Spatiotemporal Heterogeneity of Ecosystem Service Interactions and Their Drivers: Implications for Spatial Management

Investigating the heterogeneity of ecosystem services (ESs) and their interactions plays a crucial role in refining spatial management. However, little is known about the spatiotemporal heterogeneity of ES interactions. This study quantified the spatiotemporal variations in ES interactions in Northeast China. It categorized different ES bundles through self-organizing maps and ranked the importance of drivers by SHapley Additive exPlanations (SHAP), thereby proposing targeted spatial management strategies. The results indicated that between 2000 and 2020, most ES pairs exhibited synergies, whereas a few ES pairs exhibited trade-offs. And the synergistic areas for habitat quality (HQ)–net primary productivity (NPP), HQ–soil retention (SR), and NPP-SR accounted for 67.79%, 65.29%, and 75.69% of the study region, respectively. Between 2000 and 2020, the area covered by the key synergetic bundle and habitat quality bundle first increased and then decreased. SHAP analysis indicated that NPP, SR, and water retention were most influenced by biophysical indicators, while HQ was primarily influenced by anthropogenic indicators. Based on research findings regarding ES bundles and their drivers, refined strategies for ecosystem management have been proposed. This study integrated knowledge of ES interactions into spatial management, providing a basis for maximizing ES benefits and addressing future climate change.