Using Janzen-Connell to predict the consequences of defaunation and other disturbances of tropical forests

The Janzen-Connell (J-C) model of tropical tree recruitment and diversity has come of age and can now be applied to predict the consequences of defaunation and other disturbances. J-C describes a process of recruitment at a distance that results from spatially varying rates of seed dispersal and subsequent survival. The per-capita success of seeds is low under reproductive conspecifics where propagules are killed by host restricted enemies (seed predators, herbivores and pathogens). Undispersed seeds consequently experience negligible success, whereas dispersed seeds benefit from escape from enemies. At our site in Amazonian Perú, a scant rain of dispersed seeds (<1/m2-yr for common species) gives rise to a low density of seedlings (ca. 5 m2), suggesting that intracohort interactions (density dependence, competition) are weak. Defaunation and other disturbances distort or curtail the vital processes of dispersal and propagule survival. Seed dispersal is most vulnerable to hunting and other disturbances because it is disproportionately carried out by large-bodied birds and mammals that are selectively harvested by hunters or that disappear from fragments. Reduced dispersal leads more or less directly to reduced recruitment. In contrast, compensatory adjustments to missing seed predators appear to be common. Thus the escape process of J-C is more resilient to distortions in the large vertebrate community than is seed dispersal. These principles appear to be robust and can be employed to predict the consequences of defaunation and other kinds of disturbances to tropical forests.