Pedigree selection in pea ( Pisum sativum L . )

Abdel-Haleem El-Shaieny. Two pea populations (Alaska X little Marvel) and (Alaska X Dwarf Gray Sugar) were used to improve green pod yield and some yield components trait by pedigree selection for several generations (F3, F4 and F5) with check cultivar (Markado), during three winter seasons. Highly significant differences between F3 families in both populations were detected, genotypic coefficient variation were detected for selection of green pod yield per plant. After two cycles of pedigree selection the criterion of selection was increased in population 1 and 2 by 4.98, 31.53, 32.72% and 6.23, 21.49, 20.36% compared with the bulk sample, check cv., and the best parent, respectively. Heritability in broad sense in F3 populations and two pedigree selection cycles were 96.02, 95.04, 98.38 and 96.83, 88.7, 94.17 respectively, for green pod yield per plant. Families No. 1, 10, 5, 8 and 6 had a heaviest green pod weight per plant in population 1. While, families No. 1, 5 and 10 in population 2 were attained the superiority of green pod yield per plant, compared to the bulk sample, check cv. and the best parent. Therefore, these families were recommended for use as new promising lines of pea but need more field experiments to evaluating them.

Garden pea (Pisum sativum L.) is one of the most important favorable legume crops grown in Egypt, during the winter seasons. The green pod and dry seed consider essential sources for protein and vitamin which used for human food as well as animal. Increasing yielding of the unit area and the pods quality could be gained by growing new and heavy pod yield cultivars and/or improving the cultural practices. In breeding program for increasing productivity, three important decisions that faced breeder to achieve (Tammam 1995). Firstly, is to identifying germplasm for desired characteristics. Secondly, is to good choice for prospective parents should be used for hybridization. Finally, the breeder must limit which method could be used in handling the resulting segregating populations. The right decisions were easily made, when the breeding goals are to improving qualitative traits. On the other hand, seed yield controlled by multiple genes, different environmental limitations and how to handle segregating populations. Pea breeders therefore, have permission from their nations for increasing crops productivity. found that significant differences among pea genotypes. They also, reported that the differences in genotypic and phenotypic coefficients of variation (GCV and PCV) were higher in traits of plant height, biological yield, seed and number of pods/ plant. This results indicates that selection of these traits would be more effective for further breeding improvement.
Yield as being the end trait of large number of attributes, is a complex trait, it has long recognized in legume crops (Adams, 1967 suggested that most effective selection in peas should consider days to flowering, pod length, number of seeds/pod and 100-seed weight. The efficient of selection between and within segregating generations varied from case to another (Abd-Elhady 2003). Genetic variability, character association and component analysis in F 4 generation of field pea were studied by (Garima and Groopa 2012). Selection for characters viz. plant height, days to 50% flowering, number of pods/plant, pod length and days to maturity have a good impact in field pea yield enhancement. There are many other factors are held stable, an increase in these traits individually would reflect an increase in seed yield. Mass selection for traits can be done and will be more effective if it begin in later hybrid generations ( Statistical procedures:-Data were recorded for individual plants on a random sample of ten plants from each family in F 3 , F 4 and F 5 generations. The means of the ten plants were subjected to the statistical and genetic analysis for the following characters (green pod yield/ plant, 100 green seed (g), number of seeds/ pod and pod length cm). The genetic parameters were estimated in F 3 , F 4 and F 5 generations. Estimates (σ 2 p) and (σ 2 g) variance and heritability (H 2 ) estimated were calculated from estimates of mean square of variance components of the select traits. The (σ 2 p) and (σ 2 g) variance as given by Al-Jiburi et al., (1958). The phenotypic (PCV) and genotypic (GCV) coefficients of variability were estimated as outlines by Burton (1952).

Analysis of variance and means:-
The analysis of variance for each of the two population F 3 families and their parents, data for four traits studied are presented in table (2). Highly significant differences were detected among the two population F3 families and their parents. Means for green pod yield per plant (g) ranged from 92.66 for P 2 to 125.95 (g) for F 3 families with average 115.23 (g) in population 1 and ranged from 100.5 (g) for P 2 to 115.9 (g) for F 3 families with average 107.83 (g). The trait 100-green seeds (g) ranged from 30.66 (g) for (P 1 and P 2 ) to 39.91 (g) for F 3 families with average 34.07 (g) in population 1, and 30.55 for P 2 to 33.33 for F 3 families with average 31.44 (g) in population 2. Number of seeds per pods ranged from 5.66 for P 2 to 7.42 for F 3 families with average 6.69 in population 1, and from 5.77 for P 2 to 6.99 for P 1 with average 6.3 in population 2. Pod length trait ranged from 5.7 for P 1 to 8.95 for P 2 with average 7.85 in population 2. Data are shown in table (3). Similar results were previously obtained by Singh (1995)

Response to selection and genetic parameters:
The genetic variability measured as genotypic coefficient of variability, the data are presented in table (4) in both populations was sufficient for selection in the base population for the two selection criteria green pod yield per plant (g). This may be due to it is easily to identify the highest green pod yield/ plant (g) genotypes. These findings are in agreement with the reports of Guleria et al.  (5) in the two cycles, respectively. Also, in population 2, selection towards increasing of green pod yield per plan (g) by 4.38, 19.08, 21.66 and 6.23, 21.49, 20.36% and 100-seed weight by 10.27, 9..28, 24.10 and 19.75, 21.9, 25.76% in tow cycles, respectively. In population 2 compared to the bulk sample, check cv. And the best parent, and decreased by 6.19 % and 10.00 % in trait of number of seeds and pod length (cm), respectively. Compared by the check cv. Data are shown in table (5).
With regard to population 2, after the second cycles in table (5), Pedigree selection for green pod yield per plant in family no 1, 5 and 10 were higher than the bulk sample, check cv. and best parent by 13.54, 23.27, 31.05 % and 10.54, 26.48, 25.31% and 8.40, 23.97 and 28.82%, respectively. These results indicated that selection would be more effective to improve all studied traits between two populations. Finaly, the high heritable for green pod yield per plant trait (h 2 ˃ 75) was indicated that the effectiveness of selection for these trait.

Conclusion:-
The pedigree selection revealed that three families under the same selection criterion of green pod yield per plant, no. 1, 5 and 10 which evaluated in this investigation are promising lines due to higher seed yield per plant than other families.