Comparison of non-linear height and diameter functions for oriental beech (Fagus orientalis Lipsky.) in a mixed and uneven-aged Caspian forest (Case Study: Tarbiat Modares University forest research station)

Abstract

Height - diameter equations are often used to estimate tree height, when tree diameter is the only measured variable. Considering that tree diameter measuring is often easier and inexpensive than tree height, height-diameter functions are generally used to estimate tree height. In this study, nineteen non-linear height-diameter equations were fitted using data of Fagus orientalis Lipsky in the Experimental and Educational Forest of Tarbiat Modares University. The data were split into two groups: 315 individual trees for model calibration and 290 for validation. Root Mean Square Error, adjusted R2, Akaike Information Criterion and Mean Difference criteria, as well as biological evaluation were applied to select the most appropriate model. Moreover, validation datasets were divided into 5-cm DBH classes and Mean Difference was computed for each model in each class. The results showed that the used models could explain a variation amount of 0.70 – 0.76. Weibull, Schnute, and Richards appeared to be suitable models with regard to mathematical and biological features. We recommend the Richards model as the best model to predict beech height trees for its better performance than the others in the Experimental and Educational Forest of Tarbiat Modares University.

Keywords

• beech tree
• caspian forests
• growth model
• non-linear regression