山地森林群落物种多样性垂直格局研究进展

山地森林群落物种多样性垂直格局研究是揭示植被地带性分布式样的重要手段，其研究内容主要包括山地森林群落物种多样性的垂直变化模式、山地森林群落物种多样性垂直带谱和山地森林群落物种多样性及其环境解释等。目前对于山地森林群落物种多样性沿海拔梯度的格局变化模式尚无统一认识，但大致可概括为5类变化模式。垂直带谱研究从最初的单因子单目标已逐步过渡到单因子多目标，最后到多因子多目标。研究深度也从根据样地资料的简单描述和分析，发展为山地森林群落物种多样性机理假说的提出与证实，尚未见有关山地森林群落物种多样性垂直格局机理的报道。山地森林群落物种多样性垂直格局研究方法可分为取样方法和测度方法两大类。常用的取样方法有梯度格局法（或连续样带取样法）和典型群落随机取样法，其中最小取样面积的确定尤为关键。就山地森林群落物种多样性垂直格局测度和分析而言，物种多样性指数的测度和单元统计分析（回归、假设检验、方差分析、相关分析等）较为常见，随着分析手段的不断创新，植被数量分类和排序等多元统计学分析方法得到了日益广泛的应用。

Altitudinal Patterns of Species Diversity in the Montane Forest Communities: A Review

Studies on the species diversity in the montane forest communities along an altitudinal gradient is an indispensable means to reveal the patterns of vertical vegetation zonality.These studies on three aspects,i.e.,species diversity patterns in relation to altitudinal change,species diversity along the vertical vegetation spectra and the environmental interpretation for species diversity in the montane forest communities.Findings from past studies varied on species diversity in the montane forest communities along an altitudinal gradient,but they can be summarized as five models.The research on the species diversity along the vertical vegetation spectra of the montane forest communities has made great progress,which is from single factor,single target at the beginning to single factor,multi-target to multi-factor,multi-target at last.However,few studies have been conducted on the mechanism of altitudinal species diversity patterns in the montane forest communities.Methodological advances in this field of study have been great and can be described in terms of sampling methods and measurement methods.The main sampling methods include contiguous grid quadrats sampling and random communities sampling.Diversity index computation and univariate statistical methods such as regression,hypothesis testing,analysis of variance(ANOVA),and correlational analysis are commonly used methods to analyze the altitudinal patterns of species diversity in the montane forest communities.However,a series of multivariate statistical methods including numerical classification and ordination have been employed in the studies in recent years.