MERGANSER - A Geospatial Model to Predict Fish Mercury in New England Lakes

James Shanley1, Dave Evers2 Kathleen Fahey3, Craig Johnston4, John Johnston5, Neil Kamman6, Susy King7, Eric Miller8, Richard Moore4, Diane Nacci9, Alison Simcox10, Dick Smith11 and Kate Williams2

All six New England states have issued health advisories for fish consumption in specific water bodies because of mercury contamination. But how does one assess the human and ecosystem exposure risk to methylmercury in lakes that are not monitored? MERGANSER (MERcury Geo-spatial AssessmeNtS for the New England Region) is a project designed to predict fish Hg (and loon Hg) in lakes based on watershed features, local mercury deposition estimates, and other parameters for which region-wide coverages can be readily generated. MERGANSER is constructed for the 4425 lakes in New England larger than 8 ha using fish Hg measurements from 811 sites spanning the period 1996-2006 and modeled Hg deposition for 2002. By restricting consideration to this ten-year period we minimize the complication of trends. We have developed a fish Hg index normalized to body length and fish species, and will also create a loon Hg index. MERGANSER uses an empirical multiple regression approach to identify the features and factors associated with the fish Hg index. Many parameters are relatively time invariant, such as areal extent of total wetland and various wetland types, wetland in contact with lakeshore, critical load exceedance (a proxy for pH), and topographic features. Other parameters are dynamic, such as mercury deposition, population, and land cover types. MERGANSER allows prediction of fish Hg in unmonitored lakes for present conditions, and it will suggest the magnitude and direction of fish Hg change with changes in dynamic predictors such as land use and other factors. Most notably, within the model uncertainty, MERGANSER will also quantify the sensitivity of fish Hg (and loon Hg) to changes in atmospheric Hg deposition. We anticipate that the MERGANSER approach will have broad transferability to other regions.

1 U.S. Geological Survey, Montpelier, VT
2 Biodiversity Research Institute, Gorham, ME
3 NorthEast States Coordinated Air Use Management, Boston, MA
4 U.S. Geological Survey, Pembroke, NH
5 US EPA, Athens, GA
6 Vermont Agency of Natural Resources, Waterbury, VT
7 New England Interstate Water Pollution Control Commission, Boston, MA
8 Ecosystems Research Group, Ltd., Norwich, VT
9 US EPA, Narragansett, RI
10 US EPA, Boston, MA
11 U.S. Geological Survey, Reston, VA