Theories of forest dynamics rooted in physiological ecology have sought to link characteristics of individuals to stand-level dynamics, with the goal of scaling metabolic processes from the leaf level to communities of trees. A major simplification of earlier physiological tree models assume a symmetric network structure where child branches of parent branches are geometrically identical (West et al. 1997). Deviations from this pattern (asymmetry) would yield areas of the tree network with disproportionately higher volume and greater numbers of leaves, possibly affecting scaling parameters and predictions of whole-plant function. We used 3D scans of temperate and tropical trees to test an asymmetric model of tree vasculature across many tree species. Our objectives were to quantify the diversity of branching architectures at a global scale, and to test theoretical predictions for how the number of metabolic surfaces (leaf petioles) in a tree network scales with the total volume of the network. We used terrestrial LiDAR scans obtained from forest plots as well as individual cultivated trees to extract detailed scans of large numbers of individual trees across many species. Additionally, we sub-sampled branches from certain individuals to quantify how well smaller samples approximate scaling parameters of the larger network of branches.
Results/Conclusions
Symmetric and asymmetric models of tree architecture can potentially make different predictions for how organismal behavior scales with size. We compared empirically observed scaling relationships with exponents derived from models assuming symmetric or asymmetric branch structure. Both models parametrized from branching geometry tended to under-predict the slope of the scaling relationship, perhaps due to a lack of fine-scale resolution in the LiDAR scans. Symmetric and asymmetric scaling predictions were usually statistically indistinguishable, but future studies could focus on fewer focal species to increase statistical power and capture more intraspecific variability. Subsampling branches sometimes closely approximated total branching structure, and may be a viable strategy for quickly sampling architectural traits of many individuals. A general theory of forest ecosystems including important processes such as respiration and productivity is of fundamental importance to a science of the biosphere. These results could have implications for understanding how assemblages of trees aggregate to produce observed rates of NPP, whole-plant respiration, and potentially shed light on other processes such as light competition and canopy plasticity that may be mediated by variation in branching asymmetry among tree species.