Application of SWAT modeling for watershed restoration and protection in a tribal watershed setting

Abstract

Agricultural nonpoint source pollution is one of the major causes of water qualitychallenges across rural watersheds. There are more complexities involved in tribal watersheds settings where tribal sovereignty, fragmented land ownership, and climate vulnerability intersect. This study developed and applied a hydrologic and water quality model using the Soil and Water Assessment Tool (SWAT) to evaluate existing conditions in the Soldier Creek Watershed in northeast Kansas, a watershed where part of the drainage area falls within the Prairie Band Potawatomi Nation (PBPN) reservation boundary. The SWAT model was calibrated and validated for streamflow, sediments, total nitrogen and total phosphorus against the observed data. Model performance for monthly streamflow exceeded NSE of 0.7 for both calibration locations, with satisfactory agreement for sediment and nutrient loads given the sparsity of observed water quality data. Twelve BMP scenarios were evaluated, including filter strips, no-till agriculture, and regenerative agriculture (RA) assumption applied across combinations of PBPN owned and privately owned croplands within the reservation area.

Filter strips applied at the watershed scale produced the largest sediment reductions, reaching approximately 70% at the subwatershed-scale and 15–28% at the PBPN boundary outlet; however, reductions were substantially diminished when BMP application was restricted to PBPN owned parcels, reflecting the constraints imposed by the checkerboard land ownership. No-till as a standalone BMP yielded limited water quality benefits but improved performance when combined with filter strips. The combination of RA and no-till across all croplands within the reservation produced effective reductions in water quality impediments.

Climate change analysis using an 18-model CMIP5 ensemble under RCP 4.5 and RCP 8.5 pathways was then conducted to examine potential impacts on hydrological processes within the tribal area. The ensemble of models projected a consistent warming trend expected to reduce average monthly runoff by 30–40% during the historically wet season by late century (2071-2099), alongside a substantial increase in moderate to extreme drought frequency. This study found that watershed scale water quality improvement within the PBPN requires cooperative BMP adoption by both tribal and non-tribal landowners. The modeling framework developed here provides a practical tool to support conservation planning, prioritize management interventions, and develop long-term water resource resilience in the Soldier Creek Watershed within the PBPN