The objective of this work was to evaluate and compare two categories of growth and yield models of commercial plantations of eucalyptus. For that, a whole stand model, and an individual tree growth model were adjusted through simultaneous equations and artificial neural networks, respectively. Wood volume per area was estimated for tree different ages and productivity classes. Data of 63 permanent plots of unthinned clonal hybrid Eucalyptus grandis x E. urophylla were evaluated, from which those referring to 33 plots were used for model fitting and neural network training, and those referring to the other 30 plots were used for model validation. The two categories of models adjusted well to the observed data. However, for the validation of the models with independent data, the wood volume per area was better estimated with the individual tree growth model.
Eucalyptus; simultaneous equations; growth modeling; yield; artificial neural networks
