Research Temporal Bioavailability of Organochlorine Pesticides and PCBs D. SETHAJINTANIN† AND K . A . A N D E R S O N * ,†,‡ Department of Environmental and Molecular Toxicology and Food Safety and Environmental Stewardship Program, Oregon State University, Corvallis, Oregon 97331
Because PCBs and organochlorine pesticides continue to be of global concern, studies that address information gaps, such as factors and influences of spatial and temporal effects on contaminant bioavailability, are valuable. The present study focused on the spatial and temporal distribution of bioavailable organochlorine pesticides and PCBs in surface waters of a contaminated harbor. Passive sampling devices were intensively deployed adjacent to various land uses on the Willamette River, OR, including Portland Harbor and McCormick and Baxter Superfund sites, during summer and fall, extreme conditions, 2001-2004. An increase of bioavailable ΣDDTs (sum of p,p′-DDT, p,p′DDD, and p,p′-DDE) concentrations was strongly affected by the local historic production of DDTs and temporal changes in river conditions. The increase of bioavailable p,p′DDD and high DDD/DDE ratios observed during summer indicates conditions favoring anaerobic reductive processes. In contrast to ΣDDTs, the bioavailable concentrations and daily loads of dieldrin and PCBs increased during fall, especially during episodic rainstorms. On the basis of the PCB congener profiles, PCB inputs from urban runoff /sewer overflows were considered likely current sources of bioavailable PCB into the Harbor. The exceedence of the U.S. national and Oregon water quality criteria was a function of the temporal variability of each bioavailable contaminant. This illustrates the impacts associated with temporal changes of bioavailable organochlorine distributions in surface waters and the significance of considering realistic temporal, bioavailability, and site-specific conditions in risk assessment and water quality management.
Introduction The freely dissolved forms of hydrophobic contaminants are transported across biological membranes of aquatic organisms and may exert toxic effects (1, 2). A decrease in freely dissolved contaminant leads directly to reduced bioavailability and vice versa. However, there is no clear distinction between the processes that control the distribution of chemical contaminants in the environment and those that directly affect bioavailability (3). Spatial and temporal factors may modify contaminant bioavailability, thereby changing the uptake of contaminants from water by aquatic organisms. Spatial factors include sources of chemical contaminants, * Corresponding author phone: (541)737-8501; fax: (541)737-0497; e-mail:
[email protected]. † Department of Environmental and Molecular Toxicology. ‡ Food Safety and Environmental Stewardship Program. 10.1021/es052427h CCC: $33.50 Published on Web 05/12/2006
2006 American Chemical Society
location of sensitive biological resources, routes of exposure, and factors that modify contaminant mobility and availability (4). Temporal factors include seasonal changes in physical, chemical, or biological aspects of an ecosystem (4). Seasonal variation in physical or chemical parameters can modify the bioavailability of contaminants and subsequently change the nature of exposure (4). In addition to improving our understanding of chemical fate and transport in the environment, consideration of spatial and temporal distribution of contaminants can reduce uncertainty associated with chemical exposure and provide a quantitative expression of confidence in risk estimates. Despite the importance of environmental bioavailability, few studies have actually evaluated the distribution of chemical bioavailability and these influencing factors in the field. Several studies have indicated that bioaccumulation of polychlorinated biphenyls (PCBs) and DDTs (1,1-bis(4chlorophenyl)-2,2,2-trichloroethane and its metabolites) in fish and lower trophic level biota changes according to seasonal variations in PCB and DDT concentrations in surrounding water and sediment (5). A significant spatial and seasonal pattern in exposure modeling for human health risk estimates for recreational fish consumers has also been addressed (5). The influence of seasonal variation on contaminant concentration and load has been found to vary across a wide spectrum of watershed characteristics, including stormflow (6), snowmelt (7), and river discharge (8). The Willamette River in western Oregon has the thirteenth largest streamflow in the United States and yields more runoff per square mile than any other large river in the United States. The Willamette River, Portland Harbor, is heavily industrialized and has both private and municipal wastewater outfalls. Due to elevated concentrations of PCBs, organochlorine pesticides, dioxins/furans, PAHs, and heavy metals in the harbor sediment, Portland Harbor was placed in the federal National Priority List (NPL). Our previous study found that PCB and DDT residues in three recreational fish species from the Portland Harbor exceeded the US EPA fish advisory screening values and may pose health risks to recreational and subsistence fishers (9). As surface water runoffs within the harbor are controlled by climate influence and vary according to season (10), PCB and DDT concentrations in the river, particularly the bioavailable concentrations, may change temporally and alter the nature of exposure. This influence can affect both aquatic organisms and human health. These river characteristics make Portland Harbor an ideal study area for evaluating temporal bioavailability of PCBs and organochlorine pesticides in surface water. The goals of this study were to understand the temporal distribution of bioavailable PCBs, p,p′-DDT and its derivatives, and dieldrin in surface water and evaluate the environmental factors influencing their bioavailability. We focused our sampling scheme on the two seasons with the most extreme conditions: summer (with low precipitation, high temperatures, low water flow) and fall (with low temperatures, high precipitation, precipitation after extended dry periods, and high flow). Early results indicated this sampling scheme would afford us the best opportunity to see the largest temporal impact of bioavailable concentrations. A passive sampling technique using semipermeable membrane devices (SPMD) (11) was chosen to collect a timeintegrated measurement of the bioavailable fraction (12). It has been proposed that passive sampling devices (PSD) mimic key mechanisms of bioconcentration, including VOL. 40, NO. 12, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3689
FIGURE 1. Lower Willamette River at Portland, OR, showing locations of the sampling sites (RM ) river mile). diffusion through biomembranes and partitioning between organism lipid and the surrounding medium, in the absence of metabolism. In this work, the influences of temporal and episodic events, coupled with in situ water chemistry, on bioavailable transport were investigated. To screen the potential impacts associated with seasonal changes on aquatic organisms, bioavailable concentrations were compared to the U.S. national recommended water quality criteria (13) and the Oregon water quality criteria (14).
Materials and Methods Materials and Chemicals. Standard SPMDs were purchased from Environmental Sampling Technologies (St. Joseph, MO). Standards of organochlorine pesticides and PCBs were obtained from Chem Service, Inc. (West Chester, PA) and AccuStandard (New Haven, CT), respectively. Certified reference material cod liver oil (SRM 1588a, NIST, Gaithersburg, MD) was used for QC. Target PCBs of 25 individual congeners included dioxin- and non-dioxin-like congeners, listed in the Supporting Information. Study Area. The sampling areas were located in Portland Harbor on Willamette River RM 1-18 and contained both the Portland Harbor (RM 3.5 to 9.5) and the McCormick and Baxter Creosoting Co. Superfund sites (RM 7 East). The 11 sampling sites (Figure 1) were at RM 1 East (industrial area), RM 3.5 West (industrial and urban area), RM 3.5 East (industrial area), RM 7 West (Railroad Bridge; industrial area), RM 7 East (McCormick and Baxter Superfund site), RM 8 East (industrial area), RM 12 East (downtown/urban area), RM 13 West (downtown/urban area), RM 15 East (sand and gravel operation/undeveloped area), RM 17 East (golf course, park), and RM 18 West (near the mouth of Johnson Creek; urban/agricultural/residential area). Precipitation and streamflow data were collected from the National Weather Service and USGS, respectively. Calculation of loading estimates is described in the Supporting Information. Sample Collection. Deployment schemes of the SMPD have been previously described (15). Briefly, SPMDs were in deployed cages submerged 10 ft from the river bottom for 7-21 days. Water parameters (temperature, dissolved oxygen, specific conductivity, oxidation-reduction potential, pH, NH4+-N, and NO3--N) were measured using a YSI 6920 SONDE (Yellow Springs, OH). Water samples were also collected for total organic carbon (TOC), dissolved organic 3690
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 12, 2006
carbon (DOC), total suspended solids (TSS), total dissolved solids (TDS), and metals. Extraction, Analysis, and Quality Control. SPMD field extraction, laboratory cleanup, and instrumental analysis have been previously described (15, 16). Additional details are also provided in the Supporting Information. Uptake rates by SPMDs may be affected by changes in temperature, flow velocity, and buildup of periphyton and can be assessed using permeability/performance reference compounds (PRCs) (17). PCB 8, PCB 82, and endrin were used as PRCs (18). The average percent recoveries of fortified samples were as follows: PCBs 69 ( 43%, p,p′-DDT 71 ( 27%, p,p′-DDD 80 ( 21%, p,p′-DDE 66 ( 16%, and dieldrin 73 ( 18%. The average percent recoveries in certified reference material were as follows: PCBs 94 ( 26%, p,p′-DDT 98 ( 18%, p,p′-DDD 105 ( 17%, p,p′-DDE 76 ( 16%, and dieldrin 89 ( 6%. Field and laboratory quality control samples accounted for ca. 35% of all samples processed. None of the target analytes were found in field or laboratory control blanks. Data Analysis. The basic theory and models required for estimation of analyte concentrations in water from the SPMD have been described by Huckins et al. (19) and are provided in the Supporting Information. The average percent recoveries of PRCs indicated that a linear uptake model could be assumed for all target analytes. The small variations in PRC dissipation rates among sampling sites and seasons suggested the effects of membrane biofouling and flow velocityturbulence at the membrane surface were negligible. Differences in exposure temperature were corrected using established SPMD sampling rates (Rs) at multiple temperatures (20, 21). Data interpretation was performed using SPSS Version 10.0.1 (SPSS Inc., 1989-1999), Sigma Plot 2002 for Windows Version 8.0 (SPSS Inc., 1986-2001), and Microsoft Office Excel 2003 (Microsoft Corp., 1985-2003). Standard descriptive statistics, two sample t test, and linear regression techniques were used to examine temporal bioavailable organochlorine concentrations and influence variables. Statistical analyses were considered significant at p e 0.05.
Results and Discussions River Conditions and Water Chemistry. During the deployment intervals from 2001 to 2004, the river was categorized as low flow/low precipitation in summer (average river flow
< 10 000 ft3/s, total precipitation 0-1 in., average temperature g 20 °C) and high flow/high precipitation in fall (average river flow > 10 000 ft3/s, total precipitation > 1 in., average temperature < 20 °C). In addition, any intermittent precipitation that occurred was considered an episodic event. The river characteristics and water chemistry during passive sampling device deployments are summarized in the Supporting Information, Table 1. The difference in water temperature between summer and fall was approximately 10 °C. Dissolved oxygen concentrations ranged from 7.2 mg/L in summer to 12 mg/L in fall. The river pH was neutral. The oxidation-reduction potential tended to decrease in summer (p ) 0.054, t test). There was no temporal pattern observed for specific conductivity, NH4+-N, NO3--N, TOC, DOC, TSS, and TDS (Supporting Information, Table 2). The lack of temporal variation with organic carbon or total solids indicates that transport of contaminants in the Willamette River was primarily in the aqueous phase independent of season. Distributions of Bioavailable ΣDDT Concentrations and Potential Sources. Bioavailable ΣDDT concentrations (sum of p,p′-DDT, p,p′-DDD, and p,p′-DDE) from the 11 sampling sites over the 4-year study period (n ) 186) varied from 27 to 1500 pg/L with an average concentration of 190 pg/L. The highest concentrations were always observed at RM 7 West. The concentrations of ΣDDTs at RM 7 West were generally higher than for many other large rivers in the United States (22). They were comparable to or lower than the concentrations in rivers in other parts of the world (23, 24). However, few studies used passive sampling devices to study the dissolved concentrations, while other studies used conventional filtered water methodology. Unlike the passive sampling technique, the filtered waters would include DDTs associated with dissolved organic matter, which may lead to inflation or overestimation of the dissolved bioavailable concentrations. The two approaches are not directly comparable. Thus, the DDT levels measured at the Willamette River at Portland Harbor with PSD, particularly at RM 7 West, are noteworthy. DDT production and use in the United States has been banned for three decades. Surface water contamination remains if DDTs are continuously introduced through atmospheric deposition, stream runoff, and/or point and nonpoint source inputs of surrounding soils and sediment. A contribution of ΣDDTs from atmospheric deposition is unlikely to account for a particular site-specific contamination within the 18-mile stretch (Figure 2). Bioavailable concentrations of DDD, DDE, and DDT at RM 7 West were significantly higher than at any other sampling site in both summer and fall (p < 0.001; ANOVA F-test and Tukey’s HSD). The average bioavailable ΣDDT concentrations at the upstream sampling sites were approximately 10-fold less than those at RM 7 West (Figure 2). The 10-fold increases of bioavailable ΣDDTs indicated the upstream sources did not contribute significantly to the downstream sites. Our findings combined with other OC studies (25) suggest that the tributary influences on the bioavailable ΣDDT distributions in the lower Willamette River are negligible. Former DDT manufacturing and handling facilities were located approximately 1 mi. upstream of RM 7 West. Weston et al. (10) reported that sediment samples collected within 1 mi. downstream of the former DDT facilities contained ΣDDT concentrations approximately 4-10-fold higher than samples taken upstream or downstream. Historically, discharged ΣDDTs in sediment are likely recycling into the water column via the contaminated sediments or adjacent soils. Contaminants that are dissolved or associated with colloidal particles can exchange across the sediment-water interface by diffusive or advective processes. Subsequently, the sediments act as a depositional reservoir for historically dis-
charged contaminants and are an important source of contamination as they are recycled to the overlying water column. The general similarity of DDT profiles in the sediment (10) and the water bioavailable fraction measured from fall events supports the possibility of the movement of bioavailable DDT and its metabolites to the water column from historically contaminated sediments. However, the bioavailable water DDT homologue profile in summer is dramatically different than the sediment homologue profile. The sediment DDD/DDE ratio (typically e1) and the fall water bioavailable ratio (= 1) are similar. However, the summer DDD/DDE bioavailable ratio of >2.5 is significantly different (see Supporting Information, Table 3). This indicates that the summer bioavailable DDD, DDE, and DDT concentrations are not simply a function of the sediment diffusion and advection processes which dominate in fall (discussed below). Prevalence of Bioavailable p,p′-DDD in Surface Water. Figure 2 shows that p,p′-DDD (DDD) is the most abundant bioavailable DDT derivative followed by p,p′-DDE (DDE) and p,p′-DDT (DDT) at 53 ( 14%, 35 ( 11%, and 13 ( 6%, respectively. In general, the principal insecticidal ingredients of technical DDT contained p,p′-DDT 72%, o,p′-DDT 20%, p,p′-DDD 3%, o,o′-DDT 0.5%. and other 4.5%. In addition, p,p′-DDD was marketed as an insecticide in its own right. Therefore, direct input is a possibility. However, the high contribution of both p,p′-DDD and p,p′-DDE in the bioavailable fraction suggests that a large proportion of p,p′DDT has been transformed to p,p′-DDD and p,p′-DDE. The large proportion of bioavailable p,p′-DDD and the large ratio of bioavailable DDD/DDE in surface waters is unlikely to be governed by water solubility or partition coefficients (log Kow and log Koc). Water solubility of p,p′DDT, p,p′-DDD, and p,p′-DDE are 0.0055, 0.02, and 0.1 mg/L at 20-25 °C (26) and are well above those measured in this system. Values of log Kow (5.7 for p,p′-DDT, 6.1 for p,p′-DDD, 6.0 for p,p′-DDE) are similar. Also, the sorption coefficients (log Koc) (6.3 for p,p′-DDT, 5.0 for p,p′-DDD, 4.7 for p,p′-DDE at 20-25 °C) (26, 27) are similar and therefore unlikely to account for the observations. The large ratio of DDD/DDE likely indicates reductive dechlorination of DDT to DDD, which has been previously observed in sediments under flooded, anaerobic conditions (28, 29). Under anaerobic conditions, sediment DDT is mainly metabolized to DDD by reductive dechlorination either by microbial degradation or by chemical reaction. Under aerobic conditions, DDT is metabolized to DDE by dehydrochlorination (30). The Willamette River at Portland Harbor is a deep and slow moving river (10) and therefore has the potential for anaerobic reductive dechlorination of DDT to DDD rather than to DDE. Also, p,p′-DDD tends to resist further degradation as compared to p,p′-DDE (30) under anaerobic conditions: this typically leads to increasing DDD/ DDE ratios with aging contaminants. In addition to direct input of DDD, anaerobic reductive dechlorination of DDT to DDD and resistance to further degradation of DDD under anaerobic conditions in summer are most likely responsible for the enrichment of bioavailable DDD in Portland Harbor in summer. Temporal Changes in Bioavailable ΣDDT Concentrations. A 2-fold increase (p ) 0.002, t test) of bioavailable ΣDDTs in summer was measured at RM 7 West. This increase was due to higher levels of bioavailable DDD, Figure 2. In contrast, bioavailable DDT and DDE did not change or increased slightly in fall, although the difference was not statistically significant (p ) 0.58 for p,p′-DDT and p ) 0.50 for p,p′-DDE, t test.). This means that dilution was not a cause of the DDD decrease in fall. The bioavailable DDD concentration was positively correlated with water temperature (r2 ) 0.62, p ) 0.003) and negatively correlated with VOL. 40, NO. 12, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3691
FIGURE 2. Bioavailable contaminants in surface water in the Willamette River at Portland Harbor, OR, using passive sampling devices (n ) 186) deployments in summer (b) and fall (2) of 2001-2004. river flow (r2 ) 0.39, p ) 0.03), total precipitation (r2 ) 0.71, p < 0.001,) and dissolved oxygen concentration (r2 ) 0.49, p ) 0.016; see Supporting Information, Figure 1). The temporal bioavailable DDD concentration was not statistically correlated to TOC or DOC (see Supporting Information, Figure 2). Low river flow conditions during summer may allow adjacent soils and groundwater contaminants to migrate into the river. In fall the river is above the groundwater table and contaminant migration into the river is unlikely. If the groundwater and adjacent soils contain primarily DDD, rather than the other homolog, this could influence observed DDD levels. DDD and water temperature were correlated. An increase in water temperature may decrease the organic matter-water partition coefficient (Kom) of contaminants (31). However, one would expect that DDD, DDE, and DDT would be similarly affected; this was not observed. Only DDD increased in summer, while DDE and DDT remained the same or decreased in summer (Figure 2). Therefore, temperature 3692
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 12, 2006
effects on the partitioning constants do not account for the observations. Increased water temperature would be expected to increase growth and activity of benthic organisms, microorganisms, eutrophication, and bitoturbation, which would lead to increases in biodiffusion (32). In addition, some aquatic plants can take up and transform DDT to DDD (33). All these variables would increase production of DDD in summer. Anaerobic reductive conditions are necessary for reductive dechlorination of DDT to DDD by both microbial and chemical degradation reactions. Sediments quickly become anaerobic with increasing sediment depth (32), and slower moving surface waters during summer would enhance anaerobic reductive degradation. Kale et al. (28) found p,p′DDD was a major metabolite of p,p′-DDT in sediment and overlying water under flooded, anaerobic conditions. Temporal circumstances favoring reductive conditions in sediments, for example, those found in summer, would contribute to an increase of bioavailable DDD and a high ratio of DDD/
DDE in surface waters. The enhancement of anaerobic conditions in summer leads to increased bioavailable DDD concentrations in summer. Potential Risk of Temporal Bioavailable ΣDDTs on Aquatic Organisms. To assess the potential impacts of temporal distribution of bioavailable DDD, DDE, and DDT on aquatic organisms in the Willamette River, concentrations were compared to the U.S. national and Oregon recommended water quality criteria (13,14) (Figure 2). The PSDmeasured freely dissolved contaminant is likely a better estimate of the bioavailability and therefore the risk of transport across biological membranes and potential toxic effects. As shown in Figure 2, the bioavailable concentrations that exceed the national water quality criteria are strongly temporally influenced. For example, DDD often exceeds the criteria in summer and rarely in fall. At RM 7 West, near the DDT facility, ΣDDT often exceeds the national and Oregon criteria. Because aquatic organisms are ectothermic, their rate of metabolism undergoes an approximately 2-fold increase with every 10 °C rise in water temperature (34). Cairns et al. (34) suggested that increased temperature could change respiratory rate and membrane permeability and thereby lead to more rapid uptake and hasten the accumulation of a toxic dose. The effect of increasing temperature in summer may result in increased bioconcentration, further exacerbating the temporal effects of increased DDD in summer, which further illustrates the need to understand temporal variability of contaminants. These findings agreed with the bioaccumulation profiles of ΣDDTs in fish and human health risk estimates for ΣDDTs (2 × 10-6 to 8 × 10-5) from consuming fish in this area as discussed in our previous study (9). The agreement between the present study and that previous study supports the use of passive sampling devices as a surrogate for studying bioavailable distribution of organic contaminants in water. Temporal Changes in Bioavailable Dieldrin Distribution. Over the 4-year study period, the ranges of bioavailable dieldrin concentrations in the Willamette River were between 10 and 140 pg/L (see Figure 2). No discernible spatial patterns were observed. An increase in bioavailable dieldrin concentrations was found at most sampling sites during fall (p e 0.046, t test.) The largest increases were observed at RM 18, which is near Johnson Creek, an agricultural/urban creek. The temporal pattern of bioavailable dieldrin distribution may illustrate the transport behavior for widely dispersed, nonpoint source contaminants. Like DDT, dieldrin was widely used in the Willamette basin before it was banned and Johnson Creek has been previously reported to be heavily contaminated with dieldrin (35). Elevated concentrations of bioavailable dieldrin during high flow and storm events in fall suggest transport of bioavailable dieldrin from contaminated sites upstream, including erosion inputs to the river during precipitation runoff from land sources and riverbed sediment. In contrast to DDD, the U.S. national and Oregon human health water quality criteria for dieldrin were frequently exceeded in fall and only occasionally in summer. Because the concentrations are near the criteria limit, the 2-5-fold temporal differences are especially important to evaluating risk. Sampling during one temporal period, such as summer, would not adequately reflect the Willamette River risk for dieldrin, where >66% of the samples in fall exceed the criteria limit and < 20% exceed the limit in summer. None of samples exceeded the U.S. national (56 000 pg/L) or Oregon freshwater aquatic life criteria (1900 pg/L) for dieldrin. These findings are consistent with risk estimates (2 × 10-6 to 3 × 10-5) for dieldrin from consuming fish in this area in our previous human health risk assessment study (9). Distribution of Bioavailable Polychlorinated Biphenyls (PCBs). There were several former PCB using facilities located
on both sides of the river from RM 2 to 10.5. Dissolved PCBs or colloidal particle-associated PCBs associated with historically deposited PCBs in sediments may exchange to the water column. In addition, other sources of PCBs in urban rivers include direct urban runoff, urban community wastewater discharge, sewage treatment works, and combined sewer flow discharges (36). Bioavailable concentrations of ΣPCBs ranged from below detection limits (0.69 to 54 pg/L depending on congeners) to 410 pg/L from RM 1 to 18 during the 4-year study, with an average of 54 ( 55 pg/L. Elevated concentrations of bioavailable ΣPCBs usually reached a maximum at RM 3.5 East in the Superfund site. Temporal Changes in Bioavailable ΣPCBs Distributions. In contrast to ΣDDTs and DDD, a 2-fold increase in bioavailable ΣPCB concentrations was observed during the fall for most sites, with the residential site (RM 15-18) statistically significant (p ) 0.003, t test). Average bioavailable concentrations of ΣPCBs during fall were 30-100% higher than in summer for 10 of the 11 sampling sites throughout the duration of the study. Average of bioavailable ΣPCB daily loads in surface waters were significantly higher in fall: industrial area (p ) 0.002, t test), urban area (p ) 0.043, t test), and residential area (p ) 0.001, t test). Despite the potential dilution effect, bioavailable water PCB concentrations were larger during fall than summer. Bioavailable PCB Congener Profile in Surface Water. PCB 153 and PCB 138 (hexa-CB homolog) were the most frequently detected bioavailable congeners with 85% and 73% frequency, respectively (Supporting Information, Figures 3 and 4). PCB 49 (tetra-CB) was the next most abundant bioavailable congener followed by PCB 52 (tetra-CB) and PCB 153 (hexa-CB) congeners. PCB 166 and two of the coplanar, dioxin-like congeners, PCB 77 and PCB 126, were below the detection limits in all samples. Although bioavailable ΣPCB concentrations and loads were found to have strong spatial and seasonal variation, the PCB congener profiles did not. The similarity of PCB congener profiles for the industrial area at the superfund site (RM 1-8) and the urban area upstream (RM 12-13) suggests the local sources of PCBs at the Superfund site were important enough to increase significantly the concentrations of bioavailable PCBs in surface water but not change substantially congeners profile distribution relative to inputs from upstream. For RM 15-18, relative to the other sites, a slight distribution maximum shifted toward the hexa homologue (PCB 138 and PCB 153) as compared to the other sampling sites, which favored the tetra homologues (PCB 44, PCB 49, and PCB 52); this trend was only observed in summer. The importance of high river flow enhancing PCB transport in surface waters has been previously described (6). The present study suggests the sources of bioavailable PCBs in surface waters at the lower Willamette River were dependent on temporal conditions. Concentrations and daily loads of bioavailable PCBs increased during high precipitation and high river flow, especially during episodic rainstorms in November 2001 and October 2004. DOC and TOC concentrations at the lower Willamette River did not change seasonally over the 4-year study period and were not correlated with the bioavailable PCB concentrations. There was no strong evidence of seasonal or river flow-related change in PCB homologue profiles as determined from general lack of clustering in principle component analyses (data not shown). High river velocities during storms, however, can promote the transport of a relatively large fraction of coarse particles in the particulate phase, which have lower organic carbon contents (i.e., changing organic matter content) and reduced PCB sorption (37). The result suggests that storm events disrupt the interaction between PCBs and associated organic matter, resulting in remobilization of bioavailable fraction VOL. 40, NO. 12, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3693
from sediments or particulate matter. Atmospheric transport, including precipitation, is an important pathway for the transfer of PCBs to surface water (8, 38, 39). The tetrachlorinated congeners in Willamette River waters are similar to the congener patterns in rain samples in other studies (38, 39). The PCB homologue similarities suggest precipitation is a potential input during fall. Urban stormwaters and community wastewater discharges are another important source of PCB (36, 40) especially in fall due to combined sewage overflow events. Sewer overflows to the Willamette River occur frequently (i.e., 100 days in 2004), especially during periods of high precipitation, which usually start in mid October. The most abundant PCB congener found is stormwater overflow (40), and the most abundant congeners found in the fall events in this study (PCBs 52, 101, 118, 138 153) were the same (Supporting Information, Table 3). The increases in bioavailable PCB concentrations and loads were coincident with high precipitation and sewer overflows (Supporting Information, Table 1), suggesting this is a significant source of PCBs. PCBs-Associated Risks to Aquatic Community and Human Health. The national (64 pg/L) and Oregon human health water quality criteria (79 pg/L) for fish consumers were often exceeded within the Superfund site (RM 1-8). None of the samples exceeded the national and Oregon freshwater aquatic life criteria for ΣPCBs (14000 pg/L).With the exception of RM7 east, the national criteria were exceeded more often in fall than in summer. Consuming fish and shellfish from this area may pose adverse health risks to recreational or subsistence fishers based on a 10-6 increased lifetime cancer risk for ΣPCBs. This suggestion is consistent with risk estimates (9 × 10-6 to 5 × 10-3) for ΣPCBs from consuming fish in this area from our previous study (9). The extent of exceedance may have been greater if more PCB congeners had been included in the analysis. Nevertheless, this exceedance indicates that PCB contamination remains a significant problem at Portland Harbor.
Acknowledgments This study was partially funded by the SETAC Chemistry Early Career for Applied Ecological Research Award sponsored by the American Chemistry Council to K.A.A, the Oregon DEQ, and the OHSU pilot project from the NIEHS/ EPA Superfund Basic Research Grant. The Royal Thai government scholarship to D.S. is acknowledged. We appreciate assistance by R. Grove of USGS, Corvallis, OR, and E. Johnson, A. Ackerman, G. Sower, and S. Visalli from OSU.
Supporting Information Available Additional experimental details, formulas, and models used to calculate the bioavailable concentrations, physicochemical water quality data, DDD:DDE ratios and statistics, and temporal PCB congener profiles at each site are presented; concentrations of analytes for each event at each river mile are also presented. This material is available free of charge via the Internet at http://pubs.acs.org.
(4)
(5)
(6)
(7)
(8) (9)
(10) (11)
(12)
(13)
(14) (15)
(16)
(17) (18)
(19)
Literature Cited (1) Escher, B. I.; Hermens, J. L. M. Internal Exposure: linking bioavailability to effects. Environ. Sci. Technol. 2004, 38 (23), 455A-462A. (2) Suffet, I. H.; Jafvert, C. T.; Kukkonen, J.; Servos, M. R.; Spacie, A.; Williams, L. L.; Noblet, J. A. Synopsis of discussion session: influences of particulate and dissolved material on the bioavailability of organic compounds. In Bioavailability: physical, chemical, and biological interactions; Hamelink, J. L., Landrum, P. F., Bergman, H. L., Benson, W. H., Eds.; Lewis Publishers: Boca Raton, FL, 1994. (3) Landrum, P. F.; Hayton, W. L.; Lee, H., II; McCarty, L. S.; Mackay, D.; McKim, J. M. Synopsis of discussion session on the kinetics 3694
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 12, 2006
(20)
(21)
(22)
behinds environmental bioavailability. In Bioavailability: physical, chemical, and biological interactions; Hamelink, J. L., Landrum, P. F., Bergman, H. L., Benson, W. H., Eds.; Lewis Publishers: Boca Raton, FL, 1994; pp 203-219. Jenkins, K. D.; Lee, C. R.; Hobson, J. F. A hazardous waste site at the Naval Weapons Station, Concord, CA. In Fundamentals of aquatic toxicology: effects, environmental fate, and risk assessment, 2nd ed.; Rand, G. M., Ed.; Taylor & Francis: Washington, DC, 1995; pp 883-901. Linkov, I.; Burmistrov, D.; Cura, J.; Bridges, T. Risk-based management of contaminated sediments: consideration of spatial and temporal patterns in exposure model. Environ. Sci. Technol. 2002, 36 (2), 238-246. Foster, G. D.; Lippa, K. A.; Miller, C. V. Seasonal concentrations of organochlorine contaminants at the fall line of the Susquehanna River Basin and estimated fluxes to Northern Chesapeake Bay, USA. Environ. Toxicol. Chem. 2000, 19 (4), 992-1001. Soderstrom, M.; Nylund, K.; Jarnberg, U.; Lithner, G.; Rosen, G.; Kylin, H. Seasonal variations of DDT compounds and PCB in a eutrophic and an oligotrophic lake in relation to algal biomass. Ambio 2000, 29, 230-237. Bremle, G.; Larsson, P. Long-term variation of PCB in the water of a river in relation to precipitation and internal sources. Environ. Sci. Technol. 1997, 31 (11), 3232-3237. Sethajintanin, D.; Johnson, E. R.; Loper, B. R.; Anderson, K. A. Bioaccumulation profiles of chemical contaminants in fish from the lower Willamette River, Portland Harbor, Oregon. Arch. Environ. Contam. Toxicol. 2004, 46, 114-123. Roy F. Weston Inc. Portland Harbor sediment investigation report, Multnomah County, Oregon; EPA910/R-98-006; US Environmental Protection Agency: Seattle, WA, 1998. Huckins, J. N.; Petty, J. D.; Orazio, C. E.; Lebo, J. A.; Clark, R. C.; Gibson, V. L.; Gala, W. R.; Echols, K. R. Determination of uptake kinetics (sampling rates) by lipid-containing semipermeable membrane devices (SPMDs) for polycyclic aromatic hydrocarbons (PAHs) in water. Environ. Sci. Technol. 1999, 33 (21), 39183923. Huckins, J. N.; Tubergen, M. W.; Manuweera, G. K. Semipermeable membrane devices containing model lipid: a new approach to monitoring the bioavailability of lipophilic contaminants and estimate their bioconcentration potential. Chemosphere 1990, 20 (5), 533-552. United States Environmental Protection Agency. National recommended water quality criteria: 2002; EPA-822-R-02-047; Office of Water and Office of Science and Technology: Washington, DC, 2002. Fitzpatrick, M. S. Toxic compounds criteria: 1999-2003 Water quality standards review, issue paper; State of Oregon Department of Environmental Quality: Portland, OR, May 20-21, 2004. Anderson, K. A.; Sethaninintanin, D.; Krissankriangkrai, O.; Johnson, E.; Sower, G. Bioavailable contaminants in Portland Harbor, Oregon, USA. In Proceedings of Environmental Science and Technology; Lyon, W., Hong, J.; RK, R., Eds.; American Science Press: NO, USA, 2005; Vol. V1, pp 665-671. Anderson, K. A.; Johnson, E. R. Bioavailable organochlorine pesticides in a semi-arid region of Eastern Oregon, USA, as determined by gas chromatography with electron capture detector. J. AOAC Int. 2001, 84, 1371-1382. Booij, K.; Sleiderink, H. M.; Smedes, F. Calibrating the uptake kinetics of semipermeable membrane devices using exposure standards. Environ. Toxicol. Chem. 1998, 17 (7), 1236-1245. Huckins, J. N.; Petty, J. D.; Lebo, J. A.; Almeida, F. V.; Booij, K.; Alvarez, D. A.; Cranor, W. L.; Clark, R. C.; Mogensen, B. B. Development of the permeability/performance reference compound approach for in situ calibration of semipermeable membrane devices. Environ. Sci. Technol. 2002, 36 (1), 85-91. Huckins, J. N.; Petty, J. D.; Lebo, J. A.; Orazio, C. E.; Clark, R. C.; Gibson, V. L. SPMD technology tutorial, 3rd ed.; http:// wwwaux.cerc.cr.usgs.gov/spmd/SPMD-Tech_Tutorial.htm (July 5, 2004), Meadows, J. C.; Echols, K. R.; Huckins, J. N.; Borsuk, F. A.; Carline, R. F.; Tillitt, D. E. Estimation of uptake rate constants for PCB congeners accumulated by semipermeable membrane devices and brown trout (Salmo trutta). Environ. Sci. Technol. 1998, 32 (12), 1847-1852. Rantalainen, A.-L.; Cretney, W. J.; Ikonomou, M. G. Uptake rates of semipermeable membrane devices (SPMDs) for PCDDs, PCDFs, and PCBs in water and sediment. Chemosphere 2000, 40, 147-158. Pereira, W. E.; Domagalski, J. L.; Hostettler, F. D.; Brown, L. R.; Rapp, J. B. Occurrence and accumulation of pesticides and organic contaminants in river sediment, water and clam tissues
(23)
(24) (25)
(26)
(27)
(28) (29) (30) (31)
from the San Joaquin River and tributaries, California. Environ. Toxicol. Chem. 1996, 15 (2), 172-180. Zhang, Z. L.; Hong, H. S.; Zhou, J. L.; Huang, J.; Yu, G. Fate and assessment of persistent organic pollutants in water and sediment from Minjiang River Estuary, Southeast China. Chemosphere 2003, 52, 1423-1430. Norrgren, L.; Pettersson, U.; Orn, S.; Bergqvist, P.-A. Environmental monitoring of the Kafue River, located in the Copperbelt, Zambia. Arch. Environ. Contam. Toxicol. 2000, 38, 334-341. Anderson, C. W.; Wood, T. M.; Morae, J. L. Distribution of dissolved pesticides and other water quality constituents in small streams, and their relation to land use, in the Willamette River Basin, Oregon, 1996; U.S. Geological Survey Water-Resources Investigations Report 97-4268; U.S. Geological Survey: Portland, OR, 1997. Augustijn-Beckers, P. W. M.; Hornsby, A. G.; Wauchope, R. D. The SCS/ARS/CES Pesticide properties database for environmental decision-making. II Additional compounds. In Reviews of environmental contamination and toxicology; Ware, G. W., Ed.; Springer-Verlag: New York, 1994; Vol. 137. Petty, J. D.; Huckins, J. N.; Orazio, C. E.; Lebo, J. A.; Clark, R. C.; Gibson, V. L. Evaluation of the semipermeable membrane devices (SPMD) as a passive in situ concentrator of military organic chemicals in water, Final report, approved for release; US Army Medical Research and Material Command: Fort Dentrick, MD, 1997. Kale, S. P.; Murthy, N. B. K.; Raghu, K.; Sherkhane, P. D.; Carvalho, F. P. Studies on degradation of 14C-DDT in the marine sediment. Chemosphere 1999, 39 (6), 959-968. Huang, H.-J.; Liu, S.-M.; Kuo, C.-E. Anaerobic biodegradation of DDT residues (DDT, DDD, and DDE) in estuarine sediment. J. Environ. Health, Part B 2001, 36 (3), 273-288. Aislabie, J. M.; Richards, N. K.; Boul, H. L. Microbial degradation of DDT and its residues-a review. N. Zealand J. Agric. Res. 1997, 40, 269-282. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental organic chemistry; John Wiley & Sons: New York, 1993.
(32) Mackay, D. Multimedia environmental models: the fugacity approach, 2nd ed.; Lewis Publishers: Boca Raton, FL, 2001. (33) Gao, J.; Garison, A. W.; Hoehamer, C.; Mazur, C. S.; Lee Wolfe, N. Uptake and phytotransformation of o,p′-DDT and p,p′-DDT by axenically cultivated aquatic plants. J. Agric. Food Chem. 2000, 48, 6121-6127. (34) Cairns, J.; Heathe, A. G.; Parker, B. C. The effects of temperature upon toxicity of chemicals to aquatic organisms. Hydrobiologia 1975, 47, 135-171. (35) Anderson, C. W.; Rinella, F. A.; Rounds, S. A. Occurrence of selected trace elements and organic compounds and their relation to land use in the Willamette River Basin, Oregon, 1992-1994; U.S. Geological Survey Water-Resources Investigations Report 964234; U.S. Geological Survey: Portland, OR, 1996. (36) Rossi, L.; de Alencastro, L.; Kupper, T.; Tarradellas, J. Urban stormwater contamination by polychlorinated biphenyls (PCBs) and its importance for urban water systems in Switzerland. Sci. Total Environ. 2004, 322, 179-189. (37) Thibodeaux, L. J. Environmental chemodynamics: movement of chemicals in air, water, and soil, 2nd ed.; J. Wiley & Sons: New York, 1996. (38) Duinker, J. C.; Bouchertall, F. On the distribution of atmospheric polychlorinated biphenyls congeners between vapor phase, aerosols, and rain. Environ. Sci. Technol. 1989, 23 (1), 57-62. (39) Mandalakis, M.; Stephanou, E. G. Wet deposition of polychlorinated biphenyls in the Eastern Mediterranean. Environ. Sci. Technol. 2004, 38 (11), 3011-3018. (40) Durell, G. S.; Lizotte, R. D. PCB levels at 26 New York City and New Jersey WPCPs that discharge to the New York/New Jersey Harbor estuary. Environ. Sci. Technol. 1998, 32 (8), 10221031.
Received for review December 2, 2005. Revised manuscript received April 12, 2006. Accepted April 13, 2006. ES052427H
VOL. 40, NO. 12, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3695
