Natural enemy specificity and maintenance of host diversity under demographic stochasticity

Dipanjana Dalui, Ecology and Evolutionary Biology, University of Connecticut, Storrs Mansfield, CT, Annette M Ostling, Ecology and Evolutionary Biology, University of Michigan, Ann Arbor, MI and Robert Bagchi, Ecology & Evolutionary Biology, University of Connecticut, Storrs, CT

Background/Question/Methods
Understanding how species coexist is a fundamental challenge in ecology. A leading hypothesis to explain coexistence, the Janzen-Connell hypothesis, proposes that seedlings experience disproportionate mortality in areas of high conspecific densities as a result of specialised natural enemy attack. Theoretical models of coexistence to date have mostly assumed complete host specificity of natural enemies, which does not concur with empirical data on insect and pathogen host choice. Furthermore, host specificity will limit the resources available to natural enemies, potentially constraining their populations and exposing them to a higher risk of extinction. This potential for extinction of highly specialised natural enemies may be particularly important in stochastic systems with many rare hosts (e.g., tropical rainforests). The absence of stochasticity and unrealistic assumptions about natural enemy specificity might reduce the relevance of theoretical work to real-world situations.

We focus on the implications of demographic stochasticity of a community of host-natural enemies under varying host specificity to assess community persistence. We supplement the simulations with theoretical deterministic analysis that helps us isolate the contribution of natural enemies and stochasticity to maintenance of species coexistence.
Results/Conclusions
Our results indicate community persistence in a narrow range of specificities, and high host specificity does not translate into longer term stable coexistence of the community. Theoretical analysis of the deterministic model of a two-host-two-pathogen system suggests that the coexistence point is not an attractor, but adding demographic stochasticity leads to coexistence of all species. Furthermore, demographic stochasticity and natural enemy presence are able to maintain community diversity in a range of parameter space where host resource competition alone is unsuccessful in maintaining coexistence. In conclusion, our work suggests that whereas natural enemy enabled coexistence is an important mechanism, host specificity may not be as crucial as thought to be. A consideration for demographic stochasticity is key to understanding the host-pathogen interactions shaping community dynamics.