in trees, and anthropogenic hollows such as roofs and basements of houses, tunnels or cavities in dam walls, and abandoned mine shafts (Kunz and Lumsden 2003). Some members of Vespertilionidae, Emballonuridae and Molossidae roost in tree hollows. Radio-tracked Scotophilus viridis and S. dinganii occupied hollow Colophospermum mopane trees in the Kruger National Park and Combretum imberbe in Eswatini (Fenton et al. 1985, Monadjem et al. 2010a). The presence of mature trees in woodlands is essential for the persistence of tree-roosting bats, which are adversely affected by the removal of such trees (Fenton et al. 1998a). Similarly, ‘homogenisation’ of savanna landscapes through the impact of megaherbivores (particularly elephants) also acts to reduce bat diversity (McCleery et al. 2018). While a large amount of research has been carried out on tree-roosting microbats living in American forests, very little is known about such species in southern Africa.
Scotophilus dinganii and S. viridis are frequently found roosting in hollow spaces in the attics of houses, as opposed to the crevice-like roosts used by many Molossidae. Occasionally, so-called ‘cave bats’, such as Nycteridae, Rhinolophidae and Hipposideridae, may be found in larger attics or basements (Voigt et al. 2016) (Figure 15), and also aardvark (Orycteropus afer) burrows in the case of Nycteris thebaica (Monadjem et al. 2009).
Figure 15. Schematic comparison of diversity of different daylight domiciles selected by representative Chiroptera that roost in buildings: A spaces within roof timbers, Mops condylurus; B crevices in roof materials, Neoromicia capensis; C attic spaces, Rhinolophus blasii; D attic surfaces, Scotophilus dinganii; E dark, abandoned rooms, Rhinolophus clivosus; F gaps under suspended floors, Nycteris thebaica; G cellars, Coleura afra; H under crannies in roofs, Chaerephon pumilus; I hollow cement bricks, crevices and cracks in masonry, Molossidae species and Neoromicia; J sheltered window ledges, Kerivoula argentata; K under eaves, Taphozous mauritianus (modified after Brosset (1966a) with additions).
Cave-dwelling bats form the largest aggregations known of any mammal. As many as 20 million bats may occupy Bracken Caves in Texas (Tuttle 1997). In southern Africa, cave-dwelling bats include Rousettus aegyptiacus, Myotis tricolor, and species of the Rhinolophidae, Hipposideridae, Nycteridae, and Miniopteridae. South Africa’s largest known cave colony of bats, at De Hoop Guano Cave, near Bredasdorp in the Western Cape, comprises some 300,000 bats (McDonald et al. 1990a). Apart from De Hoop Guano Cave, other well-known cave bat roosts include Kogelbeen Cave in the Northern Cape (Monadjem et al. 2008), various caves within the Cradle of Humankind World Heritage Site in Gauteng, and Mission Rocks in the Greater St Lucia Wetlands Park in KwaZulu-Natal in South Africa; Arnhem Cave and several others in Namibia (Churchill et al. 1997); a cave system harbouring eight bat species on the Cheringoma Plateau, Mozambique (Monadjem et al. 2010b); Gcwihaba (Drotskys') Caves in Botswana (Smithers 1971); a series of caves in northern and central Zimbabwe (Cotterill and Fergusson 1999, Truluck 1992); and Leopard’s Hill Cave and other caves in Zambia (Whitaker and Black 1976, Ansell 1978, Kaiser et al. 1998). Where not already protected under national legislation, all these landforms hold high rank as sites of critical conservation status (Figures 16–18).
Figure 16. The locations of some of the caves that are important bat roosts in southern Africa.
Figure 17. Schematic cross section through a cave in coastal Tanzania illustrating differences in roost selection. These six species of cavernicolous bats select local microclimates contingent on aspect, temperature, humidity, ambient light, and height above the cave floor (after Hill and Smith 1984).
Figure 18. Cross section of De Hoop Guano Cave (after McDonald et al. 1990a): 1 summer roosting site for non-breeding bats and winter roost for Miniopterus natalensis, Rhinolophus clivosus and R. capensis (19°C); 2 maternity chamber for Miniopterus natalensis (31°C); 3 winter roost for Nycteris thebaica (30°C); 4 maternity roosts for Myotis tricolor (21°C); 5 roost for Rhinolophus clivosus and R. capensis (21°C); 6 incidental roost sites for all five species (21°C).
Bats use caves and other cavities for protection from predators and to take advantage of the stable microclimate (temperature and humidity). The microclimate preferences of bats change during the year. In summer, bats select warm and humid roosts to help their young to maintain their body temperature. In winter, hibernating bats select cooler roosts, which allow them to decrease their body temperature (and consequently their metabolic needs) through torpor to conserve energy while insect availability is low. On the other hand, non-hibernating bats (e.g. Nycteris thebaica) will select warmer roosts during winter to reduce energy expenses needed to maintain a constant body temperature.
The changing microclimate needs of different species of cave bats are well demonstrated by the colony of five bat species occupying De Hoop Guano Cave in the Western Cape (McDonald et al. 1990a) (Figure 18). The different species, and even sexes of one species, use this cave at different times of the year and for different reasons. For example, during summer, a colony comprising up to 290,000 Miniopterus natalensis lives there. Males occupy cooler roosting sites near the cave entrance, while females and their young inhabit a domed ‘maternity chamber’ within the cave complex, where temperature and humidity are extremely high (31°C and 91% relative humidity) as a result of the combined body temperature of the bats. During winter, all the pregnant females migrate 100–200 km from De Hoop to caves in the interior of the Western Cape, where the lower ambient temperatures are necessary for hibernation. Males and non-pregnant females remain at De Hoop during winter, without entering deep hibernation.
De Hoop is also an important maternity cave for Myotis tricolor, but the maternity roost of this species is somewhat cooler and drier (21°C and 85% relative humidity) than for Miniopterus natalensis. Like M. natalensis, the females and young migrate to colder caves during winter in order to hibernate.
On the other hand, the non-hibernating Nycteris thebaica is present at De Hoop Guano Cave in autumn and winter, but not in summer. Since it cannot hibernate or enter torpor, this species overwinters in a warm (30°C) roosting position, just beyond the ‘maternity dome’ of M. natalensis.
Two horseshoe bat species, Rhinolophus clivosus and R. capensis, roost at De Hoop Guano Cave throughout the year, giving birth in December, but occur in much higher numbers during winter when Miniopterus natalensis and Myotis tricolor are absent. They particularly favour a cold (19°C) roosting site just beyond the first bend in the cave, which results in most individuals being in a torpid state for much of winter. Although they are frequently torpid during the day in winter, R. capensis and R. clivosus do not enter a deep prolonged hibernation and actively forage throughout the year. This is because the winter rainfall climate of the Western Cape ensures a good supply of insects during that season. During summer, the entrance site of the cave is also occupied by males and non-breeding females of Miniopterus natalensis and Myotis tricolor (McDonald et al. 1990a).
Crevice-roosting bats
Whilst