Enemies in space: An examination of the Janzen-Connell model in the presence of habitat partitioning

Background/Question/Methods

Two of the major hypotheses of diversity maintenance in tropical forests are the Janzen-Connell hypothesis and habitat partitioning.  The Janzen-Connell hypothesis is that coexistence is promoted by spatially localized specialist natural enemies.  Habitat partitioning is the hypothesis that species respond differently to different habitats in a way that promotes coexistence.  Previous work has shown that coexistence mechanisms often interact.  Additionally, Janzen-Connell predicts that seedlings will have the highest survival rate far from conspecifics, while habitat partitioning predicts seedlings will do best near conspecifics, suggesting additional interactions.  These issues were investigated using a site-occupancy model, where seedlings competed for gaps.  Seedling performance was affected by specialist natural enemies and spatially varying environmental conditions.  We analyzed this model using an invasion analysis, in which the growth rate of a species was assessed at near zero density, when its competitors were at equilibrium.  If species gained an advantage at this low-density state, this meant that coexistence was promoted. 

Results/Conclusions

Surprisingly, localized natural enemies have no important interaction with the effects of habitat partitioning.  Instead, their effects combined additively.  Near-zero density species gained an advantage from Janzen-Connell because their natural enemies tended to be at low densities as well.  Near zero density species gained an advantage from habitat partitioning because they could better take advantage of good locations.  However, strong habitat partitioning can allow seedling to perform better in areas of high predator density, making Janzen-Connell effects difficult to detect by standard means.  Unlike conventional expectations, spatially localized enemies are worse at promoting coexistence than mobile or diffuse natural enemies.  This occurs because extremely rare species will still suffer high specialized predation rates if non-dispersed seeds experience a high predation risk.  Generalist natural enemies promote coexistence as long as they have a differential effect on hosts, and habitat partitioning does not alter this.