Plant Ecology 144: 231–242, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands. 231 Growth models of silver birch (Betula pendula Roth.) on two volcanic mountains in the French Massif Central Bernard Pr´ evosto 1 , Patrick Coquillard 2 & Jos´ ee Gueugnot 2 1 Cemagref, division Forêt et Agroforesterie, 24 avenue des Landais, BP 50085, 63172 Aubi` ere Cedex, France; 2 Laboratoire d’Ecologie V´ eg´ etale et Cellulaire UPRESA CNRS 6116, Universit´ e d’Auvergne, Facult´ e de Pharmacie, B.P. 38, F-63000 Clermont-Ferrand Cedex 1, France (E-mail: bernard.prevosto@cemagref.fr) Received 20 May 1998; accepted in revised form 20 January 1999 Key words: Betula pendula, Competition index, French Massif Central, Growth model Abstract Silver birch woodlands of two volcanic mountains (altitude 850 m and 1450 m) were studied in order to establish a growth model of birch. Height, radial increment and crown width were measured on both sites taking into account two situations: isolated birch or birch within a stand. For the latter case two categories were sampled considering the competition status of the tree: birch suffering the most severe competition (suppressed tree) were distinguished from trees facing the lowest competition (dominant tree). Measures of competition were also made using plots where each tree was located by its coordinates and its height, trunk circumference and crown width recorded. Examination of radial growth curves showed similar patterns for the two sites for the same category of tree. Radial growth was always inferior for the site located at the highest altitude but tree lifespan was about two- fold longer. Radial increment data were used to calculate circumference as function of tree age. Models predicting circumference with time were then established for each category of tree (dominant, suppressed or isolated) using Chapman–Richards’model. Height and crown width models were produced using circumference as the predictor. Competition indices based on vertical or horizontal angles weighted by the distance were calculated for birches in stands. Indices using vertical angles appeared to be more resistant and robust to characterize local competition. Competition index was then associated with the parameters of Chapman-Richards’ growth circumference model for the two sites and models predicting the circumference from the age of the tree and its competition index are proposed. Introduction Birch is an early successional tree with a high produc- tive capacity, a wind seed dispersal, a wide climatic and edaphic range and a limited tolerance to shade (Gimingham 1984; Perala & Alm 1990). These fea- tures explain the ability of birch to invade heathlands and grasslands after pastoral abandonment. In the French Massif Central the colonization of Calluna heathlands free of any grazing and the settlement of natural birch stands (Betula pendula Roth. and more rarely Betula pubescens Ehrh.) are frequent occur- rences. Nevertheless, the modeling of the functioning and changing of these heathlands from the initial stages – an heathland scattered with rare birch trees – to a mature birch woodland is a complex process that requires examining the three main following models: (i) The functioning model of Calluna heathland: life expectancies, age distributions, layering proba- bilities, annual growth are needed to simulate the dynamics of a Calluna stand and thus to estimate the resistance of heathland to birch invasion. Recent work in our study area gave elements for these different parameters. (ii) The recruitment model: demography and re- productive capacity of birch. Many other studies and field experiments have analyzed seed production and dispersal, germination, seedlings establishment for a various variety of climate, site and vegetation con-