4.7 Article

Influence of forest fragmentation on amphibian diversity in the nature reserve of Ambohitantely, highland Madagascar

期刊

BIOLOGICAL CONSERVATION
卷 96, 期 1, 页码 31-43

出版社

ELSEVIER SCI LTD
DOI: 10.1016/S0006-3207(00)00041-0

关键词

habitat loss; rain forest; Amphibian community; Madagascar; nested subset

向作者/读者索取更多资源

In Ambohitantely the rainforest is divided distinctly by pseudosteppe into more than 500 fragments. The amphibian faunas of seven such fragments varying ill size from 0.16 to 1250 ha were examined. Twenty-eight. amphibian species of two families (Ranidae and Microhylidae) were recorded in the rainforest fragments. The species numbers in the fragments were positively correlated with fragment area and form nested subsets. Composition and individual frequency in small fragments differed from that of the control site (1250 ha). The relative individual density (individuals found each searching hour) was negatively correlated with the fragment size. This is probably due to the fact that the density of streams and brooks was higher in small fragments, which also explains why smaller fragments contained a larger proportion of brook-dwelling amphibian species than larger fragments. Species that live Far from water and show a reproduction strategy independent of running waters or ponds, such as certain microhylids, were only present in fragments of 30 ha or more. The more common species in the control site were also found in the majority of fragments. A rainforest remnant of 1250 ha seems to be large enough to contain a large part of the original amphibian fauna, provided that there are suitable microhabitats. Compared to other taxa, amphibians generally seem to react less sensitively to fragmentation. Due to the sensitivity to microclimate changes microhylids and certain species of the subfamily of Mantellinae represent good bioindicators. (C) 2000 Elsevier Science Ltd. All rights reserved.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据