Staggered flowering in the Dipterocarpaceae: new insights into floral induction and the evolution of mast fruiting in the aseasonal tropics

The supra-annual mast fruiting of dipterocarps, dominant trees of Far Eastern lowland rain forests, has been considered an adaptation to reduce seed predation through satiation. However, this explanation raises four important questions. (1) What pollinates species that flower at intervals of more than one year? (2) How do mass-flowering species avoid competition for pollinators? (3) What is the environmental cue for floral induction that induces scores of species in several families to flower at irregular intervals of more than one year in an aseasonal climate? (4) And, most important, what factors could cause an initial phenological aggregation of species' fruiting times? Once such a phenological core of species exists, selection for predator satiation could entrain other species into the same fruiting schedule and perfect the synchrony of seed release. The essential question is how such a phenological core of species, flowering at irregular intervals in an aseasonal climate, could initially arise. The observations and analyses presented here indicate that the answers to these questions are intricately related to each other and to the El Nino-Southern Oscillation phenomenon. (1) Several mass-flowering dipterocarps are pollinated by thrips, which persist at low levels between mass flowerings and can explode in density as the dipterocarps come into flower. (2) Mass-flowering dipterocarp species in Shorea section Mutica significantly reduce the overlap in flowering times by staggering their flowering periods, thereby reducing competition for pollinators and/or the clogging of stigmas with foreign pollen. The new technique used to analyze whether a particular pattern of resource use significantly reduces overlap overcomes an error inherent in previous techniques, and it should be used whenever the total flowering season or range of resources is determined not by external conditions, but by the range of flowering times or resources used by the species in question. (3) The time of floral induction in Shorea section Mutica is derived using the linear relationship between the duration of each species' flowering period and mean flowering date. An analysis of meteorological records suggests that induction is caused by a drop of roughly 2$^\circ$C or more in minimum nighttime temperature for three or more nights. This extraordinary trigger is probably the most reliable signal associated with an invasion of the aseasonal tropics by a dry air mass. (4) Such a trigger