Independent and simultaneous adjustment to stimulate biomass above the Savanna Soil, Parque Cerrado phytophysiognomy

Abstract

Most researchers often overlook the biological consistency required when adjusting models to estimate total biomass and above-ground components. Mathematical models to estimate biomass that fail to take into account dependence relationship between tree components and the whole are scantily studied in native forests, especially for the Cerrado biome. Current investigation compares the above-ground biomass estimators by two different methods, independent and simultaneous adjustment of equations. Above-ground biomass data for the components and whole were collected. Mathematical models to estimate total biomass and components were then independently adjusted by the ordinary least squares method and simultaneously by the apparently unrelated nonlinear regression method. The equations´ biological consistency was verified by the two methods, coupled to heteroscedasticity, which were adjusted after having the weights of their coefficients calculated by variance structure. Simultaneously adjusted equation system presented biological consistency and lower prediction intervals when compared to the independent adjustment of equations, despite demonstrating similar statistical parameters. Biomass models of the components and total, for the phytophysiognomy Cerrado Park, should be estimated by equation systems to ensure additivity and precision.