Critical thresholds and mediating mechanisms of ecological security in Shenyang based on health-risk-service triadic framework

Ecological security (ES) is a core pillar of China's national security and underlies socioeconomic sustainability and human well-being. This study constructed a health-risk-service triadic framework, integrating machine learning with InVEST and Maxent models to explore critical thresholds and mediating mechanisms. Results showed that ecological resilience (ER) and landscape pattern (LP) were dominant driving factors with critical thresholds of 0.6 and 0.2, and their interaction presented a three-tiered continuum to reshape ecosystem service performance. Ecosystem health influenced ecosystem services primarily through direct effects, whereas ecological risk mediated these linkages through both partial mediating and masking effects. In particular, LP and natural disease (ND) exhibited relatively strong mediating roles. Practically, the suburban areas adopted two-phase management under the strategy of background safety maintenance and key window locking. The framework provides a quantifiable tool for sustainable land-use planning and ES management.