QUANTITATIVE ACOUSTICAL DETECTION OF LARVAE FEEDING INSIDE KERNELS OF GRAIN

An automated, computer-based electronic acoustic system was developed to quantify infestation of internally feeding larvae in a grain sample using spatial localization of insects in the sample. Localization was determined using arrival times of sounds produced by insect feeding activity as received by an array of acoustic transducers. In a test conducted with 0-3 fourth instars of the rice weevil, Sitophilus oryzae (L.), in 1-kg samples of wheat, the system overassessed the number of larvae present in 6% of the trials and underassessed the number of larvae present in 34% of the trials. When Federal Grain Inspection Service (FGIS) standards were applied in evaluating performance, the system was 92% accurate in grading ''clean'' grain and 64% accurate in grading ''infested'' grain.