Objective Soil erosion threatens ecological security and alters key ecosystem processes. Sediment connectivity serves as a vital tool for quantifying sediment transport efficiency and supporting soil and water conservation decision-making. However, existing research lacks systematic comparison of the representational performance of different connectivity indices, and quantitative analysis of long-term contributions of key drivers, such as vegetation and rainfall, remains insufficient. Therefore, this study aims to evaluate the effects of vegetation coverage and rainfall erosivity on sediment connectivity in the Longchuan River Basin and to identify the index that best characterizes sediment transport rate potential.

Methods This study was based on a 35-year dataset (1986−2020) from the Longchuan River Basin. Three sediment connectivity indices were employed: the sediment connectivity index (IC) based on the cover-management factor (C), the IC_R index based on rainfall erosivity (R_e), and a revised sediment connectivity index (IC_r) coupling both factors. Regression analysis and spatial correlation analysis were used to clarify the effects of vegetation coverage and R_e on sediment connectivity. The relative contributions of vegetation coverage and R_e to variations in IC_r were quantified using Lindeman-Merenda-Gold (LMG) analysis. Correlations between each index and measured sediment transport rates were analyzed to evaluate their representational performance.

Results 1) During the study period, the normalized difference vegetation index (NDVI) of the river basin exhibited a significant increasing trend, while the IC index showed a significant decreasing trend. A highly significant negative correlation was observed between their temporal variations (R² = 0.99), indicating that the improvement in vegetation coverage effectively reduced sediment connectivity in the river basin. Spatially, vegetation coverage was relatively lower in the northeastern part of the river basin, where IC values were correspondingly higher. Spatial correlation analysis further revealed a negative correlation between vegetation coverage and IC. 2) Both the IC_R index and R_e displayed pronounced interannual fluctuations without a significant long-term trend, and they showed strong positive correlations in both temporal and spatial dimensions. 3) The IC_r index demonstrated a significant decreasing trend, indicating an overall reduction in sediment connectivity in the river basin. Its variations were primarily controlled by vegetation coverage, which accounted for 97.10% of the contribution, while R_e contributed 2.90%. 4) Both the annual mean and annual maximum sediment transport rates in the river basin exhibited significant decreasing trends. The correlation coefficients between the IC_r index and the annual mean sediment transport rate (R = 0.55), as well as the annual maximum sediment transport rate (R = 0.39), were higher than those obtained using the IC and IC_R indices.

Conclusions This study successfully achieves its predefined objectives. At the interannual scale, vegetation restoration is the dominant factor inhibiting sediment connectivity and reducing transport potential in the Longchuan River Basin, whereas R_e mainly acts as a driver of interannual variability without forming a long-term trend. Among the three indices, the IC_r index, which simultaneously considers the inhibitory effect of vegetation and the driving effect of rainfall, more accurately represents the actual sediment transport potential of the river basin. These findings provide a quantitative basis for targeted soil erosion prevention and control in mountainous areas of Southwest China.