Development and validation of stem volume models for Pinus kesiya in Benguet province, Philippines

Abstract

Stem volume equations (overbark) were developed, using established volume equation forms, and validated using a subset of the data collected for Pinus kesiya in Benguet province, Philippines. A total of 481 trees from Pinus kesiya stands in Benguet were measured through non-destructive sampling. The data set was randomly split into two subsets for initial model development (80% of the data set) and validation (20% of the data set). The performance of the different models was evaluated using evaluation statistics: fit index (FI), root mean square error (RMSE), bias (E), absolute mean difference (AMD) and coefficient of variation (CV%). The best model forms were selected for final model development using the combined data set (100%). Overall, volume models with two independent variables (DBH and total height) performed better than models with only one (DBH). In the validation of the models, it was observed that AMD of the models converged in the DBH classes with higher sample size. Furthermore, one of the best models in this study, Model 2, also performed better when compared to the general single volume equation developed for the non-dipterocarp species in regions 1, 2, 3 and the Cordillera Administrative Region (CAR) of the Philippines. The models developed in this study may assist forest managers acquire improved stem volume predictions of Pinus kesiya in the forests of Benguet, Philippines.

Keywords: forest management, model evaluation, Pinus kesiya, tropical forests, volume equations

Southern Forests 2013, 75(3): 123–128