Despite over a century of intensive study, the mechanisms underlying patterns of species coexistence and diversity in plant communities remains a topic of debate. Recently, it has been argued that in order to reconcile competing hypotheses regarding species coexistence it is necessary to quantify the distribution of functional traits of all species within a community (McGill et al. 2004). However, if we are to fully adopt a functional trait approach to studying species coexistence, we must also quantify the magnitude of intra-specific variability in plant traits within and across populations as well as the number of samples necessary to differentiate species on the basis of function. Finally, it will be important to compare species mean values across communities in other localities as well as understand the underlying drivers of trait variation. The present study quantifies the degree of functional trait variation within and across species and individuals for tree species in a dry tropical forest in Costa Rica. Additionally, by comparing trait variation within and across populations arrayed along a microclimatic habitat gradient, this study tests the hypothesis that underlying drivers of plant trait variation can be predicted on the basis of biotic and abiotic factors.
The majority of the variation in the functional traits studied occurs among species, though this pattern differs for compound and simple-leaved species. The results show that in order to accurately differentiate species on the basis of function a higher sampling intensity is necessary when measuring plant functional traits than what has traditionally been the standard. Lastly, several emergent patterns of trait variability were found across microclimatic gradients, which suggest that the rarity of a species, on a local scale, may be a result of a plasticity constraint within particular leaf traits. In order to further understand the underlying drivers of variation as well as patterns of species coexistence and diversity, it will be necessary to quantify both genetic and phenotypic components of plant trait variation.