Spatiotemporal Dynamics and Driving Mechanisms of Soil Conservation Services (SCS) in Zhejiang Province, China: Insights from InVEST Modeling and Machine Learning

Abstract

Zhejiang Province, as a key ecological region in southeastern China, plays a vital role in ensuring regional ecological security and sustainable development through its soil conservation services (SCS). Based on remote sensing data, this study employed the InVEST model to evaluate the characteristics of SCS in Zhejiang from 2001 to 2020. Long-term trends were identified using Sen’s Slope and the Mann–Kendall test, spatial autocorrelation was assessed through Moran’s I, the contributions of driving factors were quantified using XGBoost combined with SHAP, and spatial heterogeneity was further explored using Geographically Weighted Regression (GWR). The results indicate that: (1) from 2001 to 2020, SCS exhibited a fluctuating trend of “decline followed by recovery,” with significantly higher values in the western mountainous areas than in the eastern coastal and plain regions; approximately 58% of the area remained stable, while 40% experienced degradation; (2) Spatial autocorrelation analysis showed that areas with strong SCS were concentrated in the western mountains, while low-value areas were mainly distributed in the eastern coastal and urban regions; (3) natural factors contributed the most, followed by climatic and human activity factors; and (4) the GWR model outperformed the OLS model in revealing the spatial variation in the effects of natural and anthropogenic drivers. These findings provide valuable scientific references and decision-making support for ecological conservation, watershed management, and sustainable land use in Zhejiang Province.