Abstract
Understanding the extent and causes of insect diversity in the humid tropics is one of the major challenges in modern ecology. We review some of the current approaches to this problem, and discuss how future progress may be made. Recent calculations that there may be more than 30 million species of insect on earth have focused attention on the magnitude of this problem and stimulated several new lines of research (although the true figure is now widely thought to be between five and ten million species). We discuss work based on insecticidal logging surveys; studies of herbivore and parasitoid specificity; macroecological approaches; and the construction of food webs. It is argued that progress in estimating insect diversity and in understanding insect community dynamics will be enhanced by building local inventories of species diversity, and in descriptive and experimental studies of the trophic structure of communities. As an illustration of work aimed at the last goal, we discuss the construction and analysis of quantitative host-parasitoid food webs, drawing on our work on leaf miner communities in Central America.