Tropical forest succession: Integrating theory and application in forest restoration.

摘要

During the next century most remaining unprotected tropical forests will be cleared for short-term agriculture or pasture and abandoned. Forests have the natural capacity to regrow after clearing via succession. However, forest succession is often slow or absent on degraded lands, and our limited understanding of succession hinders the design of successful restoration strategies. To address this problem I studied tropical forest succession in Kibale National Park, Uganda, within large sections of forest converted to agriculture then abandoned early last century. These areas became grasslands that persisted until mid-century when timber plantations were established in their place. Unlike grasslands, maturing plantations facilitated native tree and shrub recruitment in their understories. Plantation harvesting began in 1993, and young successional forests growing in their place were the focus of my study. First, I studied the interplay between post-disturbance animal-mediated seed dispersal and seedling recruitment. This interplay was complex and dependent on the initial onsite availability of resources for succession. Bird seed dispersal had far more impact on plant recruitment than mammal seed dispersal, though few seedlings established relative to numbers of seeds dispersed. Second, I explored the influence of plant-plant interactions (competition and facilitation) on forest succession. Correlative analyses from vegetation surveys suggested interactions between growth forms (shrubs and trees) and size classes (seedlings and saplings) were often strong. However, these interactions were complex and varied among successions. Experimental removal of shrubs or all non-tree vegetation failed to enhance community-wide tree recruitment or survival, though some species responded strongly. Third, I explored restoring natural forests on degraded tropical lands using timber plantations. Overall forest regrowth was dependent on availability of native stems and propagules surviving logging. While logging killed many native stems, it accelerated forest succession relative to that in unlogged plantations. Several attempts to enhance forest succession in logged areas were moderately successful, with some experiments suggesting species-level interventions may be more successful than community-wide interventions. Finally, fire-exclusion appears faster than using plantations to promote forest succession in grasslands. In each of these studies I propose guidelines for managing successions based on my findings.