than selective degradation. During this time period, coke production declined and was replaced by lower temperature combustion. 3814

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 19, 2001

(1) Baker, J. E.; Eisenreich, S. J.; Eadie, B. J. Environ. Sci. Technol. 1991 25, 500-509. (2) Jeremiason, J. D.; Hornbuckle, K. C.; Eisenreich, S. J. Environ. Sci. Technol. 1994, 28, 903-914. (3) Pearson, R. F.; Swackhamer, D. L.; Eisenreich, S. J.; Long, D. T. Environ. Sci. Technol. 1997, 31, 3523-3529. (4) Simcik, M. F.; Eisenrich, S. J.; Golden, K. A.; Liu, S.; Lipiatou, E.; Scackhamer, D. L.; Long. D. T. Environ. Sci. Technol. 1996, 30, 3039-3046. (5) Offenberg, J. H.; Baker, J. E. J. Great Lakes Res. 2000, 24, 342352. (6) DeVault, D. S.; Hesselberg, R.; Rodgers, P. W.; Feist, T. J. J. Great Lakes Res. 1996, 22, 884-895. (7) Hickey, J. P.; Chernyak, S. M.; Begnoche, J. J.; Quintal, R. T.; Hesselberg, R. J. Poster Presentation, 42nd Annual Meeting of the International Association of Great Lakes Research; Cleveland, OH, 1999. (8) Pearson, R. F.; Hornbuckle, K. C.; Eisenreich, S. J.; Swackhamer, D. L. Environ. Sci. Technol. 1996, 30, 1429-1435. (9) Howdeshell, M. J.; Hites, R. A. Environ. Sci. Technol. 1996, 30, 220-224. (10) Mackay, D.; Bentzen, E. Atmos. Envion. 1997, 31, 4045-4047. (11) Gong, Y. Y.; Depinto, J. V.; Rhee, G. Y.; Liu, X. Water Res. 1998, 32, 2507-2517. (12) Golden, K. A.; Wong, C. S.; Jeremiason, J. D.; Eisenreich, S. J.; Sanders, G.; Hallgren, J.; Swackhamer, D. L.; Engstrom, D. R.; Long, D. T. Water Sci. Technol. 1993, 28, 19-31. (13) Hermanson, M. H.; Christensen, E. R.; Buser, D. J.; Chen, L. J. Great Lakes Res. 1991, 17, 94-108. (14) McCusker, E. M.; Ostrom, P. H.; Ostrom, N. E.; Jeremiason, J. D.; Baker, J. D. Org. Geochem. 1999, 30, 1543-1557. (15) Stapleton, H. M.; Masterson, C.; Skubinna, J.; Ostrum, P.; Ostrum, N. E.; Baker, J. E. Environ. Sci. Technol. 2001 in press. (16) Cohen, A. R. Masters Thesis, University of Maryland at College Park, 2001. (17) Kucklick, J. R.; Baker, J. E. Environ. Sci. Technol. 1998, 32, 11921198. (18) Ashley, J. T. F.; Baker, J. E. Environ. Toxicol. Chem. 1999, 18, 838-849. (19) Glassmeyer, S. T.; Shanks, K. E.; Hites, R. A. Anal. Chem. 1999, 71, 1448-1453. (20) Appleby, P. G.; Oldfield, F. Catena 1978, 5, 1-8. (21) Crusius, J.; Anderson, R. F. J. Paleolim. 1995, 13, 143-155. (22) Kada, J.; Heit, M. Hydrobiologia 1992, 246, 231-241. (23) Urban, N. R.; Gorham, E.; Underwood, J. K.; Martin, F. B.; Ogden, J. G. III. Limnol. Oceanogr. 1990, 35, 1516-1534. (24) Cornelissen, G.; Rigterink, H.; Jerdinandy, M. M. A.; VanNoort, P. C. M. Environ. Sci. Technol. 1998, 32, 966-970. (25) McNally, D. L.; Mihelcic, J. R.; Lueking, D. R. Chemosphere 1999, 38, 1313-1321. (26) Harkness, M. R.; McDermott, J. B.; Abramowicz, D. A.; Salvo, J. J.; Flanaga, W. P.; Stephens, M. L.; Mondello, F. J.; May, R. J.; Lobos, J. H.; Carroll, K. M.; Brennan, M. J.; Bracco, A. A.; Fish, K. M.; Warner, G. L.; Wilson, P. R.; Dietrich, D. K.; Lin, D. T.; Morgan, C. B.; Gately, W. L. Science 1993, 259, 503-507. (27) Fish, K. M.; Principe, J. M. Appl. Environ. Microb. 1994, 60, 4289-4296. (28) Williams, W. A. Chemosphere 1997, 34, 655-669. (29) Flanagan, W. P.; May, R. J. Environ. Sci. Technol. 1993, 27, 22072212.

(30) Muir, D. C. G.; Omelchenko, A.; Grift, N. P.; Savoie, D. A.; Lockhart, W. L.; Wilkinson, P.; Brunskill, G. J. Environ. Sci. Technol. 1996, 30, 3609-3617. (31) Gassmeyer, S. T.; DeVault, D. S.; Myers, T. R.; Hites, R. A. Environ. Sci. Technol. 1997, 31, 84-88. (32) Strachan, W. M. J.; Eisenreich, S. J. In Long-Range Transport of Pesticides; CRC Press: Boca Raton, FL, 1990; pp 291-301. (33) Hillery, B. R.; Basu, I.; Sweet, C. W.; Hites, R. A. Environ. Sci. Technol. 1997, 31, 1811-1816.

(34) Wong, C. S.; Sanders, G.; Engstrom, D. R.; Long, D. T.; Swackhamer, D. L.; Eisenreich, S. J. Environ. Sci. Technol. 1995, 29, 2661-2672.

Received for review December 28, 2000. Revised manuscript received June 25, 2001. Accepted July 10, 2001. ES002044D

VOL. 35, NO. 19, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3815