Modeling the thickness and percentage of bark on trees of Tectona grandis L.f

Abstract

The objective of this work was to evaluate models and variables for modeling the thickness and percentage content of bark in Tectona grandis (teak) trees. The data were collected from 68 dominant teak trees with ages varying from 6 to 34 years, distributed in different regions of Mato Grosso. To estimate of trees bark thickness, two models were tested using a polynomial regression model and a simple linear regression model. In both models, age was also inserted as a random component. The percentage of bark was estimated using exponential family models. The quality of the adjustments was made through statistical criteria and a set of graphical analyzes. It was verified that the polynomial model with the independent variable relative height and age as random effects proved to be efficient for the prediction of the thickness of the bark along the stem of the teak trees. The percentage bark content of standing trees can be accurately estimated by means of a bi-exponential model using only the DBH as the regressor variable.