Abstract
Stand data from Lutz spruce (Picea × lutzii Little) forest types occurring on the Kenai Peninsula were analyzed by tree-based classification to develop a decision tree for classifying spruce beetle (Dendroctonusrufipennis Kby.) hazard. Model development and validation data sets contained 100 and 34 stand observations, respectively. The final decision-tree structure yielded seven possible hazard outcomes based on total stand basal area, percentage of total basal area composed of spruce, percentage of spruce basal area composed of trees with diameter >25 cm, stand elevation, and stand aspect. Three paths in the decision tree led to low-hazard outcomes (spruce basal area loss ≤10%); one path led to a high-hazard outcome (spruce basal area loss >40%). No paths through the decision tree led to a medium-hazard outcome (spruce basal area loss >10%, but ≤40%), but three less precise outcomes of low–medium and medium–high hazard were considered useful and retained in the final model. Results of model verification were considered very acceptable; in the worst case, predictions of high hazard were correct for 73% of the observations. Model validation results were also considered very acceptable considering the small number of observations available for this phase of analysis.

This publication has 0 references indexed in Scilit: