植物生态学报 ›› 1991, Vol. 15 ›› Issue (2): 101-113.

• 论文 •    下一篇

西藏阿里植物群落的间接梯度分析、数量分类与环境解释

张新时   

  • 发布日期:1991-02-10
  • 通讯作者: 张新时

Indirect Gradient Analysis, Quantitative Classification and Environmental Interpretation of Plant Communities in Ngari, Xizang(Tibet)

Chang Hsin-shih   

  • Published:1991-02-10
  • Contact: Lin Peng

摘要: 根据对西藏阿里地区163个植物群落样地资料进行的多元分析——排序、数量分类与环境解释,给出了该地区植被的基本类型、生态梯度及其与环境因子的定量关系。基本分析方法包括3个步骤:1)通过无倾向对应分析(DCA)的两个排序向量揭示了阿里植被的两个主要生态梯度;2)由该梯度的二维散点图及二元指示种分析(TWINSPAN)分别产生非等级制与等级制的植物群落分类系统;3)以多元回归分析将排序值与环境及地理参数相联系而给出各类型的环境指标——定量环境解释。分析表明,阿里植被类型及其分布主要取决于热量与湿度梯度,前者可通过地理参数,后者则通过土壤特征的数学表达式来定量地确定。两梯度包含的类型、种类与生境差异颇大,由低山暖性荒漠直到高山冰缘植被,从隐带性沼泽与盐生草甸到高原地带性荒漠与草原均各得其位,各有其值。表明该数量分析法对于处理高度生态多样性的植物群落生态信息是十分有效的。

Abstract: Basing on the multivariate analysis—ordination, quantitative classification, and environmental interpretation of 163 plant community samples collected from Ngari, Xizang, the primary vegetation types, ecological gradients, and their quantitative relations with environmental factors of Ngari are given. The basic analysis contains the following three steps. 1) the two principal ecological gradients are brought to light by two vectors of ordination scores produced by detranded correspondence analysis (DCA); 2) the nonhierarchical and hierarchical classification systems of plant communities are produced from the 2-dimensional scatter polt of DCA ordination and two-way indicator species analysis (TWINSPAN), respectively; 3) environmental indexes (quantitative interpretation) of various vegetation types are given by the multivariate regression analysis which connects the ordination scores with environmental and geographical parameters. It is shown by the analysis that the vegetation types of Ngari and their distribution are mainly determined by the thermal and moisture gradients. The former could be expressed quantitatively with the mathematic expression of geographical parameters, the latter with the soil characteristics. The two gradients contain quite different vegetation types, species, and habitates, ranging from low montane warm desert to alpine periglacial vegetation, and from intrazonal bog and saline meadow to zonal plateau desert and steppe with their paricular position and ordination scores. That shows that the quantitative analysing method used here for handling the ecological data of plant communities with great ecological diversity is highly efficient.