Although it has been recognized as a key parameter of wood quality and a good source of information on growth, annual wood density has been little studied within diffuse-porous trees such as beech (Fagus sylvatica Liebl.). In this paper we examine the variability encountered in beech ring density series and analyze the influences of ring age, ring width, climate and between-tree variability on density. Thirty ring sequences were sampled from 55-year-old dominant beech trees growing within the same stand; ring density and width were measured using radiography. Ring density proved to be less variable through time than ring width. The relationship between these two variables was less than observed in ring-porous trees and it showed great variation between trees. The sensitivity of ring width and density to climate was also different; width was strongly linked to soil water deficit whereas density was correlated to temperature and August rainfall. Unlike ring width, wood density showed sensitivity towards climatic characteristics of the late growing season. A large part of annual density variability remains unexplained, even using advanced modelled water balance variables. We hypothesize that a significant part of the tree ring is under internal control. We also demonstrated great inter-tree variability (the tree effect) in ring density, which has an influence on density but not on trees' response to climate.

No abstract

Wood density in relation to cambial age and ring width was studied in 23 trees of European oak (Quercuspetraea (Matt) Liebl. and Quercusrobur L.) from northeast France, and based on these relationships, a model predicting intratree wood density through cambial age and ring width was developed. Wood density is closely correlated with both cambial age and ring width, but its relationships with cambial age and ring width, to some extent, vary from tree to tree. The variation, however, does not show an appreciable pattern with tree height. The model developed is able to predict satisfactorily the average wood density of growth rings in European oak, and to explain the main part of the variation in wood density of individual trees. Furthermore, the model appears also possible to predict the average density of a given small piece of wood from any part of the merchantable bole of the trees. A general model, however, accounts for a relatively low percentage of the variation because there is large intertree variability of wood density in the species. The simulation results reveal that the dynamic silviculture applied to European oak might result in the production of denser and more uniform (in the sense of the radial variation of wood density) wood, compared with classical silviculture.

No abstract