Devils Lake rose dramatically during the 1990's, causing extensive flood damages. Because of the potential for continued flooding, the U.S. Army Corps of Engineers has been conducting studies to evaluate the feasibility of constructing and operating an outlet from Devils Lake. The occurrence of mercury in lakes, wetlands, and rivers and the potential for increased loading of mercury into the Sheyenne River as a result of a Devils Lake outlet needed to be evaluated as part of the studies.

Sixteen lake, wetland, and river sites in the Devils Lake, Sheyenne River, Red River of the North, and Red Lake River Basins were sampled and analyzed for mercury constituents and other selected properties and constituents relevant to mercury aquatic chemistry. For the lake and wetland sites, whole-water methylmercury concentrations ranged from less than 0.04 to 3.53 nanograms per liter and whole-water total mercury concentrations ranged from 0.38 to 7.02 nanograms per liter. Conditions favorable for methylation of mercury generally exist at the lake and wetland sites, as indicated by larger dissolved methylmercury concentrations in near-bottom samples than in near-surface samples and by relatively large ratios of methylmercury to total mercury (generally greater than 10 percent for the summer sampling period). Total mercury concentrations were larger for the summer sampling period than for the winter sampling period for all lake and wetland sites. A wetland site in the upper Devils Lake Basin had the largest mercury concentrations for the lake and wetland sites.

For the river sites, whole-water methylmercury concentrations ranged from 0.15 to 1.13 nanograms per liter and whole-water total mercury concentrations ranged from 2.00 to 26.90 nanograms per liter. Most of the mercury for the river sites occurred in particulate inorganic phase. Summer ratios of whole-water methylmercury to whole-water total mercury were 35 percent for Starkweather Coulee (a wetland-dominated site), near or less than 10 percent for the Sheyenne River sites, and less than 8 percent for the Red River of the North and Red Lake River sites.

Although the number of samples collected during this investigation is small, results indicated an outlet from Devils Lake probably would not have adverse effects on mercury concentrations in the Sheyenne River upstream from Lake Ashtabula. However, because discharges in the Sheyenne River would increase during some periods, loads of mercury entering Lake Ashtabula also would increase. Lake Ashtabula probably serves as a sink for suspended sediment and mercury. Thus, a Devils Lake outlet probably would not have substantial effects on mercury concentrations and loads in the downstream part of the Sheyenne River or in the Red River of the North. More substantial effects could occur for Lake Ashtabula.