Conspecific negative density-dependence (CNDD), or Janzen-Connell effects, are widespread and thought to promote tree diversity, particularly in hyper-diverse tropical forests. It has been further hypothesized that communities will be more stable if less abundant species are more susceptible to CNDD than common ones. However, that hypothesis was based on a mean-field model of a two species community, though both diversity and dispersal-limitation have been shown to alter the impact of CNDD. Here we reexamine the effect of interspecific variation in CNDD. We build a site-occupancy model, in which species compete for canopy gaps. To represent CNDD, we assume that seeds that disperse under a conspecific adult are less likely to recruit. We assume that the strength of CNDD varies between species. We explicitly model dispersal limitation.
Results/Conclusions
We find that variation in CNDD has two effects, both of which generally make communities less stable. First, in multi-species communities, the most common species are often the least regulated by CNDD; this reduces the stabilizing benefit of CNDD, since a species who becomes rare (and experiences low mortality because it is rare) has little advantage over a common species who is immune to CNDD. Second, if dispersal is limited, then seedlings who are immune to CNDD gain an advantage in capturing sites held by their parents, putting species who are susceptible to CNDD at a competitive disadvantage (i.e., this creates mean fitness-differences). Variation in CNDD also produces large disparities in abundance, making weak competitors more prone to demographic stochasticity. We use our model to reanalyze data from Barro Colorado Island, Panama, which suggest that the fitness effects of CNDD may outweigh its stabilizing effects. Thus, our model suggests that if CNDD varies, CNDD may make communities less stable than if predators are absent.