Abstract: Uncovering the relationship between species distribution and environmental factors is the basis and premise for in-depth understanding of forest communities. In this study, 14 permanent plots were set up in Qingyunshan Nature Reserve. We used twoway indicator species analysis (TWINSPAN), redundancy analysis (RDA), Monte Carlo test, and variation partitioning to analyze the relationship between plant species distribution and environmental factors. The results showed that: (1) According to the TWINSPAN, 14 plots were classified into five communities: Castanopsis fissa+Castanopsis carlesii, Pleioblastus amarus+Cryptocarya chingii, Rhododendron kwangtungense+Lithocarpus naiadarum, Schima superba+Rhododendron moulmainense, and Cunninghamia lanceolata+Phyllostachys eduliscommunities. (2) The results of RDA and Monte Carlo tests showed that elevation, soil total calcium content, soil water content, soil pH, soil cation exchange capacity and soil available phosphorus were the main environmental factors affecting species distribution. (3) Results from variation partitioning showed that these 10 tested factors could explain 85.2% variation of species distribution. Among those factors, topographical, soil factors and their interaction accounted for 16.2%, 62.3% and 6.7% of the variation, respectively. 14.8% of the variation were unexplained. This result indicated that species distribution was mainly affected by the environmental factors, and less affected by the interaction between organisms and random factors.

Key words: water environment, nitrogen cycle, stable nitrogen isotope, measured methods, application.