• 研究报告 •

### 基于GeoDA的甘肃白龙江流域景观破碎化空间关联性

1. （兰州大学资源环境学院/西部环境教育部重点实验室， 兰州 730000）
• 出版日期:2018-05-10 发布日期:2018-05-10

### GeoDAbased spatial correlation analysis of landscape fragmentation in Bailongjiang Watershed of Gansu.

ZHANG Jin-xi, GONG Jie*, MA Xue-cheng, LIU Dong-qing

1. (Key Laboratory of Western China’s Environmental Systems (Ministry of Education)/College of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, China).
• Online:2018-05-10 Published:2018-05-10

Abstract: Landscape fragmentation is one of the research hotspots in landscape ecology. The spatial distribution of landscape fragmentation can provide scientific basis for landscape pattern optimization. In this study, we analyzed the spatial correlation of landscape fragmentation of Bailongjiang Watershed in Gansu in 2014 using landscape pattern index, grid analysis, and spatial autocorrelation methods based on GeoDA and ArcGIS. The results showed that the Moran’s Ivalues of the landscape pattern indices, i.e. edge density (ED), contagion index (CONTAG), and Shannon’s diversity index (SHDI) under different grid sizes ranged from 0.478-0.501, 0.276-0.374, 0.406-0.436, respectively, indicating a stronger spatial positive correlation and a landscape fragmentation with the agglomeration. Taking 10 km×10 km as the characteristic scale to examine the spatial agglomeration effect of landscape fragmentation, we found that the areas with higher landscape fragmentation degrees were mainly distributed in the middle of Wudu District, and that the areas with lower landscape fragmentation degrees were located in southern Wenxian. Moreover, the Moran’s I value of human activity intensity with ED was 0.170, and the Moran’s I value of human activity intensity with SHDI was 0.180, with positive spatial correlation. However, the Moran’s I value of human activity intensity with CONTAG was -0.095, with negative spatial correlation, highlighting the role of human activity intensity in driving the spatialdistribution characteristics of landscape fragmentation. Our results suggest that GeoDA can be applied to landscape ecological spatial analysis.