Environ. Sci. Technol. 2009, 43, 2865–2870
Comparing Polychaete and Polyethylene Uptake to Assess Sediment Resuspension Effects on PCB Bioavailability C A R E Y L . F R I E D M A N , * ,‡ ROBERT M. BURGESS,† MONIQUE M. PERRON,† MARK G. CANTWELL,† KAY T. HO,† AND RAINER LOHMANN‡ U.S. Environmental Protection Agency, ORD/NHEERL, Atlantic Ecology Division, Narragansett, Rhode Island, 02882, Graduate School of Oceanography, University of Rhode Island, Narragansett, RI, 02882, and Harvard School of Public Health, Department of Environmental Health, Boston, MA, 02115
Received December 29, 2008. Revised manuscript received February 17, 2009. Accepted February 20, 2009.
Polyethylene sampler uptake was compared to polychaete uptake to assess bioavailability of polychlorinated biphenyls (PCBs) from resuspended sediments. New Bedford Harbor (MA, U.S.) sediment, contaminated with PCBs, was resuspended under four different water column oxidation conditions: resuspension alone, resuspension under aeration, resuspension under helium, and no resuspension (control). Residuals were tested for differences in PCB availability to the marine polychaete Nereis virens and to polyethylene (PE) passive samplers. Few significant differences between the four resuspension treatments were observed: under aeration, three of 23 PCBs analyzed showed significant increases in polychaete accumulation, while resuspension alone showed increased concentrations in PE samplers for nine of 23 PCBs. Otherwise, no differences were observed and overall we concluded that resuspension had no effect on residual PCB availability. The relationship between disequilibrium-corrected PE and lipidnormalized polychaete PCB concentrations was nearly 1:1 with a strong linear correlation (r2 ) 0.877), demonstrating PCBs are taken up similarly into PE and lipid. On average, PE samplers suggested dissolved PCB concentrations 3.6 times greater than those calculated with lipid-water partitioning, though on a congener-specific basis this was only observed for lower chlorinated PCBs; for higher chlorinated PCBs, PE-water partitioning suggested lower dissolved concentrations than those based on lipid. Organic carbon (OC)-water and OC and black carbon combined (OC+BC)-water partitioning suggested average dissolved concentrations 29 and 10 times greater, respectively, than those estimated with lipid-water partitioning. This demonstrates that PE-measured porewater concentrations can provide a more reliable estimate of bioavailability than sediment geochemistry. * Corresponding author phone: (401) 874-6268; fax: (401) 8746811; e-mail:
[email protected]. † U.S. Environmental Protection Agency. ‡ University of Rhode Island. † Harvard School of Public Health. 10.1021/es803695n CCC: $40.75
Published on Web 03/18/2009
2009 American Chemical Society
Introduction Sediments are capable of strongly sorbing hydrophobic organic contaminants (HOCs) in the marine environment (1). Over time, contaminated sediments can be a source of HOCs and a significant health risk to aquatic food webs. Many HOCs bioaccumulate in the benthos and biomagnify at higher trophic levels (2-5). Bioavailability of sedimentary HOCs is traditionally assessed by measuring HOC uptake into benthic organisms over a standard exposure time. More recently, passive samplers have been used experimentally to estimate bioavailability (6-10). In this study, we compared the availability of sedimentary HOCs to benthic polychaetes and to polyethylene (PE) passive samplers in a standard bioaccumulation test for different sediment treatments. To mitigate risk to aquatic ecosystems, contaminated sediments are commonly removed by dredging. However, the effectiveness of dredging at reducing risk has been questioned (11). In particular, it is not known whether resuspension or disruption of natural sediment redox conditions during dredging alters the bioavailability of particleassociated HOCs. Several studies have shown that aeration of sediment porewater and/or solids increases HOC partitioning from dissolved and particulate organic matter (DOM and POM) to the aqueous phase (12, 13). Since dissolved concentrations correlate to bioavailability (7, 9, 14, 15), increased dissolved HOCs would likely increase risk to biota. While some studies have investigated HOC desorption and changes in bioavailability during sediment resuspensions (12, 16), few have examined changes associated with residuals (resuspended particles that have resettled). The present study examined the effects of resuspension, under a range of redox conditions, on the bioavailability of polychlorinated biphenyls (PCBs) from field sediments. Bioavailability was assessed by measuring PCB uptake in Nereis virens, a marine polychaete, as well as in PE passive samplers. Use of PE and other passive samplers to estimate bioavailability is based on equilibrium partitioning theory (EqP). EqP proposes HOCs reach a predictable equilibrium distribution between sediment organic carbon (OC), porewater, and biota lipid (17). Since freely dissolved HOCs (i.e., HOCs available for partitioning and not associated with DOM/POM) have historically been difficult to measure, bioavailability is often estimated from OC-normalized sediment concentrations. However, this assumes a homogeneous distribution of OC and does not account for sequestration of HOCs due to black carbon (BC), aging effects, or differences in desorption rates (1, 18, 19). Nonpolar passive samplers circumvent these issues by absorbing porewater HOCs via partitioning of freely dissolved HOCs only. Conducting bioaccumulation tests with passive samplers would reduce the amount of space, time, and expense involved with using organisms and should provide more accurate bioavailability measurements than those based on sediment geochemistry. While other passive samplers such as polydimethylsiloxane (PDMS)-coated fibers and semipermeable membrane devices (SPMDs) have shown success as biomimetic tools (6-10), PDMS-coated fibers often accumulate total HOC masses that are lower than analytical detection limits, and SPMDs are difficult to work with because they contain triolein, a lipidlike material. Conversely, the analysis of PE can be conducted easily in most laboratories. A correlation between polycyclic aromatic hydrocarbon (PAH) uptake in PE and bioaccumulation in marine polychaetes has already been shown with PE uptake capable of explaining 2/3 of polychaete uptake (20). VOL. 43, NO. 8, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2865
Many sediment-dwelling organisms actively ingest sediment (21, 22), allowing organisms to equilibrate faster than PE. One way to determine whether samplers have equilibrated during an exposure is to add performance reference compounds (PRCs), nonenvironmental HOCs with properties similar to analyte HOCs, to samplers predeployment (23-26). PRCs allow an estimation of analytes’ absorption rates, assuming similar loss and uptake kinetics. We investigated an alternative method of assessing equilibrium status: a desorption study was conducted using PE impregnated with the same PCBs measured in our bioaccumulation test and clean sediment manipulated to have a range of OC. With this method we were able to calculate PCB exchange rate coefficients to correct for disequilibrium without extrapolating from similar compounds. Collectively, the main objectives of this study were to (1) determine whether sediment resuspension under various redox conditions causes differences in PCB availability as measured by marine polychaete and PE sampler uptake; (2) compare polychaete and PE uptake of PCBs to determine if PE can be used biomimetically; and (3) compare equilibrium porewater PCB concentrations calculated from biota, PE sampler, and sediment concentrations to assess whether PE samplers or sediment geochemistry provide more accurate estimates of exposure.
Materials and Methods Resuspension. Sediment from the New Bedford Harbor (NBH) U.S. EPA Superfund site (MA) was collected in November 2006 (41° 40.03′, 70° 55.04′) and stored in the dark at 4 °C until use. Four treatments were evaluated: “Resuspension”, “Resuspension with Air”, “Resuspension with Helium”, and “No Resuspension”. Each began with the addition of 3 L aliquots of homogenized NBH sediment to four separate 57 L olefin-blend cylindrical drums with sealable lids. Fifty-three L of Narragansett Bay, RI (NB) seawater (20 °C, 30 ‰) was added to each drum and lids were fastened. For each drum, a stainless steel drive shaft and impeller connected to a heavy duty mixing motor were inserted into the drum through a hole in the lid (see Supporting Information (SI) for more details). For two of the four drums, air and helium (respectively) were added to the overlying water with air stones connected to gas cylinders. Flow rates were 600-3600 mL/min for air and 300-3000 mL/min for helium, depending on dissolved oxygen (DO) measurements. Sediments were resuspended for three consecutive days for six hours/day. At the beginning of each day and at one hour intervals, measurements of overlying water temperature, air temperature, salinity, pH, DO, and turbidity were collected (data not shown). After initial measurements on the second and third days, 16 L of overlying seawater were removed from each drum and replaced with 16 L of fresh NB seawater to simulate daily tidal exchanges in the upper portion of New Bedford Harbor (27). On the first day of mixing, all treatments were mixed briefly (
