A Matrix Growth Model of Natural Spruce-Balsam Fir Forest in New Brunswick, Canada

A density-dependent matrix model was developed for spruce-balsam fir natural forest stands in New Brunswick, Canada. It predicted the number and basal area of trees for 5 species groups (spruce, balsam fir, other softwood, soft hardwood and hard hardwood) and 10 diameter classes. Upgrowth, ingrowth and mortality models were established with explanatory variables representing tree size, stand density and stand structure. The model was based on 305 sample plots with inventory periods from 2 to 9 years. The majority of the data (80%) was used for model development, and the rest (20%) was used for model validation. It was concluded that the model is a reliable and fairly accurate tool for short-term prediction of growth of spruce-balsam fir forest in Canada. Future work on refinements of the model is discussed.