The current knowledge about the influence of plant traits and climatic seasonality on the phenology of tropical rain forest plants is insufficient to understand how functional traits can explain mechanisms associated with phenological responses. We hypothesized that the density of the wood and the water stored in stem and root determine presence of phenological groups of trees. Leaf shedding, flushing, and flowering of high wood density species are related to rainfall, due to their low water-storage capacity. In contrast, leaf unfolding and flowering of low wood density species depend on photoperiod, due to their high water-storage capacity. For 12 months, we monitored vegetative and reproductive phases of 281 individuals of 46 tree species in a coastal rain forest in Northeastern Brazil. Data on rainfall, photoperiod, wood density and quantity of water stored in stem and roots were also recorded. Corroborating our hypothesis, wood density and the amount of water stored formed functional phenological groups: Species of high wood density and low water-storing capacity and low wood density and high water-storing capacity. Abiotic variables did not demonstrate relationships with the phenological events and revealed that the phenophases of leaf unfolding, flowering and fructification occurred previously in the low wood density and high water-storing capacity species. Thereby, species of high wood density and low water-storing capacity had these phenophases later. Although the hypotheses have not been fully confirmed, we observed that in species of low wood density, leaf unfolding, flowering and fruiting phenotypes occurred shortly before it did in high wood density.

Data were collected monthly for one year, in an area of Atlantic forest located in northeastern Brazil, the phenological data were processed using methods according to Fournier (1974) and Bencke and Morellato (2002) and the wood density data were processed according to Trugilho et al. (nineteen ninety). They were organized in excel for later analysis using the Oriana and Rstudio programs.

To use this data, you only need to adapt the format, according to the statistical program that will be used, in the case of Rstudio it will be necessary to save it in a txt separated by tabs.