Feijoa [Acca sellowiana (Berg) Burret] accessions characterization and discrimination by chemometrics
Keywords
Abstract
Background: Feijoa [Acca sellowiana (Berg) Burret] is a Brazilian native fruit with few commercial-level plantations and high agroindustrial potential. A genotype evaluation experiment was conducted from 1996 onward, aiming to obtain fruits based on the agronomical parameters; however, the selection based on chemical composition had not been evaluated with respect to developing a new cultivar. Accordingly, the present study aimed to discriminate seven accessions of feijoa in terms of nutritional composition, phenolic compounds and antioxidant activity using multivariate analysis (principal component analysis and multivariate contrast), targeting the potential production of a new cultivar with better nutritional value and high antioxidant capacity.
Results: Feijoa husk presented high content of ashes, lipids, proteins, carbohydrates, phenolic compounds and antioxidant activity compared to feijoa pulp. However, only feijoa pulp was selected to multivariate analysis because it is the fruit edible part. Data variability was explained in 78% and the feijoa pulp accessions were discriminated into four groups related to their characteristics. The accession 5 discrimination can be explained by the high content of ashes, carbohydrates, soluble solids, phenolic compounds and antioxidant activity. Accession 6 was also discriminated by the high content of total acidity, pH and proteins, as well as a low content of soluble solids.
Conclusion: Feijoa accessions may be indicated for increasing plant selection via hybridization with the other accessions, aiming to produce new cultivars with better nutritional composition and antioxidant capacity. For example, accession 5 is the most suited fruit for human consumption and is a potential plant with respect to becoming a new cultivar. © 2020 Society of Chemical Industry.
Keywords: Brazilian native fruits; genetic improvement; multivariate analysis; multivariate contrast; principal component analysis.
