How natural enemies affect plant population dynamics is an important basic and applied question. While we know that herbivory can affect individual plant performance and population growth, few good studies test how herbivory might contribute to plant population regulation. Addressing this question requires characterizing herbivore effects on density dependence and using mathematical models to understand how these effects influence long-term population dynamics. In this study we carried out a field experiment crossing five densities of a weedy native perennial plant (Solanum carolinense) with herbivore exclusion and inclusion treatments. We gathered demographic data on the plants in the experiment over four years (three yearly transitions) and used these data to parameterize a density dependent matrix model. Using this model we asked 1. Whether insect herbivores influenced which transitions were density dependent, 2. Which parts of the plant life-cycle are most influenced by herbivores and 3. Whether herbivores alter the predicted equilibrium population size.
Results/Conclusions
We found that the demographic transitions that were density dependent changed with the presence of herbivores, but also across years. In one of the three transitions we observed equivocal evidence for density dependence. In the other two transitions the presence of herbivores tended to strengthen density dependence of plant demographic transitions. Although in previous work we have found that herbivore damage decreases individual plant growth, we found that herbivores increased the predicted equilibrium plant population size (number of stems). This occurred because the distribution of stem sizes shifted toward a greater number of smaller stems in the presence of herbivores. While herbivores increased the predicted equilibrium number of stems, they thus decreased predicted equilibrium biomass. These results suggest that in our system herbivores do influence the structure of density dependence in the plant population. While characterizing density dependence is often logistically difficult, our study suggests that this information is important for both our basic understanding of plant ecology, and for predicting outcomes of management such as releasing biocontrol agents.