Predicting susceptibility of white fir during a drought-associated outbreak of the fir engraver, <i>Scolytus</i><i>ventralis</i>, in California

Phenotypic traits were compared with a vigor (growth efficiency) index for accuracy in predicting susceptibility of white fir, Abiesconcolor (Gord. &amp; Glend.) Lindl., during a drought-associated outbreak of the fir engraver, Scolytusventralis LeC., in the central Sierra Nevada at Lake Tahoe, California. Predictor variables were estimated for 633 firs in six forest stands in 1987. After 2 years, virtually all of the trees had been attacked by the beetle, and 230 (36.3%) had been killed or were dying. In all of the predictor variables, firs that were killed differed significantly from those that survived. Compared with survivors, firs that died averaged shorter, more ragged crowns and lower growth efficiencies. Also, firs that died were more frequently dominant or codominant in the stand canopy and, when characterized in 1987, more often evidenced signs of being under current or recent (in 1985 or 1986) attack by the beetles. But, on either an individual tree or a stand basis, predictive accuracy was inadequate. On an individual tree basis, discriminant functions using either the phenotypic traits or vigor index as predictors produced overall percentages of correct classification little or no higher than would be obtained by predicting all trees would survive. On a stand basis, regression models using stand means for either the phenotypic traits or vigor index and white fir basal area as predictors statistically accounted for at least 95% of observed variation in basal area of white fir killed. But another model, using only white fir basal area, performed as well. The "best" model, containing white fir basal area and total stand basal area, accounted for over 98% of observed variation.