Agri-driven habitat modification and small mammal species assemblage nexus in the Atewa Range and Bimpong Forest Reserves, Ghana

This study provides first-hand information on how the changing microhabitat in evergreen forests influences the assemblage of small terrestrial mammals. 30 plots of 100 × 50 metres were randomly sampled in the closed forest, open forest and farmlands habitats. Four transects, 100 m in length, spaced at 10 m, were established in each plot. Ten traps spaced at 10 m from each other were placed on each transect. In the dry and wet seasons, small mammals were captured for three consecutive nights twice a year using Sherman collapsible traps baited with oatmeal and groundnut paste. Small mammals were photographed and identified to species level in situ. The multidimensional scaling method was used to identify clusters of sites based on the composition of their communities. Differences in species composition among habitat types were tested using Permutational Multivariate Analysis of Variance. To identify the species contributions to the major differences between habitat types, a SIMPER analysis was performed. Species richness was compared by subjecting the relative abundance data to an individual-based rarefaction technique. Species diversity of the habitat types in the study areas was compared using the Renyi diversity ordering approach. A total of 507 individual small mammals of 22 species were sampled from both study sites. Our analysis revealed that the composition of small mammal species differed significantly between habitat types in the two study areas. Praomys tullbergi, Mastomys natalensis and Mus minutoides contributed 27.5%, 16.8% and 14.4%, respectively to overall community dissimilarity between the habitat types. Together these three species contributed to 54.8% and 62.8% of observed dissimilarities in Atewa and Bimpong study areas respectively. Overall observed species richness in Atewa (16 species) was higher than Bimpong (14 species). Generally, farmlands and open forests were more abundant, species-rich and diverse than the closed forest habitats in both study areas. Our results supported the hypothesis that different habitat types (farmlands, open forest, and closed forest) in the study areas will differ in small mammal assemblage due to different levels of disturbances.