A site index model for lodgepole pine (Pinus contorta Dougl. var. Iatifolia) in northern Sweden

A model was developed for predicting site index (SI) values and top height growth of lodgepole pine in northern Sweden. Data on 169 experimental plots from 61 experiments were used for model construction. Six dynamic site equations derived using the generalized algebraic difference approach (GADA) were tested. All of the equations estimate height and SI values on the basis of a known height and total age, and are base-age invariant. Unbiased parameter estimates were obtained by non-linear modelling without accounting for autocorrelation of residuals. The special formulation of the Hossfeld model and the GADA-transformed logistic function produced the most reliable SI curves. The Hossfeld model is recommended as it did not produce extreme outlying estimates for young stands. The developed model is polymorphic with variable asymptotes and can predict SI and dominant height growth for lodgepole pine between 20 and 50 years old.