Abstract: Variation and transgressive segregations were studied in the backcross generations of populations involving three crosses in long bean [Vigna sesquipedalis (L.) Fruw]. Trends were different for different crosses. In Cross 1, high mean, variance and significant transgressive segregation (STS) pattern were exhibited by BC2F1 generation for shelf life, pod yield per plant and pod weight. In Cross 2, high means were shown by BC1F1 generations for shelf life and pod yield per plant, but high STS were observed in BC2F1 generation for shelf life, BC1F1 for pod yield per plant and BC1F2 for pod weight. In Cross 3, BC1F2 generation revealed high variances for shelf life, pod yield per plant and number of pods per plant, but highest STS values were revealed for BC1F3 (selected) for shelf life and pod yield per plant, BC2F1 for number of pods per plant, and BC1F2 for pod weight. In general, shelf life was found to have exhibited high mean and STS in the BC2F1 generation. Pod length, in general did not show good transgressive segregation in the crosses. Exploitation of these transgressive segregants in breeding programmes for improvement of the character of interest is suggested.
Introduction
Long bean (Vigna sesquipedalis (L.) Fruw) is one of the most commonly grown vegetables in Malaysia, and ranks the second in importance for total vegetable consumption (Anang, 1988). The young and unripe pods are harvested for consumption. The crude protein content of the seeds and in the pods are high, about 29 and 23%, respectively (Mak, 1973; Poh, 1976). However, as in most vegetables, long bean is highly perishable, having short shelf life. Little attention has been given on genetic improvement efforts to extend its shelf life.
Introgression of genes from donor parents with long shelf life into genetic backgrounds of the recurrent parents is one of the potential means of creating variability for qualitative as well as quantitative characters. There is no available information on gene segregation in backcross generations from crosses between parents with high and low shelf lives in long bean. The aim of the present study was to evaluate the variation and pattern of the transgressive segregation revealed in populations of long bean developed from a backcrossing programme for characters of economic importance. The effectiveness of backcross breeding programmes can be improved by evaluating transgressive segregants for shelf life, and subsequently selecting for those with high yield and other related characters, before crossing them back to the recurrent parents.
Materials and Methods
Four cultivars of long bean, namely L30, KU7, KU8 and CSL19 were used as germplasm sources in the study. L30 had long shelf life and was used as donor parent. KU7, KU8 and CSL19 had shorter shelf life but with desirable pod characteristics, and were therefore used as the recurrent parent.
Using L30 as the donor parent, a total of three populations (henceforth designated as crosses) were developed. They were Cross 1 (L30 x KU7), Cross 2 (L30 x KU8) and Cross 3 (L30 x CSL19). The F1 of each cross was backcrossed to their respective recurrent parents to obtain the BC1F1 generation. Seeds of the BC1F1 generation of all crosses were sown to obtain seeds for BC1F2 generation. A total of 200 plants from each cross were grown. From these 200 plants, five plants from each cross were selected for shelf life, pod length and pod yield. BC1F3 seeds (from BC1F2 plants) from each of the five plants were grown separately along with their respective recurrent parents to be used as pollen parents. The selected BC1F3 plants of each cross were backcrossed to their corresponding recurrent parents to produce BC2F1 seeds. The resulting BC2F1 seeds from each cross were taken from plants with shelf life longer than the others. Seeds of the P1, P2, BC1F1, BC1F2 (selected) and BC2F1 generations of Cross 1, Cross 2 and Cross 3 were evaluated separately in the field using a randomized complete block design with four replications. Each generation was represented by six plants in each replication. Individual plants were maintained in single-row beds, each measuring 6.0 m long and 0.9 m wide. The distance between plants was 1.0 m. Shelf life, pod yield per plant, pods per plant, pod length and pod weight were recorded for each plant. Percentage of transgressive segregants in BC1F1, BC1F3 (selected) and BC2F1 generations were calculated according to Reddy and Singh (1990). Extreme progeny types with values that exceeded their better parent means were classified as transgressive segregants (TS) and those regarded as significantly transgressive segregants were those with values exceeding the means of their better parents, when LSD comparison at p≤ 0.05 was made.
Results and Discussion
The analysis of variance revealed that the backcross generations of all the crosses were significantly different for all the characters measured. Means, variances and percentages of transgressive segregants in Cross 1, Cross 2 and Cross 3 are presented in Tables 1, 2 and 3, respectively. Mean values of generations for shelf life differed with crosses, where in general, BC1F3 (selected) generation manifested reasonably high mean in all the crosses. BC1F2 in general recorded high variation for shelf life in Crosses 2 and 3, but not in Cross 1.
In Cross 1, BC2F1 generation produced the highest variance (0.68), while BC2F1 and BC1F1 generations produced highest STS for shelf life (both were 20.8%) (Table 1). The same trend was also true for pod weight. In the BC2F1 generation, high mean (1501.6 g), large variance (97975.3), high TS (75.0%) and high STS (62.5%) were obtained for pod yield per plant. For number of pods per plant, BC2F1 generation also gave the highest mean value, but BC1F2 produced the highest variance.
Reddy and Singh (1990) also observed high mean values and variances for number of pods per plant, and high STS for pod yield in BC1F2 generation of long bean.
Table 2 shows the variations and transgressive segregations for the five characters measured in different backcross generations in Cross 2.
| Table 1: | Mean, variation and percentage of transgressive segregants for characters measured in the backcross generations of Cross 1 ( L30 x KU7) of long bean |
| TS = transgressive segregants; STS = significant transgressive segregants (LSD at p≤ 0.05) | |
| Table 2: | Mean, variation and percentage of transgressive segregants for characters measured in the backcross generations of Cross 2 ( L30 x KU8 ) of long bean |
| TS = transgressive segregants; STS = significant transgressive segregants (LSD at p≤ 0.05) | |
The BC1F1 generation had high mean value for shelf life, revealing the effect of heterosis, while BC1F2 expectedly showed high variance for the trait. Cox (1984) showed that the genetic variance among the backcross lines derived from comparatively higher number of backcross families would be expected to be higher. TS and STS for shelf life were observed only in the BC2F1 generations (20.8 and 4.2%, respectively). High mean, large variance, high number of TS and STS for pod yield per plant and number of pods per plant were observed in BC1F1 generations. Likewise, BC1F2 had high mean, variance, TS and STS with respect to pod length. All the backcross generations in Cross 2 expressed similar mean values for pod weight (18.6, 18.9, 18.9 and 18.1 g, for BC1F1, BC1F2, BC1F3 (selected) and BC2F1, respectively). However, for pod weight, the BC1F2 generation showed high variance, high TS and high STS (Table 2).
In Cross 3 (Table 3), BC2F1 generation manifested highest mean value (3.5 days) and highest TS (39.1%) for shelf life, while BC1F3 (selected) generation recorded highest STS. As expected, BC1F2 exhibited highest variance for shelf life. BC1F2 was also observed to have high mean and high TS and STS. Interestingly, BC1F1 manifested higher variance while BC2F1 produced no TS and STS in contrast to the other backcross generations. This could have possibly been due to selection conducted earlier on the plants causing pod length of the generation shorter than the better parent, as an indirect effect. For pod yield per plant, the BC1F1 generation exhibited highest mean value (1371.7 g) and TS (83.3%), while the BC1F2 generation showed highest variance (175309.69), and BC1F3 (selected) generation showed highest STS. For number of pods per plant, highest mean was expressed by BC1F1 generation (64.4), which also revealed highest TS (41.7). Highest variance was shown by BC1F2 generation (226.80), while highest STS was revealed by BC2F1 generation (26.1%). The different backcross generations did not differ much among each other for pod weight. High variances were shown by the BC1F2 and BC2F1 generations (7.34 and 7.67, respectively). The BC1F3 (selected) and BC1F2 generations were the highest for TS (45.8%) and STS (20.8), respectively.
In general, most of the later backcross generations in the crosses did not have transgressive segregants for pod length. This could have been due to selection of plants having optimum pod length of 50 to 55 cm previously conducted on the populations.
| Table 3: | Mean, variation and number of transgressive segregants for characters measured in the backcross generations of Cross 3 ( L30 x CSL 19) of long bean |
| TS = transgressive segregants; STS = significant transgressive segregants (LSD at p≤ 0.05) | |
For most other characters measured, significant amounts of transgressive segregants were found in all the backcross generations of all the three crosses. The presence of transgressive segregants in backcross generations have also been reported for different characters in tomato (Franklin, 1987), grain sorghum (Rao et al., 1983) and soybean (Cianzio and Fehr, 1982). These transgressive segregants were successfully exploited in developing cultivars possessing combinations of various selected traits. Higher variances coupled with higher transgressive segregation for shelf life and pod yield were observed in BC2F1 generation of Crosses 1 and 2. In Cross 3, high variance for shelf life and pod yield were observed in BC1F2 generation, but the highest transgressive segregation was found in BC1F3 (selected) generation. Hence a selection of segregants with the character of interest could be exploited for further improvement of the trait in all the three crosses studied.