(Introduction):

Reliably quantifying and predicting the volume of wood is essential to assist in forest production management.

(Objective):

Generate commercial volume tables using the selected mathematical models at different site qualities for clonal plantations of Tectona grandis.

(Methodology):

Three site qualities were evaluated according to site indices (SI) 27, 24, and 21, respectively. The diameter of 263 standing trees (119 for SI27, 120 for SI24, and 28 for SI21) was measured at the stump, 1.3 m above ground level, and every 1.5 m until the total height. The volume of the tree was obtained by adding each measured section with the Smalian formula. For each site index, models were adjusted by the ordinary least squares method using the Statgraphics Centurion XVI statistical package. The Durbin-Watson test verified compliance with the assumptions of autocorrelation of residuals. The model was selected based on the sum of the normalization of the statistical parameters evaluated.

(Results):

The selected models had good adjustments (R2≥ 94 %) and low estimation errors (<3 %) and were very practical since they only use the normal diameter (DBH) as a predictor variable.

(Conclusions):

Along with the table of commercial volume, the models are an important tool to quickly and accurately quantify the commercial volume of clonal teak with bark up to 15 cm in diameter.