ABSTRACT
Mercury (Hg) concentrations in fish in boreal reservoirs have been shown to be increased for up to 3 decades after impoundment. However, the time course of increased concentrations is not well known. The purpose of this study was to determine the evolution of Hg concentrations in fish in the boreal reservoirs of northern Manitoba, Canada, and its relationship with severity of flooding. We determined total Hg concentrations in three species of fish for up to 35 years after impoundment in 14 lakes and lake basins. Postimpoundment trends depended on fish species and reservoir. In the benthivorous lake whitefish (Coregonus clupeaformis), Hg concentrations increased after flooding to between 0.2 and 0.4 microg g(-1) wet weight compared with preimpoundment concentrations between 0.06 and 0.14 microg g(-1) and concentrations in natural lakes between 0.03 and 0.06 microg g(-1). Hg concentrations in lake whitefish were usually highest within 6 years after lake impoundment and took 10 to 20 years after impoundment to decrease to background concentrations in most reservoirs. Hg concentrations in predatory northern pike (Esox lucius) and walleye (Sander vitreus) were highest 2 to 8 years after flooding at 0.7 to 2.6 microg g(-1) compared with preimpoundment concentrations of 0.19 to 0.47 microg g(-1) and concentrations in natural lakes of 0.35 to 0.47 microg g(-1). Hg concentrations in these predatory species decreased consistently in subsequent years and required 10 to 23 years to return to background levels. Thus, results demonstrate the effect of trophic level on Hg concentrations (biomagnification). Peak Hg concentrations depended on the amount of flooding (relative increase in lake surface area). Asymptotic concentrations of approximately 0.25 microg g(-1) for lake whitefish and 1.6 microg g(-1) for both walleye and northern pike were reached at approximately 100% flooding. Downstream effects were apparent because many reservoirs downstream of other impoundments had higher Hg concentrations in fish than would be expected on the basis of flooding amount.
