Tree-rings and sex: A case study from high-altitude Austrocedrus chilensis forests from southern Argentina and Chile

FIDEL ROIG; CARLOS LEQUESNE; JOSE A. BONINSEGNA; MARIANO S. MORALES

Davos, Switzerland

Conferencia; Tree rings and people: an international conference on the future of dendrochronology; 2001

WSL, SLF, University of Basel, IUFRO, Tree-ring society

Sex ratio in vascular plants assumes that female function is more costly than male function. These hypothetical differential costs between male and female individuals are interpreted in terms of different resource investment related to reproduction activities. The higher energy cost for females may cause reductions in vegetative growth that can be reflected in changes of the annual tree-ring widths. Austrocedrus chilensis is a dioecious conifer tree from the xeric-mesic habitats of southern Argentina and Chile. Populations of Austrocedrus chilensis growing on both sides of the northern Patagonian Andes were considered in this study to elucidate possible sex-related differences in the annual growth patterns of trees. To assess any differences in ring-width series related to sex expressions, the intercorrelation among each individual sample at each site and among chronologies was analyzed using principal component analysis through the 1890-1990 common interval. The arrangement of male and female Austrocedrus series with respect to the first three eigenvector axes reflects differential patterns according to sex. Each sex expression is clearly grouped in any population analyzed. This suggests that male and female trees separate owing to particular characteristics of growth, probably associated with differential reproductive costs. However, the assumption that female trees could pay a higher cost in terms of wood production is somewhat controversial since not all sites show higher growth rates for male trees. According to the large percentage of the common variance accounted for by the first eigenvector, tree-ring variations at the different sites appears to largely reflect macroclimatic signals. Eigenvectors of order two and three could reflect smaller-scale variations that likely represent secondary growth-related features associated to sex. When comparing chronologies developed by sex, the third principal component appears to reflect differences between groups located at both sides of the Andes. This study points out the importance to extend our results to other populations on both faces of the Andes in order to validate our provisional estimations about differences in growth pattern related to sex and registered at both, local and regional scales. The work discusses differences in female/male growth rate characteristics, growth cycles contained in chronologies developed by sex as well as implications of the use of these chronologies in climatic reconstructions.