During a 4-yr phenological study of the palm flora of Barro Colorado Island, Panama, palm species displayed substantial variation in reproductive phenology despite a shared uniform growth architecture. Reproduction in the majority of species was distinctly seasonal: most, but not all, initiated flowering sometime during the rainy season. In two species, individuals exhibited multiple flowering episodes. Palm species differed in the degree of within-year flowering synchrony but were relatively predictable in flowering time from year to year. Especially within asynchronously flowering species, fruiting was more synchronous than flowering, in part as a result of the failure of individuals flowering at certain times to produce mature fruit. Levels of flowering within each species were generally high except in the clonal Bactris palms, and it appeared for some species that canopy openness may stimulate flowering. Palm vegetative phenology was also seasonal and appeared more responsive to annual climatic variation than did reproductive phenology. In many species, rates of leaf expansion slowed during the dry season, an effect that was exaggerated during the prolonged El Ninio dry season of 1982-1983; however, a species' total annual leaf production was relatively constant from year to year. Leaf production correlated with individual height in some species, but was generally not related to the degree of canopy openness. The underlying significance of the variation in reproductive phenology within this palm flora remains largely unexplored. MANY TROPICAL PLANT COMMUNITIES display conspicuous seasonal patterning in vegetative and reproductive phenologies at both community and species levels (Fournier & Salas 1966, Medway 1972, Frankie et al. 1974, Monasterio & Sarmiento 1976, Foster 1982), despite the apparent absence of the very dramatic photoperiodic and climatic fluctuations that regulate temperate zone plant phenologies. Although alternative mechanisms such as rainfall patterns have been suggested as the phenological signals for tropical species (e.g., Opler et al. 1976, Augspurger 1982), the resulting phenological patterns are presumed to represent adaptive responses, whatever the proximate cues may be. Adaptive strategies for successful pollination and seed dispersal are frequently invoked to explain plant reproductive timing (e.g., Janzen 1967, Stiles 1978, Augspurger 1981); however, Borchert (1983) has cautioned that constraints of physiology and vegetative morphology must be considered before interpreting reproductive phenologies from the standpoint of biotic interactions with pollinators and seed dispersers. One group of tropical plants, the palms, offers an opportunity to evaluate morphological constraints on reproductive phenology because their growth pattern is essentially uniform. The monopodial palm stem grows continuously from a single meristem and produces new leaves sequentially, with a corresponding sequential death and shedding of old leaves (Corner 1966). A single axillary inflorescence bud can be initiated at each leaf node at the time of leaf formation, and flowering commences upon the rapid enlargement and expansion of these preformed buds. This shared growth pattern may limit the range of possible reproductive phenological patterns among palm species, and a continuous leafing phenology would provide the potential for aseasonal reproductive activity throughout the year because of the direct association between inflorescence buds and developing leaves. We examined these potential restrictions upon reproductive phenology in a 4-yr study of the palm flora of a tropical lowland forest on Barro Colorado Island, Panama. Here we address three questions about palm phenologies: (1) is vegetative growth seasonal or aseasonal; (2) is reproduction correspondingly seasonal or aseasonal; and (3) does reproductive phenology vary among species, or is it relatively consistent from the phylogenetic constraints imposed by similar growth morphology. In considering the phenologies of members of a defined taxonomic group within a community, we focus our study at an intermediate level between whole-community surveys (e.g., Frankie et al. 1974) and studies of single species (e.g., Augspurger 1981). Our analysis is performed at the levels of the individual plant and plant population; inflorescence phenology as it relates to breeding systems (e.g., Bullock 1981, Beach 1984) is not considered here. We utilize quantitative data from sample populations of each species to measure relative synchrony and seasonal constancy in phenological timing and to assess the reproductive performance of these species populations. I Received 2 November 1985, revision accepted 9 April 1986. 2 Present address: Department of Biological Sciences, University of Wisconsin, Milwaukee, Wisconsin 53201, U.S.A. 3Present address: Department of Botany, University of Florida, Gainesville, Florida 32611, U.S.A. 342 BIOTROPICA 19(4): 342-356 1987 This content downloaded from 157.55.39.132 on Thu, 15 Sep 2016 04:54:48 UTC All use subject to http://about.jstor.org/terms