Response of soil detachment capacity to landscape positions in hilly and gully regions of the Loess Plateau

Landscape position may impact soil detachment capacity (D-c) by overland flow through the imposing influences on soil properties and plant roots. Nonetheless, little knowledge existed with regard to the response of D-c to landscape position. Therefore, this study was conducted to detect the response of D-c to landscape positions and identify the primary influencing factors in hilly and gully regions of the Loess Plateau. 540 undisturbed soil samples taken from six landscape positions were subjected to scour in a hydraulic flume to determine D-c. The results demonstrated that D-c of the top ridge was the maximum (1.13 kg m(-2)s(-1)), while D-c of the footslope was the minimum (0.0026 kg m(-2) S-1). The landscape positions significantly influenced D-c, presenting a regularly decreasing trend from the top to the bottom. D-c increased with stream power (omega) by a power function. Power function was better than linear function for predicting of D-c. The partial least square regression analysis showed that soil organic matter, water stable aggregate (WSA), root mass density (RMD), and soil cohesion (Coh) dominated the variations in D-c. Significant negative relationships were detected between these primary factors and D-c. Landscape position affected Coh, WSA and RMD, which in turn controlled the regular change of D-c with landscape position. D-c of six landscape positions could be satisfactorily estimated by omega, WSA, and RMD (r(2) = 0.81; NSE = 0.81). The results are helpful to understand the vertical changes in soil erosion intensity in small watershed scale.