"Ecological studies in the humid tropics have been concerned almost entirely with forest communities or human impacts on forests. Studies of the ecology of wild species in deforested, human-dominated landscapes are rare, in contrast to the non-tropical literature. While a focus on the few remaining near-pristine sites makes sense, the almost total neglect of human-dominated landscapes does not. First, there is nothing else in an increasing proportion of the humid tropics, and regional endemics must survive, if at all, in the interstices between the predominant agricultural or urban land-uses. Second, studies of such landscapes are needed to assess the degree to which biodiversity is dependent on the protection of near-pristine habitats. Reserve management could then be targeted at those species that most needed it. Third, minor changes in current management practices may increase the conservation value of these landscapes, further reducing the burden on the reserve system. Fourth, in long-settled regions, the least useful parts of the landscape have often been abandoned. It may be possible to rehabilitate such areas and restore at least some elements of the original biota. Finally, human-dominated landscapes are where most tropical people live and most encounters between people and wild species occur. Support for conservation is more likely from people who know and appreciate their local wildlife. These points will be illustrated by examples from the highly degraded but still biodiverse landscape of Hong Kong."