Canadian Forest Service Publications

Hierarchical analysis of black spruce and balsam fir wood density in Newfoundland. 2015. Groot, A.; Luther, J.E. Canadian Journal of Forest Research 45 (7): 805-816.

Year: 2015

Available from: Atlantic Forestry Centre

Catalog ID: 37219

Language: English

CFS Availability: PDF (download)

Abstract

We used a hierarchical approach to examine patterns of black spruce and balsam fir wood density across Newfoundland. Wood density measurements were aggregated at ring, tree, and plot levels. Portions of the variance in wood density at the different levels were explained by predictor variables at multiple levels of a forest structural hierarchy (ring-, tree-, and plot-level variables). Hierarchical fixed effects models accounted for 39%, 61%, and 86% of the variance in wood density of black spruce at ring, tree, and plot levels, respectively, with RMSE values of 62.7, 34.6, and 19.4 kg·m−3. Corresponding models accounted for 31%, 38%, and 63% of the variance in wood density of balsam fir, with RMSE values of 62.4, 35, and 16.7 kg·m−3. The hierarchical analysis demonstrated consistent negative associations of wood density to radial growth rate at ring, tree, and plot levels of aggregation. Variables that act as surrogates for radial growth are thus important for understanding patterns of wood density at higher scales and for the practical application of mapping wood density across landscapes.

Plain Language Summary

Wood density is an important wood quality attribute, affecting the strength of manufactured lumber and paper products, wood-processing characteristics, and carbon storage in forests. This paper describes wood density patterns and develops models of wood density across a hierarchy of scales (ring, tree, and stand). The analysis shows that although wood density is an outcome of wood formation processes at the cellular level, wood density aggregated at different levels is influenced by ring-, tree-, and plot-level variables. The results demonstrate consistent negative associations of wood density to radial growth rate at ring, tree, and plot levels of aggregation. The results suggest that variables that act as surrogates for radial growth are important for understanding patterns of wood density at higher scales and for the practical application of mapping wood density across landscapes.

Date modified: