9420031 Coley One of the most striking features of tropical forests is their great diversity. The most widely cited explanation for the maintenance of tropical forest tree diversity is the Janzen-Connell hypothesis, which suggests that seed predators, herbivores and pathogens maintain diversity by preferentially attacking seeds or young trees that are close to conspecific adults. This difference in the rate of attack results in higher juvenile mortality near adults and increases the distance between successful conspecific recruits. Thus, the natural enemies of trees are proposed to prevent competitive dominance, which allows a greater diversity of tree species to persist. Despite the theoretical importance of the Janzen-Connell hypothesis, few studies have looked for the predicted pattern of higher seed or seedling mortality near adults. Even less attention has been directed at understanding the predicted mechanism of natural enemies centered on the adults. This study will critically evaluate this hypothesis by asking: 1) whether damage depends on distance from adult conspecifics, 2) whether damage negatively impacts growth and survival, and 3) whether the herbivore and pathogens responsible for the majority of damage are host specialists centered on adults. Understanding the mechanisms which control diversity of tropical forests has applications to management of rain forests as well as cultivated plantations. In order not to lose diversity in managed forest reserves, policies should take into account the Janzen-Connell effects. For example, encouraging recruitment through plantings or selective cutting should consider distance to nearest conspecifics. Monocultures, such as are typically found in plantations, tend to have extremely high levels of damage from insect pests. Preliminary work suggests that one method for limiting damage would be to have mixed planting, with a minimum of 50m between any two individuals of the same species.
