Growth forms and age estimation for treeline species

The correct evaluation of the age of small tree individuals is of great importance in ecological studies, and can be performed if growth units are identifiable even under strong environmental constraints. We aimed to assess the reliability of age evaluation of small individuals of four species commonly occurring at the treeline in the Alps and the Apennines (N Italy), Picea abies, Pinus cembra, Larix decidua and Fagus sylvatica. A total of 224 individuals were sampled along an altitudinal gradient up to the treeline. For each species, we assessed the growth architecture and the appearance of annual growth units; for each individual, age and evaluation uncertainty were estimated. To verify the estimation a core or a cross-section was obtained. The estimated age was compared with the tree-ring age through linear regression and reduced chi square test, and the estimation accuracy was evaluated as function of tree age, dimension and environmental parameters. A good estimation was possible for all species, even when loss of vertical growth occurred. The highest error occurred for Larix, the lowest for Pinus; the error was generally correlated with tree age and dimension, and was affected by altitude and geomorphology. Good reduced chi square values were reached considering only the youngest individuals for Fagus and Picea and the smallest for Fagus and Larix; the uncertainty was generally underestimated for Pinus. Even though reliable age estimation could be obtained, a realistic and species-specific evaluation of the estimation uncertainties is necessary to assess estimation accuracy for ecological studies.