Patterns of temporal and spatial variations of a forest mammal community in the Udzungwa Mountains National Park, Tanzania

Abstract

In tropical forests, mammals represent particularly rich communities of species with a variety of trophic groups and wide range of life history traits. This diversity plays a significant role in the functioning of the ecosystems. Yet, tropical forest mammals are negatively affected by a number of factors caused by both natural and anthropogenic activities. However, limited and often not-consistent information is available on the status or trends of mammal communities in tropical forests. For that reason, this study employed systematic camera trapping during 2009 – 2014 to determine the temporal and spatial variations of forest mammal communities. The study was conducted in Mwanihana forest within the Udzungwa Mountains National Park, south-central Tanzania, an outstanding area in Africa for biological richness and endemism, particularly the mammals. Camera trapping data were collected together with field- and remote sensing-based environmental and human disturbance variables considered influential of species’ occurrence. State-ofthe art occupancy analytical framework was used to investigate the relationships between species abundance and environmental covariates, including seasonal variations. Occupancy requires repeated temporal replications to resolve the ambiguity between species absence and non-detection when species are unobserved at sample locations. Such analysis was complemented by more classic, multiple regression analysis such as Generalized Linear Models used to analyze effects of covariates on a raw index of abundance, the camera trapping rate (RAI). Results show that a minimum of 32 mammal species are present in the target forest, and the species-specific occupancy modelling revealed novel ecological knowledge for the 11 most detected species, highlighting patterns such as ‘montane forest dwellers’, e.g. the endemic Sanje mangabey (Cercocebus sanjei), and ‘lowland forest dwellers’, e.g. suni antelope (Nesotragus moschatus). Furthermore, the habitat associations in relation with the raw index of abundance was found to be positively correlated with distance to the park border for IUCN-Endangered Abbott’s duiker (Cephalophus spadix), indicating preference for interior forest habitat and avoidance of disturbance while that of Eastern Arc-endemic Lowe’s servaline genet (Genetta servalina lowei) was positively correlated with the diversity of large trees but negatively correlated with visibility and herbaceous cover, indicating preference for mature forest habitat. As for the analysis of seasonal variation (as especially related to rainfall), only bush pig (Potamochoerus larvatus) showed a significant decrease in detectability from dry to wet season. Analysis of the variation of relative abundance with space and time revealed that anthropogenic activity such as firewood collection had a negative impact, by lowering relative species’ abundance for a number of species such as Harvey’s duiker (Cephalophus harveyi). Moreover, the relative abundance of the pooled species most targeted by poaching increased with decreased poaching activities over time. Overall this study shows that the analysis of camera trap data deployed using systematic and standardized protocols - and with account for imperfect detection - can provide robust ecological assessments of mammal communities. It further shows how these assessments can determine the various drivers of change of mammals’ occurrence (both natural and anthropogenic), hence providing insights that are of clear potential relevance to mammals’ conservation management.