TROPICAL RAINFOREST GAPS AND TREE SPECIES DIVERSITY

Evolutionary hypotheses about how so many species of tropical rainforest trees might have arisen include (a) genetic drift (71), (b) habitat specialization (8) in benign environments, or (c) repeated geographic isolation followed by remixing of species during Pleistocene climatic fluctuations (144). Ecological hypotheses about how these species continue to coexist are often cast into equilibrium or non-equilibrium frameworks: Do tropical forests comprise . . . sets of highly coevolved niche-differentiated tree species in stable or semistable floristic assemblages, or do they consist of . . . diffusely coevolved, broadly generalist species which slowly drift in relative abundance within a few large life-history guilds (67, 102)? In this context, there has been considerable recent interest in the role of adaptations by species to different regeneration sites in structuring plant assemblages in general (83, 84, 86, 141), and tropical tree communities in particular (62, 89, 90, 150). The immediacy of this interest is heightened by rising rates of deforestation throughout the tropics and a critical need for management strategies of the remaining preserves, for ecologically sound harvesting procedures, and for the tools with which to restore degraded forests. Openings in the forest canopy are widely recognized as important for the establishment and growth of rainforest trees (5, 33, 62, 90). Hartshorn (89) suggests that perhaps 75% of the tree species at La Selva Biological Station, Costa Rica are dependent on canopy opening for seed germination or for growth beyond sapling size. Similar statements are found in descriptions of forest dynamics in Queensland, Australia (185), Malaysian dipterocarp forests (195), and West African rain forest (108). Demographic studies demon-