Abstract
A comprehensive understanding of how human disturbance affects tropical forest ecosystems is critical for the mitigation of future losses in global biodiversity. Although many genetic studies of tropical forest fragmentation have been conducted to provide insight into this issue, relatively few have incorporated landscape data to explicitly test the effects of human disturbance on genetic differentiation among populations. In this study, we use a newly developed landscape genetic approach that relies on a genetic algorithm to simultaneously optimize resistance surfaces to investigate the effects of human disturbance in the Udzungwa Mountains of Tanzania, which is an important part of a universally recognized biodiversity hotspot. Our study species is the endangered Udzungwa red colobus monkey (Procolobus gordonorum), which is endemic to the Udzungwa Mountains and a known indicator species that thrives in large and well-protected blocks of old growth forest. Population genetic analyses identified significant population structure among Udzungwa red colobus inhabiting different forest blocks, and Bayesian cluster analyses identified hierarchical structure. Our new method for creating composite landscape resistance models found that the combination of fire density on the landscape and distance to the nearest village best explains the genetic structure observed. These results demonstrate the effects that human activities are having in an area of high global conservation priority and suggest that this ecosystem is in a precarious state. Our study also illustrates the ability of our novel landscape genetic method to detect the impacts of relatively recent landscape features on a long-lived species.