Barley is an important crop worldwide, and it can be affected by various fungi, among them Fusarium is one of the most relevant due to the economic losses caused by mycotoxin contamination. Enniantins (ENNs) are one of the emergent group of mycotoxins that have been found in grains around the world. Nowadays, the main analytical tools available to evaluate these contaminants are based on chromatographic techniques that are efficient but time-consuming and expensive. In this context, the present study aimed to assess the performance of near infrared (NIR) spectroscopy to detect and/or classify the enniatin (ENN) content on barley grains. Sixty samples of barley grains from three different regions of Brazil were investigated and the ENN content determined by UPLC-MS/MS. The levels found were then used to develop multivariate models based on infrared spectral data. The results indicated high incidence off ENN presence in the samples (>70 %) and the PLS-DA model determined by NIR data showed adequate values of sensitivity and sensibility (100 % and 94.2 %, respectively) distinguishing between contaminated and non-contaminated barley samples, demonstrating NIR as a promising tool to monitoring this emergent mycotoxin.
Keywords: Contaminant; Enniantin; FT-NIR; Fungi; Machine learning; PLS-DA.
Copyright © 2022 Elsevier Ltd. All rights reserved.