Udzungwa Elephant Project, Udzungwa Mountains, Tanzania

In collaboration with Trevor Jones (Anglia Ruskin University), Phyllis Lee (Stirling University) and Animal Behaviour Research Unit (Mikumi National Park)

This is the first study of elephants in the Eastern Arc montane forests of Tanzania. Elephants were presumed to have been poached to near-extinction in the Udzungwa region in the late 1980s and the African Elephant Database shows only the northeastern Udzungwa Mountains as within 'possible' elephant range. Our recent observations of elephants, dung and other signs suggest that increased protection over the last 15 years, combined with elephants having found some refugia, has enabled the resident population in Udzungwas to survive and to begin recolonising parts of its former range. However, connectivity with neighbouring elephant populations in Mikumi, Selous and Ruaha is highly threatened and human-elephant conflict is increasing.

We are counting, measuring and mapping elephant dung piles with a recce-survey transect method in remote regions and near corridor routes in the Udzungwa Mountains National Park (1990km2) and surrounding catchment forest reserves. Our aim is to map elephant distribution and habitat-use and obtain estimates of population density and structure. To examine population demography, we are using a method for estimating elephant age from dung boli diameter using a model developed in Amboseli National Park. Dung is also being examined for food parts, particularly seeds, to explore elephants' diets and role in seed dispersal in these montane forests. Ultimately, we would like to 1) describe elephant population ecology and movement following some release from human pressure, 2) forecast ecosystem changes resulting from an increase in the density of a large mammalian herbivore, 3) develop models to predict the risk of human-elephant conflict at UMNP boundaries and 4) incorporate findings into national corridor planning.