Comparison of Different Height–Diameter Modelling Techniques for Prediction of Site Productivity in Natural Uneven-Aged Pure Stands

Reliable estimates of forest site productivity are a central element of forest management. The model of height-diameter relationship of dominant trees using algebraic difference approach (ADA) is a commonly used method to measure site productivity of natural uneven-aged stands. However, the existing models of this method do not recognize site type or sample plot specific variability in height curves; thus, it cannot be effectively used to estimate site type or sample plot-related site productivity for natural uneven-aged stands. Two primary subject-specific approaches, ADA with dummy variable (DV) (ADA + DV) and ADA with combination of dummy variable and nonlinear mixed-effects modelling (CM) (ADA + CM), were proposed for height–diameter modelling. Height–diameter models developed with ADA, ADA + DV and ADA + CM were compared using data from 4161 observations on 349 permanent sample plots of four major natural uneven-aged pure stands (Spruce, Korean Larch, Mongolian Oak, and White Birch) in northeastern China. It was found that models developed with ADA + CM provided the best performance, followed by the models with ADA + DV, and the models developed with ADA performed the worst. Random effects at the plot level were substantial, and their inclusion greatly improved the model’s accuracy. More importantly, the models developed with ADA + CM provide an effective method for quantifying site type- and sample plot-specific forest site productivity for uneven-aged pure stands