| Summary: | Oriental spruce (Picea orientalis
L.) is one of the most important tree species in Turkey. Therefore, the
information is necessary about growth and yield of the species for developing
future management and planning strategies. The one of the essential building
blocks in forest growth and yield prediction models is the equations for
estimating individual tree volume. One of the most accurate and reliable
approaches for estimating stem total volume and merchantable volume is the use
of taper models. In this study, a taper model was developed by using nonlinear
mixed effects modeling approach (NLME) using data from 170 trees felled in
oriental spruce stands from Ardahan-Posof Region. An NLME approach accounted
for within-and between tree variations in stem form. In the first stage, all
possible combinations of expansion with random effects in one and two model
parameters were tested, selecting then the best one. The inclusion of random
effects was not enough to account for the existing autocorrelation between the
residuals and then the variance-covariance matrix of the error term was
modelled through a first order autocorrelation structure AR (1). In a second
step, we evaluated the response obtained by calibration (which implies
estimation of random effects for a new tree) based on upper-stem diameter
measurements at different points. The selected mixed-effects model produced the
best results both in fitting and calibration process. It was found that an
upper-stem diameter measurement at 40-80% of total height was best suited for
calibrating tree-specific predictions. As a result, model calibration should be
considered an essential criterion in mixed model selection.
|
|---|