Assessment of Natural Attenuation via in Situ Reductive

Jan 7, 2005 - Polychlorinated biphenyl (PCB)-contaminated sediment cores taken from five locations in Lake Hartwell, SC, with an increasing distance f...
0 downloads 0 Views 330KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 945-952

Assessment of Natural Attenuation via in Situ Reductive Dechlorination of Polychlorinated Biphenyls in Sediments of the Twelve Mile Creek Arm of Lake Hartwell, SC USARAT PAKDEESUSUK, CINDY M. LEE,* JOHN T. COATES, AND DAVID L. FREEDMAN Department of Environmental Engineering & Science, Clemson University, Clemson, South Carolina 29634-5002

Polychlorinated biphenyl (PCB)-contaminated sediment cores taken from five locations in Lake Hartwell, SC, with an increasing distance from the point source were evaluated for the presence of in situ reductive dechlorination of PCBs on the basis of a comparative congenerspecific analysis of PCB distribution profiles between historical (1987) and current (1998) sediments from the same sites. A layer of 1998 sediment that was equivalent to 1987 sediment was determined by direct comparison of total PCB depth profiles after correction for any sedimentation that occurred at each location since 1987. Natural capping of contaminated sediments with the continued deposition of new sediments was observed in all locations except the one farthest from the source area. The residual PCB congeners accumulated in the field samples did not vary from site to site. Certain PCB congeners (e.g., 236-24 + 34-34, 245-25, and 23-4 CB) decreased with time and with depth along with an increase in lower chlorinated PCB congeners in all sampling locations. A similarity in distribution profiles between dechlorinated PCBs in laboratory microcosms and in the field samples was observed. These results provide supporting evidence that in situ reductive dechlorination has occurred in the Twelve Mile Creek arm of Lake Hartwell. Several sediment layers, particularly the sites with highest PCB concentration, showed similar PCB distribution profiles between 1987 and 1998. An additional change in chlorine distribution between 1987 and 1998 at most “equivalent” depths was not observed. The ortho- and para-substituted congeners that accumulated during dechlorination of Aroclor 1254 after nearly 1 yr of incubation in the laboratory were the prominent residual products in all field samples. At a few locations and depths, evidence for dechlorination at surprisingly low concentrations (1-5 ppm) was observed. These results confirm that in situ reductive dechlorination of PCBs is operating at a very slow rate and may have been at a plateau since 1987 for certain depths and certain locations.

* Corresponding author phone: (864)656-1006; fax: (864)656-0672; e-mail: [email protected]. 10.1021/es0491228 CCC: $30.25 Published on Web 01/07/2005

 2005 American Chemical Society

Introduction Monitored natural attenuation (MNA) is among several methods that can be considered for remediation of polychlorinated biphenyl (PCB)-contaminated sediments because of its potential cost savings and risk reduction. However, it has not been a commonly selected remedial option for many sites because of the longer time frame that is usually required to achieve the cleanup goal. Therefore, at many PCBcontaminated sites, for example, the Upper Hudson River and the Fox River, dredging of contaminated sediment has been selected as the cleanup option (1, 2). In the case of Lake Hartwell, SC, MNA was selected as the final remedy (3, 4). The natural process monitored in the Lake Hartwell case is burial of the contaminated sediment with clean sediment from the watershed. However, Lake Hartwell provides an excellent opportunity to evaluate the presence, extent, and effectiveness of in situ reductive dechlorination of PCBs as an additional MNA process that could affect the fate and distribution of PCBs in the sediment. To determine the presence of in situ reductive dechlorination, information on the original PCB composition in an environmental sample is very helpful; however, it is quite difficult to obtain. The general lines of evidence that are required to demonstrate the presence of in situ reductive dechlorination are a shift in congener distributions in environmental samples from the original PCBs discharged or commercial Aroclor standards that have been used at the site, a selective depletion of higher chlorinated PCB congeners, and an accumulation of the lower chlorinated PCB congeners, especially the ortho-substituted congeners (57). Changes in the total concentration of PCBs or chlorines per biphenyl is not as sensitive a measure as congener-specific analysis. This approach is based on the assumption that the PCB compositions discharged were similar to those deposited at each site. This assumption is appropriate if analysis of the sediment samples shows a distribution similar to that of those Aroclors used at the site. However, the individual PCB congeners vary greatly in their Henry’s law constant (KH) and sediment-water partition coefficient (Kd), resulting in variation in their susceptibility to volatilization, solubilization, sorption, and/or microbial degradation, subsequently resulting in differences in congener compositions in a given environmental compartment (8, 9). Therefore, evaluation for the presence of in situ reductive dechlorination of PCBs that relies mainly on the assumption of similarity between the original PCB composition deposited at the site and the commercial Aroclors or the original PCBs discharged from a point source could underestimate the extent of in situ reductive dechlorination in a given system because solubilization and volatilization remove the lower chlorinated congeners, which are products of dechlorination. In the case where historical PCB data are available along with current sediment information at the same site, comparison of the existing congener distribution after correction for sedimentation will provide a better assessment of the presence of in situ reductive dechlorination in a given environment. The presence of in situ reductive dechlorination can be evaluated on the basis of the concept of comparative analysis of the historical PCB data and the current PCB data at an “equivalent” depth. An equivalent depth is defined as a layer of the currently collected sediment that can be matched to the previously collected sediment after correction for sedimentation since the previous sediment collection. The equivalent depth can be determined by either direct comparison of total PCB depth profiles or radionuclide dating (e.g., 210Pb). VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

945

FIGURE 1. Location of Lake Hartwell, the Twelve Mile Creek arm of the lake, and sampling sites for this study. Lake Hartwell, located on the border between Georgia and South Carolina as shown in Figure 1, was contaminated with PCBs from the mid-1950s to 1970s. PCB-containing effluents were discharged into Twelve Mile Creek, and subsequently drained to Lake Hartwell (10, 11). As a result of PCB contamination, the entire Twelve Mile Creek watershed and the Seneca River arm of Lake Hartwell were proposed for the National Priority List (NPL) in 1987, and became eligible for cleanup funds under CERCLA in 1990 (12). A comprehensive sediment survey to determine the extent of PCB contamination in Lake Hartwell was conducted 946

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

in 1987 (10, 13). This survey and others suggested the presence of reductive dechlorination in sediments of Lake Hartwell on the basis of the accumulation of lower chlorinated PCBs (10, 13, 14). Laboratory microcosm studies provided evidence of indigenous PCB dechlorinating populations in Lake Hartwell sediments (15, 16). With the available historical data on the PCB congener distribution for Lake Hartwell sediment, comparative congener-specific analysis of the current sediment data to the historical data at the equivalent sediment layers was used in this study to determine the presence of in situ reductive

dechlorination. Information on the presence of in situ reductive dechlorination in Lake Hartwell is valuable because so little work on in situ reductive dechlorination as a form of MNA for PCBs exists. In addition, the remediation strategy could be simply to allow this biological process to take place and yield less chlorinated products, which are less toxic (17, 18) and are susceptible to aerobic degradation (19). Comparative analysis at two different time points allows not only assessment of in situ reductive dechlorination but also determination of whether the in situ transformation process is in an active or plateau phase. A plateau phase can be determined when the dechlorination rate is very slow, where the slope of the concentration data approaches or becomes zero. An active phase is indicated by a slope distinctly greater than zero for the decreasing concentration. The information on the current phase of in situ transformation processes is a crucial step in making policy decisions about remediation strategies (20).

Materials and Methods Sediment Collection. Sediment cores were collected in July 1998, using the same approach as previously reported (1516). Sediment cores were collected from five locations in the lake, where Germann (13) collected his samples in 1987. Four sediment core samples were collected from the Twelve Mile Creek arm of Lake Hartwell, while one core was collected from a downstream location at the confluence of the Seneca River and the Twelve Mile Creek arm of Lake Hartwell. These sampling locations as designated by Germann (13) were G27, G30, G33, G46, and G49B (Figure 1). Cores were transported to the L.G. Rich Laboratory (Clemson University, Clemson, SC). Sediment cores were extruded from the sampling tubes and sectioned into 5 cm increments. To avoid contamination from the tube and adhering sediment, sediment contacting the Lexan tube was discarded. The remaining portions were transferred to a glass jar and stored at 4 °C until extraction and analysis. PCB Extraction and Analysis. PCBs were extracted from sediments with acetone and isooctane under ultrasonication (Fisher 300 sonic dismembrator, Fisher Scientific) as described previously by Germann for the 1987 cores (10, 13). Samples were spiked with octachloronaphthalene as a recovery standard to determine the extraction efficiency, which averaged 103.4% ((23.3%). The extraction results were not corrected for extraction efficiencies. Sediment extracts were analyzed for PCBs on a Hewlett-Packard 6890 gas chromatograph equipped with a 63Ni electron-capture detector (ECD) using essentially the same conditions as Germann used for the 1987 cores (10, 13). A 30 m fused silica capillary column with a 0.25 mm diameter and a film thickness of 0.25 µm (ZB-5, Phenomenex, Torrance, CA) was used with a temperature program that began at 100 °C, which was held for 2.5 min, and was ramped at 10 °C/min to 150 °C, which was held for 0.5 min, ramped at 1.1 °C/min to 225 °C, which was held for 3 min, and finally ramped at 10 °C/ min to 260 °C, which was held for 15 min. The injector was set at 250 °C and the detector at 325 °C. PCB congeners were identified and quantified using the internal standard method as previously described (15) with the daily working standards containing a 4:1 mixture of Aroclors 1016 and 1254 according to Germann (13). Reductive dechlorination was evaluated on the basis of changes in the mole percentage of each identified PCBcontaining peak and/or changes in the average number of m-chlorines + p-chlorines, o-chlorines, and total chlorines per biphenyl as observed in the 1998 sediment samples compared to its equivalent depth in the 1987 sediment samples. The mole percent distribution for each peak was determined by normalizing the congener-specific data to the total PCB concentration. The average number of chlorines

FIGURE 2. Total PCB concentration profile with depth at each sampling location: (O) 1987; (b) 1998. per biphenyl was calculated as the product of the average number of chlorines and molar concentration of each peak divided by the total molar concentration summed over all peaks (21). Coeluting congeners were assumed to be present in equal proportions, and the biphenyl moiety was assumed to remain intact (7, 21, 22). It was also assumed that between 1987 and 1998 the only losses of PCBs were due to dechlorination and not to solubilization or desorption. Data analysis supports this assumption (see mass balance analysis in the Results and Discussion). Determination of Equivalent Depths. The equivalent depths for the 1998 and 1987 cores were determined by matching the total PCB concentration profiles for the two time periods. Mass balances for the G30, G33, G46, and G49 cores indicate that the total number of moles of PCBs did not change substantially between 1987 and 1998. In effect, total PCBs served as a conservative tracer, with their shift in depth reflecting the deposition of sediment.

Results and Discussion Total PCB Profiles and Equivalent Depths. Profiles of total PCB concentrations as a function of depth are shown in Figure VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

947

FIGURE 3. Number of total chlorines, m-chlorines + p-chlorines, and o-chlorines per biphenyl at the equivalent depths in G30 and G46 cores: (O) 1987 data; (b) 1998 data. 2 for core samples taken in 1987 and 1998, at five locations (Figure 1). For G30, G33, G46, and G49, PCB profiles in the 1987 and 1998 cores are very similar, except that, in the G30, G33, and G46 cores, the distribution is shifted downward. Since no major hydraulic disturbances of the sediment (i.e., flushing events) were reported between 1987 and 1998, displacement of the profiles by increasing depth provides a method to estimate the natural sedimentation rate at each location. For the G30, G33, and G46 cores, the 1998 PCB distribution overlaps with the 1987 distribution if the 1998 profiles are shifted upward to an equivalent depth. The distance for this shift is approximately 20 cm for the G30 and G33 cores, which translates to an average deposition rate of 2 cm/yr. This estimate of sedimentation rate assumes a continuous deposition over the 11 yr between the collection of the two cores. For the G46 core, the shift to an equivalent depth is approximately 10 cm, or an average deposition rate of 1 cm/ yr. Cores taken by Brenner et al. (4) indicate that the 1987 surface lies at about 20 cm depth for a core taken in 2000 and at about 50 cm for a core taken in 2001 at the G30 location as estimated by lead-210 dating. They also report the 1987 surface is at about 10 cm for a core taken in 2000 at the G46 location (4). These depths correspond well to the profiles shown in Figure 2. Our estimated rates are in agreement with those reported by Bechtel (23) and Farley et al. (10). Brenner et al. (4) reported sedimentation rates for G30 ranging from 0.77 to 3.26 g/(cm2 yr) and for G46 of 0.23 g/(cm2 yr) on the basis of lead-210, cesium-137, and total PCB profiles. Our estimated rates are about 4.7 g/(cm2 yr) for G30 and 2.4 g/(cm2 yr) for G46, assuming a sediment density of 2.6 g/cm3. The differences are likely due to differences in the estimation 948

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

of densities, the assumption of continuous deposition, and variability of the deposition process as seen from the range reported for G30 by Brenner et al. (4) for cores taken in 2000 and 2001. For the G49B core, the 1987 and 1998 total PCB distributions essentially overlap, indicating little or no deposition occurred. This is consistent with a reported deposition rate of 0-1 cm/yr for the G49B region of the lake (23). Of the five cores analyzed, G49B is the furthest downstream from the PCB source area, and the total PCBs present were lowest there. Nevertheless, the apparent lack of capping at G49B over the 10 yr period evaluated suggests that PCBs in this area may be serving as an ongoing source for bioaccumulation and persistence in various aquatic biota. The G27 core is the only one of the five tested for which the 1987 and 1998 PCB profiles do not match. Elzerman et al. (12) reported sedimentation rates of 6 cm/yr for the Madden Bridge basin in which the G27 cores were collected. Others report high amounts of sand deposited in this area from periodic flushing of the upstream impoundments (4). At this rate, the PCB distribution in 1998 was probably located at depths of 60-120 cm. Brenner et al. (4) reported a maximum total PCB concentration for the same location at 95-105 cm. Unfortunately, the 1998 core did not extend beyond 60 cm, suggesting that the PCB distribution for 1998 was missed by not coring deeply enough. The total mass of PCBs in the G30, G33, G46, and G49B cores (determined by integrating the area under the concentration profiles in Figure 2) did not decrease significantly between 1987 and 1998, indicating little if any in situ removal of total PCBs occurred over that time span. Calculating the total moles per core by assuming a density of 0.3 g (dry

weight)/cm3 (10) shows little change between 1987 and 1998 for G33 (13.9 and 14.5 µmol), G46 (7.56 and 5.99 µmol), and G49B (0.75 and 0.82 µmol). For G30, there is a larger change (26.1 and 35.4 µmol) which is likely due to failure to capture the entire contaminated profile as can be seen in Figure 2, where the profile does not return to zero for either year. This observation is consistent with the similarity in total chlorines per biphenyl in the 1987 and 1998 cores at several depths for G30 (Figure 3A), G46 (Figure 3B), and G33 and G49B (see the Supporting Information). The average of 2.8-4.0 chlorines per biphenyl is within the range reported for the dechlorination products from Aroclor 1254 during a microcosm study that used sediment samples from the G30 and G33 locations (15). The lack of a substantial change in the magnitude of m-chlorines + p-chlorines throughout the depth of most of the cores (i.e., at equivalent depths) further suggests a lack of significant in situ reductive dechlorination between 1987 and 1998 (Figure 3C,D). However, in the deeper portions of some of the cores (i.e., >∼25 cm equivalent depth), there was a modest decrease in m-chlorines + p-chlorines, especially for G46 (Figure 3D). Even modest changes in the number of chlorines per biphenyl is significant because this measure is relatively insensitive compared to congenerspecific analysis. No significant change in the average number of o-chlorines per biphenyl occurred with depth in G30 (Figure 3E), in G49 (Supporting Information), and at shallow depths in G46 (Figure 3F). There appears to be a slight increase in the average number of o-chlorines per biphenyl with depth in G46 (Figure 3F). However, it is likely that the change is insignificant at less than 0.2 chlorine given the error in determining equivalent depths. The decreases in total chlorines per biphenyl and in m-chlorines + p-chlorines per biphenyl at the two deepest G46 levels do appear to be significant, at greater than 0.4-1 chlorine (Figure 3B,D). Changes in PCB Composition. Although Figures 2 and 3 indicate there was little if any decrease in the total amount of PCBs present in the cores between 1987 and 1998, congener-specific analysis suggests that changes in PCB composition (although not in total moles) did occur at most depths and locations. There was an increase in the mole percent of lower chlorinated congeners (including 2-2 + 26, 2-4, 4-4, 24-2, and 24-4) in the former surface of G30 cores (Figure 4). Other dechlorination products such as 2-chlorobiphenyl may have formed, but they were not included in the number of congeners determined due to separation difficulties under the selected GC conditions. The depth listed for 1998 in Figure 4 corresponds to the 1987 depth plus sedimentation. Similar results were observed in the former surface of G33 cores (see the Supporting Information) and G46 cores (Figure 5). For the G46 location the enrichment of 2-2 + 26 congeners was especially dramatic at depths of 35-40 cm (Figure 5D) and 40-45 cm (Figure 5E) in the 1998 core compared to the 1987 core. 2-2 + 26 dichlorobiphenyls were detected at very low concentrations in the G49B core collected in 1987 but were prominent in 1998 (see the Supporting Information). The increase in mole percent of lower chlorinated congeners corresponded to a decrease in the higher chlorinated congeners (e.g., 236-34 + 34-34, 24525, 23-4, and 24-24 + 246-4 + 245-2) at most equivalent depths, confirming the continuation of in situ reductive dechlorination. These observed changes in congener distribution reflecting reductive dechlorination were surprising given the low concentrations (1-5 ppm) at the G49B location. A decrease in the average number of chlorines per biphenyl (Figure 3) is apparent at the deepest portions of G46 (Figure 3B), where we also observed definitive increases in the lower chlorinated congeners (Figure 5D,E). Despite the definitive decrease in several of the higher chlorinated congeners and a corresponding increase in lower chlorinated congeners at

FIGURE 4. Mole percent distribution profiles of selected PCBs representing an active phase at four different equivalent depths in G30 cores collected in 1987 and 1998 after correction of the 1998 depth for sedimentation: (0) 1987 data; (9) 1998 data.

FIGURE 5. Mole percent distribution profiles of selected PCBs representing an active phase at five different equivalent depths in G46 cores collected in 1987 and 1998 after correction of the 1998 depth for sedimentation: (0) 1987 data; (9) 1998 data. several depths, the net sum of these changes was not sufficient to yield a detectable change in the mass concentration of total PCBs for the entire core (Figure 2). This seems reasonable given the relatively low concentration of individual congeners. While changes in the PCB congener distribution did occur in the G30, G33, G46, and G49 cores at several equivalent depths between 1987 and 1998, a lack of change in the congener distribution was also apparent at other depths, especially in the upstream locations with the highest PCB concentrations (Figure 2). The G30 cores had similar congener VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

949

FIGURE 6. Mole percent distribution of selected PCBs representing a plateau phase at three different equivalent depths in G30 cores collected in 1987 and 1998 after correction of the 1998 depth for sedimentation: (0) 1987 data; (9) 1998 data. profiles between 1987 and 1998 in the 4-10 cm (Figure 6A), 9-15 cm (Figure 6B), and 15-20 cm (Figure 6C) layers (1987 depths). Other cores showed similar results at certain depths.

These results suggest that in situ reductive dechlorination for certain depths may be at the plateau phase, when the rate of dechlorination becomes very slow or effectively zero. Variations in the dechlorination rate at different depths within a core are possible since the extent and rate of PCB dechlorination are reported to be concentration-dependent (5, 26-28). Also the supply of electron donors and other environmental conditions that affect biological reductive dechlorination are likely to vary with depth. Given that specific removal processes have been associated with locations and microbial communities (5, 6), a reasonable explanation for the slowed dechlorination rate is that congeners with susceptible chlorine positions have been removed. For further dechlorination to occur, a different dechlorination process, that is, a different microbial community, would be necessary. A recent paper by Cho et al. (29) found a threshold for individual congeners in Aroclor 1248 at a total PCB concentration of more than 40 ppm in microcosms inoculated with microorganisms from the St. Lawrence River. They divided the congeners that were dechlorinated into two groups; one required at least 40 ppm total PCBs, and the other required 60 ppm or more. These groups may represent congeners with different susceptibilities for different dechlorinating microorganisms. Comparison of PCB Congener Distribution in Core Samples and Microcosms. Figure 7 compares the PCB

FIGURE 7. Mole percent distribution of PCB congeners in Aroclor 1254 standard, dechlorinated Aroclor 1254 in a laboratory microcosm, and selected examples from sediments from G30, G33, and G46. Error bars for the laboratory microcosms are the standard deviations of three bottles; where not shown, the deviations were smaller than the size of the symbols. 950

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

congener distribution in 1998 core samples from G30 (Figure 7C), G33 (Figure 7D), and G46 (Figure 7E) at an actual depth of 35-40 cm with the dechlorination products from Aroclor 1254 (Figure 7B) measured during a laboratory microcosm study (19). The microcosms shown were prepared with a core sample from G30 and spiked with 500 ppm Aroclor 1254. After a lag phase of 50 days, a period of rapid dechlorination ensued over the next 90 days, followed by no detectable additional dechlorination during the next 113 days of incubation. Approximately 40% of the total chlorines in Aroclor 1254 were removed, yielding an average of about 3.0 chlorines per biphenyl when dechlorination approached the plateau phase at which the dechlorination rate is very slow, where the slope of the data approaches or becomes zero (15). The chlorines from the meta position were preferentially removed as seen from the accumulation of ortho- and parasubstituted PCB congeners as the final dechlorination products in the spiked microcosms (15). Similar results were obtained for microcosms prepared with sediment samples from G33. The PCB congener distribution profiles observed in the field samples at the indicated depth parallel those from the microcosms, providing further evidence that in situ reductive dechlorination of PCBs did occur in the Twelve Mile Creek arm of Lake Hartwell. Furthermore, the ortho- and parasubstituted congeners observed in the field samples were also the prominent peaks at the end of the 253 day incubation period for the G30 and G33 microcosms (19). This similarity is especially interesting because the field PCBs had incubated in the sediment for decades, while the microcosms were monitored for less than 1 yr (253 days). The results suggest that when PCBs entered Lake Hartwell, they underwent a relatively rapid rate of dechlorination (dependent on concentration and various environmental factors), followed by the current extended period of little or no detectable activity, or a plateau phase. Even if some changes in congener distribution do occur during continued incubation, the magnitude of this change (based on the 1987 and 1998 data) is not likely to result in a significant decrease in total PCBs, at least on a time scale of decades. The field evidence for a plateau phase in PCB dechlorination activity, corresponding to the microcosm results, implies that successful application of monitored natural attenuation for Lake Hartwell will hinge on capping of the recalcitrant PCBs with a sufficient amount of uncontaminated sediment to isolate them from the food chain. Capping with fresh sediment will be an especially significant challenge in areas where the rate of sediment deposition is very slow (e.g., G49B). Even though PCB levels are lower in these downstream areas, such locations may be providing an ongoing source for bioaccumulation that is keeping fish advisories in place for Lake Hartwell (30). At certain depths and locations, slow dechlorination continues at surprisingly low concentrations; however, little to no mass removal was observed, and without additional understanding of optimum sediment conditions for dechlorination, predictions about where dechlorination will take place are not possible. Geochemical parameters such as organic matter content and concentrations of electron acceptors including nitrate, sulfate, iron, and manganese will be investigated for the various depths to improve our ability to make predictions.

Acknowledgments We gratefully acknowledge support from the U.S. Environmental Protection Agency, Ecosystems Research Divisions, National Exposure Research Laboratory, Athens, GA. We also gratefully acknowledge James L. Meyers for assistance in collecting Lake Hartwell sediments. We also appreciate the insightful comments of the anonymous reviewers, which improved the manuscript greatly.

Supporting Information Available Four tables with the total PCB concentrations and total chlorines, m-chlorines + p-chlorines, and o-chlorines per biphenyl at every depth for G30, G33, G46, and G49B; three figures with data from G49B showing total chlorines, ochorines, and m-chlorines + p-chlorines and G33 and G49B showing the congener distribution at equivalent depths. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Hudson River PCBs Reassessment RI/FS Phase 3 Report: Feasibility Study. Book 1; Prepared for U.S. EPA Region 2 and U.S. Army Corp of Engineers by TAMS Consultants, Inc.: Bloomfield, NJ, 2000. (2) Sandin, J. Milawaukee J. Sentinel 2001, Oct 3. (3) EPA Superfund Record of Decision: Sangamo Weston/Twelve Mile Creek/Lake Hartwell PCB Contamination OU2, Pickens, SC; EPA/ROD/R04-94/178; U.S. Environmental Protection Agency, Region IV: Atlanta, GA, 1994. (4) Brenner, R. C.; Magar, V. S.; Ickes, J. A.; Foote, E. A.; Abbott, J. E.; Bingler, L. S.; Crecelius, E. A. Long-term recovery of PCBcontaminated surface sediments at the Sangamo-Weston/ Twelvemile Creek/Lake Hartwell Superfund Site. Environ. Sci. Technol. 2003, 38, 2328-2337. (5) Bedard, D. L.; Quensen, J. F., III. In Microbial Transformation and Degradation of Toxic Organic Chemicals; Young, L. Y., Cerniglia, C. E., Eds.; Wiley-Liss: New York, 1995; pp 127-216. (6) Bedard, D. L.; May, R. J. Characterization of the polychlorinated biphenyls in the sediments of Woods Pond: Evidence for microbial reductive dechlorination of Aroclor 1260 in situ. Environ. Sci. Technol. 1996, 30, 237-245. (7) Sokol, R. C.; Kwon, O.-S.; Bethoney C. M.; Rhee G.-Y. Reductive dechlorination of polychlorinated biphenyls in St. Lawrence River sediments and variations in dechlorination characteristics. Environ. Sci. Technol. 1994, 28, 2054-2064. (8) Burkhard, L. P.; Armstrong, D. E.; Andren, A. W. Partitioning behavior of polychlorinated biphenyls. Chemosphere 1985, 14, 1703-1716. (9) Dunnivant, F. M. Ph.D. Dissertation, Clemson University, 1988. (10) Farley, K. J.; Germann, G. G.; Elzerman, A. W. 1994. In Environmental Chemistry of Lakes and Reservoirs; Baker, L. A., Ed.; American Chemical Society: Washington, DC, 1994; pp 575-600. (11) Data Package for EPA Lake Hartwell Superfund Study (PER 34); South Carolina Department of Health and Environmental Control: Columbia, SC, 1976. (12) Elzerman, A. W.; Dunnivant, F. M.; Germann, G. G. PCBs and Related Compounds in Lake Hartwell; Report No. 131; South Carolina Water Resources Research Institute: Clemson, SC, 1991. (13) Germann, G. G. M.S. Thesis, Clemson University, 1988. (14) Ickes, J.; Brenner, R.; Magar, V. S.; Durell, G.; Johnson, G.; Crecelius, E.; Abbott, J.; McCarthy, C. P-.; Bingler, L. In Phytoremediation, Wetlands, and Sediments, The Sixth International In Situ and On-Site Bioremediation Symposium, San Diego, CA, June 4-7, 2001; Leeson, et al., Eds.; Battelle Press: Columbus, OH, 2001; Vol. 6, pp 231-236. (15) Pakdeesusuk, U.; Freedman, D. L.; Lee, C. M.; Coates, J. T. Reductive dechlorination of polychlorinated biphenyls in sediment from the Twelve Mile Creek Arm of Lake Hartwell. Environ. Toxicol. Chem. 2003, 22, 1214-1220. (16) Pakdeesusuk, U.; Jones, J. W.; Lee, C. M.; Garrison, A. W.; O’Niell, W. L.; Freedman, D. L.; Coates, J. T.; Wong, C. S. Changes in enantiomeric fractions (EF) during microbial reductive dechlorination of PCB132, PCB149, and Aroclor1254 in Lake Hartwell sediment microcosms. Environ. Sci. Technol. 2003, 37, 11001107. (17) Mousa, M. A.; Ganey, P. E.; Quensen, J. F. I.; Madhukar, B. V.; Chou, K.; Giesy, J. P.; Fischer, L. J.; Boyd, S. A. Altered biologic activities of commercial polychlorinated biphenyl mixtures after microbial reductive dechlorination. Environ. Health Perspect. 1998, 106 (Suppl. 6), 1409-1418. (18) Quensen, J. F., III; Mousa, M. A.; Boyd, S. A.; Sanderson, J. T.; Froese, K. L.; Giesy, J. P. Reduction of aryl hydrocarbon receptormediated activity of polychlorinated biphenyl mixtures due to anaerobic reductive dechlorination. Environ. Toxicol. Chem. 1998, 17, 806-813. (19) Abramowicz, D. A.; Olson, D. R. Accelerated biodegradation of PCBs. CHEMTECH 1995, 25, 36-41. VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

951

(20) Rhee, G.-Y.; Cho, Y.-C.; Ostrofsky, E. B. Microbial degradation of PCBs in contaminated sediments. J. Rem. 1999, 11, 69-82. (21) Cho, Y.-C.; Kwon, O.-S.; Sokol, R. C.; Bethoney, C. M.; Rhee, G.-Y. Microbial PCB dechlorination in dredged sediments and the effect of moisture. Chemosphere 2001, 43, 1119-1126. (22) Quensen, J. F. I.; Boyd, S. A.; Tiedje, J. M. Dechlorination of four commercial polychlorinated biphenyl mixtures (aroclors) by anaerobic microorganisms from sediments. Appl. Environ. Microbiol. 1990, 56, 2360-2369. (23) Remedial Investigation Report for the Sangamo Weston, Inc./ Twelve Mile Creek/Hartwell Lake Site Operable Unit Two, Pickens, Pickens County, South Carolina; Prepared for the U.S. EPA by Bechtel Environmental, Inc.: Oak Ridge, TN, 1993; Contract W9-0058. (24) Lake, J. L.; Pruell, R. J.; Osterman, F. A. In Organic Substances and Sediments in Water; Baker, R. A., Ed.; Lewis Publishers: Chelsea, MI, 1991; pp 173-197. (25) Lake, J. L.; Pruell, R. J.; Osterman, F. A. An examination of dechlorination processes and pathways in New Bedford Harbor sediments. Mar. Environ. Res. 1992, 33, 31-47. (26) Abramowicz, D. A.; Brennan, M. J.; Van Dort, H. M.; Gallagher, E. L. Factors influencing the rate of polychlorinated biphenyl

952

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

(27) (28)

(29)

(30)

dechlorination in Hudson River sediments. Environ. Sci. Technol. 1993, 27, 1125-1131. Fish, K. M. Influence of Aroclor 1242 concentration of polychlorinated biphenyls transformation in Hudson River test tube microorganisms. Appl. Environ. Microbiol. 1996, 62, 3014-3016. Rhee, G.-Y.; Bush, B.; Bethoney, C. M.; DeNucci, A.; Oh, H.-M.; Sokol, R. C. Reductive dechlorination of Aroclor 1242 in anaerobic sediments: Pattern, rate, and concentration dependence. Environ. Toxicol. Chem. 1993, 12, 1025-1032. Cho, Y.-C.; Sokol, R. C.; Frohnhoefer, R. C.; Rhee, G.-Y. Reductive dechlorination of polychlorinated biphenyls: Threshold concentration and dechlorination kinetics of individual congeners in Aroclor 1248. Environ. Sci. Technol. 2003, 37, 5651-5656. South Carolina Department of Health and Environmental Control. South Carolina Fish Consumption Advisories. www.scdhec.gov/eqc/admin/html/fishadv.html (accessed July 2003).

Received for review June 10, 2004. Revised manuscript received November 11, 2004. Accepted November 15, 2004. ES0491228