Journal of Conference Abstracts

Cycling of Methyl Mercury in Boreal Forest Ecosystems

J. W. M. Rudd¹ (rudd@ cc.umanitoba.ca), R. A. Bodaly¹ (Drew.Bodaly@c-a.dfo.dfo-mpo.x400.gc.ca), C. A. Kelly² (ckelly@bldgwall.lan1.umanitoba.ca), V. St. Louis² (vstlouis@starfish.acs.brockport.edu), M. Paterson¹ (Michael.Paterson@c-a.dfo.dfo-mpo.x400.gc.ca),
D. Rosenberg¹ (David.Rosenberg@c-a.dfo.dfo-mpo.x400.gc.ca), R. Harris⁵ (harrisr@idirect.com), B. Hall⁴ (britt@intouch.bc.ca), A. Heyes³ (heyes@acnatsci.org), B. Branfireun³,
T. Sellers² (TSELLERS@bldgwall.lan1.umanitoba.ca) & K. Scott² (kscott@cc.umanitoba.ca)

¹ Department of Fisheries and Oceans, Freshwater Institute, Winnipeg, Canada.

² Department of Microbiology, University of Manitoba, Winnipeg, Canada.

³ Department of Geography, McGill University, Montreal, Canada.

⁴ Department of Entomology, University of Manitoba, Winnipeg, Manitoba, Canada.

⁵ Tetra Tech Inc., Toronto, Canada.

Contamination of fisheries by Hg is a widespread problem in natural boreal forest lakes, and in manmade reservoirs, located in the boreal region and elsewhere. We have studied the production and cycling of methyl mercury (MeHg) in natural and flooded Boreal Ecosystems at The Experimental lakes Area (ELA), northwestern Ontario, over the last five yrs. Concentrations of total Hg (THg) and MeHg at ELA are low in atmospheric deposition as compared with other sites. Thus, our study was conducted in a relatively pristine site.

It was found that wetlands in the natural state are sites of MeHg production and that wetlands can be important sources of MeHg to lakes, especially in wet years, when inflow to lakes from wetlands is high. In contrast, inputs of MeHg to lakes from upland areas are very low and upland areas are sites of MeHg storage or destruction, relative to inputs from wet deposition. Sediments of lakes are important sites of MeHg production. The relative importance of in-lake versus wetlands as sources of MeHg to lake systems varies according to the percentage of wetland in the lake basin and the year-to-year variation in volume of atmospheric deposition, which controls the mass inflow of MeHg from wetlands. Natural lakes are strong sinks for MeHg (i.e. even though MeHg is produced in lakes, much less MeHg flows out of lakes than flows in). Lakes are MeHg sinks because: 1) MeHg is stored in sediments and 2) photodegradation of MeHg in the surface water is rapid. In clear water lakes where light penetration is high, photodegradation of MeHg is equal to or greater than the quantity of MeHg flowing into lakes from wetlands.

A wetland was experimentally flooded to simulate the impact of reservoir construction on MeHg production in boreal forest ecosystems. In its natural state, the experimental wetland was a site of MeHg production, but after flooding, rates of MeHg production increased by 30 fold. and concentrations of MeHg in water increased about 10 fold to a maximum of 3 ng L^-1. After flooding, in contrast to undisturbed ELA lakes, the lake within the experimental wetland became a large downstream source for MeHg and THg.

Food chain organisms and fish responded quickly to the elevated concentrations of MeHg in the ecosystem. Efforts were made to quantify the movement of MeHg into selected food chain organisms so that a predictive mathematical model of mercury cycling in reservoirs could be developed. For example, it was found that the log BAF for zooplankton did not change after flooding (and its was similar to other unflooded ELA lakes) so that the concentration of MeHg in zooplankton should be reasonably straightforward to predict from MeHg concentrations in particulate and water. Fish Hg bioaccumulation increased 3 fold after flooding and varied from year-to-year based on climate conditions. It was also demonstrated, using in situ low-level Hg techniques, that fish accumulate Hg almost exclusively from food as compared to water.

All of this food-chain information as wells as information of biogeochemical cycling of Hg in unflooded and unflooded boreal ecosystems has been incorporated into the reservoir Hg model and the model is presently being calibrated using data from the experimental reservoir. In future it will be applied to large man made reservoirs to try to predict the severity of contamination of fisheries in reservoirs before they are constructed and to predict the duration of Hg contamination problem.