Quantitative classification and ordination of shrub species and communities in a loess landscape of western Shanxi

The study region was located in Zhongyang County of western Shanxi Provice, China, and the center of this area was at 37°15′N, 111°58′E. Our study area was located across an ecotone between the loess hills and bedrock hills. Quantitative analysis of relationships between vegetation and environment was an effective method in research field of modern vegetation ecology. Quantitative classification (TWINSPAN) and ordination analysis (detrended correspondence analysis, DCA) were used to examine the ecological characteristics of all shrub species and communities in a loess landscape of western Shanxi, China. All 29 shrub species were examined in thirty-one 10×10m2 plots. From TWINSPAN analysis, we classified the plots into 8 types and all tallied species into 9 groups. The DCA analysis provided us with similar results on the shrubs and communities in the landscape. Both classification and ordination analysis of the plots reflected the spatial distribution of shrub communities. The first axis of DCA appeared to be related to moisture; the second axis to the thermal conditions; and the diagonal line in the 2-D ordination diagram to the changes of elevation, temperature and moisture the three important variables determining the distributions of shrub communities in our landscape. The DCA figure of shrub species reflected the spatial distribution of shrub ecotypes and their relationships with the environment. The first axis of DCA appeared to related to drought tolerance of species. Diagonal line of the 2-D ordination diagram of species DCA reflected the spatial distribution of species. The species which lied on the bottom right of DCA figure numerously were distributed in the loess hill area, while the species lied on the top left of the figure mainly were distributed in the bedrock area.