Exploring the relationship between soil erosion and hydraulic parameters on slopes with clipped herbaceous vegetation through Bayesian network modeling

Abstract Clipping management, which alters vegetation conditions by removing aboveground vegetation while preserving the underground root system, directly impacts hydrological processes and flow dynamics, ultimately affecting soil erosion processes. This study conducted field experiments (2 × 6 m) encompassing four clipping intensities (with vegetation coverage of 70%, 50%, 25%, and 0%), three rainfall intensities (60, 90, and 120 mm·h −1 ), and three slope gradients (10°, 20°, and 30°). The results showed that clipping treatments significantly reduced the Darcy–Weisbach resistance coefficient ( f ). Elevated levels of clipping intensity, rainfall intensity, and slope gradient led to increased flow velocity ( v ), Reynolds number ( Re ), and stream power ( ω ). The shear stress ( τ ) exhibited an upward trend with rising slope and rainfall intensity, with no significant difference in the effect of clipping intensity. Variation partitioning analysis demonstrated that the influence of steep slopes amplified the effects of rainfall intensity while diminishing the impact of clipping intensity on v values. Critically, the Bayesian network model suggested that, in comparison to the influence of raindrop splashing, soil detachment and transportation by runoff were the predominant driving mechanisms for erosion on slopes with clipped vegetation. Rainfall intensity and clipping intensity were found to raise the runoff depth and enhance the v value, which ultimately influenced soil erosion. Overall, these results contribute to an improved comprehension of the hydraulic dynamics of soil erosion on slopes with clipped vegetation and provide invaluable perspectives for developing effective clipping management strategies to ensure sustainable land utilization and judicious water resource governance.