Development and testing of volume models for <i>Pinus nigra Arn., Fagus sylvatica L.,</i> and Quercus pubescens Willd.</i>

Abstract

Volume models were developed and tested for Pinus nigra Arn., Fagus sylvatica L., and Quercus pubescens Willd. to accurately predict the woody stem volume at tree level, since despite the economic importance of these species in the Mediterranean type of forests standard models do not apply. Using a total non-destructive sample of 471 trees in mountain forests of northern Greece, two fundamental volume models, the Hohenadl–Krenn and the standard equation were fitted using a generalised least squares (GLS) estimator along with a power function, to regulate the unequal variance and the normality assumption of the residuals. The results showed that both models presented good fitting performance explaining the largest part of stem volume variance and therefore they can be used in operational forest management planning. In addition, the normality assumption of the distribution of the residuals can be satisfied through weighting techniques for reducing the potential heteroscedastic trends during the fitting procedure. The GLS may provide the necessary framework for stem volume model development on an accurate and unbiased basis.

Keywords: forest management; generalized least squares (GLS); Mediterranean mountain region; standing tree volume