基于多元回归树的公益林群落结构解析

Community Structure Analysis of a Public Welfare Forest Based on Multivariate Regression Trees

为揭示不同生境类型下公益林群落物种多样性及优势种种间关系特征,以临安市公益林125个固定监测样地为研究对象,利用多元回归树(MRT)进行群落分类,以物种多样性指数和种间联结系数为指标,对不同群落类型的结构特征进行探索。结果表明:临安市公益林125个典型样地可划分为5个类型;类型I、II、III与类型V的多样性指数之间均无显著差异,但类型III的多样性指数均最高,类型IV的物种多样性指数均显著低于其它类型,海拔是影响群落多样性水平的主要因子。种间联结初步分析显示,研究区群落中稳定性最高的层片是乔木层,且毛竹入侵极可能是影响5个类型植物群落稳定性的关键因素。研究结果可为临安市公益林分类经营管理提供理论依据,同时为区域公益林群落数量结构研究奠定基础。

To reveal the species diversity and interspecific relations among dominant species of a public forest community under different habitat types, we studied 125 fixed plots of the Public Welfare Forest of Lin＇an city by the multiple regression tree（MRT） method. Species diversity index and interspecific association coefficient were used to explore the structural characteristics of different communities. The 125 typical plots of the forest were divided into 5 types by MRT. The species diversity index did not differ among type I, type II, type III and type V. The species diversity index for type III was the highest, and that of type IV was significantly lower than the others. The slope had a more profound effect on the species diversity index than did elevation. The highest stability layer was the arborous layer in the study area, and bamboo invasion was most likely a key factor seriously affecting the stability of five types. This study can provide a theoretical basis for the classification management of a public forest in Lin＇an city and can be a reference for quantitative structure study of regional public forest communities.