Predicting mercury concentrations and fluxes in the water column and sediment of lakes with a limited dataset.

The purpose of the present study was to evaluate the mercury (Hg) Environmental Ratios Multimedia Ecosystem Sources (HERMES) model on two Ontario, Canada lakes (Harp and Dickie) and to include modifications to enable the model to estimate the major model input variables that tend to be missing for lakes with limited datasets. No significant differences were found for either sediment solid or bulk water total mercury (THg) when the HERMES model was applied to the two Ontario lakes, regardless of whether all available data were altered during application or only the 10 variables that tend to cause the most variation in model output (i.e., concentration of THg in atmosphere, water inflow THg concentration, water inflow rate, water volume, surface area, mean depth, suspended particulate matter concentration, settling rate of solids in water column, water temperature, and precipitation rate). Since measured sediment and water THg values do not exist for most lakes removed from industrial activities, empirical relationships were incorporated into the HERMES model framework to provide a method to double-check model output for lakes where this information is unavailable.

Temporal analysis of net fluvial methylmercury loading in a dystrophic and a clear water lake.

The concentration of methylmercury (MeHg) in aquatic ecosystems is the net result of the highly dynamic abiotic and biotic processes of mercury methylation and demethylation. In this study, we conduct an examination of the net fluvial loading of methylmercury (MeHg(Net)=MeHg(Watershed)-MeHg(Lake outflow)) across a 3 year time frame in both a dystrophic lake and an oligotrophic lake. A significant portion of MeHg(Net) variance in both lakes could be attributed to a seasonal pattern (11.4%, p=0.009; oligotrophic, and 27.0%, p<0.0001; dystrophic) which in both cases, was most correlated with air temperature. The dystrophic lake appeared to be a net source of methylmercury (MeHg(Net)=-1.9+/-0.3 mg MeHg d(-1)) while the oligotrophic lake appeared to be a net sink (MeHg(Net)=0.4+/-0.2 mg MeHg d(-1)), indicating that there was net methylation in the dystrophic lake and net demethylation in the oligotrophic lake. Higher MeHg loading to the lakes occurred during the summer and between seasons there was a difference in MeHg(Net) of 1.1+/-0.3 mg MeHg d(-1) and 3.1+/-0.6 mg MeHg d(-1). Seasonal patterns of MeHg(Net) in the oligotrophic lake lagged behind the dystrophic lake by 39 days. The short term variation in MeHg(Net) was dominated by precipitation (t=2.73, p=0.008; dystrophic, t=2.53, p=0.017; oligotrophic).