The goal of most tropical reforestation is to develop planting strategies for fast-growing species that will eventually produce quality timber.  There are relatively few tropical reforestation studies that focus on ecosystem restoration, and as a result very little is known about community interactions in this applied setting, such as biotic factors that affect seedling growth.  Herbivory is a pervasive force in natural and managed settings, especially in the tropics, so there is a clear need for studies of plant-insect interactions in tropical reforestation systems.  We hypothesized that well-defended seedlings will suffer less herbivory and therefore grow more successfully than traditionally used fast-growing, but less defended species in reforestation.  We also hypothesized that diverse plots will have less herbivory because greater habitat complexity will impair herbivore host-searching.  We focused our study of plant defenses on saponins because their combination of anti-herbivore and anti-fungal properties may defend plants against herbivores and leaf-cutter ants, which use leaves for fungal gardens and are the primary pest in reforestation.  We ran a two-year reforestation experiment on lands converted to cattle pasture in the Atlantic lowlands of Costa Rica and examined plant growth and herbivory in mono- and polycultures of defended and undefended tree species. 

Results/Conclusions

After two years, we found leaf-cutter ants only cut leaves from saponin-poor species, and that the growth of these species was negatively affected by herbivory.  Polycultures also experienced less herbivory than monocultures, suggesting diverse plots of well-defended species should be focused on in reforestation designs.  These results highlight the importance of basic research in applied settings and suggest that consideration of plant secondary metabolites will be useful for designing future reforestation projects.