Ecological zoning and its response to human activities based on ecosystem health and landscape ecological risk: A case study of the Yellow River Basin

Intensive human activities increase pressure on ecosystems, thereby constraining sustainable development. Developing a scientifically robust ecological zoning framework is therefore essential for environmental protection and effective territorial spatial management. Focusing on the Yellow River Basin (YRB), this study quantified the spatiotemporal dynamics of the Ecosystem Health Index (EHI), Ecological Risk Index (ERI), and Human Activity Intensity Index (HAII) at both sub-watershed and grid scales. Bivariate spatial autocorrelation analysis was employed to examine the spatial dependence between EHI and ERI. Ecological zones were delineated based on their interrelationship, and Random Forest–SHapley Additive exPlanations (RF-SHAP) and Geographically Weighted Regression (GWR) model were used to evaluate their responses to human activities. The results indicate that: (1) from 2000 to 2020, the mean changes in EHI, ERI, and HAII across both spatial scales were + 6.09%, −3.39%, and + 1.72%, respectively; (2) EHI and ERI exhibited a significant negative spatial correlation; (3) at both spatial scales, HAII is predominantly negatively correlated with EHI, while its relationship with ERI is more complex. As a negative indicator within the HAII framework, the Normalized Difference Vegetation Index (NDVI) emerged as the factor most strongly associated with ecological zoning at the grid scale; however, at the sub-watershed scale, both NDVI and the proportion of cropland contributed relatively strongly. These findings provide empirical support for the implementation of differentiated ecosystem management strategies in the YRB.