Modeling the Impact of Changing Land Use and Vegetative Cover on Hydrology, Nutrient, and Sediment Loads from an Agricultural Catchment of Lake Michigan (USA)

High nutrient loads are an indicator of pollution sources in a watershed that need to be identified and quantified. These loads in surface and groundwater have been a major concern that impacts water quality in the Midwestern US, including the Great Lakes Basin. To investigate the influence of land use change, especially at urban/rural interfaces, we used the Soil and Water Assessment Tool (SWAT) to model water, sediment, and nutrient export for the St. Joseph River Basin (SJRB), which drains an area of 12,200 km² in Southwest Michigan/Northwest Indiana and enters Lake Michigan. The SWAT models were built, calibrated and validated for monthly streamflow, groundwater, total suspended solids (TSS), total nitrogen (TN), total phosphorous (TP), nitrate (NO₃-N) and dissolved reactive phosphate (DRP; as orthophosphate), using two stream gages (Niles, USGS ID 04101500; Paw Paw, USGS ID 04102500) in Berrien County, MI. We found that monthly hydrology, sediments, and inorganic nutrients were well captured by the model with very good to excellent performance at the Niles gage, and, good to satisfactory performance at Paw Paw. The simulated average annual groundwater was 137 mm and 129 mm for Niles and Paw Paw, respectively, suggesting that on average 57% and 60% of long-term streamflow in the basin comes from groundwater and shallow subsurface flow. For water quality variables, TSS loads were strongly correlated with streamflow with R² reaching 0.90. Using this model, we investigated how land use change (e.g., agriculture), and the planting of winter cover crops in the fallow season would impact water and nutrient yields from the SJRB. We found that the impact of changing land use and applying cover crops on water quality components was significant and dependent on the selected spatial scale. The simulated outputs indicated that cover crops have no impact on hydrology but significantly reduced DRP and NO₃-N export to Lake Michigan by up to 30% and 50%, respectively. Application of this model will assist regional land and water managers in planning for future impacts of land use and climate change and their impacts on water quality and quantity, enabling stakeholders to implement conservation practices to sustain the SJRB and other similar basins in the Great Lakes region.