Selection for Optimal Fruit-Crop Size in Bird-Dispersed Plants

Recent investigations of plant-frugivore interactions have failed to demonstrate the expected peak in fruit-removal rates (plant fitness) at intermediate or high crop sizes. In three species of Costa Rican forest understory plants, I found that the likelihood of bird visitation increased with increasing crop size. Among those plants visited by birds, the absolute number of fruits removed increased with increasing crop size, whereas the proportion, or relative number removed, actually decreased. Thus, effects of crop size on the different components of the dispersal process (likelihood of visitation by birds, fruit removal by those that do visit, and postremoval behavior of dispersers) often conflict, such that no clear relationship between crop size and overall dispersal success is evident. Results presented here suggest selection for synchronous ripening of large seasonal fruit crops in these plants. Nevertheless, I propose that models of the relationship between crop size and reproductive success in trees (Howe and Estabrook 1977) are largely inappropriate for understory plants for two reasons. First, because of their small stature and small fruit crops, understory herbs and shrubs are often not detected by frugivores. Second, the fruit crops they produce are rarely large enough to encourage sedentary behavior in dispersers. Finally, I suggest that the best estimator of fitness differs for plants with different life history strategies. In most plants, fitness is best approximated by the absolute number of fruits removed. In highly iteroparous species, the relative number removed may be a more suitable estimator, but only if increased present reproductive effort results in decreased adult survival or future reproduction.