A model assessing the conservation threats to freshwater turtles of Sub-Saharan Africa predicts urgent need for continental conservation planning

Global assessment of conservation threats to several taxonomic groups are urgently needed for species living in regions of the world where field research is curtailed by logistic or economic constraints. For instance, it is now widely recognized that freshwater turtles represent one of the most endangered groups of vertebrates in the world. This situation has been particularly evident with Asian species. Although in Sub-Saharan Africa species are exposed to very similar threats (e.g., intense hunting for food and traditional medicine, habitat loss, collection for the pet trade, etc.), there is no comparable concern in the international conservation community simply because the African species are still too poorly known. Here, I propose a model for analyzing the threat levels to freshwater turtles of Sub-Saharan Africa and I test this model by taking into consideration a series of six risk variables of distribution and ecology. Each variable was categorized into five ranks (0-4) of increasing survival risk. The results, although preliminary, show a great sensitivity of most species. This strongly contrasts with the fact that apart from Malagasy endemics very few species are currently considered as threatened according to IUCN lists. There were clear differences among-families in the major threats. Softshell turtles are especially threatened because their large body size makes them a very good food source for native populations. The single podocnemidid species is also vulnerable, and is particularly sensitive to human hunting for domestic consumption. Concerning the Pelomedusidae, there are several species that are considered by the model as sensitive or highly sensitive to rarity. They are affected by a combination of threats (Extent of distribution in Sub-Saharan Africa, Habitat breadth, Habits, Likeliness of human direct/indirect persecution, and Habitat vulnerability). The conservation implications of these results are discussed.