Summary/Abstract
Grassland resilience is crucial for sustaining its ecosystem functions. Although ecosystem resilience is profoundly influenced by climate change and grazing intensity, its dominant drivers remain unclear, limiting the understanding of ecosystem resilience. This study examined resilience dynamics (2001−2021) across the Chinese grassland using remote sensing datasets and machine learning to disentangle the impacts of climate change and grazing. The results revealed a significant increase in 1-lag temporal autocorrelation (TAC), indicating a general decline in China’s grassland resilience, most notably on the Tibetan Plateau. Sensitivity analysis showed that climate warming and intensified precipitation variability exerted negative impacts on resilience across approximately 79% and 82% of the grasslands, respectively, though this impact was partially buffered by increases in vegetation density. Grazing also caused resilience to decline across 62% of the study area, exhibiting greater sensitivity to increases in grazing pressure than to comparable decreases. Critically, the contribution analysis found that climate change dominated the resilience decline, contributing approximately 8.39 times more than grazing. These findings underscore the sensitivity of ecosystem stability to climate pressures, suggesting that management should prioritize climate adaptation and mitigation strategies.