Of all anthropogenic pressures, urbanisation is one of the most damaging, and is expanding in its influence throughout the world. In Australia, 90% of the human population live in urban centres along the eastern seaboard. Before European settlement in the early 1800s, much of the Australia's East coast was dominated by forests. Many of the forest dependent fauna have had to adapt to forest fragmentation and habitat loss resulting from clearing for urbanisation. However, relatively few studies have investigated the impact of urbanisation on biodiversity. This is especially true for the remaining fauna in large metropolitan areas, such as Melbourne, Sydney and Brisbane. The physical and conceptual context of this thesis is the increasing impact of urbanisation and the potentially threatening factors to forest dependent fauna. Bats were selected because they comprise a third of Australia's mammal species, and therefore form a major component of Australia's biodiversity. Very little is known about the ecology and conservation biology of hollow-dependent bats in general, but particularly in urban environments. The study was conducted in Brisbane, south-east Queensland, one of Australia's most biodiverse regions. More than a third of Australia's bat species occur in this region. A large insectivorous bat, the white-striped freetail bat (Tadarida australis), was selected to study two key resources in this urban area - hollow availability and foraging habitat. This thesis also examined if artificial roost habitat could provide temporary roosts for white-striped freetail bats and other insectivorous bats and assessed whether these bat boxes can be used as a conservation tool in urban environments where natural hollow-availability is limited. The white-striped freetail bat is an obligate hollow-dweller and roosted largely in hollows of old or dead eucalypts throughout Brisbane's urban matrix. These roost trees harboured significantly more additional hollow-dependent species compared to control trees of similar age, height, and tree diameter. Roost cavities inside trees often exceeded 30 cm in diameter. Furthermore, maternity colonies used cavities of hollow trunks, which often extended into major branches, to roost in big numbers. Therefore artificial alternatives, such as small bat boxes, may provide temporary shelter for small roosting groups, but are unlikely to be suitable substitutes for habitat loss. Although five bat species used bat boxes during this study, the white-striped freetail bat was not attracted into bat boxes. Roost-switching behaviour was then used to quantify associations between individual white-striped freetail bats of a roosting group. Despite differences in gender and reproductive seasons, the bats exhibited the same behaviour throughout three radio-telemetry periods and over 500 bat-days of radio-tracking: each roosted in separate roosts, switched roosts very infrequently, and associated with other tagged bats only at a communal roost. Furthermore, the communal roost exhibited a hub of socialising between members of the roosting group especially at night, with vocalisation and swarming behaviour not found at any of the other roosts. Despite being spread over a large geographic area (up to 200 km2), each roost was connected to others by less than three links. One roost (the communal roost) defined the architecture of the network because it had the most links. That the network showed scale-free properties has profound implications for the management of the habitat trees of this roosting group. Scale-free networks provide high tolerance against stochastic events such as random roost removals, but are susceptible to the selective removal of hub nodes, such as the communal roost. The white-striped freetail bat flew at high speed and covered large distances in search for food. It foraged over all land-cover types found in Brisbane. However, its observed foraging behaviour was non-random with respect to both spatial location and the nature of the ground-level habitat. The main feeding areas were within three kilometers of the communal roost, predominantly over the Brisbane River flood plains. As the only mammal capable of flight, bats can forage above fragmented habitats. However, as this study showed, hollow-dependent insectivorous bats, including free-tailed bats, are specialised in their roosting requirements. The ongoing protection of hollow-bearing trees, and the ongoing recruitment of future hollow-bearing trees, is essential for the long-term conservation of these animals in highly fragmented landscapes. Furthermore, loss of foraging habitat is still poorly understood, and should be considered in the ongoing conservation of bats in urban environments.

Day-roosts are an essential resource for tree-hole roosting microbats (Microchiroptera), providing shelter, protection from predators and an appropriate microclimate for energy conservation and reproduction. Microbats often make use of multiple roosting sites, shifting between roosts frequently. Conservation of tree-hole roosting microbats requires an understanding of roost selection and fidelity to enable the protection of sufficient suitable roosting sites. In Australia, as in other countries, habitat loss, particularly in the form of large hollow-bearing trees, is threatening the survival of microbat populations. In addition, the renewal of natural roosts in Australia is very slow, as trees may need to be 100 years old for hollows to form. Where roosting resources are limited, such as in urbanised areas, batboxes may be used as a substitute. As bat-boxes are also accessible to researchers, these roosting sites can help to improve our understanding of roosting behaviour. / This thesis investigates the roosting behaviour of two sympatric microbat species: Gould’s wattled bat (Chalinolobus gouldii) and the white-striped freetail bat (Tadarida australis). These are insectivorous tree-hole roosting species, which naturally occur in urban Melbourne, Australia. Both species make use of bat-boxes at three sites in Melbourne, often sharing roosts with members of the other species. This provided an opportunity not only to study their use of bat-boxes for conservation management purposes, but to investigate factors influencing bat roost selection and fidelity. This study incorporated PIT tags (microchips) and a detector array at the bat-boxes, in addition to monthly manual bat-box inspections, as a method for monitoring roost-use. This approach enabled the collection of long-term, fine-scale roosting data. These data, along with captive and field-based experiments were used to examine the influence of parasites, microclimate and social structure on roost selection patterns and roost fidelity. The specific questions posed were whether tree-hole roosting bats: select roosts based on physical characteristics; perceive a cost of carrying ectoparasites and avoid infested roosts; select roosts to maintain social associations; and select for specific beneficial microclimates. / The patterns of roost selection, ectoparasite diversity, social structure, and the selection of roost microclimate differed between the two species. Microclimate of the bat-boxes was a strong influence on roost selection for both species, as it is for microbats generally. White-striped freetail bats preferred warmer roosts with stable humidity. For Gould’s wattled bats, the selection of roost microclimate differed between the sexes and even between separate, but adjacent, roosting groups. Patterns of preference indicated that individuals had knowledge of the available roosting sites. / The presence of parasites had no obvious influence on roost selection patterns in either species. The white-striped freetail bat was found to support lower ectoparasite diversity, which may be influenced by characteristics of the pelage and may partially explain why parasite load was not a useful predictor of roost selection in this species. In contrast, Gould’s wattled bat supported a larger diversity of ectoparasites, which showed clear patterns of distribution through the bat populations, and intra-specific and spatial variability. A radio-tracking study indicated that parasites in the roost and on the Gould’s wattled bat may influence their roosting behaviour. Additionally, experimental assessments of the bats’ grooming response to parasites indicated that the perceived costs of these parasites differed with parasites that remained permanently attached to the host eliciting a stronger response than those also found in the roost. The defensive mechanism against parasites that completed part of their life-cycle in the roost was expected to be avoidance behaviour, yet, in both captive and field experiments, these parasites did not strongly influence roost selection or fidelity. / Social associations among white-striped freetail bats appeared to be random, and did not explain roosting patterns. This may reflect the restricted sampling of roosting sites, and the possible role of the bat-boxes in this study as ‘satellite’ roosts, separate from a larger communal roost, likely to be in a large tree-hollow. Unlike white-striped freetail bats, Gould’s wattled bats showed fission-fusion social structure, driven by stronger female associations. The distribution and abundance of parasites was correlated with the social structuring of the host species, and host selection appeared to facilitate transmission. These patterns suggest that female Gould’s wattled bats, in particular, are choosing roosts based on the benefits of social association despite the cost of increased parasite risk, and may provide an explanation for sexual segregation in temperate tree-roosting bats. / This study demonstrates the species-specificity of roosting behaviour, and the importance of investigating several factors that influence roost selection, to better understand roost requirements. It also highlights the inherent complexity in roost selection by tree-hole roosting microbats, which may be making trade-offs between the benefits of social associations and the cost of parasitism, as well as choosing an optimal microclimate. Further investigation into interactions between these factors will greatly advance our understanding of roost selection and fidelity in tree-hole roosting bats.