试用多元分析方法研究植物形态结构与生态环境的关系

Application of multivariate analysis on studying relationship between plant morphological structure and ecological environment

应用多元分析方法对栎属高山栎组不同环境条件下的３３个植物样品的形态解剖性状的１７个指标进行了综合分析．结果表明，二元指示种分析（ＴＷＩＮＳＰＡＮ）聚类的结果能很好地把相似生态环境条件下的样品聚在一起，对样品进行无趋势对应分析（ＤＣＡ）和主成分分析（ＰＣＡ）都得到了满意结果．植物形态解剖性状的聚类、排序分析也可用来研究个体性状之间变异的连续性和间断性，进而有助于分类学的研究．

In this paper, multivariate analysis is used to study the morphological and anatomical features of Quercus suber under different environmental conditions. 33 plant samples are taken, and 17 indices are selected. The results show that two-way indicator species analysis(TWINSPAN) can better cluster the samples under similar ecological environments, and detrended correspondence analysis (DCA) and principal components analysis(PCA) can give satisfied results. The cluster and ordination analyses of plant morphological and anatomical features can also be used to study the continuity and discontinuity of variations of individual features, and be useful to plant classification.