Modelling the radial and longitudinal variation in wood density and its sensitivity to climate in black spruce (Picea mariana (Mill.) B.S.P.)

Abstract

Wood density variation affects structural timber performance and is correlated with different sources (genetic diversity, stand dynamics and climatic effects) which, to date, have not been comprehensively quantified in black spruce (Picea mariana (Mill.) B.S.P.). Controlling the confounding factors of ontogeny and stand dynamics allowed the effects of climate on annual wood density to be examined. The objectives of this study was to (1) to describe the variation in average ring density in black spruce as a function of cambial age, stem height and growth rate; (2) to identify climatic effect on density indicators (i.e. average ring density, minimum ring density, and maximum ring density) and determine the density-climate relationship according to tree size; and (4) to determine the effect of disc height on the density-climate relationship. The dataset utilized densitometric data on a total of 35,127 annual rings from 450 radial strips collected in 107 black spruce trees sampled across 13 sites in northwest Ontario, Canada. Different aspects of the whole dataset were selected for various sub-studies. We observed, when averaging the pith-to-bark profiles for all discs, that ring density (RD) was high near the pith and decreased rapidly in the first 12 growth rings, i.e. declined from 591.0 kg m-3 to a minimum of 473.8 kg m-3 between rings 10 and 15, followed by a slow increase until a consistent value was reached between rings 25 and 60. This average pattern till ring 60th corresponds to the type II pith-to-bark profile. However, thereafter, the dataset was separated into two groups. It was observed that 106 out of 450 radial samples showed a gradual decline in RD near the bark. These declining trends in annual ring density near the bark were found to be more common in old and slow-growing trees. We hypothesise that such trends reflect a gradual reduction in tree vigour over the life of the tree. The climatic factors (monthly temperature and precipitation) were found to be significantly affecting density indicators across all dominance levels (dominant, co-dominant and intermediate trees). The density-climate relationships observed appears to be mediated according to dominance levels. This difference was assumed to be a result of dominance level related factors such as, thermal stress stratification and transpiration rate. Additionally, it was evidenced that these density-climate relationships differed not only among dominance levels, but also along the length of the stem. The phenomenon of the various climatic sensitivities according to stem heights could be related to top-down auxin and carbon distribution, hydraulic effects and respiration rate among stem heights. This work has provided a promising modelling method to disentangle the confounding sources of variation of wood density. It has also highlighted some of the future challenges which should be addresses on this research topic. Furthermore, this work has offered a starting point for studies on the effects of monthly climatic factors (temperature and precipitation) on density indicators and how these vary according to dominance level and stem height.