Simulation of the development and restoration of old-growth northern hardwood forest

Individual-tree models for stem and crown growth were developed and evaluated for Acer saccharum, Fraxinus americana, and Tilia americana to simulate forest stand and gap dynamics in northern hardwoods. The use of crown variables resulted in minimal improvement in model accuracy compared to diameter-based variables except for white ash. A variety of different variables for measuring crowding or competition were all about equally useful.; Crown growth equations had moderate R2 values (0.20–0.48). Lateral growth rate of exposed branches in sugar maple was clearly greater than shaded branches as indicated by highly significant coefficients for the crown exposure variable. This feature of the crown growth equation allows for prediction of asymmetric crown development and more rapid closure of canopy gaps by lateral growth of border tree crowns.; Three alternative crown-based models had similar predictive accuracy, with a root mean square error for basal area increment of 6–9 cm2/yr. Several validation measures for predicted basal area increment were evaluated and revealed good fits for all models and species. Long-term projections for individual tree growth and stand growth made over a 300-year period agreed closely with tree and stand characteristics in actual old-growth stands.; A computer simulation of old-growth restoration treatments was evaluated for two different northern hardwood stands, using several different approaches to thinning and crown release. Three heavy thinning treatments at 15-year intervals substantially reduced the time to reach threshold old-growth conditions from 74 years to 31 years in the younger stand. However, repeated heavy thinning in the older stand caused a 5–11 year delay in reaching target conditions because some relatively large trees had to be cut to provide the intended degree of crown release or stocking reduction. Surprisingly, total coarse woody debris volume showed only small differences among treatments and the control. Overall, the results suggest that old-growth structural features can be restored in many second growth stands in 30 years or less.