期刊
JOURNAL OF APPLIED ECOLOGY
卷 53, 期 4, 页码 1225-1235出版社
WILEY-BLACKWELL
DOI: 10.1111/1365-2664.12650
关键词
biodiversity; body size; camera trap; diet; grasslands; hierarchical Bayesian models; human disturbance; multi-species modelling; protected areas; species richness
资金
- Fulbright Study/Research Grant
- PEO Scholar Award
- Wilderness Wildlife Trust
- Columbus Zoo
- Cleveland Metroparks Zoo
- Wild Felid Association
- Idea Wild
- Temenos Foundation
- Rufford Foundation
- Virginia Tech
Emerging conservation paradigms have shifted from single to multi-species approaches focused on sustaining biodiversity. Multi-species hierarchical occupancy modelling provides a method for assessing biodiversity while accounting for multiple sources of uncertainty. We analysed camera trapping data with multi-species models using a Bayesian approach to estimate the distributions of a terrestrial mammal community in northern Botswana and evaluate community, group, and species-specific responses to human disturbance and environmental variables. Groupings were based on two life-history traits: body size (small, medium, large and extra-large) and diet (carnivore, omnivore and herbivore). We photographed 44 species of mammals over 6607 trap nights. Camera station-specific estimates of species richness ranged from 8 to 27 unique species, and species had a mean occurrence probability of 032 (95% credible interval=021-045). At the community level, our model revealed species richness was generally greatest in floodplains and grasslands and with increasing distances into protected wildlife areas. Variation among species' responses was explained in part by our species groupings. The positive influence of protected areas was strongest for extra-large species and herbivores, while medium-sized species actually increased in the non-protected areas. The positive effect of grassland/floodplain cover, alternatively, was strongest for large species and carnivores and weakest for small species and herbivores, suggesting herbivore diversity is promoted by habitat heterogeneity.Synthesis and applications. Our results highlight the importance of protected areas and grasslands in maintaining biodiversity in southern Africa. We demonstrate the utility of hierarchical Bayesian models for assessing community, group and individual species' responses to anthropogenic and environmental variables. This framework can be used to map areas of high conservation value and predict impacts of land-use change. Our approach is particularly applicable to the growing number of camera trap studies world-wide, and we suggest broader application globally will likely result in reduced costs, improved efficiency and increased knowledge of wildlife communities.
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。
推荐
暂无数据