Spatiotemporal evolution of ecosystem health and its spatially non-stationary response to human activities in valley-type cities under topographic constraints

Ecological degradation is an increasingly prominent global challenge. Particularly in topographically constrained urbanized areas, intensive human activities pose a severe threat to ecosystem health. Using Lanzhou, a typical valley city in the arid region of Northwest China, as a case study, this study evaluated the spatiotemporal evolution of ecosystem health and its spatially non-stationary response to human activities from 2000 to 2020. A localized, high-resolution Vigor-Organization-Resilience-Ecosystem Services (VORS) framework was constructed to quantify the ecosystem health index (EHI), and human activity intensity (HAI) was characterized using human footprint datasets. Subsequently, bivariate spatial autocorrelation analysis and a geographically weighted regression (GWR) model were employed to reveal the interactive mechanisms between human activities and ecosystem health. The results indicated the following: (1) The EHI in Lanzhou initially declined but subsequently recovered, presenting a spatial polarization gradient characterized by high values in the peripheral mountains and low values in the central valley; meanwhile, HAI increased continuously, expanding linearly along the valley axis. (2) A significant and strong global negative spatial correlation existed between EHI and HAI, with this correlation initially strengthening and then weakening over time. (3) Under the strict constraints of the “two mountains flanking a river” topography, the bivariate spatial clustering was dominated by antagonistic spatial patterns of Low EHI-High HAI and High EHI-Low HAI. (4) The impacts of human activities on ecosystem health demonstrated a significant non-linear valley-mountain spatial differentiation. In the topographically restricted valley, high-intensity human activities exerted a severe negative stress effect, whereas the peripheral mountains exhibited a “positive synergistic effect”. Notably, after 2010, the spatial agglomeration of negative stress areas decreased, indicating a disintegration of the stress pattern, while positive synergistic zones expanded into the valley-mountain transition zone. This study enriches the fine-scale app lication of the VORS framework, providing a scientific basis for formulating differentiated, sustainable spatial management strategies in topographically constrained regions worldwide.