| 163 plant community samples and climatic data were collected from Ngari, Tibet. By means of multivariate analysis with Detrended Correspondence Analysis (DCA), Reciprocal Averaging (RA), and Two-way Indicator Species Analysis (TWINSPAN) of the Cornell Ecology Programs, the ordination and classification for community data of Ngari were produced. Two ecologically significant ordination axes are: a thermal or altitudinal axis and a soil moisture axis. A hierarchical classification of community types is produced from TWINSPAN and a nonhierarchical one from DCA axes. 15 zonal community types and 9 intrazonal (phreatic) community types were divided.; The relationships between climate, soils, topography, and vegetation along the altitudinal gradient and soil moisture gradient in Ngari are analyzed. The ordination value of community types and their distribution gradients are further correlated to environmental data with multiple regression. Then, the environmental interpretation for community types is presented in an objective and quantitative mode.; The abundance (coverage%) of 7 commonest species is also related to environmental indices with multiple regression analysis. Then, the matrixes calculated by a set of site parameters of the samples and a set of the constants from the obtained regression equations for the 7 species produces a set of the predicted abundance of the 7 species for the samples.; Thus, the distribution patterns and community structure of vegetation in Ngari and their environmental indexes could be quantitatively presented and predicted. |
