A predictive model for detection of Agrilus planipennis (Col., Buprestidae) larvae in girdled ash (Fraxinus spp.)

Abstract

Abstract Agrilus planipennis (emerald ash borer, Coleoptera: Buprestidae) is a pest of ash (Fraxinus spp.) in North America and has caused mortality of ash throughout its introduced range. One technique used for detection of A. planipennis is the establishment and peeling of girdled trap trees. In an effort to reduce the search effort and target detection survey efforts within ash trap trees, a predictive model was created using data from 2007 and validated using data from 2008. In 2007 and 2008, ash trap trees were established, harvested, peeled and inspected for A. planipennis larvae. Gaussian curves were fit to describe the relationship between stem diameter and relative proportion and frequency of larvae. The observed and predicted 2008 relative proportion and frequency of larvae did not differ in paired t‐tests. Within the relative proportion and frequency Gaussian models, the curves peaked at approximately 10 cm diameter signifying the greatest proportion and frequency of A. planipennis larvae occurred at 10 cm stem diameter. This peak was then bracketed by 2 cm on each side creating a target stem section with a top diameter of 8 cm and a bottom diameter of 12 cm. A simple linear regression was fit to describe the relationship between the larvae count within the targeted 8–12 cm section of tree and the larvae per cm3 of phloem per tree for 2007. The observed and predicted 2008 larvae per cm3 did not differ in a paired t‐test. Targeting the 8–12 cm diameter section of the trap tree stem will reduce the amount of the tree peeled to less than 45% with more than 50% of A. planipennis larvae within the tree encountered in this targeted section. This reduction in the amount of tree peeled will greatly increase the number of trees and area surveyed for A. planipennis detection and population delimiting surveys.