Threshold Effects of Hydrothermal and Human Activities on Vegetation Dynamics in the Forest-Steppe Ecotone of the Da Hinggan Mountains

The forest-steppe ecotone is a critical transitional ecosystem that is highly sensitive to climate change and anthropogenic disturbances. Despite large-scale ecological restoration efforts over the past two decades that have substantially improved regional vegetation conditions, the threshold responses of vegetation to hydrothermal conditions and anthropogenic disturbances remain poorly quantified. This knowledge gap has limited our understanding of ecosystem stability and constrained the development of adaptive management strategies for fragile ecological transition zones. Based on kNDVI data (2000–2020) derived from the Google Earth Engine platform, this study systematically examined spatiotemporal vegetation dynamics, driving mechanisms, threshold effects, and future trajectory in the forest-steppe ecotone of the Da Hinggan Mountains in Northeast China. The results revealed an overall greening trend, with natural factors exerting a stronger influence than anthropogenic activities. Importantly, both hydrothermal and anthropogenic factors exhibited significant nonlinear thresholds that regulated ecological transitions from greening to browning. For forests, precipitation and temperature thresholds were identified at 386 mm and -0.6°C, respectively, whereas the corresponding thresholds for grasslands were 308 mm and 0.4°C. Anthropogenic thresholds for livestock density were identified at 24 SU/km²for forests and 90 SU/km²for grasslands. Future projections based on CMIP6 (Coupled Model Intercomparison Project Phase 6) scenarios suggest that continued warming may increase the risk of vegetation browning, particularly under high-emission scenarios. This study highlights the pivotal role of climatic and anthropogenic thresholds in regulating ecosystem stability and provides a scientific foundation for adaptive management and vegetation restoration in fragile ecological transition zones.