Modeling mercury biomagnification (South River, Virginia, USA) to inform river management decision making.

Mercury trophic transfer in the South River (VA, USA) was modeled to guide river remediation decision making. Sixteen different biota types were collected at six sites within 23 river miles. Mercury biomagnification was modeled using a general biomagnification model based on delta(15)N and distance from the historic mercury release. Methylmercury trophic transfer was clearer than that for total Hg and, therefore, was used to build the predictive model (r(2) (prediction) = 0.76). The methylmercury biomagnification factors were similar among sites, but model intercept did increase with distance down river. Minimum Akaike's Information Criterion Estimation (MAICE) justified the incorporation of distance in the model. A model with a very similar biomagnification factor to the South River (95% confidence intervals [CI] = 0.38-0.52) was produced for a second contaminated Virginia river, the North Fork Holston River (95% CI = 0.41-0.55). Percent of total Hg that was methylmercury increased monotonically with trophic position. Trophic models based on delta(15)N were adequate for predicting changes in mercury concentrations in edible fish under different remediation scenarios.