Modelling the tree height-diameter relationship of Macedonian pine (Pinus peuce Gris.) forests in North Macedonia

Abstract

The most crucial individual tree variables in forest management are diameter at breast height (DBH) and height of trees (H). Projection of stand development over time relies on accurate height-diameter functions. The main aim of this paper is to define the best usable model for tree height prediction from diameter at breast height. We explore the place where the Macedonian pine was discovered in the Baba Mountain in the Pelister National Park in North Macedonia. Thus, we established 48 experimental plots (EP) with a circle shape, a radius of 12.62 m and an area of 500 m2 each. The EP were established in pure Macedonian pine stands in an elevation gradient between 950 m a.s.l. and 1 700 m a.s.l. Every tree in the EP was attributed with data for diameter at breast height, tree height, and stand characteristics data (elevation, slope, aspect, coordinate) as well. For predicting the tree height, we used 40 models from many authors which are the most usable in forest practice. Also, we prepared evaluations and tests for all models, in order to choose the best responsive model for the Macedonian pine forest. For predicting the best tree height model for Macedonian pine, we decided on Mamoun's equation, with a high correlation value of 0.85 with 73% out of the observed data. Also, this model showed a lower root mean square error of 32.65, a lower model prediction accuracy of 6.77, and a lower mean absolute percent error of 11.73%. Finally, it can be concluded that the nonlinear connection between DBH and tree height is the most responsive regression model.