Rutelonis, Wes2017-11-272017-11-272017-05https://hdl.handle.net/11299/191229University of Minnesota M.S. thesis. May 2017. Major: Earth Sciences. Advisors: Mike Berndt, David Fox. 1 computer file (PDF); viii, 49 pages.The St. Louis River Watershed in Northeastern Minnesota has been studied extensively to determine the degree to which sulfate loading from the Mesabi Iron Range affects microbial methylation and bioaccumulation of mercury. Recent studies have identified natural processes unrelated to mining, most often in non-mining portions of the region, as the primary source of methylmercury loading to the river. Here, we further evaluate those contributions by interpreting water chemistry (DOC, THg, MeHg and Fe) from seven St. Louis River tributaries and three main channel sampling sites with the Riparian Flow-Concentration Integration Model (RIM) which was developed for interpretation of stream chemistry in boreal streams in Sweden. This model assumes that riparian wetland soil, the last substrate that porewater encounters before becoming river water, controls the chemistry of local groundwater recharging the river. In locations that contain mixed mining and non-mining contributions, a watershed model (Hydrologic Simulation Program – Fortran: HSPF) was used to estimate the relative groundwater and point source contributions. The RIM approach with soil temperature incorporated as a time-varying parameter is more physically based compared to regression-based methods that have been used previously to interpret stream loads in the region. The comparison of Nash-Sutcliffe model efficiency calculations for both RIM and regression-based models indicate that RIM offers a significant improvement in model predictive power. Since stream flow and temperature are the main drivers, RIM reduces the necessity for widespread, repetitive methylmercury sampling efforts to estimate methylmercury loads.enDOCHSPFmercurymethylmercuryRIMSt. Louis RiverThesis or Dissertation