LAUSR

Tree growth is one of the main variables needed for forest management planning. The use of simple models containing traditional equations to describe tree growth is common. However, equations that incorporate different factors (e.g. site quality of the stands, crown classes of the trees, silvicultural treatments) may improve their accuracy in a wide range of stand conditions. The aim of this work was to compare the accuracy of tree diameter growth models using () a family of simple equations adjusted by stand site quality and crown class of trees, and ( ii) a unique global equation including stand and individual tree variables. Samplings were conducted in 136 natural even- aged Nothofagus antarctica (Forster f.) Oersted stands in Southern Patagonia (Argentina) covering age (20-200 years), crown class and site quality gradients. The following diameter growth models were fitted: 16 simple equations using two independent variables (age and one equation for each stand site quality or crown class) based on Richards model, plus a unique global equation using three independent varia bles (age, stand site quality and crown class). Simple equa- tions showed higher variability in their accuracy, explained between 54% and 92% of the data variation. The global model presented similar accuracy like the better equations of the simple growth models. The unification of the sim- ple growth models into a unique global equation did not greatly improve the accuracy of estimations, but positively influenced the biological response of the model. Another advantage of the global equation is the simple use under a wide range of natural stand conditions. The proposed global model allows to explain the tree growth of N. antarctica trees along the natural studied gradients.