Files
Abstract
Tropical wet forests support one of the most diverse terrestrial ecosystems, yet are currently subject to forest fragmentation globally. Despite such prevalence, studies in the tropics come predominantly from lowland forests. Forest fragmentation effects may differ regionally and applying insights from a few well-studied sites elsewhere may be misleading. Inequality also exists in fragmentation effects on different functional groups. Regeneration of animal-dispersed, shade-tolerant tree species are considered most vulnerable to forest fragmentation but light requirements may change ontogenetically, and our understanding of how forest fragmentation may affect different growth stages is limited. I studied various stages of regeneration of five animal-dispersed tree species in tropical premontane wet forest fragments in Costa Rica. First, I assessed forest fragmentation effects on various early regeneration stages and demonstrated that not all stages are equally affected, but rather effects are limited to particular times. I also compared germination of seeds from small and medium-sized fragments and showed consistently higher germination from small fragments, in contrast to results from most previous studies. Then, I ranked relative light requirements of species whose life-history traits were unknown and tested whether species with lower light requirement are more negatively affected by forest fragmentation for all size classes. There was evidence for ontogenetic effects of forest fragmentation. The general prediction that species with low light requirement are more vulnerable to forest fragmentation did not apply at an individual-species level. Finally, I assessed the biological aspect of species regeneration by testing the Janzen-Connell hypothesis in forest fragments. The Janzen-Connell hypothesis was supported for all non-pioneer species but only for the smallest size class, and the underlying mechanisms proposed by the Janzen-Connell hypothesis were largely rejected. While mixtures of species with different life histories may result in complex spatiotemporal plant dynamics, the spatiotemporal dynamics of some species may be predictable based on positive distance-dependent survivorship. To conclude, my results emphasize the importance of studying various regeneration stages and individual sizes in underrepresented study systems. Such studies should provide better understanding of the species and underlying mechanisms for species regeneration in fragmented landscapes, allowing significant interface between biological understanding and conservational applications.