Low-level experimental selenite additions decrease mercury in aquatic food chains and fish muscle but increase selenium in fish gonads.

We investigated whether low-level addition of selenium (Se) could decrease mercury (Hg) in freshwater fish without imposing Se toxicity. Using a regression design, selenite was added to large mesocosms in a lake to achieve target concentrations ≤1.6 µg/L. (198)Hg (spike Hg) was added to mesocosms to determine changes in Hg bioaccumulation. Adding Se decreased spike total Hg (THg) in fish muscle, ambient THg in fish liver, and bioaccumulation of spike THg in muscle and spike methylmercury (MeHg) in zooplankton and Chironomid larvae relative to controls. Se decreased Hg in the food web but not in water, indicating that the dominant effect of Se on Hg cycling occurs in the food web. Concentrations of Se in gonads of fish were positively correlated with Se concentrations in water but did not exceed reproductive toxicity thresholds after 8 weeks. We conclude that low-level addition of Se decreases MeHg bioaccumulation and increases Se in gonads of fish; however, additions of Se to freshwater systems to decrease Hg in fish should be treated with caution because Se in fish gonads were likely to exceed toxic concentrations if exposed to increased Se for a longer period of time.

Strategies to lower methyl mercury concentrations in hydroelectric reservoirs and lakes: A review.

Mercury (Hg) concentrations in fish in lakes are elevated due to increased global cycling of Hg. A special case of elevated Hg concentrations in fish occurs in new hydroelectric reservoirs because of increased rates of converting Hg in the environment into methyl mercury (MeHg). People and wildlife that eat fish from hydroelectric reservoirs have an elevated risk of accumulating too much MeHg. Demand for electrical energy is leading to the creation of new reservoirs. In 2005, Canada derived 60% of its electricity from hydroelectric reservoirs. As a result, hydroelectric companies and governing agencies are exploring strategies to lower MeHg contamination. Strategies may involve lowering the source of Hg before flooding, the rate of Hg methylation, or MeHg bioaccumulation and biomagnification. Possible strategies reviewed in this article include selecting a site to minimize impacts, intensive fishing, adding selenium, adding lime to acidic systems, burning before flooding, removing standing trees, adding phosphorus, demethylating MeHg by ultraviolet light, capping and dredging bottom sediment, aerating anoxic bottom sediment and waters, and water level management. A preventative strategy is to limit the flooded area, especially wetland areas. Flooded upland areas that contain less carbon produce MeHg for a shorter time than wetland areas. Run-of-the-river reservoirs contain lower MeHg concentrations than reservoirs that flood vast areas, at the cost of exporting MeHg downstream. Managing water levels to flush systems during times of peak MeHg production may have benefits for the reservoir, but also transports MeHg downstream. Intensive fishing can lower MeHg in food webs by increasing fish growth rate. Additions of selenium can lower MeHg bioaccumulation, but the mechanisms are not well established and excess selenium causes toxicity. Liming can lower fish Hg concentrations in lakes acidified with sulphuric and nitric acid. Burning before flooding can lower the production of MeHg, but MeHg bioaccumulation may increase. The most promising strategy will be one that is agreeable to all affected people.

The burning question: does burning before flooding lower methyl mercury production and bioaccumulation?

Production of methyl mercury (MeHg) is elevated in new hydroelectric reservoirs because organic carbon stimulates methylation of inorganic mercury (Hg) stored in the terrestrial system. This can cause adverse health in fish and in organisms that eat fish. We expected that burning vegetation before flooding would decrease the amount of Hg and organic carbon and thereby lower MeHg production. We conducted a replicated field experiment to investigate the effects of burning vegetation and soil before flooding on MeHg production and bioaccumulation. Vegetation and soil were added to mesocosms in the following combinations: unburned vegetation and unburned soil (Fresh treatments), burned vegetation and unburned soil (Partial Burn treatments), and burned vegetation and burned soil (Complete Burn treatments). Controls had no added vegetation or soil. During combustion with propane torches, a large percentage of the total Hg (THg) and MeHg was lost from vegetation and soil. THg and MeHg concentrations were highest in the surface water of Fresh treatments, lower in Partial Burn treatments and lowest in Complete Burn treatments and controls. Differences in concentrations of MeHg in biota were consistent among treatments, but did not follow aqueous concentrations. On the final sample date, MeHg concentrations in biota of Controls and Partial Burn treatments were greater than in Complete Burn and Fresh treatments. The lack of relationship between MeHg in biota and MeHg in water may have been due to modification of the bioavailability of MeHg by dissolved organic matter as the ratios of MeHg in biota to water were inversely correlated with concentrations of dissolved organic carbon. Although burning before flooding decreased MeHg concentrations in the water, it did not lower MeHg accumulation in the lower food web.