Pre-breeding of architectural traits related to direct harvesting in dry bean.

Abstract

Dry bean (Phaseolus Vulgaris L.) is an important crop for direct human consumption worldwide. In South Africa, it is a major source of plant proteins and income among growers. The crop is mainly grown by commercial producers for the market in Mpumalanga, Gauteng, Northwest, Free State, Limpopo, KwaZulu-Natal and the Eastern Cape provinces. Dry bean is known to be one of the most labour intensive field crops, especially the harvesting operation. In order to maximise profits, commercial producers prefer using the direct harvesting system that has the advantages of labour efficiency and saving time, compared to the pull/cut and window followed by threshing harvesting system. However, the success of the direct harvesting system requires a cultivar with a suitable upright architecture. The upright bean architecture suitable for direct harvesting is a complex trait that may only be improved through indirect selection for sub-component morphological traits associated with the trait. It is therefore important to evaluate diverse genotypes for morphological traits associated with the upright bean architecture to improve the trait. This study was aimed at (1) evaluating architectural traits related to direct harvesting and establishing relationships amongst traits on selected genotypes from the Andean gene pool, (2) estimating the phenotypic and genotypic variation, heritability, and genetic gain of traits related to direct harvesting on selected genotypes from the Andean gene pool, and (3) evaluating the architectural traits related to direct harvesting and establish the relationships amongst traits in local South African genotypes. The evaluation of architectural traits related to direct harvesting and establishing relationships amongst traits, was carried out on 30 Type I genotypes from the Andean gene pool. The trial was evaluated under field conditions in a 10 × 3 alpha lattice design at two sites. The traits collected were the days to physiological maturity, upright plant score, lodging, stem diameter, plant height, shattering, number of branches per plant, number of pods per plant, seed weight and seed yield. The analysis of variance showed highly significant differences (p ≤ 0.001) on genotypes across all traits, except for the number of branches, which was moderately significant (p ≤ 0.05). Superior genotypes were identified in each trait based on the grand mean. The stem diameter was identified to be important for multiple selections, because of its correlations with the upright plant architecture score, lodging, the days to physiological maturity and seed yield. Three principal components were extracted, accounting for 78.55% of total variation. Selection for the seed yield, stem diameter and plant height would be essential for improving the suitability to direct harvesting. The genotypes ADP 35, ADP 166, ADP 211, ADP 36, ADP 395, ADP 436, ADP 455, ADP 458, ADP 661, Mbomvu and Ukulinga were found to have thicker stem diameters and non-shattering, and therefore, they may be useful in improving both seed yield and the upright architecture suitable for direct harvesting. The estimation of the phenotypic and genotypic variation, heritability and genetic gain of traits related to direct harvesting was also carried out on 30 Type I genotypes from the Andean gene pool. The trial was evaluated in a 10 × 3 alpha lattice design at two sites. The traits collected were the days to physiological maturity, upright plant score, lodging, stem diameter, plant height, shattering, number of branches per plant, number of pods per plant, seed weight and seed yield. The analysis of variance showed highly significant differences among genotypes on all traits except the number of branches per plant, which showed moderately significant differences. The highly significant differences indicate the presence of genetic variability in the data set. A significant interaction of genotype with the site was observed on days to physiological maturity, lodging, plant height and seed weight. The phenotypic coefficient of variation was slightly higher than the genotypic coefficient of variation for all traits, showing little environmental influence on the expression of traits. Generally, a high variability was observed in the population. Broad sense heritability estimates ranged from moderate to high, except for the number of branches per plant, which recorded a low heritability value of 29%. The moderate to high heritability and expected genetic gains observed in the population could be exploited through selection and hybridisation during the improvement of the upright bean architecture suitable for direct harvesting. The evaluation of architectural traits related to direct harvesting, and establishing relationships amongst traits on local South African genotypes, was carried out on twenty four genotypes. The trial was laid out in a 6×4 alpha lattice design at Ukulinga Research Farm. The traits collected were, the days to 50% flowering, days to physiological maturity, upright plant score, lodging, stem diameter, plant height, shattering, number of branches per plant, number of pods per plant, seed weight and seed yield. The analysis of variance showed highly significant differences for the number of days to physiological maturity, days to 50% flowering, lodging, number of branches per plant, plant height, number of pods per plant, seed weight and the upright plant score, while it was very significant for the first pod insertion height and moderate for stem diameter, shattering and seed yield. Superior genotypes were identified using the grand means of the different traits and these genotypes may be used in a breeding programme to improve the suitability to combine harvesting. The correlation analysis showed that plant height would be useful in selection. However, an optimum height, to reduce lodging and improve the upright architecture should be selected for. The factor analysis revealed that seed yield, upright plant score, first pod insertion height, plant height and lodging had more influence on the variation in the data set and as such may be considered during selection. The cluster analysis grouped genotypes in two groups, while two genotypes were stand alone and the four principal components accounted for 78.55% of variation.