Combined analysis of variance of partial diallel crosses of soybean for yield and yield-related traits across locations revealed that the two environments were significantly different from each other (Table 2). Entries (parents and crosses) were significantly different from each other for all the traits studied, indicating sufficient genetic variability among the parents and crosses generated (
Kose, 2019). Significant interaction between environment and entries was observed in all the measured traits, except for number of pods/plant and seed yield/plant. There were significant general combining ability (GCA) effects across environments for almost all the measured traits, except for days to poding. Similarly, the specific combining ability (SCA) effects across environments were significant for days to poding, number of pods/plant and seed yield/plant, exhibiting that variability in the breeding material can be attributed to both additive and non-additive gene effects. Highly significant GCA × environment interactions for almost all the measured traits showed that the performance of parents used in the study was influenced by environment and thus, testing under different environments will ensure selection of stable parents that can perform to the potential of that environment
(Machado et al., 2009).
Greater magnitude of GCA compared to SCA was observed for all measured traits, which reveals the prevalence of additive gene action, indicating and that selection will be effective to improve the traits
(Gravina et al., 2004; Nazim et al., 2014). The GCA: SCA ratio close to unity for all measured traits showed that the parents contributed mostly to the performance of the crosses observed, and influence of the environment was minimal and thus, there is preponderance of additive gene action controlling traits studied
(Murtadha et al., 2018). Adsul et al., 2016 reported that additive gene action was found predominant in the inheritance of 100-seed weight and yield/plant in segregating population of soybean. Also,
Umar et al., (2017) reported the importance of additive gene action in inheritance of days to 50% flowering, days to maturity, number of pods/plant and 100-seed weight.
Nassar (2013) also observed high GCA: SCA ratio for earliness and number of pods/plant in soybean.
The estimates of GCA effects of parental lines used in the study across Ibadan and Fashola (Table 3), revealed that TGx 1988-5F had desirable negative and significant GCA effect for days to flowering and poding. The parent TGx 1448-2E gave significant and positive GCA effect for number of pods/plant and seed yield/plant, showing the importance of this parent in improving these traits. Soybean parental lines with significant GCA have been reported by
Durai and Subbalakshmi (2009). Good GCA in soybean for yield and its related traits have also been reported earlier by
Srivastava et al., (1978) and
Sharma and Phul (1994).
The SCA effects of the cross TGx 1835-10E × TGx 1989-10F was negative and significant and desirable for days to flowering (Table 4). Two crosses, TGx 1988-5F × TGx 1989-19F and TGx 1485-1D × TGx 1835-10E exhibited significant and positive SCA effects for days to poding. Significant good specific combining ability for reduced plant height was observed in cross TGx 1987-62F × TGx 1988-5F. Crosses having positive and significant SCA effects for number of pods/plant are TGx 1988-5F × TGx 1987-10F and TGx 1485-1D × TGx 1448-2E. Significant and positive SCA effect for seed yield/plant was observed in the cross TGx 1988-5F × TGx 1989-19F.
Datt et al., (2011) have also reported crosses with good SCA for earliness and grain yield/plant in soybean. Crosses showing good specific combining ability for traits studied have either parent as good or average combiners. According to
Kenga et al., (2004), cross combinations with favorable SCA estimates and involving at least one of the parents with good GCA estimate would likely enhance the concentration of favorable alleles to improve traits of interest.
Parental varieties of soybean used in the study had higher GCA than SCA showing preponderance of additive gene action controlling seed yield and its related traits. Hence, selection for measured traits at early growth of segregating populations might be effective. As evidenced by their significant GCA effects, the parental line TGx 1988-5F can be used to improve earliness, while TGx 1448-2E can be used to improve seed yield in soybean breeding program.