MODELS FOR ESTIMATING MACRONUTRIENTS IN Mimosa scabrella Bentham

The aim of this work was to adjust and test different statistical models for estimating macronutrient content in the above-ground biomass of bracatinga (Mimosa scabrella Bentham) The data were collected from 25 bracatinga trees, all native to the north of the metropolitan region of Curitiba, Parana state, Brazil To determine the biomass and macronutrient content, the trees were separated into the compartments leaves, branches < 4 cm, branches >= 4 cm, wood and stem barks Different statistical models were adjusted to estimate N, P, K, Ca and Mg contents in the tree compartments, using dendrometric variables as the model independent variables Based on the results, the equations developed for estimating macronutrient contents were, in general, satisfactory The most accurate estimates were obtained for the stem biomass compartments and the sum of the biomass compartments In some cases, the equations had a better performance when crown and stem dimensions, age and dominant height were included as independent variables