Clarifying the spatio-temporal evolution characteristics of eco-environment quality (EEQ) under land use/cover change (LUCC) and its coordinated relationship is of great importance for formulating reliable environmental protection strategies and measures to promote regional sustainable development. Most studies have emphasized the importance of LUCC for regional ecological quality. However, deeply unraveling the complex interrelationships between them remains a significant challenge, particularly in ecologically fragile regions like the Li River Basin. Therefore, based on the historical land use data and the remote sensing ecological index (RSEI) of the Li River Basin from 1990 to 2020, we analyzed the spatio-temporal evolution characteristics of EEQ and LUCC, and explored the influences and non-linear effects between them by using the bivariate spatial autocorrelation and XGBoost model. The key findings are as follows: (1) Land use/cover (LUC) in the Li River Basin was predominantly characterized by forestland and cropland, which together accounted for approximately 97% of the region. The interconversion between forestland and cropland represented the primary form of regional LUCC, while built-up land demonstrated a growth trend by encroaching on cropland. (2) The EEQ exhibited a volatile upward trend within the research period, with an average RSEI value of 0.5891, indicating a generally favorable ecological condition. (3) A significant negative spatial correlation was observed between land use intensity (LUI) and the RSEI, characterized by H–L, L–H, and non-significant clusters. (4) There was a distinct non-linear relationship that existed between LUCC and the RSEI, underscoring that appropriately regulating regional land use scale can help maintain ecological balance. These findings provide a scientific basis for optimizing land spatial management models and formulating policies to improve ecological environment quality, while also offering a new framework and reference for further ecological research on EEQ influencing factors and driving mechanisms.

Rapid urbanization and tourism-driven economic development have accelerated land-use/land-cover change (LUCC), reshaped landscape structure and increased landscape ecological risk (LER). Unraveling the spatiotemporal evolution of LER and its drivers is critical for regional ecological restoration and sustainable development. Using LUCC data for the Li River Basin (LRB) during 2000–2020, this study constructed an LER assessment model based on landscape pattern indices and employed an XGBoost–MGWR framework to identify key natural and socioeconomic drivers, as well as their nonlinear effects and spatial mechanisms. The results showed that: (1) forestland and cropland dominated the basin throughout the study period. Cropland and built-up land expanded, whereas forestland, grassland, and water areas contracted, with significant mutual conversions between forestland and cropland. (2) Relatively low- and middle-risk areas dominated the LER structure and exhibited a spatial pattern of “higher in the center and lower in the periphery”. Middle- and relatively high-risk areas expanded, while low and high-risk areas contracted, indicating a shift toward middle to relatively high risk levels. (3) Land-use intensity (LUI) and the digital elevation model (DEM) were the core drivers of LER change, while tourism intensity (TRI) showed an increasing influence over time. These findings provide a scientific basis for regional ecological management and sustainable development.