Spatial turnover and knowledge gap of African small mammals: using country checklists as a conservation tool.

Comparing species checklists across countries can be important for determining the relative uniqueness of each country, which can be conveniently defined on the basis of the number of species occurring only in that country or, at most, in one of its neighboring countries. Production of accurate country checklists is complicated by the fact that, especially in scientifically neglected regions, the knowledge of the distribution of many species is unsatisfying. When distribution of a given species is insufficiently known, typically there may be apparent gaps in its distribution range. These species are defined here as 'gap species'. In this paper, we analyze the country checklists for rodents and insectivores of the African continent with the aims of (i) identifying the countries having a higher taxonomic uniqueness; (ii) highlighting countries where more research is needed; (iii) producing a list of gap species; and (iv) determining the ecological correlates of being a gap species. For both mammal groups, the important countries because of their low numbers of shared species were D.R. Congo, Cameroon, Sudan, Kenya, Tanzania, and South Africa. The countries with highest percentages of endemic taxa were Kenya, South Africa, Somalia and Tanzania for insectivores, and Ethiopia and South Africa for rodents. The number of gap species per country was 0-5 for both insectivores and rodents, with the only exceptions of Togo (12) and Benin (15). Apart from Togo and Benin, the main gap countries for rodents were Nigeria, Chad, Gabon, Burundi, and Rwanda, and for insectivores were Niger and Chad. In both groups, the number of gap species per country was independent on the country area, and both range and body sizes did not influence the probability for a species to have distribution gaps. However, most gap species were tropical forest inhabitants. The biogeographic and conservation implications of these data are discussed.