Characterization of colored maize seed fractions using fluorescence spectroscopy and multivariate analysis

Abstract

Application of fluorescence spectroscopy combined with chemometrics algorithms provides rapid and non-destructive screening method in seed quality estimation, widely used in the agricultural industry and crop breeding. Fluorescence spectroscopy is a technique capable of detecting differs fluorophores among various colored maize seed cultivars and through different seed fractions. In the present study, we used the Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) algorithm to analyse the excitation-emission matrices (EEMs) of various cultivars of colored maize (Zea mays L.) seeds and its fractions. The EEMs were recorded as a set, with the excitation ranging from 280 nm to 330 nm and the emission spectra ranging from 300 nm to 550 nm. The MCR-ALS analysis yielded two major fluorescence components for all of the analysed samples. Both position and shape of component 1 (C1) varied among the samples. On the other hand, the position and shape were similar for component 2 (C2). C1 could be used as a marker for the discrimination of colored seeds and their fractions. The observed variations in C1 between the analysed seeds may be due to the presence of their individual fluorophores, assigned to anthocyanins, proteins, and phenolics. In conclusion, the MCR-ALS analysis of the seed emission spectra has a great potential for the rapid and non-expensive characterization of various cultivars of colored seeds.