Tuesday, August 5, 2008 - 4:00 PM

COS 41-8: Three years of observations of the flowering phenology of the tropical dry forest tree Enterolobium cyclocarpum (Guanacaste)

Oscar J. Rocha, Kent State University, Carlos Gomez, Universidad de Costa Rica, and James L. Hamrick, University of Georgia.

Background/Question/Methods . We studied the flowering phenology of the dry forest tree Enterolobium cyclocarpum (Guanacaste) for three consecutive years in Costa Rica. We monitored the flowering phenology of 94 trees using a semi-quantitative scale, where trees were given a score every week based on the percent of the crown bearing flowers (0 - no flowers, 1 - open flowers in less than 25% of the crown, 2 - open flowers in 25% to 50% of the crown, 3 - open flowers in 50% to 75% of the crown, and 4 - open flowers in more than 75% of the crown). Our data revealed significant differences in the flowering behavior among trees. We arbitrarily defined four classes of phenological profiles on basis of the magnitude and duration of the flowering episode. It was considered that trees bearing flowers in more than 50 % of the crown were more likely to attract pollinators, and hence received and exported more pollen than trees producing flowers in less than 50 % on the crown. Therefore, trees were grouped according to the duration of time during which they beard flowers in more than 50% of the crown. We examined the consistency of the flowering profiles exhibited by each tree during the three flowering episodes.

Results/Conclusions . Our analysis revealed that there are significant differences between the expected and the observed number of trees that showed the same profile for three years, the number of trees that showed the same profile for at least two of the three years, and the number of trees that showed a different profile each year (Table 3) (χ2 = 42.08, d. f. = 5, p < 0.001). Overall, there were significantly fewer trees showing a different profile each year, significantly more trees showing the same profile all three years, and significantly more trees showing the same profile at least two of the three years. Our data also indicated that there were significant differences in fruit production among trees, and we found a significant correlation between the phenological profiles observed in 2005 and the fruiting score observed in 2006 (r2 = 0.38, p < 0.0001, n = 93) and between the phenological profiles observed in 2006 and the fruiting score observed in 2007 (r2 = 0.29, p = 0.005, n = 93). Overall, trees with moderate and long flowering profiles tend to produce more fruits than trees with weak and short flowering profiles.