Abstract
[Display omitted]
•Quinoa and amaranth seeds differ in sucrose, K, Ca, UFA, and ω6/ω3 ratio.•Multivariate analyses allowing us to discriminate samples.•PCA can separate samples based on fatty acid, sugar, mineral, and fiber contents.•CDA multivariate models can be used to classify quinoa and amaranth samples.
Univariate (Analysis of Variance_ANOVA) and multivariate (Principal Component Analysis (PCA) and Canonical Discriminant Analysis (CDA)) analyses were performed in order to classify and authenticate the seeds from different varieties of quinoa (Chenopodium quinoa Will.), and amaranth (Amaranthus cruentus L.). The univariate analysis showed differences between species for sucrose, K, Ca, unsaturated fatty acids, and the ω6/ω3 ratio. Nevertheless, to strengthen this classification, a PCA was applied separating the samples in 2 groups; group 1, formed by quinoa seeds, presented higher contents of margaroleic, eicosadienoic, behenic, erucic, linolenic, linoleic, and gadoleic acids, proteins, sucrose, and total sugars. Group 2, formed by amaranth seeds, showed positive values for Mn, Mg, Fe, P, Zn, Ca, fiber, glucose, and ω6/ω3 ratio. Furthermore, the CDA models developed resulted in a probability of event of 100% when classifying the samples in the groups quinoa or amaranth, highlighting the good sensitivity of the models used.