COS 62-6
How do leaf traits differ in their patterns of variation across scales?

Wednesday, August 13, 2014: 9:50 AM
Regency Blrm D, Hyatt Regency Hotel
Julie Messier, Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ
Brian J. McGill, School of Biology and Ecology / Mitchell Center for Sustainability Solutions/Mitchell Center for Sustainability Solutions, University of Maine, Orono, ME
Brian J. Enquist, Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ
Martin J. Lechowicz, Department of Biology, McGill University, Montreal, QC, Canada
Background/Question/Methods

In a previous study, we found that two key leaf functional traits showed surprising patterns of trait variation across scales: they were both highly variable within species and similar within sites.  Since the two traits are known to be highly correlated and functionally associated, these results raise the question of whether all leaf traits show similar patterns of variation across scales. Here, we measure four new traits on the same dataset and quantify how the six leaf traits vary across six biological scales. Further, we assess whether the strength of correlation, or functional association, between two traits is predictive of their similarity in pattern of variation across scales. 

Results/Conclusions

We find two main differences among traits in their variance pattern across scales. First, traits differ in their degree of species-specificity, as shown by the relative importance of interspecific variance. Second, traits differ in the scale and type of response to environmental variation: they either show variation at large scales, consistent with a site-level filtering response to large-scale environmental gradients or variation within within trees, consistent with a plastic response to small-scale environmental gradients. The results suggest that leaf traits, even functionally associated ones, are affected by different drivers of variation. Further, we find that the strength of association between two traits is not predictive of their  similarity in patterns of variation across scale.