Modeling dominant height growth of radiata pine (Pinus radiata D. Don) plantations in north-western Spain

A model for predicting the height growth of radiata pine plantations in Galicia (north-western Spain) was developed. Data from stem analysis of 161 trees were used for model construction. Seven dynamic site equations derived with the generalized algebraic difference approach (GADA) were tested, which constitute compatible site index and height models in one common equation. All of the equations directly estimate height and site index from any other height and age, and are base-age invariant. To obtain unbiased parameter estimates the base-age invariant dummy variables method was used, which recognizes that each measurement is made with error and, thus, fits a curve to the observed individual trends in the data by considering one parameter to be site-specific and the remaining parameters common for all individuals. A second-order continuous-time autoregressive error structure was used to correct the inherent autocorrelation of the longitudinal data used in this study, which allows the model to be applied to irregularly spaced, unbalanced data. The GADA formulation derived on the basis of the Bertalanffy-Richards model by considering the asymptote and the initial pattern parameters as related to site productivity resulted in the best compromise between biological and statistical considerations, producing the most adequate site curves. This model showed superior behavior when compared to the model currently in use. It is, therefore, recommended for height growth prediction and site classification of radiata pine plantations in Galicia.