Towards conserving regional mammalian species diversity: a case study and data critique

Abstract

Species richness maps were derived for the Transvaal region from two different databases, namely a primary point database based on actual survey records and a generalized distribution map database. It is shown that sixteenth degree grid square (= QDS in Lombard 1995) species richness maps based on these two data sets for the region are highly disparate, which may be attributed to overestimation of species distributions by distribution maps and uneven sampling reflected in the primary point database. The limitations and problems associated with the two databases are discussed. Of the 10% most species-rich grid squares based on distribution maps and primary point data, 33,3% and 12% respectively are fully encompassed by existing conservation areas and are well represented in the sixteenth degree grid square networks selected by two iterative reserve selection algorithms.