Research Article

Multivariate analysis of backcross progeny of Passiflora L. (Passifloraceae) for pre-breeding genotype selection

Published: November 30, 2015
Genet. Mol. Res. 14 (4) : 15376-15389 DOI: 10.4238/2015.November.30.15

Abstract

The Ward-MLM procedure was used to evaluate genetic variation in four backcross progenies and in their parents, hybrid F1 HD13 and donor parent Passiflora sublanceolata. Sixteen quantitative descriptors and five qualitative characteristics of relevance to ornamental flower production were assessed. Using the pseudo-F and pseudo-T² criteria, we identified four groups among these plants in two evaluation periods. In both evaluations, the BC1 plants showed greater dissimilarity to their recurrent parent, but showed high genetic similarity with the P. sublanceolata parent. The first two canonical variables produced by the Ward-MLM procedure accounted for over 90% of the variation in both evaluation periods, enabling the representation of diversity through two-dimensional graphics. Groups II and IV were formed in the first assessment period. Groups I and IV formed in the second period and showed plants with selection potential. We found that it was essential to use both qualitative and quantitative variables for this analysis. Assessments of quantitative descriptors indicate that the selection of BC1 plants can be performed in any of the four progenies. Because of the similarities observed for some floral descriptors between BC1 and the P. sublanceolata parent, a second generation backcross was not recommended. However, the selection of BC1 plants for evaluation and direct use as an ornamental cultivar, or as a resource in other breeding programs, can be recommended.

The Ward-MLM procedure was used to evaluate genetic variation in four backcross progenies and in their parents, hybrid F1 HD13 and donor parent Passiflora sublanceolata. Sixteen quantitative descriptors and five qualitative characteristics of relevance to ornamental flower production were assessed. Using the pseudo-F and pseudo-T² criteria, we identified four groups among these plants in two evaluation periods. In both evaluations, the BC1 plants showed greater dissimilarity to their recurrent parent, but showed high genetic similarity with the P. sublanceolata parent. The first two canonical variables produced by the Ward-MLM procedure accounted for over 90% of the variation in both evaluation periods, enabling the representation of diversity through two-dimensional graphics. Groups II and IV were formed in the first assessment period. Groups I and IV formed in the second period and showed plants with selection potential. We found that it was essential to use both qualitative and quantitative variables for this analysis. Assessments of quantitative descriptors indicate that the selection of BC1 plants can be performed in any of the four progenies. Because of the similarities observed for some floral descriptors between BC1 and the P. sublanceolata parent, a second generation backcross was not recommended. However, the selection of BC1 plants for evaluation and direct use as an ornamental cultivar, or as a resource in other breeding programs, can be recommended.