Network cascading effects reveal thresholds and nonlinearity in ecological vulnerability

Anthropogenic land-use change is a primary driver of global ecosystem degradation. However, its systemic impact on ecological vulnerability across complex basin networks remains poorly quantified. A complex network framework was developed in this study to assess how land-use transitions influence ecological vulnerability in nine major river basins of China and the nation as a whole. A directed transition network of 26 land-use types for 2020–2023 was constructed and cascading failure simulations with changes in ecosystem service value change and ecological load entropy (ELE) were integrated together to quantify vulnerability. This approach reveals a nonlinear response of ecological vulnerability to land-use perturbations. Crop dryland, medium coverage grassland, and forest consistently emerge as critical hubs in both the land-use transition network and for vulnerability indicators, maintaining systemic stability. The results show that reallocating cropland generally increases ecosystem service value (ESV) at low disturbance levels before reaching a stability threshold, whereas disturbances to hydrological and forest land-use types cause the most severe losses. In arid regions, however, conversion of desert related land-use types initially increases ESV before declining after the threshold is exceeded. Analysis of recent changes (2020–2023) indicates that, on average, land-use transitions have improved ESV but have increased exposure to cascading failures in critical ecological nodes. A national stability threshold for medium coverage grassland at 0.5, alongside thresholds for other key land-use types across the basins were identified in this study. These findings provide a mechanistic and network perspective on land system vulnerability, offering scalable insights for designing spatially targeted conservation and strategies in China and other rapidly developing regions.