Abstract: The community types of Masson pine forest of Dayuan Creek were classified by using flexible beta clustering, DCA and CCA. Meanwhile, the species diversity in the Masson pine forest was studied. The results showed that: (1) 27 sample plots containing Masson pine were classified into 8 community types and the classification showed explicit distribution boundaries in DCA by flexible beta clustering, so the dominant species of tree layer was an ideal basis of the classification to forest community. (2) Monte Carlo Permutation Test showed that the result of CCA ordination could not accept that the environmental factors explained the species distribution in plots, because the forest in the research area was seriously disturbed by human activities. However, combined with the eigenvalue of order axes, it was inferred that soil available nitrogen content was strongly correlated with species distribution. (3) The importance value of Masson pine had a marginally significant correlation with elevation, soil available nitrogen and available phosphorus contents, while it was negatively related to the diversity of the tree layer and herbaceous layer. (4) The analysis of variance showed that, compared with mixed broadleafconifer forest and evergreen broadleaf forest, the species diversity of tree layer in Masson pine forest was significantly lower while the differences were not obvious for the shrub and herb layers.

Key words: floristic quality assessment, conservatism coefficient, habitat quality, floristic quality index.