Spatiotemporal Heterogeneity in Runoff and Sediment Reduction by Vegetation Restoration Under Extreme Rainfall

ABSTRACT In recent years, frequent extreme rainstorm events have increasingly emerged as the primary external driver of erosion in the Loess Plateau, while vegetation restoration combated the erosive energy of these extreme storms by mitigating rainfall runoff processes and reinforcing soil through erosion resistance. While previous studies have focused on runoff and sediment production processes on slopes with varying vegetation types, cover and spatial patterns, the dynamic role and main driving mechanisms at different stages of natural vegetation during extreme rainfall remained understudied. Therefore, this study examined typical fallow slopes on the Loess Plateau, through simulated scouring experiments; we quantified the runoff and sediment reduction effects at distinct stages of succession and identified shifts in the main drivers of sediment production. Over a 25th year successional period, vegetation exhibited the greatest runoff reduction impact at 11th years (average peak runoff reduction rate of 25%) and the highest sediment reduction at 25th year (average sediment peak reduction rate 88%), indicating temporal heterogeneity of flow and sediment reduction patterns. Increased flushing discharge diminished vegetation's capacity in runoff reduction while enhancing its sediment control efficacy. On slopes with 1st year of succession, the correlation between cumulative denudation (Dr) and cumulative runoff energy consumption (Ec) showed consistent variation, with fitted coefficients of determination above 0.96. This correlation weakened as succession progressed, driven by increased vegetation diversity and abundance. Initially, the independent explanation of the dynamic factor was 0.91, while soil and vegetation factors increased by 212% and 68%, respectively. The final successional stage. The final successional stage highlighted the nonlinear regulation of vegetation‐soil coupling (soil × vegetation = 0.21) and the higher‐order interaction of the three factors (dynamic × soil × vegetation = 0.27). This study assessed the erosion control efficacy of natural vegetation restoration under extreme events from the time dimension, and clarified the dominant controlling factors at each successional stage. The results provided valuable scientific insights for optimizing soil and water conservation strategies.