As one of the first potential bottlenecks to tree regeneration, fruit production fundamentally influences tree species composition. Reproductive output may be affected by the size of the individual and both abiotic and biotic environmental factors. The goal of this study is to investigate how tree size, soil nutrient availability, local crowding, and local dominance by conspecific trees influence fruit production among 11 northern hardwood forest species. To examine these factors, over two growing seasons I visually measured fruit production in crowns of approximately 1700 trees located across a natural fertility/productivity gradient in northwest lower Michigan. Following this method, fruit production for individual trees is measured prior to fruit dispersal and predation and thus more closely approximates reproductive effort than seed rain or seedling density. Influences on fruit production were tested by calibrating individual-based models of fruit production as functions of tree size, neighborhood crowding, local conspecific dominance, and soil resource availability.

Results/Conclusions

The smallest diameter at which fruit production occurs varies by species, ranging from 10.2 cm in Acer rubrum to 28.6 cm in Fraxinus americana, but was not related to species shade tolerance or soil fertility association. For the six species with substantial reproductive activity, diameter was a significant predictor of individual two-year fruit production, with the relationship being strongest in Fagus grandifolia. However, a great deal of the variation in fruit production remained unexplained, and many individuals did not reach the fruit production level that would be predicted by diameter. In Quercus rubra and Q. velutina, which were treated together since they naturally hybridize, conspecific relative basal area and soil ammonium were positively correlated with fruit production, suggesting that pollination efficiency and soil resources are important for reproduction. Except in Q. rubra/velutina, soil resources (base cations, phosphorus, and nitrogen), and neighborhood density (the number of trees within 5 m of each individual) were not correlated with fruit production. We offer the caveat, however, that none of the tree species that we examined mast-fruited during the first two years of the study; it remains likely that resources, neighborhood densities, and local conspecific density are strongly influential during years of peak reproduction.