Individual tree basal area growth models for Chir pine (Pinus roxberghii Sarg.) in western Nepal

Abstract

The individual tree growth models are important decision-making tools in forestry. Age dependent and age independent individual tree basal area growth models were developed for Chir pine (Pinus roxberghii Sarg.) in one of the western districts, Rukum district, in Nepal. Data from thirty-five destructively sampled trees, which were representative of all possible stand densities, site productivities, age classes, and size classes of Chir pine forests in the district, were used. Sample trees were felled and diameters and ages were measured on the cut surface of the stump (at 30 cm above the ground). Since measurements from the same stump of a tree were strongly correlated, the autoregressive error structure modelling approach was applied while specifying the model in order to reduce bias. All parameter estimates of the models were significant (P < 0. 01) and the models described most of the variations of basal area growth (R2adj > 0.86). Residual graphs showed no serious systematic bias for all observed age classes and diameter classes. The age independent growth model showed relatively better fit statistics (R2adj = 0.8751, RMSE = 4.8494) than its age dependent counterpart (R2adj = 0.8668, RMSE = 5.0158). Because of being more precise and simpler, the age independent model is recommended to apply to both even-aged and uneven-aged stands of Chir pine in the district.

Keywords

• autoregressive error structure
• age dependent model
• age independent model