https://www.euro-online.org/enog/inoc2007/Papers/mac-slots.html https://www.euro-online.org/enog/inoc2007/Papers/m https://www.euro-online.org/enog/inoc2007/Papers/mac-slots.html

Roasting chestnuts, not insects

No one wants to bite into a fruit to find that it contains a maggot or other insect larvae. Unfortunately, detecting such insect infestation before actually biting into the fruit can be quite difficult, especially for chestnuts. Now, though, US and Italian scientists have shown that NIR spectroscopy appears to be up to the job.

Chestnuts are mainly infested by the larvae of beetles such as weevils, which burrow into the fruit and then feed on it for up to 45 days, before exiting the fruit and dropping to the ground. The resultant 1mm hole in the chestnut is a clear sign that the chestnut has been infested, but while the larvae is still inside the chestnut there is little obvious, visible sign of infestation.

Up to now, the only of way for growers to detect infestation without damaging the chestnut has been to see whether it floats in salt water, with any that do float assumed to be infested. But this technique is not very accurate and results in many perfectly healthy chestnuts being discarded. So the team led by Riccardo Massantini at Tuscia University in Viterbo, Italy, decided to see whether NIR spectroscopy could do any better.

As reported in Postharvest Biology and Technology, they analysed 15kg of chestnuts with NIR spectroscopy and then cut the chestnuts open to find out which ones contained insect larvae. Analysing the resultant spectral data with principal components analysis revealed that the infected and non-infected chestnuts naturally split into two broad, slightly overlapping groups. However, a linear discriminant analysis model built from this data could accurately identify infected chestnuts from their NIR signal, although it worked best when the flat side of the chestnut was probed by NIR spectroscopy. When they compared this model against the conventional floatation system, they found it was 55% more accurate at identifying infected chestnuts.

Specific NIR wavelengths from 1100nm to 2300nm were preferentially absorbed by the infected chestnuts. From these wavelengths, Massantini and his team think that NIR spectroscopy is detecting both the chitin in the larvae and the higher water content that results from the damage caused by the larvae.

Blog tags: 
Techniques: