Skip to main content
U.S. flag

An official website of the United States government

Factors affecting species richness and distribution spatially and temporally within a protected area using multi-season occupancy models

February 22, 2019

Exploring trends in species richness and the distribution of individual species over time as well as the factors affecting these trends informs conservation priorities in protecting species and ecosystems as a whole. We used data from 41 park-wide line transect surveys in 2009 and 2014 and multi-season occupancy models with multi-species data to explore trends in species richness and distribution of individual species and factors affecting these trends in Nyungwe National Park (NNP), Rwanda. Mammalian species richness and the distributional range of 5 of the 7 species increased between 2009 and 2014 in NNP. The probability of colonization of a species into a new area in 2014, where it was not present in 2009, was highest in sites with a lower probability of poaching activity, close to tourist trails, and at lower elevations. The probability of colonization with no poaching activity was about 50% but dropped to about 10% with a 100% chance of poaching activity. Duiker species had the largest increase in distribution during the study, while there was a decrease in the distribution of the eastern chimpanzee and blue monkey. Our results suggest that increased patrols should be implemented in areas of the park with low species richness and areas with a low probability of occurrence for species of conservation concern to combat poaching activity and thus increase the probability of a species moving into a new area. Our use of a single multi-season model for multiple species explicitly accounts for imperfect detection and species-specific identities, while allowing for inferences to be made about rarely detected species by sharing covariates with common species. These results can be used to improve conservation planning in NNP for species management and ranger patrol protocols and our modelling framework is broadly applicable to any protected area with presence/absence species field data.