REPRODUCTIVE PHENOLOGY OF A TROPICAL DRY FOREST IN MUDUMALAI, SOUTHERN INDIA

1 Flowering and fruiting phenologies of a tropical dry forest in Mudum alai, southern India, were studied between April 1988 and August 1990. Two sites, a wetter site I receiving 1100mm and a drier site II recei ving 600mm of rainfall annually, are compared. A total of 286 trees fr om 38 species at site I and 167 trees from 27 species at site II was m arked for phenological observations. There were 11 species common to t he two sites. Several hypotheses relating to the evolution of reproduc tive phenology are tested. 2 Frequency of species flowering attained a peak at site I during the dry season but at site II, where soil moist ure may be limiting during the dry months, the peak was during the wet season. At both sites a majority of species flushed leaves and flower ed simultaneously. Among various guilds, the bird-pollinated guild sho wed distinct dry season flowering, which may be related to better adve rtisement of large flowers to pollinators during the leafless dry phas e. The wind-pollinated guild flowered mainly during the wet season, wh en wind speeds are highest and favourable for pollen transport. The in sect-pollinated guild showed no seasonality in flowering in site I but a wet season flowering in site II. 3 Fruiting frequency attained a pe ak in site I during the late wet season extending into the early dry s eason; a time-lag correlation showed that fruiting followed rainfall w ith a lag of about two months. Site II showed a similar fruiting patte rn but this was not statistically significant. The dispersal guilds (a nimal, wind, and explosive passively-dispersed) did not show any clear seasonality in fruiting, except for the animal-dispersed guild which showed wet season fruiting in site I. 4 Hurlbert's overlap index was a lso calculated in order to look at synchrony in flowering and fruiting irrespective of climatic (dry and wet month) seasonality. In general, overlap in flowering and fruiting guilds was high because of seasonal aggregation. Among the exceptions, at site II the wind-pollinated flo wering guild did not show significant overlap between species although flowering aggregated in the wet season. This could be due to the need to avoid heterospecific pollen transfer. 5 Rarer species tended to fl ower earlier in the dry season and this again could be an adaptation t o avoid the risk of heterospecific pollen transfer or competition for pollinators. The more abundant species flowered mainly during the wet season. Species which flower earlier have larger flowers and, having i nvested more energy in flowers, also have shorter flower to fruit dura tions.