Reductive Dechlorination of Polychlorinated Biphenyls: Threshold

Reductive dechlorination of individual PCB congeners in Aroclor 1248 was investigated using sediment microorganisms from the St. Lawrence River (NY). ...
0 downloads 10 Views 121KB Size
Download PDF
Environ. Sci. Technol. 2003, 37, 5651-5656

Reductive Dechlorination of Polychlorinated Biphenyls: Threshold Concentration and Dechlorination Kinetics of Individual Congeners in Aroclor 1248 YOUNG-CHEOL CHO,† R O G E R C . S O K O L , ‡,§ ROBERT C. FROHNHOEFER,† AND G - Y U L L R H E E * ,†,§ Wadsworth Center, New York State Department of Health, Albany, New York 12201, Center for Environmental Health, New York State Department of Health, Troy, New York 12180, and School of Public Health, State University of New York at Albany, Albany, New York 12201

Reductive dechlorination of individual PCB congeners in Aroclor 1248 was investigated using sediment microorganisms from the St. Lawrence River (NY). No dechlorination was observed at Aroclor concentrations below 40 ppm [137 nmol (g of sediment)-1]. Above this threshold, congeners could be divided into three categories: group A, congeners that dechlorinated above 40 ppm; group B, congeners that dechlorinated only at high concentrations above 60 ppm [206 nmol (g of sediment)-1]; and group C, lower chlorinated congeners that increased in concentration. The dechlorination rate of congeners in groups A and B was a linear function of their initial sediment concentration. For group A congeners, the concentration intercepts of this linear function were the same as their concentrations in the Aroclor at the threshold concentration, and these therefore represented the threshold values. However, the intercepts of group B congeners were significantly higher than their levels at the threshold Aroclor concentration and were equivalent to their concentrations in Aroclor 1248 at about 75 ppm [258 nmol (g of sediment)-1]. The final concentrations of group A and group B congeners at the end of dechlorination were the same, regardless of their initial concentrations. These final concentrations were significantly lower than their threshold values. The accumulation rate of group C congeners was a linear function of their initial concentrations, and the total accumulation was greater at higher Aroclor concentrations in sediments.

Introduction Earlier studies of dechlorination kinetics of polychlorinated biphenyls (PCBs) have shown that dechlorination is concentration-dependent (1-4). These studies have also indicated that there is a threshold concentration below which no dechlorination takes place. Most of these investigations * Corresponding author telephone: (518)473-8035; fax: (518)4862697; e-mail: [email protected]. † New York State Department of Health, Albany. ‡ New York State Department of Health, Troy. § State University of New York at Albany. 10.1021/es034600k CCC: $25.00 Published on Web 11/11/2003

 2003 American Chemical Society

were based on the concentration of Aroclors. However, there is evidence that the dechlorination rate and the threshold concentration may vary among individual congeners (5, 6). In the dechlorination of Aroclor 1248 by sediment microorganisms from the St. Lawrence River, which was contaminated by this Aroclor, the lowest concentration at which dechlorination occurred was 155 nmol (g of sediment)-1 [45 µg of Aroclor 1248 (g of sediment)-1 (ppm)], and no dechlorination was found below 120 nmol (g of sediment)-1 (35 ppm), the next lower concentration examined, indicating that the threshold was between these two concentrations (3). The pattern and the extent of dechlorination varied widely among individual congeners, suggesting congener specificity of dechlorination, including the threshold concentration. Studies of a limited number of PCB congeners also indicated that the threshold concentration was congener-specific (6). The threshold concentration of Aroclor 1248 appears to be related to the growth of dechlorinating microorganisms. When dechlorinating microorganisms were investigated using the most probable number (MPN) method, a statistical method to estimate bacterial numbers (7), no growth was found below 137 nmol (g of sediment)-1 (40 ppm) (4). Evidence indicates that dechlorinator populations are diverse, judging from the existence of distinct dechlorination pathways and products (3, 8, 9). Our studies of PCB-contaminated sediments using a combination of the dilution fractionation and the MPN methods clearly demonstrated multiple types of dechlorinating microorganisms with distinct dechlorinating capabilities (10). It also appears that a given dechlorinator may be effective for more than one congener, probably because dechlorination is not just specific to substitution positions but depends on the pattern of Cl substitution on the biphenyl ring (11-13). Such characteristics of dechlorinating microorganisms or consortia may well contribute to differences in the dechlorination kinetics of individual congeners. In the present study, we have investigated the dechlorination kinetics of PCB congeners in Aroclor 1248, including the threshold concentrations, by incubating sediment microorganisms from the Aroclor 1248-contaminated St. Lawrence River in clean sediments spiked with the same Aroclor.

Materials and Methods Sediment Slurry Preparation. All experiments were carried out using stringent anaerobic techniques. Sediment collection, experimental setups, and PCB analysis were described earlier (3, 4). Briefly, PCB-free sediments were collected from the Grasse River, an uncontaminated tributary of the St. Lawrence River, NY. PCB concentrations in these sediments were below the detection limit. These sediments were spiked with Aroclor 1248 at 14 concentrations, ranging from 3.44 to 687 nmol (g of sediment)-1 on a dry weight basis (1-200 ppm). PCB-spiked sediments (20 g dry weight) were made into slurries by the addition of 90 mL of reduced St. Lawrence River water in 100-mL serum vials. These vials were capped with Teflon-coated butyl rubber stoppers and aluminum crimp seals. The headspace was filled with N2:CO2 (80:20). The vials were then inoculated with microorganisms eluted from St. Lawrence River sediments and incubated statically at the room temperature. Uninoculated vials of sediments at concentrations of 155, 206, and 412 nmol (g of sediment)-1 (45, 60, and 120 ppm, respectively) served as the control. The time course of dechlorination was monitored every 2 weeks by removal of a 3-mL sample of the slurry while it was continuously mixed. The incubation vials were set up in duplicate. VOL. 37, NO. 24, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5651

PCB Analysis. Sediment extraction for PCBs and congenerspecific analysis of PCBs were carried out as described previously (3, 14). Briefly, sediments were Soxhlet-extracted for 13-15 h with a 1:1 hexane:acetone (v/v) solvent mixture. The solvent extract was phase-separated by the addition of distilled water (50 mL), and the hexane layer was placed into a flask containing sodium sulfate. The hexane extracts were treated with a tetrabutylammonium sulfite reagent and were cleaned up on a 4% deactivated Florisil column to remove elemental sulfur and other chlorinated contaminants. Congener-specific PCB analysis was performed on a Hewlett-Packard 5890 gas chromatograph (Hewlett-Packard, Avondale, PA) equipped with a 63Ni electron-capture detector, autosampler, splitless injector, and a computerized data acquisition system (Chrom Perfect, Justice Innovations, Mountain View, CA). Aroclor 1248 was separated on a 60-m Rtx 5 capillary column (Restek, Bellefonte, PA). The gas chromatographic conditions were identical to those described earlier (4). The PCB congeners in the extract were identified and quantitated using a calibration standard containing a 1:1:1:1 mixture of Aroclors 1016, 1221, 1254, and 1260 (0.2 µg mL-1 of each in hexane). The calibration standards were run every sixth sample for recalibration as a part of quality assurance/quality control. A dilute-to-match procedure was used to ensure that all samples were analyzed within the linear range of the calibration standard (15). Method blanks were included with each set of samples extracted. The PCB congener numbering system in the text uses a slash to represent the separation of rings to permit an easier visualization of the chlorination substitution pattern (e.g., 2,3′,4′,5-chlorobiphenyl [CBP] will appear as 25/34-CBP).

Results Dechlorination of Aroclor 1248. The time course of dechlorination showed an initial lag period, followed by dechlorination, and then a plateau with no further concentration decrease through the remainder of a 58-week incubation period (see ref 3). No ortho dechlorination was detected. Dechlorination was not found in any of the sediment microcosms that had an Aroclor 1248 concentration at or below 120 nmol (g of sediment)-1 (35 ppm). Evidence of PCB dechlorination was detected at the next higher concentration of 155 nmol (g of sediment)-1 (45 ppm) (3, 4). Therefore, 137 nmol (g of sediment)-1 (40 ppm) was considered as the threshold concentration. Above the threshold, the dechlorination rate [nmol (g of sediment)-1 d-1] increased linearly with Aroclor concentration up to 687 nmol (g of sediment)-1 (200 ppm), the highest concentration investigated (3). A subsequent study (4) found that this linearity extended to 3090 nmol (g of sediment)-1 (900 ppm). Dechlorination of Individual Congeners. When individual congeners were examined, they could be grouped into three categories according to the trend of concentration changes over time: group A, congeners that dechlorinated above an Aroclor concentration of 137 nmol (g of sediment)-1 (40 ppm) and showed a continuous decrease until the plateau phase; group B, congeners that only dechlorinated at higher Aroclor 1248 concentrations above 206 nmol (g of sediment)-1 (60 ppm) and showed an initial increase in concentration before dechlorination; and group C, congeners that showed a continuous increase over time as a result of the dechlorination of higher chlorinated ones (Table 1, Figure 1). Here, an increase or a decrease reflects the net concentration since it was not possible to separate dechlorination and accumulation if they occurred at the same time. The rate of dechlorination or accumulation of each congener was calculated as the slope of decrease or increase in concentration between lag and plateau phases. This rate was higher at higher initial congener concentrations (Figure 2). 5652

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 24, 2003

Threshold Concentration. When the dechlorination rates of individual congeners in groups A and B were plotted against their initial concentrations, the relationship was linear within the concentration range investigated in this study (Figure 2a,b) and could be expressed as

-

dC ) kd(C - Ct) dt

(1)

where C is the congener concentration [nmol (g of sediment)-1]; kd is the slope; and Ct is the intercept on the concentration axis. Since -dC/dt ) 0 at Ct, the intercept is theoretically the threshold concentration. When the Ct values of group A congeners were plotted against the concentrations of the same congener at the observed threshold concentration of Aroclor 1248 [137 nmol (g of sediment)-1 or 40 ppm], the relationship was linear (P < 0.0002, linear regression) with the value of the slope close to unity (0.87 ( 0.16). These results show that the theoretical threshold values obtained as the concentration intercept (Ct) are indeed the true threshold concentrations (Figure 3a). The sum of these Ct values for congeners in this group was also the same as the total of the same congeners in the Aroclor at its threshold concentration (Table 1). The same type of plot for group B congeners at Aroclor concentrations at 275 nmol (g of sediment)-1 (80 ppm) and above also showed a linear relationship (P < 0.0001, linear regression) between the Ct values and the concentrations of the same congeners in the Aroclor at its threshold level [137 nmol (g of sediment)-1 or 40 ppm] (Figure 3b). However, the slope was significantly greater than unity (1.84 ( 0.06). These results indicate that the intercept values or the threshold concentrations are significantly higher than their concentrations in the Aroclor at its threshold concentration (Table 1, Figure 3b). The sum of the threshold values of group B congeners [71.9 ( 9.5 nmol (g of sediment)-1] was also significantly higher than that of their concentrations in Aroclor 1248 not only at 137 nmol (g of sediment)-1 (40 ppm), which is 38.4 ( 1.9 nmol (g of sediment)-1 but also at 206 nmol (g of sediment)-1 (60 ppm), which is 57.6 ( 2.9 nmol (g of sediment)-1 (Table 1). These results explain why this group was not dechlorinated at 206 nmol (g of sediment)-1 (60 ppm). A calculation showed that the threshold concentrations of group B congeners are equivalent to their concentrations in Aroclor 1248 concentrations at about 258 nmol (g of sediment)-1 (75 ppm). These results illustrate why dechlorination occurred at 275 nmol (g of sediment)-1 (80 ppm), but not at 206 nmol (g of sediment)-1 (60 ppm) (Figure 1b). There was an initial increase in concentration, notably in 515 and 687 nmol (g of sediment)-1 (150 and 200 ppm) sediments before dechlorination. This increase was probably because initial rates of dechlorination, if occurred, were slower than their accumulation rates. Final Concentration of Dechlorination and Threshold Concentration. The final concentrations of group A congeners during the plateau phase after dechlorination was the same regardless of their initial concentrations (Figures 1a and 4). This was also true for group B congeners in sediments with Aroclor concentrations at and above 275 nmol g sediment-1 (80 ppm) (Figures 1b and 4). The concentration of group B congeners did not change in sediments with an Aroclor concentration of 206 nmol (g of sediment)-1 (60 ppm). These results may indicate that dechlorinating microorganisms for these congeners were unable to grow at this low level (3, 4, 10). When the final concentrations of individual congeners at the plateau phase were compared to the threshold concentrations of these congeners in both groups A and B [at Aroclor concentrations above 206 nmol (g of sediment)-1 (60 ppm)], the final concentrations were significantly lower (Table 1).

TABLE 1. Kinetic Parameters of Congeners in Aroclor 1248 congenera

Group A 23/4 234/2 236/4 245/4 25/34 24/34 236/25 236/24+234/4+23/34 235/25+236/23+235/24 245/25 245/24 236/34+34/34 245/34 234/34

Group B 25/2 2/34 23/26 25/25 24/25 23/25 236/3+23/24+34/4 23/23

Group C 2/2+26 24+25 2/3 2/4+23 26/2 24/2+4/4 236+26/3 23/2+26/4 25/3 24/3 25/4+24/4 25/26+24/26 24/24

Sum group A group B group C total

concn in 40 ppm of Aroclor 1248b

X-interceptc

final concnd

ka or kde

1.54 ( 0.13 f 4.61 ( 0.51 4.93 ( 0.12 4.35 ( 0.57 9.75 ( 0.30 10.4 ( 0.23 6.58 ( 0.69 4.52 ( 0.19 3.27 ( 0.44 3.44 ( 0.17 1.89 ( 0.07 7.49 ( 0.30 2.46 ( 0.14 2.76 ( 0.51

1.61 ( 0.17g 7.70 ( 1.34 6.32 ( 1.00 4.49 ( 1.22 6.61 ( 1.53 9.82 ( 1.02 8.83 ( 0.89 3.24 ( 0.67 4.50 ( 0.64 4.15 ( 0.46 2.15 ( 0.61 10.0 ( 2.25 2.44 ( 0.56 3.06 ( 0.52

0.47 ( 0.18 f 2.62 ( 0.86 1.30 ( 0.55 0.69 ( 0.14 1.36 ( 0.25 3.63 ( 1.00 5.49 ( 0.95 1.70 ( 0.58 2.31 ( 0.75 1.50 ( 0.37 0.53 ( 0.18 2.17 ( 0.98 0.51 ( 0.23 0.08 ( 0.05

0.012 ( 0.001 (0.004) 0.018 ( 0.005 (0.064) 0.010 ( 0.001 (0.002) 0.009 ( 0.001 (0.004) 0.009 ( 0.001 (0.000) 0.006 ( 0.000 (0.000) 0.009 ( 0.001 (0.001) 0.007 ( 0.000 (0.000) 0.006 ( 0.001 (0.021) 0.008 ( 0.001 (0.001) 0.009 ( 0.001 (0.007) 0.009 ( 0.001 (0.007) 0.009 ( 0.001 (0.002) 0.008 ( 0.001 (0.001)

5.10 ( 0.29 5.03 ( 0.31 0.65 ( 0.16 8.54 ( 0.33 5.07 ( 0.36 6.95 ( 0.16 4.97 ( 0.17 2.09 ( 0.17

9.77 ( 0.76 9.45 ( 1.63 1.16 ( 0.09 16.0 ( 1.48 9.26 ( 1.63 13.1 ( 2.28 8.62 ( 0.84 4.50 ( 0.77

1.37 ( 0.33 6.52 ( 1.54 0.63 ( 0.50 6.08 ( 1.27 5.54 ( 2.61 2.67 ( 0.99 1.68 ( 0.33 2.42 ( 1.04

0.067 ( 0.006 (0.007) 0.018 ( 0.003 (0.030) 0.028 ( 0.002 (0.005) 0.037 ( 0.004 (0.009) 0.031 ( 0.005 (0.028) 0.030 ( 0.006 (0.031) 0.026 ( 0.002 (0.007) 0.027 ( 0.006 (0.049)

0.74 ( 0.15 0.17 ( 0.05 0.38 ( 0.11 1.63 ( 0.27 0.12 ( 0.16 2.11 ( 0.24 0.06 ( 0.07 2.38 ( 0.31 0.27 ( 0.08 0.01 ( 0.02 12.3 ( 0.74 0.14 ( 0.20 1.92 ( 0.06

1.17 ( 0.09 0.32 ( 0.04 0.36 ( 0.01 2.15 ( 0.54 0.16 ( 0.02 3.24 ( 0.14 0.09 ( 0.01 3.15 ( 0.60 0.24 ( 0.05 0.01 ( 0.00 10.9 ( 2.19 0.24 ( 0.03 2.68 ( 0.49

69.2 ( 60.8 13.8 ( 12.1 20.6 ( 11.3 34.2 ( 24.9 13.8 ( 10.6 43.3 ( 32.4 14.8 ( 5.91 19.4 ( 12.0 9.34 ( 1.96 10.2 ( 2.53 66.8 ( 17.0 4.76 ( 5.14 12.5 ( 8.57

0.575 ( 0.038 (0.001) 1.036 ( 0.143 (0.005) 0.173 ( 0.002 (0.000) 0.059 ( 0.010 (0.009) 0.674 ( 0.050 (0.001) 0.101 ( 0.004 (0.000) 2.126 ( 0.241 (0.003) 0.020 ( 0.002 (0.004) 0.235 ( 0.025 (0.011) 2.337 ( 0.210 (0.008) 0.008 ( 0.001 (0.010) 0.173 ( 0.020 (0.003) 0.024 ( 0.003 (0.004)

67.9 ( 4.4 38.4 ( 1.9 22.2 ( 2.4 129 ( 8.7

74.9 ( 12.9 71.9 ( 9.5 24.7 ( 4.2 172 ( 26.6

24.4 ( 7.1 26.9 ( 8.6 333 ( 205 384 ( 221

a Congeners with concentrations >1 mol % at any time point. b Congener concentration [nmol (g of sediment)-1] in Aroclor 1248 at its threshold concentration, 137 nmol (g of sediment)-1 (40 ppm). c X-intercepts [nmol (g of sediment)-1] are the calculated threshold concentrations of group A and group B congeners by eq 1 and the estimated concentrations of group C congeners in Aroclor 1248 at its threshold concentration by eq 2. d Final congener concentration [nmol (g of sediment)-1] at the end of dechlorination. e The slope (d-1) of eqs 1 and 2. Values in parentheses are P values. f Standard deviation. g Standard error of regression intercepts.

The sum of the group A plateau values was 24.4 ( 7.1 nmol (g of sediment)-1, whereas the sum of threshold concentrations for the same congeners was 74.9 (12.9 nmol (g of sediment)-1. The corresponding values for group B were 26.9 ( 8.6 and 71.9 ( 9.5 nmol (g of sediment)-1 (Table 1). Although these differences seem contradictory, they may be explained if the threshold values represent the minimum concentrations for the growth of dechlorinating microorganisms or consortia, whereas the final concentrations represent the level below which active metabolism ceases. Alternatively, it is possible that an accumulation of other “utilizable” congeners by the same microorganism may keep the total effective concentration above the threshold even when the final concentration of a particular congener is lower than its threshold. Group B congeners are characterized by a high amount of m-Cl relative to p-Cl, with a meta/para molar ratio of 3.8, whereas in group A, the relative concentration of meta is only slightly higher than that of para, with a molar ratio of 1.2. The meta/para ratio of group B decreased from the original 3.8 to 2.8 at the plateau, indicating that dechlorination involved largely m-Cl. On the other hand, the meta/para ra-

tio of group A increased from 1.2 to 1.6. This increase indicates that there was more para- than meta-dechlorination. First-Order Rate Constant for Dechlorination. Equation 1 indicates that the concentration-dependent change in the dechlorination rate is essentially first-order. When kd, the first-order rate constant or the specific dechlorination rate, was determined as the slope of the linear plot for group A congeners, it varied from 0.006 (235/25-+236/23-+235/24CBP) to 0.018 d-1 (234/2-CBP), with a corresponding halftime (t1/2 ) ln 2/kd) of 116-39 d (Table 1). An average value of kd for the group was 0.009 d-1, with a half-time of 75 d. The rate constants for congeners in group B were also congener-specific, and the value ranged from 0.018 (2/34CBP) to 0.067 d-1 (25/2-CBP), with a half-time from 38 to 10 d, respectively (Table 1). The average group B value is 0.033 d-1, with a half time of 21 d. The average rate constant for group B is higher than that for group A. However, the congeners in group B are dechlorinated only following the initial phase of net accumulation and also after a longer lag period (Figure 1b). VOL. 37, NO. 24, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5653

FIGURE 1. Time course of dechlorination of typical congeners in (a) group A (25/34-chlorobiphenyl) and (b) group B (25/2-chlorobiphenyl) in sediments with various initial concentrations of Aroclor 1248. The shaded area indicates the threshold concentration ((SD) of each congener. (c) Time course of accumulation of a typical end product in group C (2/4-+23-chlorobiphenyl) at various initial concentrations of Aroclor 1248. The shaded area is Ci ((SD) in eq 2. The value of Ci was identical to the congener’s concentration at the threshold concentration of Aroclor 1248 [137 nmol (g of sediment)-1 or 40 ppm]. Accumulation of Dechlorination Products. For group C congeners, the accumulation rate increased linearly with their initial concentrations (Figure 1c, Table 1), and similar to dechlorination rate, it can be expressed as

dC ) ka(C - Ci) dt

(2)

where ka is the slope and Ci is the concentration intercept. 5654

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 24, 2003

FIGURE 2. Initial congener concentration vs the dechlorination rate for three typical congeners in (a) group A and (b) group B. The concentration intercepts indicate threshold concentrations. (c) Initial congener concentration vs the accumulation rate of three typical group C congeners. The concentration intercepts were not significantly different from their concentrations at the threshold concentration of Aroclor 1248 [137 nmol (g of sediment)-1 or 40 ppm]. Since dC/dt ) 0 at Ci, theoretically Ci should be the same as the congener concentration at the threshold concentration of Aroclor 1248 [137 nmol (g of sediment)-1 or 40 ppm]. Indeed, when Ci values were determined for group A congeners, they turned out to be identical to the concentration of the same congeners at the Aroclor threshold concentration. A plot of these two parameters showed a linear relationship (P < 0.0001, linear regression), and the value of

FIGURE 4. Aroclor 1248 concentrations in sediments vs the total congener concentrations of groups A-C at the end of dechlorination.

FIGURE 3. Congener concentrations at the threshold concentration of Aroclor 1248 [137 nmol (g of sediment)-1 or 40 ppm] vs the calculated threshold concentrations of the same congeners in (a) group A and (b) group B. The dotted line represents a slope of 1. the slope was close to unity (0.89 ( 0.04). The sum of the intercept values was also the same as the total of the concentrations of the same congeners at the threshold level of Aroclor 1248 (Table 1). When the ka value was calculated as the slope of the linear function, they also varied by congeners, ranging from 0.008 (25/4-+24/4-CBP) to 2.337 d-1 (24/3-CBP), equivalent to half times of accumulation from 86 to 0.3 d. Unlike groups A and B, the final concentrations of the congeners in group C at the plateau phase of dechlorination were higher at higher initial concentration (Figure 4, Table 1). This is because they are the end products of dechlorination with no further dechlorination occurring.

Discussion It is remarkable that the threshold concentrations of congeners in group A, as calculated by the regression of the dechlorination rate against the initial congener concentrations, were identical to their concentrations in Aroclor 1248 at 137 nmol (g of sediment)-1 (40 ppm), the experimentally determined threshold level for the Aroclor. The fact that the Ci values of group C congeners calculated by eq 2 were also identical to their concentrations at the threshold concentration of the Aroclor confirms the validity of the regression

method to determine the threshold concentrations for dechlorination. Of 35 peaks in Table 1, 12 are coeluting ones containing more than one congener. Although we used all values in our analyses, including those of coeluting congeners, the conclusions remain the same even when coeluting congeners are excluded. In the absence of pure cultures of dechlorinating microorganisms, it is difficult to determine whether the different threshold levels for various congeners are due to the involvement of different dechlorinating microorganisms or to different substrate affinities of the same dehalogenating enzymes in the same microorganism. However, existing evidence suggests that different values are probably due to the involvement of different dechlorinators. For example, when methanogenesis was inhibited by an addition of bromoethanesulfonate, dechlorination of some of group B congeners did not take place (16). Investigations of dechlorinating microorganisms in St. Lawrence River sediments using a combination of the dilution fractionation and MPN methods clearly demonstrated two different kinds of dechlorinators. One kind was specific for the dechlorination of some group B congeners, although its population was about an order of magnitude smaller than the population of the other (10). The threshold values and the specific dechlorination rates may not be universally applicable to all contaminated sediments because they may also be a function of physical, chemical, and biological factors such as the sediment composition, age of contamination, and/or sediment microbial community. For example, even for a given Aroclor or a congener, the threshold level could change with the organic content and particle size distribution. It could also vary with the age of contamination even in the same sediments. Of course, the composition of sediment microbial community can also affect the values, as suggested by earlier investigations (3). Therefore, the numerical values of the threshold concentration for Aroclor 1248 and individual congeners in the present study should be considered to only apply to the sediments, microorganisms, and experimental conditions that we used. The congener-specific kinetic constants might change if different PCB congeners, even non-PCB haloaromatic compounds, are also present. Dehalogenating enzymes have broad substrate specificities (17-19). For PCBs, the pattern of Cl substitution on the biphenyl ring appears to be a factor VOL. 37, NO. 24, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5655

underlying the specificity (11, 14). This was also indicated by the findings that “priming” of contaminated sediments with certain congeners enhanced the dechlorination of different PCB congeners (20, 21). An addition of chlorophenols, chlorobenzenes, or chlorobenzoates to PCB-contaminated sediments enhanced various types of PCB dechlorination (22). Therefore, it is possible that the threshold concentration of a specific congener may change in the presence of cocontaminants. The final concentrations of individual congeners during the plateau phase were significantly lower than their threshold values, which seems contradictory. However, this may be explained if there exist two different kinds of threshold level. One is a threshold concentration for growth for dechlorinating microorganisms or consortia below which they cannot grow. In our previous study, this concentration for Aroclor 1248 was the same as the threshold for dechlorination, indicating that the dechlorination threshold was linked to the growth of dechlorinating microorganisms (4). There is also a significant correlation between the maximum level of dechlorination and the population size (23, 24). The other threshold is for active metabolism. Below this threshold, dechlorinating organisms die or remain dormant. Between the two threshold levels, dechlorinating microorganisms may still be able to utilize PCBs although their growth has ceased and the population may decrease. It was not possible to calculate the threshold concentrations for group C congeners because their concentrations increased during the dechlorination phase. This increase may be due to the absence of further dechlorination. It could also result if the dechlorination rate was slower as compared to the accumulation rate. If there was no dechlorination, this may stem either from the absence of dechlorinating microorganisms for these congeners or from the threshold concentrations for growth that are higher than the final concentrations. It appears that the existing PCB congener pattern in contaminated sediments may be a useful predictor of future dechlorination potential in those sediments. When the congener pattern of St. Lawrence River sediments was examined, the extent of dechlorination varied from site to site (8). Of all sites that we analyzed, GM showed the most extensive PCB dechlorination. At this site, all remaining congeners belonged to group C; no significant amounts of group A and group B congeners were present. These results suggest that dechlorination here was approaching its end. The Reynolds 002/003 site showed congeners belonging to both groups B and C, but no significant amount of group A congeners. Thus, this site may represent an intermediate stage of dechlorination. On the other hand, at Alcoa 002, significant amounts of group A and group B congeners were found in addition to group C congeners, which indicated potential for further dechlorination or a condition that is inhibitory to dechlorination. Since the final concentrations of group A and group B congeners were independent of their initial concentrations (Figure 4), the relative concentrations (mol %) would be lower for higher initial concentrations. On the other hand, the relative proportions of group C congeners would increase with higher initial concentration, since those of groups A

5656

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 24, 2003

and B decreased. These results mean that the relative concentrations of individual congeners at the end of dechlorination in contaminated sediments would vary depending on the initial concentrations of the contaminants.

Acknowledgments This work was supported by grants from the Hudson River Foundation (005/97A), the U.S. Environmental Protection Agency (R825449), and the National Institute of Environmental Health Science Superfund Basic Research Program (ES04913). We thank Charlotte M. Bethoney for her technical assistance.

Literature Cited (1) Abramowicz, D. A.; Brennan, M. J.; Van Dort, H. M.; Gallagher, E. L. Environ. Sci. Technol. 1993, 27, 1125-1131. (2) Fish, K. M. Appl. Environ. Microbiol. 1996, 62, 3014-3016. (3) Sokol, R. C.; Bethoney, C. M.; Rhee, G.-Y. Environ. Toxicol. Chem. 1998, 17, 1922-1926. (4) Rhee, G.-Y.; Sokol, R. C.; Bethoney, C. M.; Cho, Y.-C.; Frohnhoefer, R. C.; Erkkila, T. Environ. Toxicol. Chem. 2001, 20, 721-726. (5) Rhee, G.-Y.; Sokol, R. C.; Bush, B.; Bethoney, C. M. Environ. Sci. Technol. 1993, 27, 714-719. (6) Sokol, R. C.; Bethoney, C. M.; Rhee, G.-Y. Water Res. 1995, 29, 45-48. (7) Clesceri, L. S.; Greenberg, A. E.; Eaton, A. D. Standard Methods for the Examination of Water and Wastewater, 20th ed.; American Public Health Association: Washington, DC, 1998. (8) Sokol, R. C.; Kwon, O.-S.; Bethoney, C. M.; Rhee, G.-Y. Environ. Sci. Technol. 1994, 28, 2054-2064. (9) Ye, D.; Quensen, J. F., III; Tiedje, J. M.; Boyd, S. A. Appl. Environ. Microbiol. 1992, 58, 1110-1114. (10) Cho, Y.-C.; Kim, J.; Sokol, R. C.; Rhee, G.-Y. Can. J. Fish. Aquat. Sci. 2000, 57, 95-100. (11) Rhee, G.-Y.; Sokol, R. C.; Bethoney, C. M.; Bush, B. Environ. Sci. Technol. 1993, 27, 1190-1192. (12) Rhee, G.-Y.; Bush, B.; Bethoney, C. M.; DeNucci, A.; Oh, H.-M.; Sokol, R. C. Environ. Toxicol. Chem. 1993, 12, 1033-1039. (13) Wu, Q. Z.; Sowers, K. R.; May, H. D. Appl. Environ. Microbiol. 2000, 66, 49-53. (14) Rhee, G.-Y.; Bush, B.; Bethoney, C. M.; DeNucci, A.; Oh, H.-M.; Sokol, R. C. Environ. Toxicol. Chem. 1993, 12, 1025-1032. (15) Kimbrough, D. E.; Chin, R.; Wakakuwa, J. Analyst 1994, 119, 1293-1301. (16) Kim, J.; Rhee, G.-Y. Environ. Toxicol. Chem. 1999, 18, 26962702. (17) Mohn, W. W.; Kennedy, K. J. Appl. Environ. Microbiol. 1992, 58, 1367-1370. (18) Struijs, J.; Rogers, J. E. Appl. Environ. Microbiol. 1989, 55, 25272531. (19) Townsend, G. T.; Suflita, J. M. Appl. Environ. Microbiol. 1996, 62, 2850-2853. (20) Bedard, D. L.; Bunnell, S. C.; Smullen, L. A. Environ. Sci. Technol. 1996, 30, 687-694. (21) Bedard, D. L.; Van Dort, H. M.; DeWeerd, K. A. Appl. Environ. Microbiol. 1998, 64, 1786-1795. (22) Cho, Y.-C.; Ostrofsky, E. B.; Sokol, R. C.; Frohnhoefer, R. C.; Rhee, G.-Y. FEMS Microbiol. Ecol. 2002, 42, 51-58. (23) Cho, Y.-C.; Sokol, R. C.; Rhee, G.-Y. Environ. Toxicol. Chem. 2002, 21, 715-719. (24) Kim, J.; Rhee, G.-Y. Appl. Environ. Microbiol. 1997, 63, 17711776.

Received for review June 13, 2003. Revised manuscript received September 25, 2003. Accepted October 2, 2003. ES034600K