Search

The relationships between climate and wood density components, i.e. minimum, maximum and average ring density, have been characterized mainly at breast height, while little information is available for other stem heights. This work aimed to (1) identify the significant climate variables related to ring density components in black spruce (Picea mariana (Mill.) B.S.P.) and (2) assess whether their effects vary significantly along the stem. Densitometric data were obtained from 68 black spruce trees from northwestern Ontario, Canada. Six discs along the stem, from stump to base of the live crown, were sampled. For each sample height and density component, we first developed mixed-effects models using ring width and cambial age as predictors. Monthly climate variables were later integrated into the models. The candidate climate variables were pre-selected using the boosted regression trees method. In addition to observed density–climate relationships, a significant effect of sample height was found. Findings revealed that the relationships were similar in nature but varied in strength and timing along the stem. The wood density variation showed delayed sensitivities to some climate variables among sampling heights. Such delayed responses may be caused by the axial progression of water requirement and of hormonal and photosynthetic allocation along the stem throughout the growing season.

The relationships between climate and wood density components, i.e. minimum, maximum and average ring density, have been characterized mainly at breast height, while little information is available for other stem heights. This work aimed to (1) identify the significant climate variables related to ring density components in black spruce (Picea mariana (Mill.) B.S.P.) and (2) assess whether their effects vary significantly along the stem. Densitometric data were obtained from 68 black spruce trees from northwestern Ontario, Canada. Six discs along the stem, from stump to base of the live crown, were sampled. For each sample height and density component, we first developed mixed-effects models using ring width and cambial age as predictors. Monthly climate variables were later integrated into the models. The candidate climate variables were pre-selected using the boosted regression trees method. In addition to observed density–climate relationships, a significant effect of sample height was found. Findings revealed that the relationships were similar in nature but varied in strength and timing along the stem. The wood density variation showed delayed sensitivities to some climate variables among sampling heights. Such delayed responses may be caused by the axial progression of water requirement and of hormonal and photosynthetic allocation along the stem throughout the growing season.