MAPPING VULNERABILITY OF AQUATIC ECOSYSTEMS ACROSS THE CONTIGUOUS UNITED STATES

About 20 years ago, researchers at a few locations across the globe discovered high levels of mercury in fish from remote settings lacking any obvious mercury source. We now know that for most locations atmospheric deposition is the dominant mercury source, and that mercury methylation is the key process that translates low mercury loading rates into relatively high bioaccumulation levels in aquatic food webs. Presently, almost all US states have advisories for elevated levels of mercury in sport fish, and as a result there is considerable public awareness and concern for this nearly ubiquitous contaminant issue. In some states, “statewide” advisories have been issued because elevated fish mercury levels are so common, or the state has no effective way to monitor tens of thousands of lakes, reservoirs and wetlands. As such, resource managers and public health officials have limited options for informing the public on of where elevated mercury concentrations in sport fish are more likely to occur than others. This project provides, for the first time, a national map of predicted (modeled) methylmercury concentrations, which is the most toxic and bioaccumulative form of mercury in the environment. The map is the result of over two decades of research that resulted in the formulation of conceptual models of the mercury methylation process, which is strongly governed by environmental conditions – specifically hydrologic landscapes and water quality. The map shows clear regional trends in the distribution of predicted elevated methylmercury concentrations. East of the Mississippi, the Gulf and southeastern Atlantic coast, the northeast, and the upper Midwest states are predicted to have generally higher methylmercury levels. In the west, however, regional patterns are much less clear, with although numerous watersheds are predicted to have elevated methylmercury levels. Although this map is being presented at the continental US scale, the principals used to generate the modeled results can easily applied to data sets that represent smaller and larger areas.