Characterisation and pre-selection of Acca sellowiana genotypes by multivariate analysis

Abstract

Feijoa (Acca sellowiana) is a native Brazilian fruit with a peculiar flavour, a considerable amount of bioactive compounds and antioxidant activity. Even though this fruit tree is currently cultivated in several countries around the world, in Brazil, the process of domestication is underway, and the selection and breeding of new genotypes that are more productive and with better fruit quality is necessary. The objective of this work was to evaluate phenotypic diversity among and within progeny and to study the correlations among the quality variables of Feijoa fruit, seeking to select individuals with superior characteristics using principal component analysis. The parents who formed the progeny (families) were selected from a participatory breeding programme. We observed that individuals 47 and 93 had a combination of desirable fruit characteristics for selection, and individuals 15, 910, 98 and 410 should be selected for future crossings, as they had a high total fruit mass and soluble solid content or the highest percentage of pulp and rounded fruit shape. Larger fruit, in general, had a lower percentage of pulp. Principal component analysis is a viable tool in the pre-selection of new genotypes and potential progenitors for Feijoa breeding programmes.