摘要：We examined attributes of piñon pine (Pinus edulis) associated with the probability of infestation by piñon ips (Ips confuses) in an outbreak in the Coconino National Forest, Arizona. We used data collected from 87 plots, 59 infested and 28 uninfested, and a logistic regression approach to estimate the probability of infestation based on plot and tree-level attributes. Piñon pine stand density index was a good predictor of the likelihood of infestation by piñon ips at the plot level, and a cross-validation analysis confirmed that the model correctly classified 82% of the cases. Diameter at root collar and piñon dwarf mistletoe infestation level were good predictors of individual tree infestation, and a cross-validation analysis indicated that the model correctly classified 72% of the cases. Results suggest that the occurrence of piñon ips infestations may be related to stress factors associated with increased stocking and piñon dwarf mistletoe infestations.
