Figure 45. The configuration of palatal ridges in three Pteropodidae species: (a) Epomophorus wahlbergi, (b) E. crypturus, and (c) E. angolensis.

      Associations between food plants and fruit bats, worldwide, are important to the dispersal and reproduction of many species of tropical angiosperms. In the Usambara Mountains, Tanzania, where at least seven species of pteropids occur, approximately 20% of the submontane tree flora is bat-dispersed (Seltzer et al. 2013). Globally, fruit bats are known to pollinate at least 528 species of flowering plants (Fleming et al. 2009). In Africa, the pollination of selected plants by fruit bats has been studied in West African fruit bats only (Rosevear 1965). an inaugural study revealed that fruit bats pollinate the baobab, Adansonia digitata (Jaeger 1945), and subsequent studies focused on pollination of the sausage tree, Kigelia africana, and other West African trees by pteropids (Baker and Harris 1957, Harris and Baker 1958, 1959). A recent study of the genus Rousettus across its Afrotropical range invokes the co-evolutionary relationships between pteropids and many tropical angiosperms; it suggests that seed dispersal by Rousettus bats, which have the ability to colonise dry habitats, could have led to the expansion of forest refugia in Africa and Asia during the Pleistocene (Stribna et al. 2019). In light of the latter, it is interesting to note the persistence of the isolated outlying population of Eidolon helvum, confined to the Aïr Massif, northern Niger, deep within the Sahara Desert (Bergmans 1990, Cotterill 2001c). These interesting subjects of the ecology and co-evolution of Afrotropical plants and pteropids await deserving study – improving our understanding of bat–plant interactions in pollination and seed dispersal in tropical landscapes has important significance for biodiversity conservation (Bermingham et al. 2005, Fleming et al. 2009). Such exigency and the dearth of knowledge on these subjects challenges evolutionary biologists – in southern Africa especially – with interesting opportunities.

      The phylogenetic relationships within the Pteropodidae have recently been clarified using molecular techniques (Almeida et al. 2011, 2016). Previous studies based on morphology suggested that African fruit bats comprised two subfamilies: Epomophorinae (including the tribes Epomophorini, Myonycterini, Plerotini and Scotonycterini) and Rousettinae (including the genera Rousettus and Eidolon) (Bergmans 1997), with the genus Pteropus restricted to offshore islands (O’Brien et al. 2009). Molecular phylogenies do not support this division (Almeida et al. 2016). For a start, the genus Eidolon represents an independent lineage not closely related to other fruit bats, and is placed in its own subfamily Eidolinae. All other mainland African fruit bats (i.e. excluding Pteropus) group together as an independent lineage represented by the subfamily Rousettinae, with the Asian genus Eonycteris the only non-African member of this taxon (Giannini and Simmons 2003, Almeida et al. 2011, 2016). Thus, the highly distinctive genus Megaloglossus with its unique, extensible tongue and nectivorous diet, does not group with other Asian nectivorous fruit bats (e.g. Macroglossus), but is embedded within the Rousettinae. Fruit bats have colonised the African mainland at least four times (Almeida et al. 2016), with the genus Pteropus having independently colonised offshore Indian Ocean islands on three occasions (O’Brien et al. 2009).

      In summary, African fruit bats can be divided into the following seven tribes: (1) Eidolini (Eidolon); (2) Scotonycterini (Scotonycteris, Casinycteris); (3) Rousettini (Rousettus); (4) Stenonycterini (Stenonycteris); (5) Epomophorini (Epomophorus, Epomops, Hypsignathus, Micropteropus, Nanonycteris); (6) Myonycterini (Myonycteris, Megaloglossus); and (7) Plerotini (Plerotes).

      The Epomophorini include bats with distinctive white ear patches and shoulder epaulettes (but absent in Hypsignathus). The Scotonycterini also have white facial markings. The remaining genera have less striking faces and lack shoulder epaulettes, rendering these bats far drabber in appearance. Some genera (e.g. Rousettus and Myonycteris) have a short tail (Figure 44). The taxonomic position of Lissonycteris was in dispute for a long time – this taxon was previously placed in its own genus, and prior to that considered a subgenus of Rousettus. This species has now been confirmed to belong to the genus Myonycteris (Nesi et al. 2013). Before the advent of molecular techniques, relationships within the Pteropodidae were inferred from a number of morphological characters; one of the most important of these was the configuration of palatal ridges (Figure 45). If not particularly useful for understanding systematic relationships, the palatal ridges are still useful in species identification.

      Hypsignathus and Nanonycteris are monotypic genera, and their taxonomic status is not currently disputed. Micropteropus, Epomophorus and Epomops, however, are not as clearly distinct and there is still debate on the position of species within them (Almeida et al. 2016). Epomophorus grandis was originally described as Micropteropus grandis (hence the species epithet), and may still be a member of the latter genus but this awaits molecular analysis; the palatal ridges in this species appear to be intermediate between that of the two genera. Similarly, Epomops dobsonii (but not the other two Epomops species, E. franqueti and E. buettikoferi) has recently been transferred to Epomophorus based on molecular analysis (Almeida et al. 2016); we also suggested this in the first edition judging by its biogeography and its intermediate palatal ridges (Monadjem et al. 2010b). Externally, Micropteropus and Nanonycteris are almost identical (to the point that they cannot be identified with certainty on external features alone), but they are morphologically (palatal ridges) and genetically quite different, justifying generic separation. (Note that Nanonycteris does not occur in southern Africa.) Finally, the relationship between Micropteropus and Epomophorus is not yet clearly established and requires further molecular studies, particularly the inclusion of M. grandis, M. intermedius and most of the species of Epomophorus (Almeida et al. 2016).

      TABLE 5.

      TABLE 5. IDENTIFICATION MATRIX FOR GENERA WITHIN THE FAMILY PTEROPODIDAE