Derivation of wildlife values for mercury

A procedure has been developed to estimate surface water concentrations of toxicants ("wildlife values") that will protect the viability of wildlife p opulations associated with aquatic resources. This procedure was designed p rimarily to protect piscivorous birds and mammals from compounds that bioac cumulate in fish and was used in the Great Lakes Water Quality Initiative ( GLI) to calculate wildlife values (WV) for mercury, DDT/DDE, total polychlo rinated biphenyls (PCBs), and 2,3,7,8-tetrachlorodibenzodioxin (TCDD). Publ ished in 1995, and expressed as total mercury in unfiltered water, the fina l wildlife value (WVf) for mercury derived in the GL1 was 1300 pg Hg/L. Thi s value was selected as the wildlife criterion (WC) for mercury in the Grea t Lakes basin. A second WV, for mercury was derived in 1997 as part of a Co ngressionally mandated report on airborne mercury emissions. These calculat ions were based upon mercury speciation data that were largely unavailable when the GLI was developed. important features of the WV, in the Report to Congress include its calculation on a dissolved methylmercury basis and a r eliance on field data to estimate fish bioaccumulation factors. Calculated as methylmercury in filtered water, the WV, derived in the report is 50 pg Hg/L (equivalent to 54 pg MeHg/L). A comparison of WV in the CLI and the Re port to Congress requires that average values be specified for mercury spec iation in natural systems. Based on this information, the WV, given in the report corresponds to a value of 910 pg Hg/L, as total mercury in unfiltere d water, or about 70% of the WV, derived in the GLI. in this article we des cribe the algorithm used to derive WV in the GLI and the Report to Congress and review its application to mercury. Scientific uncertainties in derivin g WV, particularly as they apply to mercury, are critically examined.