Combining Spectroscopic techniques with Multivariate statistical approaches for discrimination of Vicia seeds

Legume Seeds of Vicia species are cultivated and consumed worldwide for their nutritional value and bioactive compounds. Notably, Vicia faba (fava bean) seeds, with their many cultivars or varieties, are deeply rooted in cuisines of the Middle East and across the globe. In this work, simple and fast spectroscopic techniques, including UV and FT-IR spectroscopy, were used in combination with multivariate statistical techniques not only to discriminate between different varieties of fava beans but also to distinguish them from other Vicia legumes such as Vicia sativa and Vicia monantha. In addition, the total phytochemical phenolics and flavonoids and in vitro radical scavenging activity were assessed. Preliminary exploratory data analysis using PCA on both UV and FT-IR spectra was capable of distinguishing the seeds of fava bean varieties from other Vicia species. On the other hand, the FT-IR was limited in distinguishing between the varieties of fava beans compared to the UV spectra. Therefore, UV spectra were subjected to unsupervised techniques, PCA and HCA, and supervised classification techniques, SIMCA and PLS-DA, to construct useful discrimination models for eight varieties of fava beans. PCA and HCA successfully segregated the eight fava bean varieties into three informative clusters: the first cluster for the five traditional commercial Egyptian varieties, the second cluster for the two new Egyptian varieties Maryoute 2 and 3, and the third cluster for the Spanish variety Luz de otoño. Furthermore, SIMCA and PLS-DA models demonstrated well separation between these three classes of fava beans with 100% accurate classification of the validation set samples. In addition, the varieties of fava beans and other Vicia species showed a diverse content of Phenolics, flavonoids, and radical scavenging capacity, with the traditional Egyptian varieties of Sakha4 and Giza 843, as well as Vicia sativa and Vicia monantha, being the best. In conclusion, for the first time, UV spectroscopy combined with multivariate techniques could serve as a simple and fast method to distinguish between some Vicia seeds. Additionally, Vicia sativa, Vicia monantha, and the Sakha 4 and Giza 843 fava bean varieties might be superior to others in developing functional foods and phytopharmaceuticals.