Spatial Distribution Patterns of Invasive Alien Species in China
Jing Chen,Fangzhou Ma,Yanjing Zhang,Chenbin Wang,Haigen Xu
DOI: https://doi.org/10.1016/j.gecco.2020.e01432
IF: 4
2021-01-01
Global Ecology and Conservation
Abstract:The spatial distribution of invasive alien species is driven by various factors, and should be comprehensively evaluated based on multiple invasion indices and different taxonomic groups. In our study, we constructed three sets of invasion indices: (1) absolute invasive alien species (AIAS) richness; (2) weighted invasive alien species (WIAS) richness; (3) relative invasive alien species (RIAS) richness. First, we used principal component analysis (PCA) and k-means clustering to distinguish the invasion risk levels of 28 provinces in China. Then, the ordinary least squares (OLS) and partial least squares (PLS) regression methods were used to explore the factors predicting the distribution patterns of three sets of invasion indices for total species, plants and animals. Finally, the residuals’ regressions of alien plant species richness on alien animal species richness were used to evaluate the effect of environmental and anthropogenic factors on spatial congruence. Results showed that 17 provinces were identified as high invasion risk regions (Liaoning, Hebei, Shandong, Anhui, Jiangsu, Zhejiang, Hunan, Hubei, Jiangxi, Guangdong, Guangxi, Fujian, Taiwan, Hainan, Guizhou, Yunnan and Sichuan). For environmental factors, the potential evapotranspiration (PET) or actual evapotranspiration (AET) could better predict AIAS and WIAS richness, and area better predicted RIAS richness. For anthropogenic factors, the gross domestic product (GDP) or human population explained more variation in AIAS or WIAS richness, and transport volume or GDP could better explain the variation in RIAS richness. Generally, the environmental and anthropogenic factors showed identical prediction ability. The results of PLS regressions indicated that one set of predictors showed similar effects to the alien plants and animals within a model, and indicated the spatial distribution congruence of them. Removing the environmental and anthropogenic factors, to some extent, will weaken the congruence, but the congruence of plants and animals still exists for AIAS and RIAS richness. Our study provides meaningful information for the prediction and management of the biological invasions.