Abstract: Exploring the spatiotemporal changes of landscape ecological risk and assessing the ecological risk situation is important for maintaining regional ecological security and sustainable economic and social development. Landscape pattern index, spatial autocorrelation analysis and geodetector analysis methods were used in this study to explore spatial-temporal variation characteristics and factors influencing landscape ecological risk in Jiangjin District of Chongqing at the optimal spatial scale based on remote sensing images in 2000, 2009 and 2018. The results reveal that: (1) With changes in grain size and extent, different regularities were exhibited in landscape indexes. A pixel size of 90 m and a grid size of 3 km better reflected the spatial heterogeneity of landscape patterns; (2) Spatial patterns of landscape ecological risk in the study area changed substantially, the overall threat degree of landscape ecological risk increased and the medium-risk area gradually became the dominant type of the region from 2000 to 2018; (3) There was marked spatial concentration of landscape ecological risk in each period of the study area, and the spatial distribution of the landscape ecological risk index mainly exhibited a high-high aggregation pattern and a low-low aggregation pattern; (4) Human activity had an increasingly strong impact on regional landscape patterns and ecosystems. Population density and gross domestic product were important driving factors affecting the spatial and temporal distribution of landscape ecological risks. These results could provide decision-making references for the construction of important ecological barriers in the upper reaches of the Yangtze River.

Key words: scale effect, landscape pattern, landscape ecological risk, spatial autocorrelation, geodetector, Jiangjin District