Revealing scale-dependent imbalance between landscape ecological risk and ecosystem services: Multi-scale analysis and driving mechanisms.

Unreasonable exploitation of natural resources has caused the degradation of ecosystem services (ESs) and elevated landscape ecological risks (LER). Existing studies have mostly focused on the imbalanced relationships between ESs and human well-being, while insufficient attention has been paid to the correlations between LER and ESs. This study developed an integrated multi-scale analysis framework. Taking Henan Province as a case study and considering globality, spatiotemporal dynamics, and topography, this study selected four scales-1 km grid, county, sub-watershed, and topographic gradient-to analyze the interrelationships between LER and ESs. The eXtreme Gradient Boosting-SHapley Additive exPlanations (XGBoost-SHAP) method was employed to explore their driving factors. The results show that: (1) Both LER and ESs exhibited declining trends from 2000 to 2023. The area of the lowest LER increased at an average annual rate of 4.71%, while the lower LER class dominated, covering 67.78% of the total study area. ESs were generally higher in the southwest and lower in the northeast. (2) Across multiple scales, the relationship between LER and ESs exhibited a significant negative correlation. The relationships at the 1 km grid, county, and sub-watershed scales fluctuated but varied in magnitude. At the topographic gradient scale, a synergistic correlation was observed in low-altitude areas, whereas an imbalance emerged in high-altitude areas. (3) Among the key factors influencing their correlation, natural factors exerted stronger effects than social factors at the 1 km grid, sub-watershed, and topographic gradient scales. However, social factors played a more prominent role at the county scale. This study clarifies the correlation mechanism between LER and ESs and provides scientific support for regional ecological management and policy formulation.