Ten-year estimation of Oriental beech (<i>Fagus orientalis</i> Lipsky) volume increment in natural forests: a comparison of an artificial neural networks model, multiple linear regression and actual increment

Abstract Determining forest volume increment, the potential of wood production in natural forests, is a complex issue but fundamental importance to sustainable management. increment through growth and yield models necessary for proper management future prediction characteristics (diameter, height, volume, etc.). Various methods have been used determine productive capacity amount acceptable harvest forest, each has advantages disadvantages. One these involves artificial neural network techniques, which can be effective resource due its flexibility potentially high accuracy prediction. This research was conducted Ramsar forests Mazandaran Province Iran. Volume estimated using both an regression methods, were directly compared with actual 20 one-hectare permanent sample plots. A sensitivity analysis inputs employed had most effect predicting increment. The average annual beech 4.52 m3ha?1 yr?1, predicted 4.35 4.02 yr?1 best developed regression, respectively. results showed that estimate relatively well method, method able higher than traditional models. standing at beginning measurement period diameter trees greatest impact on variation