The "arid corridor" distribution in Africa: a search for instances among amphibians

Abstract

Marked similarities exist between the flora and fauna of the 5outhwestern and northeastern coiners of Africa. Among amphihianr, a tendency towards this distribution is shown by vicariating species of the Rrrfb verrehralis group, by the R. ,qrrmlnt~i/p.seudngnrtnaniJpr~u~~~'i complcx, and by Tomnp~emn cr-vptoiis. In all of these, greater taxonomic complexity is present in the sourh, and ranges include the mois~er sout1ie;lst. Marked similarities between the fauna and flora of the southwestcm and northeastern corners of Africa have been commented on by several authors (e.g. Balinsky 1962: Werger 1978: Kingdon 1990). A strip of relatively country still connects the two areas through western Tanzania, and particularly during dry phases of the Quaternary there probably existed a substantial drought corridor'' or arid corridor (Balinsky 1962: Werger 1978) allowing xeric fauna1 and floral interchange between the southwest and northeast. Several species of mammals, birds and plants currently show populations isolated in the southwestern and northeastern corners of the continent (Werger 1978; Kingdon 19901, but amphibians do not appear in such lists. They might indeed not be expected to show this pattern on account of a supposed vulnerability to drought. Neverthekss, during the course of a review of distribution patterns of African amphibians. centred in the Natural History Museum, London, a search has been made for distribution patterns that suggest connections, present or past, between the dry southwest and northeast areas of the continent. This paper reports on two groups of bufonids and a ranid species (possibly a species complex) which have been found to show at least a tendency towards an arid corridor dish bu t ion. The genus Bufo could be expected to exhibit the clearest instances of a southwest to northeast connection, since many Bufo species are canfined to parts of Africa. It has to be said that the detemination and definition of African amphibian species is still in a rudimentary state, with the result that the groundwork has barely been set for phylogenetic analyses. In (he caseofBufo, species grouping has been attempted in various ways other than phylogenetic analysis, without complete agreement being reached (Tandy & KeiUl1972; Poynton & Broadley 1988). Nevertheless, it is noticeable that some generally accepted groups within the genus are composed entirely of species distributed in mm parts of the continent and, in some cases, the adjoining Arabian Peninsula and beyond. Species in Africa that arc assignable to these groups are reviewed first, after which consideration is given tootherBufospecies that occur mainly in areas, but which are assignable to ecologically more heterogeneous groups.