Weight equations for determining biomass fractions of young hardwoods from natural regenerated stands

Abstract

This study gives an account of dry weight equations in biomass fractions of seven hardwoods from 90 plots in a hemiboreal forest zone in southeast Norway. The following species were investigated: goat willow (Salix caprea L.), grey alder (Alnus incana (L.) Moench.), black alder (Alnus glutinosa (L.) Gaertn.), silver birch (Betula pendula Roth), rowan (Sorbus aucuparia L.), aspen (Populus tremula L.), and european ash (Fraxinus excelsior L.). Sample trees were collected in stands—pure in species—from abandoned land, forest sites, and brook valleys below maximum shoreline height since the last glaciation (marine limit). The highest predictive equations appeared to be natural logarithms with base e. The equations have been generated from three models depending on their ability to explain the whole variation in the regression. In most cases over 80% of the variation in the dependent variables can be explained by the independent variables in the linear model. For various biomass fractions, the coefficient of determination (R²) decreases in this order of rank: dry weight of stemwood including bark, branchwood including bark, and foliage. Analyses of regression coefficients showed highly significant differences between most of the hardwoods with regard to tree components. However, no difference was found between goat willow and grey alder for branchwood including bark. For these species, there was no statistical difference in the intercept at the 5% significant level. Percentage bias increased from stemwood including bark (1.47%) to foliage (32.71%). Types of equations and selection of parameters in the regressions are discussed.