Development of a SWAT+ model to improve conservation practice efficacy for protecting stream health in karst pasturelands

Authors: MOUSER, JOSHUA - Department Of Fish And Wildlife; CIPARIS, SERENA - Us Fish And Wildlife Service; White, Michael; ANGERMEIER, PAUL - Us Geological Survey; CZUBA, JONATHAN - Virginia Tech

Submitted to: Journal of the American Water Resources Association
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/24/2026
Publication Date: 3/23/2026
Citation: Mouser, J.B., Ciparis, S., White, M.J., Angermeier, P.L., Czuba, J.A. 2026. Development of a SWAT+ model to improve conservation practice efficacy for protecting stream health in karst pasturelands. Journal of the American Water Resources Association. 62. Article e70104. https://doi.org/10.1111/1752-1688.70104.
DOI: https://doi.org/10.1111/1752-1688.70104

Interpretive Summary: In areas with karst landscapes—where water flows mostly underground—agriculture can have a big impact on stream water quality. We used a SWAT+ hydrologic model to predict which conservation practices would work best in a pasture-heavy area of southwest Virginia. Underground water flow was the main way pollution moved, and much of the sediment in streams came from eroding streambanks, not farm fields. To improve water quality, fencing cattle out of streams and managing fertilizer use were the most effective.

Technical Abstract: Placing appropriate conservation practices in the proper locations can benefit stream health, especially in karst environments that are susceptible to agricultural pollution. Watershed models such as the Soil and Water Assessment Tool+ (SWAT+) can be used to optimize both practice selection and placement in the landscape, but to our knowledge, no studies have tested this application in karst pasturelands. We built a SWAT+ model for a watershed in southwest Virginia within the southern Appalachians — an area with extensive pasturelands and karst topography. SWAT+ predicted streamflow estimates were most affected by the available water capacity and hydraulic conductivity of the soil, revealing that water, and associated pollutants, are moving primarily through subsurface pathways due to the karst topography of the region. Predicted sediment yield was negatively associated with the extent of agricultural land cover and were strongly influenced by channel erodibility — indicating that the predominant source of sediment may be from streambanks rather than upland fields. The model unsatisfactorily predicted total nitrogen and total phosphorus loads. Based on the pollutant dynamics revealed by SWAT+, practices that stabilize stream banks (e.g., fencing cattle out of streams) would be most effective at reducing sediment loads, and practices that stop pollutants at their upland sources (e.g., nutrient management plans) would be most effective at reducing instream nitrogen. The utility of SWAT+ for cattle grazing operations in karst regions could be improved by more accurately representing effects of cattle grazing on streambank erosion and nitrogen and phosphorus movement through the environment.