Study On Simulating The Geographical Distributions Of Common Tree Species In China Based On Generalized Models And Classification And Regression Tree

The analysis of species, vegetation-environment relationships has always been a central issue in ecology. Recently, with the rise of new powerful statistical techniques and GIS tools, the development of predictive species distribution models has rapidly increased in ecology. Such models are static and probabilistic in nature, since they statistically relate the geographical distribution of species or communities to their present environment. A wide array of models has been used in the studies of biogeography, conservation biology, plant community, biodiversity and climate change research. In this paper, based on four models: Generalized Linear Models (GLM), Step Generalized Linear Models (SGLM), Generalized Additive Models (GAM) and Classification and Regression Tree (CART), we simulated the geographical distributions of common tree species in China. In order to select a suitable model for simulating and predicting the geographical distributions of tree species under climate change in the future. 20 tree species were selected for model calibration and evaluation. As a result, the four models could simulate the geographical distributions of tree species very well except Pinus tabulaeformis and Quercus liaotungensis, and GAM is the best one among them. Combining with geographical information system (GIS), the simulating effects for tree species distributions of Cyclobalanopsis glauca and Pinus tabulaeformis were given. The results showed that the four models could respond the geographical distributions of Cyclobalanopsis glauca very well, and could not respond the geographical distributions of Pinus tabulaeformis, which are good fit to the model's evaluation results. The relative performance of different models was discrepant among species suggesting that the most accurate model varies between tree species and distribution regions. Moreover, the potential geographical distributions of five typical tree species: Pinus massoniana, Pinus tabulaeformis, Pinus koraiensis, Cyclobalanopsis glauca and Quercus mongolica under climate change were studied based on GLM, SGLM, GAM and CART. The four models showed a north and west expansion for Pinus massoniana. Addition to a north shift, Pinus tabulaeformis showed an expand trend to northeast and southwest by SGLM, GAM and CART. For Pinus koraiensis, based on SGLM, GAM and CART, its distribution area largely in northeast of Liaoning, west of Jilin and middle north of Heilongjiang provinces would disappear. The predicting results of Cyclobalanopsis glauca by GLM, SGLM and GAM were similar, and showed its geographical distributions would expand to the west and the north, but CART indicated that its geographical distributions in southern of Guangdong and Guangxi provinces would disappear. The predicting results of Quercus mongolica's geographical distributions by the four models were similar, and they would shift to the west in the future.