Modeling flexural properties in white spruce (Picea glauca) and jack pine (Pinus banksiana) plantation trees

Abstract

Mixed models combining random coefficient effect and covariance patterns were used to investigate mechanical property variations in jack pine (Pinus banksiana Lamb.) and white spruce (Picea glauca (Moench) Voss) trees. Modulus of elasticity (MOE) and modulus of rupture (MOR) were measured by conducting three-point bending tests on small defect-free samples selected from different radial positions and at a height of 2.5 m above ground within the stems. The objective of the paper was to build statistical predictive models describing the radial variations in stems for wood mechanical properties using easily measurable explanatory variables that are typically available in the wood manufacturing industry: distance from pith, tree height and diameter, and spacing. The explanatory variables integrated into the models explained MOE adequately, whereas MOR appeared harder to predict with only these variables and at this resolution. For white spruce, the best mixed-effects models explained 80% and 61% of the variation in MOE and MOR, respectively. For jack pine, it was 51% and 33% for the same response variables. These results are a step toward models that could be used in sawing simulation software designed to estimate the internal properties of sawlogs and, as a result, better predict lumber and pulp chip quality.