Influence of abiotic and biotic factors in measuring and modeling soil erosion on rangelands: state of knowledge.

Abstract

The first standardized soil erosion prediction equation used on rangelands was the Universal Soil Loss Eguation (USLE). The Revised Universal Soil Loss Equation (RUSLE) was developed to address deficiencies in the USLE by accounting for temporal changes in soil erodibility and plant factors which were not originally considered. Improvements were also made to the rainfall, length, slope, and management practice factors of the original USLE model. The Water Erosion Prediction Project (WEPP) model was developed to estimate soil erosion from single events, long-term soil loss from hillslopes, and sediment yield from small watersheds. Temporal changes in biomass, soil erodibility, and land management practices, and to a limited extent, spatial distribution of soil, vegetation, and land use are addressed in the WEPP model. To apply new process-based erosion prediction technology, basic research must be conducted to better model the interactions and feedback mechanisms of plant communities and landscape ecology. Thresholds at which accelerated soil erosion results in unstable plant communities must be identified. Research is needed to determine the confidence limits for erosion predictions generated by simulation models so that the probability of meeting specified soil loss values (kg ha-1 yr-1) for given management systems can be calculated at specific significance levels. As the technology for modeling soil erosion on rangelands has improved, limitations with the techniques of parameter estimation have been encountered. Improvements in model parameterization techniques and national databases that incorporate vegetation and soil variability are required before existing erosion prediction models can be implemented.