BIOMASS ALLOMETRY FOR TREE SPECIES OF NORTHWESTERN MEXICO

Abstract

Tree biomass plays a key role in sustainable forest management since it is the basis for estimating stocks and fluxes of several biogeochemical elements, the amount of energy stored in biomass, and other conventional goods and services. The most common mathematical model takes the form of the logarithmic equation where biomass is estimated as a function of diameter at breast height with the scaling coefficients a and B. In this study, I answered the following questions related with the allometric model: a) Is it important to develop biomass equations at the species scale or at the site-specific scale?; b) What is the least number of data required for fitting an allometric equation?; and c) Is it possible to develop allometric equations with few or null biomass data without loosing accuracy in biomass estimation? I employed a biomass data source collected in northwestern Mexico for nine different forest species, collected in six different sites from southern Chihuahua to southern Durango, Mexico to answer these questions. Results showed that by fitting site-specific biomass equations there is a net gain of 5% and close to 20% in the coefficient of determination and the standard error, respectively in contrast to fitting an equation at the species level. The minimum number of observations needed is 60 harvested trees to calculate parameters with the least variance and with high consistency. I present two alternate restrictive methods of biomass estimation: a) restricting the number of harvested trees to three to fit equations available in the scientific literature and b) a non-destructive model to fit equations with the same level of accuracy that display conventional allometric models. Both methods estimate biomass within the confidence bounds imposed on the B coefficient of the conventional allometric model.