Modelling juvenile-mature wood transition in Scots pine (Pinus sylvestris L.) using nonlinear mixed-effects models

Abstract

-Nonlinear mixed-effects-models are applied successfully to estimate the cambial age of juvenile-mature wood transition in Scots pine sample trees from slow-grown stands. Till now segmented regression models are applied separately for each pith-to-bark-profile of wood density. The nonlinear mixed-effects-model overcomes this limitation while consistently and efficiently estimating the transition point for the whole sample. Furthermore standard errors can be calculated and impacts of stand and tree variables on the shape of pith-to-bark-curves can be tested. Mean ring density, earlywood, and latewood density profiles from 99 trees were determined by X-ray densitometric analysis of disks taken at 4-m stem height. The cambial age of transition from juvenile to mature wood is described according to nonlinear mixed-effects-models based on latewood density profiles. The time-series nature of the data are taken into account. The segmented quadratic-linear model shows the transition at cambial age of 21.77, which vary with the probability of 0.95 within the interval of [18.31; 26.85]. Impacts of tree variables or stands on the location of the transition point were not found, but impacts of stands on the shape of pith-to-bark-curves.Pinus silvestris / microdensitometry / nonlinear mixed effects model / pith-to-bark wood density profile / juvenile-adult transition point Résumé -Modélisation du passage bois juvénile-bois adulte chez le pin sylvestre (Pinus sylvestris L.) à l'aide de modèles mixtes non linéaires. Des modèles mixtes non linéaires ont été utilisés avec succès afin d'estimer l'âge (compté depuis la moelle) du passage bois juvénile-bois adulte pour des pins sylvestres provenant de peuplements à croissance lente. Jusqu'à présent, des modèles de régression segmentés étaient ajustés individuellement à chaque profil de densité du bois de la moelle à l'écorce. Le modèle mixte non linéaire permet de dépasser cette limitation en estimant de manière efficace et cohérente l'âge du point de passage pour l'ensemble de la population échantillonnée. En outre, des variances peuvent être estimées et les impacts des peuplements et des caractéristiques des arbres sur la forme des profils de densité du bois de la moelle à l'écorce peuvent être testés. À cette fin, les profils de densité moyenne de cerne, de densité du bois initial et de densité du bois final de 99 arbres ont été mesurés par exploration microdensitométrique de clichés radiographiques obtenus à partir d'échantillons prélevés dans des disques découpés à 4 mètres de hauteur. L'âge compté depuis la moelle du passage bois juvénile-bois adulte a été identifié à partir de modèles mixtes non linéaires appliqués aux profils de densité du bois final. La nature longitudinale des données a été prise en compte. Le modèle segmenté linéaire quadratique retenu permet d'identifier un âge moyen de passage du bois juvénile au bois adulte de 21,77 ans assorti d'un intervalle de confiance à 5 % de 18,31 à 26,85 ans. Si les impacts des variables "caractéristiques des arbres" ...

Physical properties of wood in thinned Scots pines (Pinus sylvestris L.) from plantations in northern Spain

“…In the case of volumetric shrinkage, the variation explained by the fixed effects was 36.4%, with an additional 19.6% explained by the random effects. These results indicate the importance of considering the random effects in modelling physical properties variation along the stem, as considered for Scots pine and other species (Mutz et al, 2004). It should be considered that, although no direct effect of height in the stem was considered in these models, the negative effect of ring width and the increase in basic density outward from the pith of the stem result in a necessary decrease in density from the base to the tree top, as considered in other models (Kellomäki et al, 1999).…”

“…Scots pine is considered to show a clear transition between juvenile and mature wood, which has been estimated at a cambial age of 22 years (Mutz et al, 2004;Zobel and van Buijtenen, 1989).…”

Comparison of the hygroscopic behaviour of 205-year-old and recently cut juvenile wood fromPinus sylvestrisL.

“…The application of this model determined that the age of the pieces of old wood was from 25 to 30 years. Other studies on the same wood presume it to be mature at this age [25]. However, as the cambial age is quite variable between trees of the same forest and between different regions of provenance, particularly in conifers, the old wood samples were considered to be of juvenile wood in accordance with the specific preliminary study on the forest from which they came, in terms of the number of rays and the length of the tracheids, which enabled it to be established that up to ring 40, with a 5% confidence level, the juvenilemature cambial age did not occur [22].…”

Analysis of coniferous species to identify and distinguish juvenile and mature wood

“…In the opinion of the authors, density of late wood leads to an overestimation of transition age in relation to ring density and the share of late wood in annual increments in diameter. Based on studies of density of late wood zones conducted on discs collected from 99 stems of Scots pine (Pinus sylvestris L.) coming from the height of 4 m, Mutz et al (2004) identified the boundary between juvenile and mature wood between the 18 th and 27 th annual rings. After several statistical analyses were conducted and a square linear model was specified, the transition boundary was established at cambial age of 21 years.…”