Detection of External and Internal Insect Infestation in Wheat by NearInfrared Reflectance Spectroscopy

Abstract

Near-infrared spectroscopy (NIR) in the reflectance mode was investigated for the rapid, automatic and non-destructive detection of insect stored-grain pests external or internal to wheat kernels. Convincing calibration performance was obtained for external adultOryzaephilus surinamensis(L) (saw-toothed grain beetle) in unmilled samples including varieties Beaver (soft wheat) and Mercia (hard wheat) at two moisture contents. With this large substrate variability, the method could differentiate between uninfested samples and samples with approximately 270 insects kg−1 or more. Milling made no improvement. Large spectral differences were observed between uninfested kernels and kernels infested internally withSitophilus granarius(L) (grain weevil) larvae or pupae, arising from both a changed chemical composition and physical structure. Single uninfested and infested kernels were discriminated by their second derivative (d2) spectra. For both external and internal infestation there was substantial evidence that insect protein and/or chitin and moisture were being detected. NIR should be useful as a rapid method of detection.