Variable density yield model for teak plantations in the northeast of Thailand was developed based on 85 temporary sample plots located in the 7 provinces of the northeast i.e., Nakhon Ratchasima, Khon Kaen, Sakon Nakhon, Loei, Si Sa Ket, Ubon Ratchathani and Yasothon. The plots ranged in site index from 14 m to 30 m (dominant tree height at 30 years) and were measured for stand growth 1−4 times during the year 2000−2009. All of measurement plots provided 157 data sets for constructing the model.
The estimation data consisted of 36 growth periods of stands ranging in age from 3 to 40 years. Multiple linear regression method was used to simultaneously fit prediction equations for stem volume, stand basal area, stand volume and stand density. An existing Mitscherlich model for dominant height growth fitted with the same estimation data was used to estimate site quality index of the stands. The models were evaluated both quantitatively and qualitatively. Correlation among error components of the prediction equation for stem volume as well as the equations for stand density, stand basal area and stand volume were strong and statistically significant and the mean biases for those estimators were positive. The presented system of models could serve for constructing variable density yield table of teak plantations that required stand level input data.