Integrating and mapping spatial heterogeneity of the ecosystem: a novel approach to indicating environmental threats

The spatial characterization of ecosystems and environmental threats has garnered significant attention. However, quantitative analyses of the spatial matching relationship between these threats and ecosystem landscape patterns are rarely explored. This research examines the spatiotemporal changes in ecosystems and their landscape indices, evaluates the sensitivity and spatial distribution of major environmental issues: soil erosion, land desertification, and rocky desertification, and reveals the relationship between ecosystem changes and environmental threat sensitivity through spatial autocorrelation analysis. This study proposes a Landscape Adjustment Index (LAI) based on environmental threats using XGBoost. Our findings demonstrate evident spatial heterogeneity in the matching effects between ecosystem changes and environmental threats, with varying responses. The R² values for LAI and three environmental threats were 0.2632, 0.4670, and 0.1364, respectively, highlighting significant differences. LAI exhibited a distinct spatial differentiation pattern, characterized by "low in the west and high in the east" relative to the Hu-line. The ecological zones of the Loess Plateau and the Hengduan Mountains predominantly fell into the "high" and "relatively high" categories, aligning with areas highly sensitive to soil erosion. Conversely, the relatively low and low categories were mainly concentrated in the Inland Northwest area, Inner Mongolia Plateau, and the western part of the Qinghai-Tibet Plateau, demonstrating clear spatial relationships with vulnerable land degradation areas. The areas highly sensitive to rocky desertification coincide with regions of high LAI values. This study explores the relationship between environmental threats and the static and dynamic evolution of ecosystems, proposes LAI based on environmental threats, identifies characteristic driving patterns for different threats, and presents a multi-threat prevention strategy. These findings provide empirical support for ecological restoration efforts in China.