Capturing diversity and interspecific variability in allometries: A hierarchical approach

There is growing recognition of the role of mechanistic scaling laws in shaping ecological pattern and process. While such theoretical relationships explain much of the variation across large scales, at any particular scale there is important residual variation that is left unexplained among species, among individuals within a species, and within individuals themselves. Key questions remain on what explains this variability and how we can apply this information in practice, in particular to produce estimates in high-diversity systems with many rare and under-sampled species. We apply hierarchical Bayes statistical techniques to data on crown geometry from diverse temperate forests in order to simultaneously model the differences within and among species. We find that tree height, canopy depth, and canopy radius are affected by both successional stage and wood mechanical strength, while tree height conforms to the predicted 2/3 power relationship. Furthermore, we show that hierarchical modeling allows us to constrain the allometries of rare species much more than traditional methods. Finally, crown radius was shown to vary substantially more within individuals than among individuals or species, suggesting that the capacity for local light foraging and crown plasticity exerts the dominant control on tree crowns.