Multi‐Level Interactions Arising from Herbivory: A Simulation Analysis of Deciduous Forests Utilizing Foret

We examined potential effects of herbivory on temperate forest ecosystems using FORET, a forest succession simulation model with the capacity for treating various hierarchical levels for long time periods. Two species of trees were chosen for evaluating herbivory effects: white oak (Quercus alba), a relatively slow—growing shade—tolerant species, and tulip poplar (Liriodendron tulipifera), a fast—growing shade—intolerant species. Both are dominants in the Southeastern U.S. forest system selected as a basis for this study. The study focused on four hierarchical levels, covering individual, phenotypic, interspecific, and community interactions. Simulations of herbivory were arrived at by stochastically varying annual incidence that herbivores were present on each simulation plot over a 500—yr period, and by modifying the proportion of energy allocated to either tree growth, or to its defense from herbivore attack. Two hypothetical tree phenotypes were considered, one that allocated specified amounts of energy to herbivore defense mechanisms at all times, and the other that made the allocation only when herbivores were present according to the stochastic determination made for a specific study period. Thus, one phenotype was fixed in its life history strategy; the other was given a facultative strategy where it switched its growth and defense tactics as a function of variation in herbivore presence. The results of the simulations suggest how deciduous forests may respond to long—term variations in the intensity of herbivore stress on two dominant tree species, and show the importance of hierarchical relationships within the community. White oak tended to show a greater sensitivity to interspecific interactions; tulip poplar showed a higher sensitivity to intraspecific interactions. Changes in growth rates associated with the switching strategies (an ultimate factor) were more important in answering variations in productivity than was impact imparted by annual changes in incidence of herbivore presence (a proximate factor) for these two species, although there were important differences in several statistical interactions. While our results suggest that herbivore stress can explain a larger degree of the variation in long—term community dynamics, ecological interactions between herbivore and climate effects must be more closely linked in such long—term studies.