A Study on Numerical Classification and Dynamics of Forest communities

本文对林分的定量和有效分类方法进行了研究。对冈山大学研究林中的单株树木的性质进行了调查，从中产生了16个结构量。对16个变量的因子分析表明，只有乔木层和下乔木层的叶生物量、物种多样性和林龄4个因子对林分的特征有贡献，然后对这4个因子进行聚类分析，用数量化理论Ⅲ对沙地特征的分析表明，林分类型中，按物种组成可分为三种类型。这两种分析的组合很容易合成七个林分类别，其中的特点是可视化的示意图与个人电脑。

A study on the procedure to categorize the forest stands quantitatively as well as efficiently was carried out. Properties of the individual trees in the stands in the Research Forest of Okayama University were investigated, from which sixteen structural quantities were produced. The factor analysis with the sixteen variables elucidated that just four factors, the leaf biomass of tree layer as well as lower tree layer, species diversity and forest age, would contribute to the characteristics of the stands.Then the cluster analysis with the four factors was done, which brougth forth three categories of the forest.Other analysis on the characteristics of the sands with the quantification theory III showed that the stands might be classified into three types in relation to the species composition. The combination of the both analyses readily synthesized seven stand categories, of which characteristics were visualized as schematic diagrams with a personal computer.