Jump to Main Content
Wood Quality Attribute Models and their Utility when Integrated into Density Management Decision-Support Systems for Boreal Conifers
- Newton, P.F.
- Forest ecology and management 2019
- Picea mariana, Pinus banksiana, boreal forests, conifers, decision making, decision support systems, equations, models, modulus of elasticity, prediction, probability, stand density, surface area, tracheids, trees, wood density, wood quality
- The objectives of this study were to (1) develop wood quality attribute prediction models for a suite of commercially-relevant jack pine (Pinus banksiana Lamb.) fibre attributes (wood density (Wd), microfibril angle (Ma), modulus of elasticity (Me), fibre coarseness (Co), tracheid wall thickness (Wt), tracheid radial (Dr) and tangential (Dt) diameters and specific surface area (Sa)), (2) given (1), incorporate the parameterized equations within structural stand density management models (SSDMMs), and (3) given (2), exemplify their utility in silvicultural decision-making via the comparative assessment of attribute outcomes arising from operationally-relevant crop plans. Analytically, the equations were developed deploying Silviscan-determined attributes derived from transverse breast-height radial xylem sequences obtained from 61 trees sampled from 2 geographically-separated thinning experiments located in the central region of the Canadian Boreal Forest Region. Hierarchical mixed-effects regression modeling combined with cross-validation procedures were used to specify, parameterize and evaluate the attribute-specific prediction equations. Overall, the results revealed that the attribute trajectories were size-dependent and the resultant models were adequate in terms of their goodness-of-fit characteristics (e.g., I2 values of 75, 71, 71, 66, 60, 55, 49 and 38% for Co, Me, Wt, Sa, Wd, Ma, Dr and Dt respectively), lack-of-fit indicators (e.g., temporally invariant patterns of absolute and relative errors devoid of evidence of systematic bias), and predictive performance (e.g., 95% probability that 95% of all future relative Dt, Dr, Sa, Wd, Co, Wt, Me and Ma errors would be within ±5, ±9, ±11, ±12, ±12, ±13, ±35 and ±43% of their true values, respectively). Incorporating the jack pine equations along with a similar suite of functions previously developed for black spruce (Picea mariana (Mill) B.S.P) into the SSDMM analytical framework, yielded a pair of enhanced stand-type-specific (natural-origin and planted stands) decision-support systems for each species. These systems enabled the estimation of attribute-specific developmental trajectories for the rotational tree population from which diameter-class and stand-level wood quality performance measures were derived. As exemplified, the development of size-dependent fibre attribute prediction models and their subsequent integration within SSDMMs provides forest managers with a decision-support platform for evaluating and comparing end-product-related consequences of selected crop plans.