Submitted to: Journal of Economic Entomology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/1/2008
Publication Date: 6/1/2008
Citation: Mankin, R.W., Smith, M.T., Tropp, J.M., Atkinson, E.B., Jong, D.Y. 2008. Detection of Anoplophora glabripennis (Coleoptera: Cerambycidae) larvae in different host trees and tissues by automated analyses of sound-impulse frequency and temporal patterns. Journal of Economic Entomology. 101:838-849. Interpretive Summary: Scientists at the Center for Medical, Agricultural, and Veterinary Entomology and the Beneficial Insect Introduction Research Unit collected and analyzed sounds produced by the Asian longhorned beetle in black locust, black willow, and red maple trees at cool and warm temperatures and different moisture conditions. The larvae of these quarantine pests are difficult to identify because they feed hidden inside the tree trunks. The signals were variable, but it was possible to distinguish between larval sounds and background noise with good reliability. It was found that analysis of the patterns of moving and chewing helped distinguish insect sounds from background more reliably than previously used analyses that relied primarily on the comparisons of frequencies alone. These results may be helpful in development of new methods to identify infested trees in quarantined areas.
Technical Abstract: Anoplophora glabripennis, an invasive pest quarantined in the U. S., is difficult to detect because the larvae feed unseen inside trees. Acoustic technology has potential for reducing costs and hazards of tree inspection, but development of practical methods for acoustic detection requires the solution of technical problems involving transmission of resonant frequencies in wood and high background noise levels in the urban environments where most infestations have occurred. A study was conducted to characterize sounds from larvae of different ages in cambium, sapwood, and heartwood of bolts from three host-tree species. Larval sounds in all of the tested trees and tissues consisted primarily of trains of brief, 3-10-ms impulses. There were no major differences in the spectral or temporal pattern characteristics of signals produced by larvae of different ages in each tissue, but larval sounds in sapwood often had fewer spectral peaks than sounds in cambium and heartwood. A large fraction, but not all background sounds could be discriminated from larval sounds by automated spectral analyses. In 3-min recordings from infested bolts, trains containing impulses in patterns called bursts occurred frequently, featuring 7-49 impulses separated by small intervals. Bursts were rarely detected in uninfested bolts. The occurrence of bursts was found to predict infestations more accurately than previously used automated spectral analyses alone. Bursts and other features of sounds that are identifiable by automated techniques may ultimately lead to improved pest-detection applications and new insights into pest behavior.