Mercury on a landscape scale—Balancing regional export with wildlife health

Abstract

The Cosumnes River watershed requires a 57–64 percent reduction in loads to meet the new Delta methylmercury (MeHg) total maximum daily load allocation, established by the Central Valley Regional Water Quality Control Board. Because there are no large point sources of MeHg in the watershed, the focus of MeHg load reductions will fall upon non-point sources, particularly the expansive wetlands considered to be a primary source of MeHg in the region. Few management practices have been implemented and tested in order to meet load reductions in managed wetlands, but recent efforts have shown promise. This project examines a treatment approach to reduce MeHg loads to the Sacramento-San Joaquin River Delta by creating open-water deep cells with a small footprint at the downstream end of wetlands to promote net demethylation of MeHg and to minimize MeHg and Hg loads exiting wetlands at the Cosumnes River Preserve. Specifically, the deep cells were were located immediately up gradient of the wetland’s outflow weir and were deep enough (75–91 centimeter depth) to be vegetation-free. The topographic and hydrologic structure of each treatment wetland was modified to include open-water deep cells so that the removal of aqueous MeHg might be enhanced through (1) particle settling, (2) photo-degradation, and (3) benthic microbial demethylation. These deep cells were, therefore, expected to clean MeHg from surface water prior to its discharge to the Cosumnes River and the downstream Delta. Our goal was to test whether the implementation of the deep cells within wetlands would minimize MeHg and total Hg export. Further, we sought to test whether continuous flow-through hydrology, would lower MeHg concentrations in resident biota, compared to traditional wetland management operations. The dominant practice in seasonal wetlands management is the “fill-and-maintain” approach, in which wetlands are filled with water and the water levels maintained without substantial draining until drawdown. Our approach was to create and characterize replicate treatment wetland complexes, in conjunction with monitoring of hydrologic, biologic, and chemical indicators of MeHg exposure for two full annual cycles within winter-spring flooded seasonal wetlands. In addition to the creation of deep cells within treatment wetlands, hydrology was manipulated so that there was a constant flow-through of water, while the control wetlands utilized the fill-and-maintain approach. Specifically, the treatment wetlands were maintained in a flow-through manner, while the control wetlands were maintained in a fill-and-maintain manner from September through May, to test the hypothesis that the flow of water through the seasonal wetland can lower fish bioaccumulation through dilution of MeHg-concentrated water within the wetland by constant inflows of water into the wetland. The major tasks of this study included: (1) field design and implementation, (2) water and wetland management, (3) hydrologic monitoring and water quality sampling, (4) MeHg export and load estimates, (5) caged fish experiments for examining MeHg bioaccumulation, (6) site and process characterization to improve understanding and transferability of results, (7) adaptive management, transferability, and outreach, and (8) reporting of results and conclusions. This report summarizes the key findings of this study, which focuses on MeHg load estimates from control and treatment wetlands, quantification of three MeHg removal mechanisms (particulate settling, benthic demethylation, and photo-demethylation) in the deep cells within the treatment wetlands, and MeHg bioaccumulation in wetland fishes. Key findings include: Over two years of study, mean whole-water MeHg load decreased 37 percent in deep cells, when comparing inlet of check weir flows to outlet. Of the 37 percent MeHg load removed within the deep cell, photodegradation accounted for 7 percent and particle flux to the benthos accounted for 24 percent of the mass removed, with the remaining 6 percent apparent MeHg loss unexplained. Benthic MeHg degradation did not appear to be a major MeHg removal process in the deep cells, as changes in the ambient MeHg pool over 7-day bottle incubations showed that the surface sediment exhibited net MeHg production in the majority (87 percent) of incubation experiments. In only 13 percent of the incubations (3 out of 24) was net MeHg degradation observed. Estimates of benthic diffusive flux of MeHg across the sediment/water interface were small relative to particulate flux and variable (positive or negative), suggesting this is likely a minor term in the overall MeHg budget within the deep cells. Although the deep cells served as net MeHg sink overall, MeHg export from the flow-through treatment wetlands (shallow and deep combined) exceeded export from the fill-and-maintain managed control wetlands, because of the differences in hydrologic management between the two wetland types. Shallow wetlands under flow-through conditions generated a net export of MeHg. Most of the annual MeHg export from the treatment wetlands occurred within the first 3 months of flood up (September to November), shortly after hydrologic management began. Despite the effectiveness of the deep cell in lowering MeHg export concentrations, total mercury (THg) concentration did not decrease in biosentinel fish ( Gambusia affinis , Mosquitofish) between the deep cell inlet and outlet. Mosquitofish THg concentrations were higher in treatment wetlands than in control wetlands during the first year of study, likely because of an associated increase in MeHg availability immediately following wetland construction activities. Mosquitofish THg concentrations declined in the treatment wetlands during the second year of study, and fish THg concentrations in treatment wetlands were no different from those in the control. Similarly, the increased hydrologic flow rates in the treatment wetlands did not lower fish THg concentrations nor aqueous MeHg concentrations in the shallow cells, suggesting that MeHg flux from the sediment to water column exceeded the flow-through flushing rate in the shallow portion of the treatment wetlands. Reductions in MeHg concentrations of surface water and fish may require higher flow rates than used in the study to achieve the region’s regulatory goals. However, the flow rates necessary may not be feasible for these managed wetlands because of limited water supply and the associated costs for water and pumping. The use of deep cells in seasonal wetlands were effective in lowering MeHg exports under continuous water flow-through hydrology. However, fill-and-maintain hydrology had lower exports overall, because of a single major drainage event at the end of the flood season. Future studies focused on limiting MeHg export should consider combining deep cells with the fill-and-maintain or fill-and-trickle hydrologic management approach.