RESUMEN
A one hectare pond on the headwaters of a mercury-contaminated creek in Oak Ridge, Tennessee acted as a biochemical reactor for the production of methylmercury, increasing waterborne methylmercury concentrations in the stream below the pond discharge. The flow of the creek was diverted around the pond in order to eliminate this input. Waterborne total mercury, methylmercury, and mercury in fish, were monitored in the pond and stream before and after bypass. Waterborne methylmercury concentration in the creek downstream from the pond decreased over 800% following diversion of streamflow around the pond, but mercury in redbreast sunfish in the pond tailwater did not decline similarly. Within the pond, now isolated from fresh waterborne mercury inputs from the stream, methylmercury concentrations in the water column remained similar to levels present before bypass. However, mercury concentrations in sunfish in the pond decreased approximately 75% following bypass, despite the continued presence of highly contaminated sediments (approximately 50 mg Hg/kg dry weight). We concluded that a decrease in the fraction of 'dissolved methylmercury' in the isolated pond relative to pre-bypass conditions explained the decrease in mercury in fish within the pond. That observation also indicates that mercury associated with pond sediments was relatively unavailable for eventual bioaccumulation when compared to 'fresh' mercury contributed by upstream sources. The lack of a post-bypass decrease in mercury concentrations in tailwater fish was also likely to be associated with the particle-associated nature of waterborne methylmercury exported from the pond.