A new method to estimate the cover and management factor for soil loss prediction on the Loess Plateau in China: A case-study using a soybean field

The cover management factor (C-factor) has become one of the most sensitive and complex parameters in the revised universal soil loss equation (RUSLE) because of its sensitivity to anthropogenic activities. However, due to the regionally limited scope of conventional models, direct applications of the C-factor prediction have been hindered in China. Alternative C-factor estimation approaches often consider vegetation coverage alone. This can decrease the accuracy of the estimation to an extent. In this study, several cropping systems were established to investigate the coupling effect of surface changes under different management practices and crop cover in preventing soil loss at the plot scale. Five experimental (CP, cropped; SSR, roughened; SC, crusted; CP-SSR, cropped-roughened; CP-SC, cropped-crusted) plots were set, compared to bare plot (BP). The impact of cropping systems on soil loss was estimated by SLR (soil loss ratio), according to the universal soil loss equation (USLE). The SLR estimation models were provided via identifying the related sub-factors, subsequently, combined with the distribution curve of rainfall erosivity to allow the novel C-factor estimation. The results showed that compared with SLR_CP, the decrease in SLR_{CP − SSR} varied from 12% to 43% with a mean of 25.50%; and the decrease in SLR_{CP − SC} varied from 24% to 51% with a mean of 34.75%. Crop coverage, plant height, root weight density, < 0.5 mm root length density (both at 0–5 cm depth), initial roughness, and initial crust thickness were significantly correlated with SLR. For cropped plot, SLR_{CP − estimated} can achieve more accurate simulation result than conventional models obtained from literature findings, with RMSE (root mean square error) coefficient of 0.17. For a comprehensive understanding of the C-factor, multiple crops should be considered, and more experiments in the experimental group to better verify SLR_{CP − SSR − estimated} and SLR_{CP − SC − estimated} are recommended.