Ramazan Özçelik, Krishna P. Poudel, Mehmet Denizhan Balcı, Şerife Kalkanlı Genç
Upper stem diameter and volume predictions of Brutian pine (Pinus brutia Ten.) trees in Western Türkiye
Dendrobiology 2026, vol. 95: 96-115
https://doi.org/10.12657/denbio.095.007
Abstract:
Individual tree volume prediction is one of the most important components of growth and yield modeling systems. This is often accomplished by using taper equations because of their flexibility. In this study, stem taper models were developed to estimate upper stem diameters, merchantable volume, height, and total stem volume for natural Brutian pine (Pinus brutia Ten.) in the Akhisar region of western Türkiye. Performance of 11 commonly used taper models was analyzed using data collected from 159 Brutian pine trees representing a wide range of diameter and height classes. A second-order autoregressive error structure was specified to account for inherent autocorrelation in hierarchical data. The models of Jiang et al. (2005), Bi (2000), and Kozak (2004) had the best estimation performance for diameter, merchantable volume, height, and total stem volume estimations, respectively. If the diameter at 5.30 m is not known or cannot be estimated, either the models of Bi (2000) or Kozak (2004) are preferred for the merchantable volume, stem diameter, and total volume of Brutian pine trees in the study area. The best performing model (Jiang et al., 2005) was further improved using a non-linear mixed-effects modeling approach. Results showed that adding two random effects further improved model’s estimation performance, decreased the error variance, and had better residual properties than the fixed effects model. The mixed-effects modeling approach is recommended for calibration when preliminary information for diameter measurements is available.
Keywords: diameter estimation, autocorrelation, nonlinear mixed-effects, relative ranking