Land snail biogeography and endemism in south-eastern Africa: Implications for the Maputaland-Pondoland-Albany biodiversity hotspot

A few OGUs such as Drakensberg-Eastern Cape escarpment, northern Natal, southern Middleveld, Waterberg, Highveld-upper Karoo and Inhambane that were inappropriately placed in the cluster dendrogram with comparatively low species richness and endemism may indicate gaps in biogeographic knowledge or more likely artefacts of sampling biases. Northern Natal and Waterberg have recorded none of the selected species while only two species have been recorded from Highveld-upper Karoo, which may indicate gaps in sampling. It is appar- ent that those areas need prioritised attention with respect to malacofaunal inventorying surveys. The northern Natal and Transkei Midlands (adjacent to Drakensberg-Eastern Cape escarp- ment) were also recognised as gaps for vertebrate endemism [22], the latter supporting the idea of a Transkei Gap [107]. However, even if such a real biogeographical gap exists for land snails along the Transkei Midlands/escarpment, it certainly does not extend to the coastal belt, instead identifying the extended Pondoland as one of the most prominent COEs (see Figs 4 and 5). The MPA hotspot, the greater MPA region of endemism and conservation implications While the hotspot status of the MPA is further confirmed by its malacofauna, the greater MPA region provides a zoogeographically delimited broader region of land snail endemism with five COEs serving as local conservation priorities. A remarkable 72% of the species restricted to COEs are narrow endemics, highlighting the habitat heterogeneity provided within the

greater MPA region of endemism. Endemism-based intuitive malacofaunal subregions pre- sented for eastern SA [13] captured similar (if not finer) geographical entities congruent with our Natal and extended Pondoland COEs as well as an eastern escarpment (sections of the great escarpment in Mpumalanga and Swaziland) extension, further supporting the greater MPA region of land snail endemism. Further, the six AOEs derived from PAE and the nine biotic elements yielded from BEA are congruent to those five COEs, well supporting our pro- posals for the greater MPA region of land snail endemism and the Knysna transition zone between the greater CFR and the greater MPA region. Table 4. Comparison of centres and areas of land snail endemism in south-eastern Africa (SEA), defined in the present study through UPGMA clustering, PAE and BEA with similar areas defined for Sciobius weevils of south-eastern Africa using various methods employed in previous studies. Reference Taxa Units Method Comparison of Results Present study Land snails of SEA COEs UPGMA Sky islands Extended Maputaland Natal Extended Pondoland Albany-Knysna AOEs PAE Soutpansberg Extended Maputaland Natal Extended Pondoland Albany-Knysna Wolkberg BEs BEA Element 4 Element 8 Elements 5 & 6 Element 1 Elements 2 & 7 Morrone [59] Sciobius weevils of SEA AOEs PAE - Area 1 (partly) Area 1 Area 3 Area 2 Szumik et al. [104] AOEs AE Sets 10 & 11 (partly) Sets 2, 10 & 11 Sets 1, 4, 5, 6, 7, 8 & 9 Set 7 Sets 3, 12, 13 & 14 Mast & Nyffeler [105] AOEs PAE - Area 1 (partly) Areas 1 Area 3 Area 2

Hausdorf & Hennig [63] BEs BEA Element 1 (partly) Element 1 Element 2 Element 4 Element 3 Escalante [106] AOEs AE Set 2 (partly) Set 2 Sets 0, 2 & 3 Set 3 Set 1 (UPGMA—phenetic hierarchical clustering with UPGMA algorithm with the Jaccard index; PAE–Parsimony analysis of endemicity; BEA–Biotic element analysis; AE— Analysis of Endemicity; BEs–Biotic Elements). https://doi.org/10.1371/journal.pone.0248040.t004 PLOS ONE Land snail endemism in the Maputaland-Pondoland-Albany biodiversity hotspot PLOS ONE | https://doi.org/10.1371/journal.pone.0248040 March 4, 2021 20 / 26