Predicting red heartwood formation in beech trees (Fagus sylvatica L.)

Beech-forests (Fagus sylvatica L.) represent the natural vegetation cover for the majority of sites in Germany. Although coniferous forests actually dominate, beech is the most important among the deciduous tree-species in Germany, when considering the area and the economics. While the wood of the younger beech normally shows a whitish colour (white beech), the timber of older beech may form coloured heartwood (facultative). This heartwood is highly variable in its colour. In most cases a reddish colour predominates, but sometimes the heartwood can be brown, grey, green or even purple. Usually the coloured heartwood is called „red heartwood”. As the uses of beech with „red heartwood“ are restricted, mainly for aesthetical reasons, its presence severely reduces the timber-quality. However, the existence of “red heartwood” cannot be distinguished until the tree is felled. In this context the paper presents a possible new approach within the „red heartwood“ research. The object is to predict the future „red heartwood“ formation in beech trees, while they are actually still showing a bright colour. This information is relevant to improve the beech forest management. In order to provide this information the data of 392 felled beech trees were utilised. By means of a logistic regression analysis the probability of the occurrence of “red heartwood” was first analysed. Specific tree characteristics were tested as independent variables (age, diameter at 1.3 m height, crown base height, average diameter increment, injuries to the bark, forking). Transition probabilities (probabilities with which a recent “red heartwood” is formed) were then derived based on the statistical logistic regression function. Manuskriptversion erschienen in Ecological Modelling 169 (2003), S. 163-179 2 In order to predict the transition probabilities for still standing beech trees a model was formed comprising of tree characteristics as independent variables, which significantly influenced the „red heartwood“ probability. Among the independents the age of the tree and the average diameter growth rate were the most important. The distance to the soil surface at which the heartwood was measured, and the number of injuries to the bark (dead branches, knobs, big scars), through which oxygen may enter the stem significantly influenced the transition probability. The presence or absence of a stem fork was also significant.