Assessing lumber grade recovery improvement potential of black spruce (Picea mariana (Mill.) B.S.P.) by modelling knot characteristics

The objective of this study is to establish a methodology to assess black spruce (Picea mariana (Mill.) B.S.P.) lumber grade recovery improvement potential, with respect to knots, within a tree growth and sawing simulation framework. Models were developed to predict knot size, shape, and location in black spruce trees. A two stage model was developed to predict knot size. Using weighted nonlinear regression, maximum knot size per whorl is a function of tree diameter at breast height, depth in crown and whorl age. Using a transformed Weibull model, relative knot size is a function of relative whorl position, maximum knot size and depth in crown. The knot frequency model was developed using a general linear model with the Poisson family distribution and the final model form is a function of tree diameter at breast height, depth in crown, whorl length and maximum knot size. Relative whorl position and horizontal knot angle were fitted to beta and uniform circular distributions respectively.;A method to assess lumber grade recovery improvement potential at the log level was developed using two knot patterns (edge and centreline) as defined by visual lumber grading rules. Minimum difference in horizontal knot angle was established for each knot pattern based on knot size, lumber dimensions, and log top diameter. Knot models and lumber grade assessment method were incorporated into an existing tree growth and sawing simulation framework. Results show an indirect influence of initial tree spacing on knot size and frequency through crown recession and an improvement in lumber grade recovery potential with consideration for knot size and location prior to processing.