Abstract

Reliable information on stand dynamics and development is needed to improve management decisions on mixed forests, and essential tools for this purpose are forest growth and yield (G) models. In this study, stand-level G models were built for cohorts within the natural mixed second-growthNothofagus-dominated forests in Chile. All currently available (but limited) data, consisting of a series of stratified temporary and permanent plots established in the complete range of this forest type, were used to fit and validate these models. Linear and nonlinear models were considered, where dominant stand age, number of trees, and the proportion of basal area ofNothofagusspecies resulted in significant predictors to project future values of stand basal area for the different cohorts (withR(2)> 0.51 for the validation datasets). Mortality was successfully modeled (R-2= 0.79), based on a small set of permanent plots, using the concept of self-thinning with a proposed model defined by the idea that, as stands get closer to a maximum density, they experience higher levels of mortality. The evaluation of these models indicated that they adequately represent the current understanding of dynamics of basal area and mortality ofNothofagusand companion species in these forests. These are the first models fitted over a large geographical area that consider the dynamics of these mixed forests. It is suggested that the proposed models should constitute the main components of future implementations of G model systems.