Predicting crown class in three western conifer species

Assessing the crown class (dominant, codominant, intermediate, suppressed) of a tree is a subjective procedure. Most definitions of crown class are based on the relative height of a tree and (or) the amount of light that is incident on the tree crown. With this research, we devised a classification scheme, based on easily measured tree variables, to assign a crown class to trees. Our data consisted of tree measurements, including crown class, from four stem-mapped 0.05-ha sample plots with buffers. The light model tRAYci was used to assess the light incident on each tree crown. These data gave us field-based and light-based assessments of crown class. The classification and regression tree technique — with diameter at breast height (DBH), height, relative DBH, relative height, height to the crown base, and crown depth as variables — was used to classify the trees. Accuracy rates of 91% and 82% were achieved for the field-based and light-based assessments of crown class, respectively.