Modeling Rangeland Watershed Erosion Processes

Most common soil erosion models were developed for agricultural conditions and then modified for rangeland, forest, and other applications. Rill and interrill erosion generally dominate upland agricultural erosion on recently tilled fields. In rangelands, soil is not tilled, rainfall is sporadic, infiltration is high, and the soil surface is covered by vegetation residues and sometimes rocks. Recent research has shown that as rangeland plant species change from the tall grass species that once dominated many plant communities, to short grass species through overgrazing, the hydraulic conductivity of the soil decreases and runoff increases. The net impact of the rangeland surface and plant changes is that upland erosion may be low, and the dominant source of sediment from rangelands may be from channels. This paper describes a study in southeastern Colorado on sedimentation from rangeland watersheds. The observed rates are compared to predicted erosion rates from a locally developed erosion model, RUSLE, the WEPP model with typical rangeland vegetation and soil scenarios, and the rangeland predictions from the Forest Service WEPP interface and database. The local regression model and the Forest Service WEPP interface best predicted sedimentation from small rangeland watersheds. RUSLE overpredicted sediment yields by a factor of 3, and the WEPP rangeland predictions were 1/6 the observed sediment yields. In a sensitivity analysis with the FS WEPP interface, as cover increases, differences in runoff and erosion between soils decreases.