Single and multilevel linear and nonlinear mixed-effects model approaches were applied to model Eucalyptus stand growth. Most of the data set had a longitudinal, unequally spaced and unbalanced structure. By modeling heteroscedasticity, autocorrelation and correlation among random effects, the precision of these models in explaining data variation was increased. This approach has flexibility in selecting fixed effects which will be significantly associated with random effects, for both linear and nonlinear cases. In the linear case, the logarithm of basal area was linearly associated with some stand covariates. In the nonlinear case the three parameter logistic equation was used to explain dominant height variation over age.|A tree profile equation was developed based on a four parameter logistic equation, including covariates associated with random and fixed effects. The solid of revolution technique was applied to obtain individual tree volume having the individual tree profile. The robustness property of the mixed-effects model was used to estimate tree profiles with information on dbh and total height only. Using this technology, individual tree volume can be estimated with high precision with minimal information.
