Despite increasing publicity, habitat loss and fragmentation remain a serious threat to biodiversity. The main objectives of this research were (i) to study the effects of forest fragmentation on the distribution and abundance of resident birds in the fragmented littoral forests of southeastern Madagascar and temperate woodlands of southeastern Australia at three spatial scales (patch, landscape and regional) and (ii) to place the results of these case studies within the realms of the equilibrium theory of island biogeography and its descendent theories, to ascertain whether it is appropriate to use these theories to derive conservation scenarios within these threatened regions. Deforestation of Madagascar's remaining forests is considered a global concern due to both its current high intensity and the small amount of forest claimed to be remaining on the island. Surprisingly, very few studies have considered the impacts of forest fragmentation on bird diversity in Madagascar. A multi-scale study on the effects of littoral forest fragmentation and degradation on birds is therefore a major step forward for bird conservation on the island. Furthermore, prior to this study no known work has been conducted on the avifauna within the threatened littoral forests of eastern Madagascar. My results indicated that (i) the littoral forests contained bird species assemblages that were unique when compared to neighbouring forest habitats, (ii) many forest-dependent bird species were significantly affected by habitat structure and especially proximity to forest edge and (iii) many forest-dependent species were affected by landscape factors such as remnant shape and remnant size. No relationship was found between measures of landscape composition, remnant 'isolation' and bird distribution within littoral forest remnants.