Performance and comparison of stand growth models based on individual tree and diameter-class growth

Abstract

A distance-dependent individual tree based growth model (FOREST) was compared with a diameter-class growth model (SHAF) for describing changes in stand density and structure. Projections of Lake States' northern hardwood stand development were made by each model for 5–26 years over a range of stand conditions and harvest treatments. Results from numerous performance tests and comparisons of actual and predicted diameter distributions, basal areas, and numbers of trees, indicate the individual tree model was considerably more sensitive to harvest treatments and reproduction response than the diameter-class model. Conversely, the latter was much less expensive to operate. Prediction of species and individual tree growth with the individual tree model appeared to provide sensitivity nearly equal to that observed for predictions of the stand as a whole. Long-term projections (120 years) for reserve (no cut) and clear-cut stand conditions further suggest the potential and limitations of the models for management analyses.