Modelling primate abundance in complex landscapes: a case study from the Udzungwa mountains of Tanzania

With persistent degradation of tropical forests creating fragmented landscapes, the study of patterns of primate responses to habitat changes is of increasing conservation relevance. We modeled primate abundance in the Udzungwa Mountains of Tanzania through a landscape approach, i.e., one that includes a representative range of discrete forest blocks. The area is internationally recognized for biological endemism and is a primate hotspot in Africa. We targeted three predominantly arboreal monkeys: Udzungwa red colobus (Procolobus gordonorum), Peters’ Angola colobus (Colobus angolensis palliatus), and the Tanzania Sykes’ monkey (Cercopithecus mitis monoides). In each of the four forests (12–522 km2 in size), we counted primate groups along a grid of line transects (267 km walked) and sampled canopy trees in vegetation plots along the same transects (N = 408) to derive structural and floristic forest parameters and proxies of human impact. We found that elevation and the percentage of climber coverage on trees consistently emerged as significant predictors of primate abundance for all three species in spite of their differences in feeding habits, with a negative effect of elevation and a positive effect of climber coverage. This pattern held despite large variations in elevation, forest habitat, and human disturbance across the four forests surveyed. We conclude that arboreal primates in the Udzungwas are dependent on lowland and medium-elevation forests (ca. 300–1200 m a.s.l.) and show considerable resilience to moderate forest disturbance. However, agricultural intensification causes rapid forest degradation, with detrimental effects on primates that need to be prevented through increased protection and community conservatio