PCB Decline in Settling Particles and Benthic Recycling of PCBs and PAHs in Lake Superior JEFF D. JEREMIASON, †,‡ STEVEN J. EISENREICH,* ,† JOEL E. BAKER, ‡ AND BRIAN J. EADIE § Departm ent of Environm ental Sciences, Rutgers University, 14 College Farm Road, New Brunswick, New Jersey 08903, Chesapeake Biological Laboratory, P.O. Box 38, 1 William s Street, Solomons, Maryland 20688, and Great Lakes Environm ental Research Laboratory, NOAA, 2205 Com m onwealth Boulevard, Ann Arbor, Michigan 48105 Sediment traps were deployed at five sites in Lake Superior at multiple depths during lake stratification in 1987 and 1991. Mass, organic carbon, PCB, and PAH fluxes were determined. PCB concentrations on settling solids declined from 1984 to 1991 with a first-order rate constant of 0.26 yr -1 similar to reported water column concentration decreases (0.20 yr -1 ). Total PCB settling fluxes from the upper 35 m of water averaged 121 ( 40 ng/m 2 ‚d in 1987 and 48 ( 23 ng/m 2 ‚d in 1991. Settling fluxes are greater than reported wet and dry deposition fluxes (2.8 ng/m 2 ‚d) and demonstrate the intense recycling of PCBs within the lake. A large fraction (>50%) of the total Lake Superior water PCB burden is transported each year by settling particles to within 5 m of the lake bottom, but only 2-5% of settling PCBs accumulate in bottom sediments. Thus, most of the PCBs are recycled in the benthic region, possibly representing a major entry point for PCBs into higher trophic levels through the benthic food web. Benthic recycling of PAH compounds with three and four rings occurred, but a larger fraction of these settling PAHs accumulated in bottom sediments (8-33%). No consistent temporal trends were observed in PAH concentrations on settling particles from 1984 to 1991. Introduction Relatively few studies have employed sediment traps to evaluate the biogeochemistry of settling solids and its role in organic contaminant cycling in aquatic systems (1-8). Settling particles play an integral role in aquatic food chains and contaminant cycling in aquatic systems (1, 3, 6, 7, 9, 10). The collection of settling solids is a powerful tool in determininghowcontaminantsare processed within aquatic systems and can provide insight into how and where contaminants are introduced to the food web. The cycling of organic contaminants such as polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs) is intimatelylinked to production and cyclingoforganic matter (3, 11, 12). Organiccarbon produced byprimaryproduction can be cycled many times through the pelagic food chain or transported quickly from surface to bottom waters and introduced to the benthic food web where it may undergo further recycling (13-15). Thus, a small percentage of primaryproduction in Lake Superior accumulatesin bottom sediments(16, 17). However,efficient carbon recycling,such as occurs in Lake Superior, results in efficient contaminant transfer through the benthic food web. Lake Superior has anomalously high PCB concentrations in lake trout relative to other small regional lakes that have even higher water PCB concentrations (18, 19). PAHs are not biomagnified in food webs, but many are suspected carcinogens. PCB concentrations in Lake Superior surface waters decreased from 2.4 to 0.18 ng/L at a first-order rate of 0.20 yr -1 between 1980 and 1992 (20). Volatilization was the dominant removal mechanism over this time period, while permanent sediment burial was of minor importance. Despite burial being an insignificant removal mechanism, settling solids efficiently transport PCBs to bottom waters (3). However, most PCBs are recycled back into the water column and not incorporated into bottom sediments. Intense internal PCB recycling is also reflected by settling fluxesbeing10-50timesgreaterthan atmosphericdeposition (21). The current study suggests that settling particles efficientlytransport PCBs and PAHs from the surface waters, providing a direct and effective link to the benthic food web. The present studybuilds on the 1984-1985 Lake Superior trap study (3). Results from 1984 to 1985 demonstrated that settling solids were enriched in PCBs, phenanthrene, and fluorene relative to suspended solids collected by filtration. PCBs were efficiently recycled in the water column with less than 1% of settling PCBs accumulating in the sediments. Lower molecular weight PCBs, fluorene, and phenanthrene were recycled to a greater extent and settled faster than the higher molecular weight PCBs and PAHs (3). The 1984- 1985 study extended over both stratified and unstratified periods. In the current study, traps were deployed only during the stratified period. The objectives of the current study were (i) to quantify trends in PCB concentrations on settling solids since 1984; (ii) to reexamine PCB, PAH, and organic carbon recycling behavior in Lake Superior;and (iii) to examine the potential of recycling enhancing PCB concentrations in the food web of Lake Superior. Methods Field Sampling. Sediment traps were deployed during the stratified period in 1984 (3), 1987, and 1991 in Lake Superior at five locations and multiple depths (Table 1). The 4 or 8 in. diameter Plexiglass traps had height-to-width ratios of 5:1 (3, 13). Samples were collected in a 500-mLpolyethylene bottle below a funnel with a 1 in. diameter opening. Chloroform (20 mL) was placed in each bottle to prevent degradation of trapped material. Sodium azide (30 mM) was used as a preservative in each duplicate trap depth in 1987 (22). However, Lee et al. (23) later showed that sodium azide was not a recommended preservative for organic matter at five times this concentration. Overlying water was siphoned offupon trap recovery, and the remaining sample containing residual beads of chloroform was air-dried at 60 °C, weighed, and stored frozen. Total sample masses collected ranged from 15 mg to several grams. Laboratory. Samples were placed in Pyrex extraction thimbles and extracted for 24 h in Soxhlet extractors with approximately150 mLofdichloromethane (DCM). ABuchi Rotavap (model RE111) was used to switch the solvent to hexaneandreducethesamplevolumeto ∼1mL. The extracts were transferred to 4-mLamber vials and quantitativelysplit *To whom correspondence should be addressed. Phone: (732)- 932-8575;fax: (732)932-8644;e-mail: eisenreich@aesop.rutgers.edu. † Rutgers University. ‡ Chesapeake Biological Laboratory. § Great Lakes Environmental Research Laboratory. Environ. Sci. Technol. 1998, 32, 3249-3256 S0013-936X(98)00143-6 CCC: $15.00  1998 American Chemical Society VOL. 32, NO. 21, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 3249 Published on Web 09/17/1998