Allometric Scaling Across Environmental Gradients

Abstract:

Developing a better understanding of the controls on biomass allocation in forested systems and the consequences for carbon stocks and fluxes is required for improved ecosystem and climate modeling. A simple model, based largely on resource distribution networks, was presented by West, Brown and Enquist (1999). Their model predicts that the exponents of allometric relationships between many forest structural and functional properties will be constants, irrespective of environment or species. In this research we assess the validity of model predictions across the United States and examine their independence with respect to environment. We focus on two relationships with particular importance to biomass:

Ht ∝ DBH^2/3

nDBH ∝ DBH^-2

where Ht is height, DBH is Diameter at Breast Height, and nDBH is the number of trees in a given DBH size class.

We obtained DBH and height data from the U.S. Forest Inventory Analysis (FIA) dataset, and fit an exponent to each relationship for every FIA plot across the US. We extracted environmental data from the FIA plots (forest maximum height, species type, age, topography) and the North American Regional Reassessment dataset (precipitation, temperature, PAR) and performed random forest regression to estimate observed exponents as a function of environment. We found that forest height, age, and forest type were the most important drivers of allometry, explaining about 40% of observed variability. We found that for both relationships, as forest height and age increase, exponents constrain to the theoretical predictions presented by WBE. This suggests that WBE predictions are valid and may be useful constraints in biomass mapping and ecosystem flux models. However, they deviate from predictions in younger, shorter stands where forests have not had time to develop a complex size structure. Additionally, there is a significant difference in both relationships between conifer and deciduous-dominated stands, suggesting that species type must be considered in the further development and application of this work.