NIR spectroscopy vs. food pests: The case of stored rice. Issue 5 (August 2019)

• Record Type: Journal Article
• Authors: Biancolillo, Alessandra; Firmani, Patrizia; Bucci, Remo; Magrì, Andrea; Marini, Federico
• Is Part Of: NIR news. Volume 30:Issue 5/6(2019)
• Page Start: 18
• Page End: 21
• Publication Date: 2019-08
• Subjects: Food pests -- stored cereals -- NIR spectroscopy -- partial least squares discriminant analysis; Near infrared spectroscopy -- Periodicals; 543.5
• DOI: 10.1177/0960336019854284
• Languages: English
• ISSNs: 0960-3360

Abstract

Rice is one of the most widely consumed cereals, and it represents a staple food for several populations all over the world. One of the main peculiarities of this food commodity is that it can be stored for relatively long periods, maintaining its nutritional properties and its organoleptic attributes. Nevertheless, it is not uncommon that pests infest granaries, altering the characteristics of cereals, making them less valuable and/or un-edible. Avoiding the presence of insects in storehouses is a task difficult to accomplish; consequently, different methods for detecting pests' infestation in food commodities have been developed. In general, they are based on physical sensors, or they exploit analytical techniques such as high-performance liquid chromatography, nuclear magnetic resonance spectroscopy, enzyme-linked immunosorbent assays or other bio-chemical devices to detect insects' byproducts. Despite these approaches providing accurate results, they all present the same inconvenience: they are destructive. In the light of these considerations, the present work aims at developing a non-destructive NIR-based strategy to detect the presence of Plodia interpunctella, one of the most common intruders of granaries, in rice parcels. In order to achieve this goal, 1525 samples of rice have been analysed by NIR spectroscopy and then partial least squares discriminant analysis was used to discriminate the edible grains from the infested ones. The proposed methodology provided extremely good results, properly assigning 484 test samples over 500.