This study focused on characterizing air-surface mercury Hg exchange for individual surfaces (soil, litter-covered soil and plant shoots) and ecosystem-level flux associated with tallgrass prairie ecosystems housed inside large mesocosms over three years. The major objectives of this project were to determine if individual surface fluxes could be combined to predict ecosystem-level exchange and if this low-Hg containing ecosystem was a net source or sink for atmospheric Hg. Data collected in the field were used to validate fluxes obtained in the mesocosm setting. Because of the controlled experimental design and ease of access to the mesocosms, data collected allowed for assessment of factors controlling flux and comparison of models developed for soil Hg flux versus environmental conditions at different temporal resolution (hourly, daily and monthly). Evaluation of hourly data showed that relationships between soil Hg flux and environmental conditions changed over time, and that there were interactions between parameters controlling exchange. Data analyses demonstrated that to estimate soil flux over broad temporal scales (e.g. annual flux) coarse-resolution data (monthly averages) are needed. Plant foliage was a sink for atmospheric Hg with uptake influenced by plant functional type and age. Individual system component fluxes (bare soil and plant) could not be directly combined to predict the measured whole system flux (soil, litter and plant). Emissions of Hg from vegetated and litter-covered soil were lower than fluxes from adjacent bare soil and the difference between the two was seasonally dependent and greatest when canopy coverage was greatest. Thus, an index of plant canopy development (canopy greenness) was used to model Hg flux from vegetated soil. Accounting for ecosystem Hg inputs (precipitation, direct plant uptake of atmospheric Hg) and modeled net exchange between litter-and-plant covered soils, the tallgrass prairie was found to be a net annual sink of atmospheric Hg.