A multivariate approach to the selection of pea (Pisum sativum L.) lines obtained by the single seed descent technique

Abstract

Grain legumes, especially pea and lupine, are important agricultural crops in Central and Northern Europe. Because of the high level of protein in their seeds, these species constitute an alternative to soya meal imported from South America. The breeding of new cultivars in a short time is a major problem in pea breeding, particularly since the production of homozygous lines in the DH (doubled haploid) system, as used for cereals, for example, is very difficult. An alternative approach may be the single seed descent (SSD) method. The materials for this study were lines of pea obtained by that method combined with in vitro culture of embryos. This approach enabled 9 generations to be obtained in 2.5 years. First, SSD lines (F10) were investigated in a field experiment conducted in a complete randomized design with two replications, in which the dates of flowering and ripening and seed yield per plant were observed. Seventy seven lines out of 120 (64%) were selected for a subsequent experiment in which F4 generations of each cross combination were also included. In the second field experiment SSD lines were evaluated for earliness, plant height, seed yield per plot, and 1000-seed weight. The data were processed by multivariate analysis of variance and related methods, leading to the selection of 16 lines for further breeding experiments. Multivariate analysis can be helpful for breeders in seeking plants with a new favorable complex of yield-forming traits. <br><br><font color="red"><b> This article has been corrected. Link to the correction <u><a href="http://dx.doi.org/10.2298/GENSR1703127E">10.2298/GENSR1703127E</a><u></b></font>