Relationships Between Environmental Factors And Endangered Species Richness Of Internationally Important Wetland

Macro-scale spatial patterns in species diversity and their underlying mechanisms are central to macro-ecology and biogeography. A series of hypothesis have been raised by domestic and foreign scholars. Energy hypothesis, interruption hypothesis and habitat heterogeneity hypothesis are the three well-received ones. This study was carried out on internationally important wetlands of the world, which is a significant for endangered species protection as habitat, breeding and wintering sites. The driving factor of wetlands endangered species richness was analyzed by using the spatial autoregressive model at the distance of 300,600 and 1200 km. The study results indicate that (1) the spatial autoregressive model has some effect of removing spatial correlation. Overall, the distribution of IUCN endangered species is random. The endangered species richness of internationally important wetlands has no clear spatial autocorrelation at global scale, so the standard multiple linear regression models perform better than spatial autoregressive model. SAR model become fittest at the distance of 300 km. (2) SAR model and OLS model both show that human interruption hypothesis has the highest explain rate. Energy hypothesis and habitat heterogeneity hypothesis are decreased systematically. The R2 is 25.1%, which means the geographic pattern of endangered species richness at internationally important wetlands influenced by different hypothesis. There still exists quiet a lot unexplained rates. (3) The R2 is increasing along with the decrease of endangered level, which means the higher species number is, the fitter of this driving model is. (4) Main impact factors are annual range of temperature, log-transformed wetland area and GDP per capita (p<0.001). The two formers function as positive and the last one function as negative. In addition, Annual precipitation, GDP and population density are all key important driving factors (p<0.01). (5) Because there is no spatial correlation for the residue of SAR model, some of the explanatory variables are not significant. Some predicted value based on the neighborhood values (spatial autoregressive term), so the value of SAR model is less than OLS model. The study result can provide better understanding of driving factors and models for endangered species richness of internationally important wetlands, as well as wetlands endangered species protection and environmental protected area design.