Drivers of primate richness and occurrence in a naturally patchy landscape in the Brazilian Amazon

We assessed the spatial extent at which the species-landscape relationship is strongest (i.e. the scale of effect—SE) on primate occurrence ( Alouatta belzebul , Saguinus midas , Saimiri sciureus , and Sapajus apella and Cebus olivaceus , the last two considered together in the analysis) and species richness and evaluated which landscape, patch, and human variables influence primate distribution in a savanna ecosystem in Brazil. We used nested buffers to measure the landscape attributes, and used these data to assess the SE of the species-landscape relationships. We explored the relative contributions of landscape, patch, and human variables to species richness and occurrences by using Generalized Linear Mixed Models and logistic regression. We found that the SE did not differ between primates, but did between two regions with different matrix composition. At the landscape level, occurrence of all species was higher as the distance to the nearest block of continuous forest decreased, but was lower as the amount of water bodies and anthropogenic cover in the matrix increased. The occurrence of S. apella , C. olivaceus and A. belzebul was positively related to forest cover, and all species but A. belzebul had higher occurrence in taller forest. The occurrence of S. apella , C. olivaceus and A. belzebul decreased closer to the city, and S. apella and C. olivaceus presence increased with the number of residents. Richness was negatively related to the number of residents and anthropogenic cover, but positively to forest height. We concluded that conservation planning for primates should follow a “functional landscape” perspective, by maintaining higher forest cover and minimizing the anthropogenic alterations in the matrix.