Biodiversity loss and fragmentation - matrix effects and ecosystem consequences using a hyperdiverse Malagasy amphibian assemblage as an example

碎片化是一个可能导致生物多样性丧失的过程，即导致物种枯竭的群落碎片化。然而，物种丧失的影响很可能超出有关动物类群的群落特性（例如群落结构）的变化，而是可能影响各种生态系统特性以及生态系统功能。因此，一个破碎的景观代表了一个适当的系统来研究一般的社区-生态系统依赖关系。营养相互作用和动物群落的营养结构在大多数这些过程中起着重要作用，其研究可以直接了解生态系统的功能。本项目旨在了解破碎景观中两栖动物多样性和局部灭绝模式如何依赖于多样性的功能成分、食物网特性、形态特征和系统发育距离。此外，还将分析基质的组成和结构（即嵌入碎片的非森林生境）是否是群落和食物网特性变化的重要驱动因素。这将通过在物种、功能、群落和生态系统的水平上测试一组与多样性模式变化有关的假设来完成。拟议的项目将在马达加斯加东部Ranomafana地区的一个山地雨林生态系统的蛙类群落中进行。这是一个非常适合的模型区，具有独特的两栖动物多样性，在这里，无脊椎动物的系统分类已经基本确定，并且存在这些生物的大型分子，系统发育，分布和生物声学数据库，并将纳入本研究。各种已建立的现场方法将与最先进的实验室分析相结合。食物网结构将使用稳定同位素进行比较评估，这是一种成熟的技术，但迄今尚未以如此全面的方法应用于两栖动物群落。

Fragmentation is a process that may lead to the loss of biodiversity, i.e. to species-depleted communities in fragments. However, the effects of the loss of species may well go beyond the change in community properties of the animal groups in question (e.g. community structure) but may affect various ecosystem properties as well as ecosystem functioning. Hence, a fragmented landscape represents an appropriate system to study general community-ecosystem dependencies. Trophic interactions and the trophic structure of animal communities play important roles in most of these processes, and its study can provide direct insights into ecosystem functioning. The present project aims at understanding how patterns of amphibian diversity in a fragmented landscape and local extinctions depend on functional components of diversity, food web properties, morphological traits and phylogenetic distance. Furthermore, it will be analyzed if the composition and structure of the matrix (i.e. the non-forest habitat in which the fragments are embedded) are important drivers of changes in community and food web properties. This will be done by testing a set of hypotheses related to diversity pattern changes at the level of species, function, communities, and ecosystems. The proposed project will be conducted on anuran (frog) communities in the Ranomafana area of Eastern Madagascar, a montane rainforest ecosystem. This is a highly suitable model region with an unique amphibian diversity, where anuran systematics are largely settled and large molecular, phylogenetic, distributional and bioacoustic databases on these organisms exist and will be incorporated in this study. A variety of established field methods will be combined with state-of-the-art laboratory analyses. Food web structure will be comparatively assessed using stable isotopes, a wellestablished technique which, however, has so far not been applied to amphibiancommunities in such a comprehensive approach.