DEVELOPMENT AND VALIDATION OF A FIXED-PRECISION SEQUENTIAL SAMPLING PLAN FOR ESTIMATING BROOD ADULT DENSITY OF DENDROCTONVS PSEUDOTSUGAE (COLEOPTERA: SCOLYTIDAE)

Abstract

AbstractThe Douglas-fir beetle, Dendroctonus pseudotsugae Hopkins, attacks Douglas-fir, Pseudotsuga menziesii (Mirb.) Franco (Pinaceae), throughout western North America. Periodic outbreaks cause increased mortality of its host. Land managers and forest health specialists often need to determine population trends of this insect. Bark samples were obtained from 326 trees distributed over 21 stands during a 2-year period in late winter to early spring of 1997 and 1998 in the Colorado Front Range. The variance to mean relationship of brood adults was examined using the Taylor power law, and a fixed-precision sampling plan was developed using Green’s method. Stop lines and minimum number of samples required to estimate brood adult density per 0.046 m2 with precision levels of 0.1, 0.2, and 0.3 were calculated. A resampling simulation conducted with an independent data set indicated that desired precision levels were not met. Theoretical precision levels were adjusted until desired precision levels were achieved. Average number of samples needed to estimate brood adult densities up to 25.1 adults per 0.046 m2 with precision levels of 0.09, 0.2, and 0.3 were 91, 20, and 8, respectively. For densities greater than 25.1 brood adults per 0.046 m2, conservative estimates are obtained with 72, 15, and 6 samples for precision levels of 0.09, 0.2, and 0.3, respectively. An emergence ratio can be obtained by dividing the estimated density of brood adults by twice the number of gallery starts. This system provides the user with an immediate assessment of the population trend of Douglas-fir beetle. The data collected compare favorably with data from other Douglas-fir beetle outbreaks reported in the literature. The use of this plan outside the Colorado Front Range, or by sampling at a different height, should be cautioned until additional data from other locations and sampling heights are examined.