The construction of diameter-height models based on nonlinear mixed effect regression equations for oriental spruce located in mixed oriental spruce-scotch pine stands at Trabzon Forest District

Abstract

Because of hard-working and time-consuming measurements for tree height, heights for all tree in sample plots can't be measure in many forest applications. These unmeasured tree heights can be predicted by height-diameter models relating tree height and diameters. In modeling the relationships between tree height and diameter, autocorrelation or serially correlations can be occurred in nested sampling systems for tree data measuring forest ecosystems. In this study, it is proposed to predict Nonlinear Mixed Effect Regression Models to alleviate these autocorrelation problems for constructing stand height-diameter curves in Oriental spruce (Picea orientalis (L.) Link) trees growing admixtures with scotch pine (Pinus sylvestris L.). The Schnute (1981), Huang et al. (2009), Wykoff et. al. (1982) nonlinear mixed effect model structure were predicted and compared for model height-diameter relationships in this study. In comparing these models, The Schnute (1981)'s nonlinear mixed effect model produced the best prediction results based on accuracy statistics. The Akaike (AIC) and Bayesian (BIC) Information Criterion were calculated as 15242 and 15257 and Schnute (1981) model predicted the % 95 percent of variability for height (R2=0.95). Schnute (1981) model with significant explanatory at variability of height and without serial correlation in diameter and height data will be use reliably to obtain predictions for oriental spruce tree heights. Keyword: Relationships between height and diameter, Oriental spruce, Nonlinear mixed effect regression model