97th ESA Annual Meeting (August 5 -- 10, 2012)

COS 78-3 - Coordination of plant functional traits in predicting relative abundance of 42 tropical tree species during secondary succession in Central Panama

Wednesday, August 8, 2012: 8:40 AM
F151, Oregon Convention Center
Dylan James Craven1, Michiel van Breugel2, Jefferson Hall3, Mark S. Ashton4 and Graeme P. Berlyn4, (1)Centre for Ecosystem Modeling and Monitoring, Universidad Mayor, Chile, (2)Smithsonian Tropical Research Institution, Panama, (3)Smithsonian Tropical Research Institute, Balboa, Panama, (4)Forestry and Environmental Studies, Yale University, New Haven, CT
Background/Question/Methods

Functional traits delineate important dimensions of ecological variation in plant communities and, in doing so, reveal mechanisms that drive ecological processes. In tropical secondary forests, changes in species composition during succession have been well documented and are assumed to reflect shifts in adaptive ecological strategies, from fast growth and shade intolerance to slow growth and shade tolerance. In the present study, we examine interspecific variation in the relative abundance of 42 abundant woody species using an extensive suite of leaf and plant functional traits in a tropical secondary forest in Central Panama. Specifically, we address:  a) the extent to which plant functional traits explain interspecific differences in relative abundance of 42 abundant woody species during secondary succession using structural equation models and b) which plant functional traits best predict changes in relative abundance during secondary succession using a random decision-tree model.

Results/Conclusions

During the early stages of secondary succession (0 – 12 y), wood density, maximum tree height and leaf functional traits were positively correlated with interspecific variation in relative abundance through direct and indirect causal paths (Squared Multiple Correlation, 22 – 48 %).  For the older secondary forest (13 – 20 y), the data supported  a simpler causal path model in which maximum tree height explained 41 % of variation in relative basal area and wood density explained 18% of variation in relative stem density.

Species-specific changes in relative abundance during secondary succession were best predicted by maximum tree height, mass-based photosynthetic capacity, and leaf cellulose content.  In general, the random decision-tree model robustly predicted whether relative abundance increased or decreased during secondary succession (Somer’s D = 0.77 - 0.79). However, misclassified species were predominantly those with functional trait values associated with late-successional species, but whose relative abundance decreased during secondary succession.  

Our findings provide evidence that both acquisitive and conservative ecological strategies allow species to obtain high relative abundances during the early stages of secondary succession, yet due to changes in the abiotic and biotic environment, only conservative ecological strategies are adaptive in the later stages of secondary succession. Further, our results reveal that coordinated variation in a subset of functional traits, each representing different aspects of plant response to environmental conditions, modulated contrasting patterns of relative abundance.