八大公山森林群落物种多度及空间分布对种-面积关系的影响

Species-area relationship (SAR), as one of the few "genuine laws" in ecology, describes the increments of the numbers of species with the sampling area, and reflects the spatial turnover of species. It is always one of the oldest patterns and hottest topics in biodiversity and biogeography. SAR bridges the gaps between the species richness at different scales, and represents the co-existence of species in the community. It is used to determine the minimal sampling area in the community, to estimate the effects of disturbance, to define the appropriate size for natural reserves in the conservation of biodiversity. During the period of my PhD candidate, I analyzed the causes of geographical variation in SARs based on 32 plots (ca. 1 ha.) settled in almost all forests types in China, and found that the SARs were also significantly influenced by the species-abundance distribution (SAD) and species-spatial distribution (SSD). However, we could not obtain actual SAD and SSD data of each species in each plot precisely, as the study was carried out at a regional scale. Effects of SAD and SSD on SAR were derived from simulated data. However, census of plants in large-scale forest plot made actual SAD and SSD data available, and thus provides us an ideal platform to explore their effects on SAR. This project was proposed on the basis of my previous study in PhD period. Based on the 25-ha stem mapped subtropical forest plot in the Mt. Badagongshan, we first construct SAR using nested sampling method and quadrates combination method, respectively. Then, combining actual SAD and SSD data and computer simulation, we will analyze their effects on SAR. Further, the different significances of SAD and SSD influences on SAR will be discussed at three scales (1 ha, 5 ha, 25 ha). Our aims are to explore the mechanism of the SAR and the relative importance of the SAD and SSD on the SAR at different scales. We believe that accomplishment of this project will promote the development of methods for the relevant studies in biodiversity, and enhance our understanding of the formation and maintenance of the biodiversity patterns. In practice, this project could also help us to determine hotspots for biodiversity conservation, to design reasonable area for the natural reserves, and to utilize biodiversity sustainably.

种-面积关系是群落生态学中历久而弥新的研究热点,具有重要的理论和实践意义。申请人在博士期间系统研究了中国森林群落种-面积关系的区域分异机制，并发现种-多度分布与种-空间分布均对种-面积关系存在显著影响。本项目在申请人博士期间工作的基础上提出，拟在八大公山25 ha常绿落叶阔叶混交林定位样地的基础上，分别使用巢式样方法和组合样方法构建种-面积关系，利用群落内实测种-多度分布和种-空间分布数据，结合数据模拟，分析二者对种-面积关系的影响；并在不同尺度上（1 ha、5 ha和25 ha）探讨二者对种-面积关系的相对影响程度，旨在阐明不同尺度上种-面积关系的形成和维持机制。本项目的实施将促进对生物多样性维持机制的理解，推动生物多样性研究方法的发展，并可以为我国生物多样性保护和可持续利用及保护区的设计提供理论指导。

种-面积关系是生态学中物种组成格局研究的最好表现工具之一，具有重要的理论和实践意义。本文拟利用湖南八大公山25ha大样地实地调查的地形数据、土壤数据以及树木清查数据，分析种-多度分布和种-空间分布对种-面积关系的影响，进而研究种-面积关系形成和维持机制。八大公山样地中几乎所有的物种在小尺度上均呈聚集分布；剔除生境因子干扰后依然有近60%物种呈聚集分布。我们用同质泊松模型、异质泊松模型、同质托马斯模型与异质托马斯模型来考察种-空间分布对种-面积关系的影响。结果表明，同质泊松模型显著高估了大部分尺度上的物种多样性；异质泊松模型稍好于同质泊松模型，但仍然在很多尺度上高估物种多样性，说明生境异质性对种-面积关系的解释力不高；同质托马斯模型相较于前两个模型，对种-面积关系的解释明显提高，说明扩散限制是造成八大公山样地内物种分布格局的主要因素；异质托马斯模型的对种-面积关系的解释并没有优于同质托马斯，说明相较于生境异质性，扩散限制决定了样地中的物种分布格局。样地物种多度分布分析结果显示，样地内偶见种众多，个体数少于25个稀有种占物种总数的43.28%。包含全部物种在内时，种-多度分布对种-面积关系形状的影响远远大于种-空间分布，其原因是全部物种中包含了很多的个体数较少的偶见种，使得种-多度分布极不均匀，必然会对种-面积关系形状造成大的影响。当去除群落中个体数目最少的100个物种之后，种-多度分布对种-面积关系的影响明显降低，和种-空间分布的影响数量相当。该结果表明，群落中偶见种对种-面积曲线影响巨大，尤其是中尺度和大尺度上种-面积关系的形状。常见种会在种-面积关系前段显著的提升曲线，而偶见种对种-面积曲线的影响表现在曲线后期。当环境变量只包含地形因子不包含土壤因子时，环境因子和空间距离对beta多样性的总解释率为为60.4%，其中环境因子的解释率为15.7%，而空间因子的解释率为44.7%，这个结果表明相较于生境异质性，扩散限制对物种组成变异的影响更大。在增加了土壤数据作为环境因子后，环境和空间变量解释beta多样性空间变异的60.7%，但是环境因子和空间距离的单独解释率变化明显。其中环境变量可以解释群落组成变化的27.0%，空间距离可以解释群落组成变化的34.0%，说明生境异质性和扩散限制对八大公山常绿落叶阔叶混交林群落结构的影响都十分重要。