Long-Term Recovery of PCB-Contaminated Sediments at the Lake

King Drive, Cincinnati, Ohio 45268, and Battelle Memorial. Institute, 505 King ... (8) resembled. Processes C and H′, as identified by Bedard and Qu...
2 downloads 0 Views 541KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 3548-3554

Long-Term Recovery of PCB-Contaminated Sediments at the Lake Hartwell Superfund Site: PCB Dechlorination. 2. Rates and Extent V I C T O R S . M A G A R , * ,† RICHARD C. BRENNER,‡ GLENN W. JOHNSON,§ AND JOHN F. QUENSEN, III| U.S. Environmental Protection Agency, National Risk Management Research Laboratory, 26 West Martin Luther King Drive, Cincinnati, Ohio 45268, and Battelle Memorial Institute, 505 King Avenue, Columbus, Ohio 43201

This paper reports on extensive polychlorinated biphenyl (PCB) dechlorination measured in Lake Hartwell (Pickens County, SC) sediments. Vertical sediment cores were collected from 18 locations in Lake Hartwell (Pickens County, SC) and analyzed in 5-cm increments for PCB congeners. The preferential loss of meta and para chlorines with sediment depth demonstrated that PCBs in the sediments underwent reductive dechlorination after burial. Notably, ortho chlorines were highly conserved for more than 5 decades; since the first appearance of PCBs, ca. 1950-1955. These dechlorination characteristics resulted in the accumulation of lower chlorinated congeners dominated by ortho chlorine substituents. Dechlorination rates were determined by plotting the numbers of meta plus para chlorines per biphenyl molecule (mol of chlorine/mol of PCB) with sediment age. Regression analyses showed linear correlations between meta plus para chlorine concentrations with time. The average dechlorination rate was 0.094 ( 0.063 mol of Cl/mol of PCB/yr. The rates measured using the 2001 cores were approximately twice those measured using the 2000 cores, most likely because the 2001 cores were collected only at transects O, L, and I, which had the highest rates measured in 2000. An inverse of the dechlorination rates indicated that 16.4 ( 11.6 yr was required per meta plus para chlorine removal (ranging from 4.3 to 43.5 yr per chlorine removal). The rates determined from this study were 1-2 orders of magnitude lower than rates reported from laboratory microcosm studies using Hudson River and St. Lawrence River sediments, suggesting that dechlorination rates reported for laboratory experiments are much higher than those occurring in situ.

Introduction Risk reduction is the long-term goal of contaminated sediment management. In aquatic environments affected * Corresponding author present address: ENVIRON, International Corporation, 123 N. Wacker Dr., Suite 250, Chicago, IL 60606; phone: (312)853-9430; e-mail: [email protected]. † Battelle Memorial Institute. ‡ U.S. Environmental Protection Agency. § Permanent address: University of Utah, Energy and Geoscience Institute, Civil and Environmental Engineering Dept., 423 Wakara Way, Suite 300, Salt Lake City, UT 84108. | Permanent address: Michigan State University, 528 Plant and Soil Science Bldg., East Lansing, MI 48824. 3548

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 10, 2005

by contaminated sediments, risk management strategies focus on either removing the contaminated material or interrupting exposure pathways by which contaminants might pose an ecological or human health risk over time (pathway interdiction). This is generally achieved by dredging, isolation (capping), or natural recovery. Natural recovery results in risk reduction either by pathway interdiction through natural processes such as contaminant burial and sequestration (1-6) or by mass removal via chemical weathering and biodegradation. The National Risk Management Research Laboratory (NRMRL) of the U.S. Environmental Protection Agency (U.S. EPA) has an ongoing interest in developing effective, inexpensive remediation technologies for contaminated sediments. Their interest includes the study and development of field monitoring tools to evaluate natural recovery processes in aquatic sediments. This study is part of an evaluation of the recovery of PCB-contaminated lake sediments at the Sangamo-Weston/Twelvemile Creek/Lake Hartwell Superfund site (Pickens County, SC; Figure 1). The Sangamo-Weston plant manufactured capacitors from 1955 to 1978 using PCB-containing dielectric fluids containing Aroclors 1016, 1242, and 1254 (7). Although the SangamoWeston plant no longer exists, waste disposal practices dating back to earlier operations led to PCB-contaminated sediments in Lake Hartwell, downstream of the Sangamo-Weston plant. In a previous paper (1), we reported that natural burial processes led to measurable PCB concentration reductions in surface sediments at Lake Hartwell, and in our companion paper (8), we use fingerprinting analyses of ambient data to characterize extensive dechlorination in Lake Hartwell sediments; fingerprinting was conducted using a polytopic vector analysis (PVA) multivariate receptor modeling method (9) and via direct comparison of surface and buried congener distribution patterns with literature reported patterns. Dechlorination patterns reported by Magar et al. (8) resembled Processes C and H′, as identified by Bedard and Quensen (10). Under anaerobic conditions, such as those typically found in Lake Hartwell sediments, the primary metabolic pathway for PCBs is reductive dechlorination in which chlorine removal and substitution with hydrogen by bacteria result in a reduced organic compound with fewer chlorines (1012). Reductive dechlorination of PCBs preferentially removes chlorines from the meta and para positions of PCBs (10), which has been shown to lead to the conservation of biphenyl rings and ortho chlorines in laboratory dechlorinating enrichment cultures (13, 14). Dehalogenation at ortho positions has been reported (15) but is not a common process in the environment. Generally, higher chlorinated biphenyls are preferentially dechlorinated over lower chlorinated congeners, resulting in the accumulation of mono-, di-, and trichlorobiphenyls (16). Dechlorination of meta and para chlorines can result in relative detoxification through the elimination of coplanarlike congeners and aryl receptor-mediated toxicity (17-19) and through the transformation of generally more toxic higher chlorinated congeners to generally less toxic lower chlorinated congeners (20). Dechlorination rates have been reported in the literature but are derived solely from laboratory microcosms; to our knowledge, dechlorination rates measured directly in the field have not been reported. Abramowicz et al. (21) reported a laboratory-measured dechlorination rate of 0.0133 chlorine/ biphenyl/week for unspiked, aged Hudson River sediments (20 ppm t-PCBs) and a rate of 0.0348 chlorines/week for 10.1021/es0486216 CCC: $30.25

 2005 American Chemical Society Published on Web 04/14/2005

freshly spiked Hudson River sediments (80 ppm t-PCBs using a 7:2:1 mix of Aroclors 1242/1254/1260). Laboratorymeasured rates using freshly St. Lawrence sediments spiked with 200 ppm Aroclor 1248 ranged from 9 nmol of Cl/g of sediment/d (∼0.013 chlorines/biphenyl/d) (22) to 19 nmol of Cl/g of sediment/d (∼0.027 chlorine/biphenyl/d) (23). A threshold concentration of approximately 40 ppm was idenfitied, below which no dechlorination was apparent (22, 23, 24). Zwiernik et al. (25) demonstrated a linear correlation between estimated aqueous-phase PCB concentrations and PCB dechlorination rates for laboratory flask studies using original data and data published by Abramowicz et al. (21) and Rhee et al. (26). The reported relationship is shown in the following equation (25):

MDR ) 1.16Ce + 6.37

(r2 ) 0.960)

(1)

where MDR (nmol of Cl-/g of sediment/week) is the maximum dechlorination rate and Ce (µg/L) is the aqueousphase PCB concentration; aqueous-phase PCB concentrations were estimated based on equilibrium-partitioning coefficients, fraction of organic carbon, and total PCB (tPCB) concentrations. Though Zwiernik et al. reported that this predictive relationship “may be valid over an extended range of conditions” (25), the t-PCB concentrations in the experiments used into develop eq 1 ranged from 20 to 800 mg/kg, which are higher than concentrations commonly found in the environment and much higher than most target cleanup concentrations. Thus, it is uncertain whether this relationship extends to the much lower PCB concentrations often required to meet risk-based cleanup standards (27). At Lake Hartwell, for example, t-PCB concentrations ranged from trace concentrations to ∼60 mg/kg; the cleanup requirement in the record of decision (ROD) is 1 mg/kg in the upper 10 cm, with long-term goals of 0.4 and 0.05 mg/kg (7). This paper focuses on PCB dechlorination rates in Lake Hartwell/Twelvemile Creek sediments. We report on the magnitude, extent, and meta plus para dechlorination rates with sediment depth and time. Sediment age dating, using lead-210 (210Pb) and cesium 137 (137Cs) (1), and linear relationships between meta plus para chlorine concentrations and sediment age made it possible to determine dechlorination rates. These rates are compared with literaturereported rates (21-24) and to rates calculated using eq 1 by Zwiernik et al. (25). The reader is referred to Brenner et al. (1) for a detailed site description, sediment core locations, vertical t-PCB concentration profiles, sediment age dating results, and estimated surface sediment recovery rates.

Materials and Methods Sediment collection methods were described previously (1, 8). Briefly, 18 sediment cores were collected during two annual events conducted in 2000 and 2001 (Figure 1). During the 2000 event, sediment cores were collected from transects O, N, L, J, I, T6, T16, W7, Q, and P; of these 10 cores, the data produced from analysis of the cores from transects Q, O, N, L, I, and T6 are reported in this paper. Transect J core is not shown because it had relatively poor recovery (20 cm) (8) and showed little dechlorination. The data for the cores from transects T16, W7, and P are not shown because they closely resembled the results shown for core Q. During the 2001 sampling event, eight additional sediment cores were collected: three from transect O, three from transect L, and two from Transect I. Sediment cores were collected and extruded as described previously (1, 8). Sediments from each interval were analyzed for 107 PCB congeners, particle size distribution (PSD), total organic carbon (TOC), moisture content, and radioisotopes

FIGURE 1. Lake Hartwell and Town Creek. Lake Hartwell sediment cores were collected at 10 transect locations. Surface sediment samples were collected in Twelvemile Creek and in Town Creek above and below the former Sangamo-Weston plant. (210Pb and 137Cs) to age date the sediments (1). On-site decontamination and chain-of-custody followed U.S. EPA approved Quality Assurance Project Plans (QAPPs) (28, 29). In this paper, core locations for the 2000 sampling event are identified by transect (T) as follows: T-Q, T-O, T-N, T-L, T-I, and T-T6. For the 2001 sampling event, transects perpendicular to the direction of flow were collected at T-O, T-L, and T-I are labeled T-OA, T-OB, and T-OC; T-LA, T-LB, and T-LC; and T-IA and T-IB, respectively. Sediment segments are identified by depth beginning at the sediment/water interface. Sediment PCB Analyses. Sediment PCB analyses were conducted at the Battelle Ocean Science Laboratories (Duxbury, MA) using modified U.S. EPA SW-846 Method 8270. Methods are described previously (1, 8). Sediment extracts were analyzed for the concentration of 107 PCB congeners using modified Method 8270 (see Table SI-1, Supporting Information for the list of congeners), and total PCB (t-PCB) concentrations were determined as the sum of the 107 PCB congeners. Congener method detection limits ranged from 1 to 2 ng/g. Surrogate internal standard (PCB 14, PCB 34, PCB 104, and PCB 112) recoveries averaged 80 ( 49% for 2000 and 65 ( 16% for 2001; PCB concentrations were not surrogate corrected. (See Tables SI-2 and SI-3, Supporting Information, for t-PCB concentrations with sediment depth.) Sediment Age Dating. Sediment dating was conducted using 210Pb and 137Cs isotopes, which are relatively common in sediments and can be used to determine sediment age over years or decades (1, 30-33). 210Pb and 137Cs isotopes were analyzed in accordance with methods described previously (1, 33). The age dates used in this paper to determine PCB dechlorination rates were reported by Brenner et al. (1). TOC, PSD, and Moisture Content Analyses. TOC and PSD measurements were performed by Soil Technology, Inc. (Bainbridge, WA), in accordance with U.S. EPA Method 9060s Total Organic Carbon and the American Society for Testing and Materials (ASTM) D422 (34), as described by Brenner et al. (1). Moisture content was determined gravimetrically using ASTM Method D2216 (35). VOL. 39, NO. 10, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3549

FIGURE 2. Numbers of ortho (9) and meta plus para (4) chlorines per biphenyl molecule (mol of chlorine/mol of PCB 1) with sediment depth for sediment cores collected in 2000 (see ref 2 for t-PCB concentration profiles). Figures include cores from transect Q (a), transect O (b), transect N (c), transect L (d), transect I (e), and transect T6 (f). Shaded areas were used to calculate dechlorination rates with time (see Figure 4).

Results and Discussion In the Lake Hartwell sediments, reductive dechlorination was characterized by the preferential loss of meta and para chlorines, the conservation of ortho chlorines with sediment depth and age, and from the historical transformation of higher-chlorinated PCB congeners (congeners with four or more chlorines) to mono-, di-, and trichlorobiphenyl congeners with sediment depth and time (8). Sediment age dating results (2) are used to determine dechlorination rates, which are compared with literature-reported rates (21-25). Dechlorination of meta and para Chlorines and Conservation of ortho Chlorines. The magnitude and extent of PCB dechlorination was measured by plotting ortho chlorines and meta plus para chlorines per biphenyl molecule (i.e., mol of chlorine/mol of PCB) with sediment depth for the 2000 and 2001 sediment cores (Figures 2 and 3, respectively). Coeluting congeners made it impossible to calculate the numbers of meta and para chlorines per biphenyl molecule separately. For example, PCB 56 (2,3,3′,4′-TeCB) and PCB 60 (2,3,4,4′-TeCB) coeluted; PCB 56 has two meta and one para chlorine, and PCB 60 had one meta and two para chlorines. The loss of meta and para chlorines and the conservation of ortho chlorines with sediment depth shows that the PCBs in the sediments underwent reductive dechlorination after burial. Notably, ortho chlorines were remarkably wellconserved for more than 5 decades, since the first appearance of PCBs (ca. 1950-1955). The preferential dechlorination of meta plus para chlorines and the conservation of ortho 3550

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 10, 2005

chlorines is consistent with reports by others (13, 14) and resulted in the accumulation of lower chlorinated congeners dominated by ortho chlorine substituents. These results also are consistent with microcosm study results reported by others using Lake Hartwell sediments (36, 37) where only meta and para dechlorination reactions were observed during 250-d incubation periods for congeners PCB 132 (234-236hexachlorobiphenyl [HCB]) and PCB 149 (245-236-HCB) and for Aroclor 1254. Homologue shifts were quantified in both the 2000 and 2001 sediment cores (see Supporting Information, Tables SI-4 and SI-5, for homologue congener shifts). On a molar basis, the average surface sediment PCBs for all 18 cores consisted of ∼30% mono-, di-, or trichlorobiphenyl congeners; 65-69% tetra- through hexachlorobiphenyls; and less than 5% hepta- through decachlorobiphenyl homologues. Dechlorination resulted in a substantial shift from tetrathrough decachlorobiphenyl congeners to mono- through trichlorobiphenyl congeners. In buried sediments and after maximum dechlorination, the tetra- through hexachlorobiphenyl homologues averaged ∼20% and the mono- through trichlorobiphenyls comprised ∼80% of the PCB concentrations, representing a ∼50% shift a molar basis from tetrathrough decachlorobiphenyl to mono- through trichlorobiphenyls. Dechlorination Extent. Though dechlorination tended to be very extensive in most of the cores, it was not always consistent from core to core or at various depth intervals

FIGURE 3. Numbers of ortho (9) and meta plus para (4) chlorines per biphenyl molecule (mol of chlorine/mol of PCB 1) with sediment depth for sediment cores collected in 2001 (see ref 2 for t-PCB concentration profiles). Figures include cores T-OA (a), T-OB (b), T-OC (c), T-LA (d), T-LB (e), T-LC (f), T-IA (g), and T-IB (h). Shaded areas were used to calculate dechlorination rates with time (Figure 4). within a single core. This inconsistency was most evident in the 2001 cores. For example, more extensive dechlorination occurred in the transect L cores than in the transect O cores. Little dechlorination was apparent in the upper 40 cm of the 2001 transect O cores (Figures 3a-c), though it was apparent in the upper 40 cm of the transect L cores (Figures 3d-f). In the transect L cores, dechlorination reached its maximum extent at approximately 30, 50, and 20 cm depths, but at various other depth intervals the dechlorination extent decreased, such as at the 50-55 cm depth interval in core T-LA and at the 50-60 cm and 90-100 cm depth intervals of core T-LC. The transect I cores (T-IA and T-IB; Figures 3g,h) exhibited similar behavior to the transect L cores; the 35-40 cm depth intervals of T-IA and T-IB demonstrated virtually no change in the numbers of meta and para chlorines per biphenyl when compared to the surface 0-5 cm depths,

suggesting that negligible dechlorination occurred at this depth interval. Reports by others suggest that dechlorination rate and extent may be related to t-PCB concentrations (21-26). At first glance, the decreased dechlorination between ∼90 and 100 cm in cores T-OA and T-LB might be explained by the low PCB concentrations at these depths; t-PCB concentrations in the bottom two segments of these cores were 0.9 and 0.5 mg/kg in core T-OA and 4.0 and 3.2 mg/kg in core T-LB. However, dechlorination was seen in core segments with comparable and even lower t-PCB concentrations, including the 60-70-cm interval in core T-OB (0.1-0.7 mg/kg) and the 35-50-cm interval in core T-LA (0.38-9.9 mg/kg). Furthermore, several segments where dechlorination was slowed or interrupted exhibited relatively high t-PCB concentrations. Poor regressions (r2 < 0.35) were observed when t-PCB VOL. 39, NO. 10, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3551

FIGURE 4. Numbers of ortho (9) and meta plus para (4) chlorines per biphenyl molecule (mol of chlorine/mol of PCB 1) with sediment age for the transect L sediment core collected in 2000 (see ref 2 for sediment age dating results). Linear regressions were used to calculate meta plus para dechlorination rates (mol of Cl/mol of PCB/yr).

TABLE 1. Measured Dechlorination Rates Using the Number of meta plus para Chlorines (per mol) over Measured Time Intervals for Buried Sedimentsa depth range (cm)

core

date range (year)

no. of samples

dechlorination rate (no. of Cl- (mol correlation of PCB)-1 yr-1) coefficient

2000 Transect Cores T-O 0-40 T-N 0-40 T-L 0-35 T-I 0-60 T-T6 0-20 average ( SD

1981-1999 1958-1997 1979-1999 1980-1999 1960-1998

T-OC 20-100 T-LA 15-35 T-LB 0-60 T-IA 0-20 T-IB 0-15 average ( SD average ( SDb

1952-2000 1962-1984 1987-2001 1993-2000 1996-2000

7 8 7 11 4

0.088 0.046 0.082 0.085 0.039 0.068 ( 0.024

0.959 0.923 0.971 0.950 0.985

2001 Transect Cores 9 4 9 4 3

0.023 0.067 0.111 0.170 0.232 0.120 ( 0.083 0.094 ( 0.063

0.729 0.998 0.946 0.922 0.790

a Dates were determined using 210Pb and 137Cs age dating techniques and are reported in Brenner et al. (1). b All cores.

concentrations were compared to homologue distributions, on the basis that homologue distributions reflect the extent of dechlorination in each core segment. TOC concentrations exhibited no evident correlation with the extent of PCB dechlorination; correlations between TOC and homologue distributions had r2 values less than 0.10. These results suggest that such factors as PCB concentration or TOC alone could not explain the extent of PCB dechlorination in the Lake Hartwell sediments. Dechlorination Rates. Dechlorination rates were determined for the Lake Hartwell site by plotting the numbers of meta plus para chlorines per biphenyl molecule (mol of chlorine/mol of PCB) with sediment age (see ref 1 for sediment age dating results). Figure 4 shows a typical graph and a linear regression used to determine the rate of meta plus para chlorine removals over time. Results of the linear regression analyses and corresponding meta plus para dechlorination rates measured for 2000 and 2001 cores are shown in Table 1. Three things made it possible to calculate meta plus para dechlorination rates in this paper. The first is that our companion paper (8) showed that the source material was relatively constant in Lake Hartwell surface sediments. The second is the observation that ortho chlorines were well-conserved with sediment depth and time (the number of ortho chlorines would change with depth if the input composition had changed with time). And the third is the linear relationship observed between the number of meta plus para chlorines and time. 3552

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 10, 2005

The data ranges used to calculate a rate for each sediment core are identified by shading in Figures 2 and 3. The correlation coefficients in Table 1 suggest linear relationships between the rate of meta plus para dechlorination reactions and time within the time periods examined. The average dechlorination rate was 0.094 ( 0.063 mol of Cl/mol of PCB/ yr for both years. An inverse of the dechlorination rates indicates the time required per chlorine removal ranged from 4.3 to 43.5 yr and averaged 16.4 ( 11.6 yr. The rates measured using the 2001 cores were approximately twice those measured using the 2000 cores, which appears to be due to the very high rate measured in core T-IB, and the fact that rates were highest at transects O, L, and I. The reason for the high T-IB rate is unknown. However, the average 2001 rate without core T-IB was 0.093 ( 0.063, which was much closer to the 2000 rate for transects O, L, and I (0.085 ( 0.003 mol of Cl/mol of PCB/yr). These rates were an order of magnitude slower than laboratory-measured rates for aged Hudson River sediments (0.69 Cl/biphenyl/yr) and freshly spiked Hudson River sediments (1.8 Cl/biphenyl/ yr) (21), and they were 2 orders of magnitude lower than the rates measured using freshly spiked St. Lawrence River sediments (4.7-9.9 Cl/biphenyl/yr) (22, 23). The higher laboratory rates are likely due to laboratory artifacts such as microcosm bacterial enrichments, higher temperatures, and much higher PCB concentrations in the freshly spiked laboratory microcosms. To compare the dechlorination rates measured in this paper with those reported by Zwiernik et al. (25), eq 1 was used to calculate rates for Lake Hartwell. The rates were calculated using two assumed sediment PCB concentrations, namely, 1 mg of PCB/kg of sediment and 60 mg of PCB/kg of sediment; these concentrations approximately span the range of concentrations measured at Lake Hartwell. Because porewater aqueous-phase concentrations were not measured at Lake Hartwell, they were estimated using eq 2:

Ce )

Cs Cs ) Kd Koc foc

(2)

where Ce is the aqueous-phase PCB concentration (µg of PCB/L), Cs is the solid-phase concentration (1000 or 60 000 µg of PCB/kg of sediment), Kd (L/kg) is the sediment/ porewater equilibrium coefficient, Koc (L/kg) is the organic carbon equilibrium coefficient, and foc is the fraction of organic carbon (3%). We assumed log Koc ) 5.0, consistent with Zwiernik’s calculations. The corresponding Kd value was 3000 L/kg. For the 1 and 60 mg/kg of sediment PCB concentrations, corresponding aqueous-phase porewater concentrations (Ce) were 0.33 and 20 µg/L, respectively. Corresponding MDR values were 6.57 and 29.7 nmol of Cl/g of sediment/week. (At 1 mg/kg, the MDR value asymptotically approaches the 6.57 y-intercept value.) The average dechlorination rate of 0.094 mol of Cl/mol of PCB/yr for this study corresponds to 6.9 × 10-6 mol of Cl/g of PCB/week, assuming 261 g/mol for Aroclor 1242 (the MDR equation was calculated using experiments with Aroclor 1242 experiments) (25). Multiplying by 1 mg/kg (1000 ng of PCB/g of sediment) and 60 mg/kg (60 000 ng of PCB/g of sediment), the average dechlorination rates at 1 and 60 mg/kg sediment concentrations equated to 0.007 and 0.41 nmol of Cl/g of sediment/week, respectively, which were 3 and 2 orders of magnitude lower than those estimated using eq 1, respectively. The higher rates determined using eq 1 were not surprising because they were based on laboratory reported rates (21, 26). The higher laboratory rates are likely due to higher incubation temperatures, greater bioavailability of freshly spiked PCBs, or the enrichment of PCB-dechlorinating microorganisms by the higher PCB concentrations used.

The empirical evidence of dechlorination in this study was not limited by laboratory artifacts such as PCB concentration, whether sediments were freshly spiked, test duration, sample representativeness, terminal electron acceptor conditions, and microcosm preparation. The rates reported in this paper reflect actual field conditions and suggest that in-situ rates may be much slower than those measured in the laboratory. Although this paper reports slower in situ dechlorination rates as compared to rates measured in the laboratory, it nonetheless demonstrates that dechlorination does occur even at relatively low in situ concentrations and shows that rates can be measured under such conditions. Notably, dechlorination was observed at concentrations much lower than the 40 mg/kg Aroclor 1248 concentration threshold reported by others (22-24). At PCB-contaminated sediment sites, dechlorination may be an important component of the natural recovery of the sediments. However, the positive impact of dechlorination (e.g., reduced mass and reduced toxicity) is tempered by the fact that dechlorination increases with sediment depth and age. Thus, near surface sediments, which typically pose the greatest risk of environmental exposure, are likely to exhibit the least dechlorination because they will have had the least amount of time in situ. Consequently, PCB dechlorination is unlikely to provide short-term risk reduction and should be evaluated for the long-term detoxification of PCBcontaminated buried sediments. More immediate recovery must be provided by sediment burial (38).

Acknowledgments We thank Craig Zeller (U.S. EPA, Region 4, Athens, GA), the Remedial Project Manager for the Lake Hartwell Superfund Site, for his assistance and access to the site. We also thank Battelle Ocean Sciences Laboratory (Duxbury, MA) staff (Greg Durell and Carole Peven-McCarthy) and Battelle Marine Sciences Laboratory (Sequim, WA) staff (Eric Crecelius and Linda Bingler) for their contribution to the PCB analytical chemistry and sediment age dating analyses, respectively, and their exceptional attention to quality assurance/quality control. This work was supported by U.S. EPA NRMRL under Contract 68-C5-0075, Work Assignment 4-30, and Contract 68-C-00-159, Task Order 09; Battelle was the prime contractor for this work. All work was conducted in accordance with a Level III, U.S. EPA/NRMRL-approved QAPP (24, 25). This paper does not necessarily reflect U.S. EPA views. No official endorsement should be inferred from this paper.

(6)

(7)

(8)

(9)

(10)

(11) (12) (13)

(14)

(15)

(16)

(17)

(18)

(19)

Supporting Information Available

(20)

Analytical congener list (Table SI-1), t-PCB concentrations for each sediment segment in 2000 (Table SI-2) and 2001 (Table SI-3), and homologue shifts observed in 2000 (Table SI-4) and 2001 (Table SI-5). This material is available free of charge via the Internet at http://pubs.acs.org.

(21)

Literature Cited (1) 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 PCB contaminated surface sediments at the Sangamo-Weston/ Twelvemile Creek/Lake Hartwell Superfund site. Environ. Sci. Technol. 2004, 38, 2328-2337. (2) U.S. EPA. The Incidence and Severity of Sediment Contamination in Surface Waters of the United States. National Sediment Quality Survey, 2nd ed.; Office of Science and Technology, Standards and Health Protection Division: Washington, DC, 2001. (3) van Metre, P. C.; Callender, E. Water-quality trends in White Rock Creek Basin from 1912-94 identified using sediment cores from White Rock Lake Reservoir, Dallas, Texas. J. Paleolimnol. 1997, 17, 239-249. (4) van Metre, P. C.; Mahler, B. J.; Furlong, E. T. Urban sprawl leaves its signature. Environ. Sci. Technol. 2000, 34, 4064-4070. (5) van Metre, P. C.; Wilson, J. T.; Callender, E.; Fuller, C. C. Similar rates of decrease of persistent, hydrophobic and particle-reactive

(22)

(23)

(24)

(25)

(26)

contaminants in riverine systems. Environ. Sci. Technol. 1998, 32, 3312-3317. Schneider, A. R.; Stapleton, H. M.; Cornwell, J.; Baker, J. E. Recent declines in PAH, PCB, and toxaphene levels in the northern Great Lakes as determined from high-resolution sediment cores. Environ. Sci. Technol. 2001, 35, 3809-3815. U.S. EPA. Superfund Record of Decision: Sangamo-Weston/ Twelvemile Creek/Lake Hartwell Site, Pickens, GA: Operable Unit 2; 1994; EPA/ROD/R04-94/178. Magar, V. S.; Johnson, G.; Brenner, R. C.; Quensen, J. F., III; Foote, E. A.; Durrell, G.; Ickes, J. A.; Peven-McCarthy, C. Longterm recovery of PCB-contaminated sediments at the Lake Hartwell Superfund site: PCB declorination. 1. End-member characterization. Environ. Sci. Technol. 2005, 39, 3538-3547. Johnson, G. W.; Ehrlich, R.; Full, W. Principal components analysis and receptor models in environmental forensics. In An Introduction to Environmental Forensics; Morisson, R., Murphy, B., Eds.; Academic Press: San Diego, 2002; pp 461-515. Bedard, D. L.; Quensen, J. F., III. Microbial reductive dechlorination of polychlorinated biphenyls. In Microbial Transformation and Degradation of Toxic Organic Chemicals; Young, L. Y., Cerniglia, C., Eds.; Wiley-Liss Division, John Wiley & Sons: New York, 1995; pp 127-216. Mohn, W. W.; Tiedje, J. M. Microbial reductive dechlorination. Microbiol. Rev. 1992, 6, 482-507. Magar, V. S. PCB treatment alternatives and research directions. J. Environ. Eng. 2003, 129 (11), 961-965. Tiedje, J. M.; Quensen, J. F., III; Chee-Sanford, J.; Schimel, J. P.; Boyd, S. A. Microbial reductive dechlorination of PCBs. Biodegradation 1993, 4, 231-240. Quensen, J. F., III; Tiedje, J. M.; Boyd, S. A. Dechlorination of four commercial polychlorinated biphenyl mixtures (Aroclors) by microorganisms from sediments. Appl. Environ. Microbiol. 1990, 56, 2360-2369. van Dort, H. M.; Bedard, D. L. Reductive ortho and meta dechlorination of a polychlorinated biphenyl congener by anaerobic microorganisms. Appl. Environ. Microbiol. 1991, 57, 1576-1578. Quensen, J. F., III; Tiedje, J. M. Evaluation of PCB dechlorination in sediments. In Methods in Biotechnology: Bioremediation Protocols; Sheehan, D., Ed.; Humana Press: Totowa, NJ, 1997; pp 257-273. McFarland and Clarke. Environmental occurrence, abundance, and potential toxicity of polychlorinated biphenyl congeners: considerations for a congener-specific analysis. Environ. Health Perspect. 1989, 81, 225-239. National Research Council (NRC). A Risk-Management Strategy for PCB-Contaminated Sediments; National Academy Press: Washington, DC, 2001. Quensen, J. F.; 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 microbial dechlorination. Environ. Toxicol. Chem. 1998, 17 (5), 806-813. Agency for Toxic Substances and Disease Registry (ATSDR). Toxicological Profile for Polychlorinated Biphenyls (PCBs); Agency for Toxic Substances and Disease Registry, Division of Toxicology, Toxicology Information Branch: Atlanta, 2000. Abramowicz, D. A.; Brennan, M. J.; Van Dort, H. M.; Gallagher, E. L. Factors influencing the rate of polychlorinated biphenyl dechlorination in Hudson River sediments. Environ. Sci. Technol. 1993, 27, 1125-1131. Sokol, R. C.; Bethoney, C. M.; Rhee, G.-Y. Effect of Aroclor 1248 concentration on the rate and extent of polychlorinated biphenyl dechlorination. Environ. Toxicol. Chem. 1998, 17 (10), 19221926. Rhee, G.-Y.; Cokol, R. C.; Bethoney, C. M.; Cho, Y.-C.; Frohnhoefer, R. C.; Erkkila, T. Kinetics of polychlorinated biphenyl dechlorination and growth of dechlorinating organisms. Environ. Toxicol. Chem. 1998, 20 (4), 721-726. 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 (24), 56515656. Zwiernik, M. J.; Quensen, J. F., III; Boyd, S. A. Residual petroleum in sediments reduces the bioavailability and rate of reductive dechlorination of Aroclor 1242. Environ. Sci. Technol. 1999, 33, 3574-3578. Rhee, G.-Y.; Bush, B.; Bethoney, C. M.; DeNucci, A.; Oh, H.-M.; Sokol, R. C. Anaerobic dechlorination of Aroclor 1242 as affected

VOL. 39, NO. 10, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3553

(27)

(28) (29)

(30)

(31) (32)

(33)

by some environmental conditions. Environ. Toxicol. Chem. 1993, 12, 1033-1039. Magar, V. S.; Ickes, J. A.; Foote, E. A.; Abbott, J. E.; Brenner, R. C.; Zeller, C. Natural recovery of the Sangamo-Weston/ Twelvemile Creek/Lake Hartwell Superfund site. In Proceedings of Dioxin 2003; Boston, 2003. Battelle. Final Addendum 01 to the Level III Quality Assurance Project Plan, QAPP No. 163-Q4-0; U.S. Environmental Protection Agency: Cincinnati, 2001. Battelle. Level III Quality Assurance Project Plan: Natural Attenuation of Persistent Organics in Contaminated Sediments at the Sangamo-Weston/Twelvemile Creek Lake Hartwell Superfund Site, QAPP No. 163-Q2; 2000. Lefkovitz, L. F.; Cullinan, V. I.; Crecelius, E. A. Historical Trends in the Accumulation of Chemicals in the Puget Sound; U.S. Department of Commerce and National Oceanic and Atmospheric Administration: Washington, DC, 1997. Bloom, N. S.; Crecelius, E. A. Distribution of silver, mercury, lead, copper, and cadmium in central Puget Sound sediments. Mar. Chem. 1987, 21, 377-390. Crecelius, E. A.; Bloom, N. Temporal trends of contamination in Puget Sound. In Processes in Marine Pollution Series, Vol. 5: Urban Wastes in Coastal Marine Environments; Wolfe, D. A., Ed.; Robert Krieger Publishers: Malabar, FL, 1989; Chapter 14, pp 149-155. Brenner, R. C.; Magar, V. S.; Ickes, J. A.; Abbott, J. E.; Stout, S. A.; Crecelius, E. A.; Bingler, L. S. Characterization and fate of

3554

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 10, 2005

(34)

(35)

(36)

(37)

(38)

PAH-contaminated sediments at the Wyckoff/Eagle Harbor Superfund site. Environ. Sci. Technol. 2002, 36, 2605-2613. American Society for Testing and Materials. Standard Method for Particle Size Analysis of Soils, Method D422-63; ASTM: West Conshohocken, PA, 1988. American Society for Testing and Materials. Standard Method for Total Organic Carbon Analysis, Method D2216; ASTM: West Conshohocken, PA, 1998. Pakdeesusuk, U.; Jones, W. J.; Lee, C. M.; Garrison, A. W.; O’Niell, W. L.; Freedman, D. L.; Coates, J. T.; Wong, C. S. Changes in enantiometric fractions during microbial reductive dechlorination of PCB132, PCB149, and Aroclor 1254 in Lake Hartwell sediment microcosms. Environ. Sci. Technol. 2003, 37, 11001107. Pakdeesusuk, U.; Lee, C. M.: Coates, J. T.; Freedman, D. L. Assessment of natural attenuation via in situ reductive dechlorination of polychlorinated biphenyls in sediments of the Twelve Mile Creek arm of Lake Hartwell, SC. Environ. Sci. Technol. 2005, 39 (4), 945-952. Magar, V. S. Editorial: Natural recovery of contaminated sediments. J. Environ. Eng. 2001, 127 (6), 473-474.

Received for review September 3, 2004. Revised manuscript received March 22, 2005. Accepted March 22, 2005. ES0486216