FIGURE 3 in Vertebrate endemism in south-eastern Africa numerically redefines a biodiversity hotspot

FIGURE 3. (a) Numerical refinement of the Greater Maputaland-Pondoland-Albany (GMPA) region, based on the cross-taxon consensus on a region maximally congruent to its qualitative delimitation proposed by Perera et al. (2011) recovered from phenetic relationships of OGUs for each taxonomic group, with further consensus from both Phenetic Cluster Analysis and Parsimony Analysis of Endemicity for all taxa combined. See text for calculation of the percentage consensus. C = core of the numerically refined GMPA region; E = an extension to the numerically refined GMPA region; *according to consensus, the Knysna OGU can be regarded as a part of the core region of GMPA, but treated as an extension due to the transitional nature of its biota between the GMPA and the Cape Floristic Region biodiversity hotspot (see text for details); (b) The core region (thick boundary) and extensions (dashed boundary) of the numerically refined GMPA region of vertebrate endemism in relation to the proposed qualitative delimitation of the same, and the Maputaland-Pondoland-Albany biodiversity hotspot; (c) Congruence of the numerically refined boundary of the GMPA to its qualitative delimitation, from Phenetic Cluster Analyses for individual vertebrate groups, for macro-ecologically releted groups (see text for details), and for all vertebrates combined, as well as for Parsimony Analysis of Endemicity for all vertebrates combined.