The rare, endemic and forest-dependent Abbott's duiker Cephalophus spadix of Tanzania has been in decline for several decades, leading, in 2008, to the raising of its threat status on the IUCN Red List from Vulnerable to Endangered. Previously occurring in several highland areas in Tanzania, the local extirpation of this antelope from many of these sites over the last 100 years has slowly become apparent. The 2008 assessment estimated a maximum of 1,500 mature individuals across four disparate mountain ranges: Kilimanjaro, Usambara, Southern Highlands and the Udzungwa Mountains of southern Tanzania. The latter area was thought to hold the largest numbers and offer the greatest hope for the long-term survival of the species. Following extensive surveys across the Udzungwas, using camera traps and molecular analysis of dung to confirm results from transects and questionnaires, three new populations have now been confirmed.
From 2007 to 2010 we surveyed 24 sites in 10 forests of varying size (5–522 km2). At each site our primary survey method was linear transects for dung. Over 300 km of transect were walked and all dung and other sign of mammals recorded. However, the Udzungwa forests harbour up to five sympatric antelope species and our previous work has shown identification of antelope dung to species to be unreliable even for experienced fieldworkers and hunters (Conservation Genetics, 10, 251–255). Where present, Abbott's duikers are usually at a low density, very shy and rarely seen (in many years of fieldwork we have glimpsed the species only five times). We therefore used a combination of camera trapping and molecular analysis of dung samples to confirm the presence of the species following our transect surveys.
For the 2008 Red List assessment six forests in the Udzungwa Mountains were known or strongly suspected to still contain Abbott's duiker: Mwanihana, Luhomero-Ndundulu and Ukami (where the species was locally common); Nyumbanitu (where the species was scarce); Uzungwa Scarp and Matundu (where the species was rare). We have confirmed all of these records, including extensions of the species' known range in the large Luhomero-Ndundulu (221 km2) and Matundu (522 km2) forests, and also obtained photographic or molecular evidence of the continued presence of the species in New Dabaga Ulang'ambi and Kising'a-Rugaro Forest Reserves and Iwonde forest within the Udzungwa Mountains National Park.
Camera-trapping and identification of dung to species through analysis of DNA are now vital tools for surveying and monitoring low-density, secretive forest animals, particularly those species that cannot be reliably identified in the field from dung. We are using our results to develop a reliable protocol for monitoring Abbott's duiker at its remaining Tanzanian sites. All the nine Udzungwa forests now known to harbour the species are discrete, separated from each other by either montane grassland or farmland, with little prospect of connecting these forest patches. Further research will be necessary to obtain population estimates and assess the viability of each of these populations.
This revised distribution information alone does not, however, warrant a reassessment of the species' Red List status. Several of the nine forest patches are under increasing pressure, with the western forest reserves of Kising'a-Rugaro, New Dabaga-Ulang'ambi and Uzungwa Scarp having already lost many large mammal species because of hunting (T. Jones, unpubl. data). Uzungwa Scarp is of particular concern: a report published earlier this year highlighted declines and potential extinctions of species of monkey and antelope linked to unsustainable hunting for bushmeat (http://www.conservation.org/newsroom/pressreleases/Pages/CEPF-Report-Bushmeat-Hunting-Tanzanian-Crisis.aspx). Nevertheless the new records of Abbott's duiker, together with the discovery in 2006 of a small population in the Rubeho Mountains just north of the Udzungwas (Oryx, 42, 4–5), provide renewed optimism for the future of this beautiful antelope and reaffirm the importance of the Udzungwa Mountains for its survival.
These surveys, camera traps and molecular analysis were funded by the Wildlife Conservation Society Tanzania Programme, the Zoological Society of London's Erasmus Darwin Barlow Conservation Expeditions Programme, Anglia Ruskin University and the Whitley Wildlife Conservation Trust. We are especially grateful to Athumani Mndeme, Richard Laizzer and Amani Mahundu for help in the field. Francesco Rovero and Martin Nielsen helped us access additional samples.
