A seven-year study of individual variation in fruit production in tropical bird-dispersed tree species in the family Lauraceae

Lifetime patterns of fruit production, like other features of a plant's reproductive biology, have been molded by natural selection by seed dispersers over thousands of generations. At least, that is what most researchers interested in frugivory and seed dispersal assume. It is daunting to recognize, however, that we make such an assumption in the absence of crucial information on the heritability of reproductive traits, or on the strength of selection imposed by different kinds of interactions with seed dispersers. We know very little about the scheduling of reproduction or the magnitude of annual and individual variation in fruit production in most tropical tree species (Janzen, 1978). Seldom do we know if we are witnessing a 'normal' year, or even what a 'normal' year is in terms of plant reproduction. The earliest systematic studies on flowering and fruiting in tropical plants were conducted only recently (McClure, 1966; Medway, 1972; Frankie et al., 1974; Hilty, 1980; Opler et al., 1980). The work was directed chiefly at determining broad withinyear patterns such as .the number of species flowering or fruiting within a given month or forest stratum. From these descriptive studies, the focus of research shifted to the question of how competition for pollinators or seed dispersers might select for staggered phenologies within years (Frankie, 1975; Stiles, 1977; for a review see Wheelwright, 1985a) or how the timing of fruit production related to the Abstract. Fruit crop sizes varied from year to year among 22 sympatric, bird-dispersed tree species in the Lauraceae. Each species in the lower montane forests of Monteverde, Costa Rica fruited at a characteristic season, but there was wide yearto-year variability in the porportion of each population that produced fruit and in the average size of fruit crops. Over a 7-year period (1979-1985), overall fruit production was high during three nonconsecutive years and low during four years. Within genera, tree species displayed distinct fruiting schedules. Even within populations, individual trees ~ometimes fruited in different years or failed to fruit altogether. Yearly rainfall and temperature patterns did not explain annual variation in fruit production. Unexpectedly, neither did previous reproductive histories: there was little correlation between an individual tree's fruit production in a given year and its fruit production the previous year. On the other hand, vegetative growth was negatively correlated with reproduction in 12 of 15 species. Lauraceous fruits make up 60-80% of all fruits eaten by bird specie$ such as Three-wattled Bellbirds and Resplendent Quetzals. These birds may respond to annual variation in the availability of lauraceous fruits by migrating locally, by expanding their diets to include previously ignored foods or unripe fruits, or by delaying breeding.