This common bean study was carried out to improve the understanding of the several strategies that can be used to obtain segregant populations in self-pollinating plant breeding programs. Eight inbred lines were crossed in a pyramidal form, and four single, two double and one multiple hybrid were obtained. The potential of the different hybrids was investigated based on the performance of the F 2 and F 2 plus F 3 segregant populations. The potential for obtaining superior inbred lines was estimated by (m + a), which is the average of the F ∞ inbred lines, and by the probability of a given population producing inbred lines that outperform the yield of the Pérola cultivar by 20%. It was shown that the use of multiple crosses is not advantageous when breeders have well defined objectives and experimental conditions to assess and identify the most promising segregant populations.

-

Three segregating bulk populations derived from the cross between one early (Manteigão Fosco 11) and three normal cycle (Carioca MG, Ouro and Milionário) dry bean cultivars were assessed for natural selection effects. The populations were advanced from F 2 to F 13 . A seed sample was kept to represent the generation and the remnant used to obtain the following generation. Populations from each cross and their respective parents were assessed in randomized complete blocks with four replications, in three sowing seasons (February, July and November). The traits growth habit (determinate or indeterminate), weight of 100 grains (g), reaction to Colletotrichum lindemuthianum and grain yield were scored. Natural selection effects were observed on grain yield (the increase was 4.8% per generation), growth habit and weight of 100 grains in all crosses. Individuals with indeterminate growth habit and smaller seeds were predominantly kept. The reaction to C. lindemuthianum was neutral to natural selection.

The common bean cultivar "BRSMG Realce", recommended for the state of Minas Gerais, has high yield potential and resistance to powdery mildew, bacterial wilt and to anthracnose pathotypes 65, 73, 77, 81, 91, 475 and 479. The plants and pods are high, resulting in lower loss during mechanical harvest.

In an attempt to identify strategies for choosing common bean segregating populations, the diallel parameters, the contribution of homozygous (m+a') and heterozygous (d) loci, the heritability ( h a 2 ), the genetic distance (RAPD), and the selection potential (Z) were estimated in a study carried out in two stages. In the first, 30 hybrid combinations (partial diallel) were assessed in the F 2 and F 3 generations. Increasing grain yield was due dominance effect (d) although, both GCA and SCA were significant. d varied greatly and gave an information like SCA, however h a 2 alone was not indicative of success in a breeding program. In the second stage, five populations with lesser d and five with higher were selected, and twenty-seven families were taken from each, which were assessed with the parents and seven controls. The estimated parameters were used to check those from the first stage. There was partial agreement among the estimates and the percentages of lines superior to Pérola cultivar, which suggests that two or more estimates are needed to identify the promising segregating populations.

No abstract

Recombination can be a limiting factor in recurrent selection programs, especially in species such as common bean where artificial hybridization reduces seed setting. Therefore, the development of new techniques to improve recombination efficiency is important. Experiments were conducted in two sowing periods, July (autumn-winter season) and February (dry growing season), using the ESAL 686 and Small White cultivars as the female parents and the Carioca, Ouro Negro and ESAL 686 or Small White cultivars as male parents, depending on the female parent. Five technicians with different hybridization experience and two hybridization methods (with and without emasculation) were tested. A mixture of pollen from the three male parents was used in the pollinations, which were made in greenhouse. The resulting F 1 seeds were sown in the field and the male parent identified by morphological markers. The greatest pollination success (68%) occurred in the autumnwinter season without emasculation. It was also observed that the technicians differed in efficiency, but no method x technician interaction was detected. The pollen mixture was efficient to improve recombination.

No abstract