Analysis of variance
The mean performance and analysis of variance indicated that all the genotypes differ significantly for all the traits in both environments, as well as in pooled analysis. Polled analysis of variance revealed that the environmental differences were statistically significant for all the traits except awn length and grain weight per spike. The differences due to pooled analysis among the wheat genotypes were statistically significant for plant height, peduncle length, spike length, awn length, spikelets per spike, flag leaf area, 1000 grain weight and it was non-significant for days to 75% heading, days to maturity, tillers per plant, grains per spike, grain weight per spike, grain yield per plant and harvest index. The mean squares due to genotype x environment (G x E) interactions were significant for all the traits except peduncle length, indicate the major contribution of G X E interaction than genotypic component in the expression of these traits. The significant mean square due to genotype, environment and G X E for plant height, spike length, spikelets/spike, flag leaf area and 1000 grain weight suggested the importance of both genotype and environment components for these traits. Therefore, inference of the present study and also the previous work (Ahmed et al., 2007 and Almeselmani et al., 2011)revealed that the high yielding wheat genotypes with water stress tolerance cannot simply developed by crossing between the water stress tolerant and high yielding genotypes without referring to environmental effect.
Analysis of variance for combining ability
The analysis of variance for combining ability was performed for fifteen yield and its contributing traits in irrigated (E1) and rainfed (E2) conditions. The mean squares due to crosses were partitioned into mean squares due to testers, due to lines and line x tester interaction components. The mean square due to lines were significant for ten characters in E1, days to 75% heading, days to maturity, plant height, spike length, peduncle length, awn length, tillers per plant, 1000 grain weight, grain weight per spike and biological yield per plant. In E2, ten characters exhibited significant mean squares namely days to 75% heading, plant height, spike length, peduncle length, awn length, spikelets per spike, flag leaf area (cm2), 1000 grain weight, grain yield per plant and biological yield per plant.
Mean squares due to testers were significant for five characters namely plant height, spike length, spikelets per spike, flag leaf area (cm2) and grains per spike in E1, while in E2 conditions four characters viz. spike length, awn length, spikelets per spike and flag leaf area (cm2).
Mean squares due to line x tester interaction were found to be significant for all fifteen characters namely days to 75% heading, days to maturity, plant height, spike length, peduncle length, awn length, tillers per plant, spikelets per spike, flag leaf area (cm2), 1000 grain weight, grain weight per spike, grain yield per plant, grains per spike, biological yield per plant and harvest index in E1 and twelve characters in E2 namely days to 75% heading, plant height, peduncle length, awn length, spikelets per spike, flag leaf area (cm2), 1000 grain weight, grain weight per spike, grain yield per plant, grains per spike, biological yield per plant and harvest index.
Higher magnitude of line x tester component than either due to lines or testers indicated the predominant role of non-additive gene action (dominance). The results from the present study revealed that line x tester interaction are significant for most of the traits in both environments. Line x tester interaction is far more important than line or tester alone in deciding the hybrid performance. Similar findings were also reported by Srivastava et al., (2012) and Farooq et al., (2019).
Variance components of combining ability
In irrigated condition, highest general combining ability variance (σ2gca) was observed for biological yield per plant followed by plant height. Rest of the characters showed relatively smaller amount of σ2gca. However, in rainfed condition, highest gca variance was observed for harvest index followed by biological yield / plant while rest of the characters showed relatively smaller amount of σ2gca. In irrigated condition, maximum variance for specific combing ability (σ2sca) was observed for the trait biological yield per plant followed by harvest index. The estimate of σ2gca/σ2sca indicated predominance of non-additive gene action for all the characters. However, in rainfed condition, maximum σ2sca was observed for the trait grains per spike followed by biological yield / plant. The estimates of σ2gca/σ2sca indicate predominance of non-additive gene action for yield and its contributing characters condition as in E1. Results presented by Rajesh et al., (2002), Majeed et al., (2011), Srivastava et al., (2012) and Farooq et al., (2019) also indicated preponderance of non- additive gene effects in the expression for grain yield and its attributes.
Considering the above results, it may be concluded that there was predominant role of non-additive gene action in the inheritance of all the traits in both irrigated and rainfed condition. The most efficient way for utilizing the non-additive genetic variance is through the exploitation of heterosis and the selection process for identification of superior plant type should be postponed to further generations like F4 or F5. Since it was observed that sca was the predominant contributor to genetic variance, thus, it is suggested that selection of sca is likely to be the most effective method to exploit hybrid vigour.
Estimates of combining ability effect
In combining ability effects, the general combining ability (gca) effect represent the additive gene action and specific combining ability (sca) effects represent the non- additive gene action interactions. The non-additive gene effects contribute in the improvement of grain yield in self-pollinated crops by the commercial exploitation of heterosis. In self-pollinated crops, however, the additive x additive type of interaction is also feasible in later generations and can be exploited for the improvement of grain yield and its attributes. The results of gca effect of parentes and sca effect of all cross combinations are given in Table1 and 2.
In irrigated condition (E1), VL3001 was identified as good general combiner for maximum number (nine) of traits viz., plant height, peduncle length, awn length, spikelets per spike, flag leaf area, grain weight/spike, grain yield/plant, grains/spike, biological yield/plant followed by FRANCOLIN#1/BAJ#1 and WH1142 for eight traits. FRANCOLIN#1/BAJ#1 was identified as good general combiner for days to 75% heading, days to maturity, plant height, peduncle length, awn length, flag leaf area, 1000 grain weight and harvest index while, WH1142 was a good general combiner for spike length, peduncle length, awn length, spikelets per spike, flag leaf area, grain yield/plant, grains/spike and biological yield/plant.In rainfed condition (E2), KACHU*2//WHEAR/SOKOLL, FRANCOLIN #1/BAJ#1 and VL3001 were identified as good general combiner for maximum number (seven) of traits. KACHU*2//WHEAR/SOKOLL was identified as a good general combiner for plant height, spike length, peduncle length, spikelets per spike, flag leaf area, grain weight/spike, grains/spike, whereas, FRANCOLIN #1/BAJ#1 for days to 75% heading, awn length, productive tillers/plant, spikelets per spike, flag leaf area, grains/spike and harvest index, VL3001 was good for days to maturity, plant height, peduncle length, flag leaf area, 1000 grain weight, grain yield/plant and harvest index.VL3001 was identified as a good general combiner for maximum number (nine) traits in pooled condition viz., plant height, peduncle length, spikelets per spike, flag leaf area, grain weight/spike, grain yield/plant, grains/spike, biological yield/plant and harvest index followed by BOW/VEE/5/ND/VG9144//KAL/ BBB/YACO/4/CHIL/6/CASKOR/3/… for days to 75% heading, days to maturity, spikelets per spike, flag leaf area, 1000 grain weight, grain weight per spike and grains per spike and FRANCOLIN#1/BAJ#1 for days to 75% heading, days to maturity, plant height, peduncle length, awn length, flag leaf area and harvest index.
In specific combining ability, cross BECARD/KACHU × WH1080 appeared as good specific combination for maximum number (eight) of traits viz., plant height, spike length, peduncle length, awn length, productive tillers per plant, grain yield per plant, grains per spike, biological yield per plant followed by KACHU*2//WHEAR/SOKOLL × WH1080 for days to 75% heading, awn length, productive tillers/plant, spikelets/spike, grain yield per plant biological yield per plant and harvest index and C306 × WH1142 for spike length, peduncle length, awn length, flag leaf area, grain weight per spike, grains per spike and biological yield per plant in irrigated condition (E1). In rainfed condition (E2) BOW/VEE/5/ND/VG9144//KAL/BBB/YACO/ 4/CHIL /6/CASKOR/3/… × WH1080 was identified as good specific combination for maximum number (eight) of traits viz., spike length, productive tillers per plant, flag leaf area, 1000 grain weight, grain weight per spike, grain yield per plant, grains per spike and biological yield per plant followed by VL3001 × WH1080 for days to 75% heading, productive tillers per plant, flag leaf area, 1000 grain weight, grain yield per plant, grains per spike and biological yield per plant, C306 × WH1142 for days to 75% heading, spike length, peduncle length, productive tillers/plant, spikelets/spike, grain weight per spike and biological yield per plant. In pooled analysis, C306 × WH1142 was identified as good specific combination for maximum number (nine) of traits viz., spike length, peduncle length, awn length, productive tillers/plant, spikelets/spike, flag leaf area, grain weight per spike, grains per spike and biological yield per plant followed by BECARD/KACHU × WH1080 for plant height, spike length, awn length, grain weight per spike, grains per spike, biological yield per plant and harvest index, cross BOW/VEE/5/ND/VG9144//KAL/BBB/YACO/4/CHIL/6/CASKOR/3/… × WH1080 for days to maturity, spike length, productive tillers per plant, flag leaf area, 1000 grain weight, grains per spike and biological yield per plant.
The results for gca and sca effects were also close conformation with the observation of Mandal and Madhuri (2016) and Ishaq et al., (2018) for days to 75% heading; for days to maturity with the observation of Singh et al., (2013), Pawar et al., (2014), Mandal and Madhuri (2016) and Ishaq et al., (2018); for plant height with the observation of Khan et al., (2007), Saxena and Rawat (2011), Hei et al., (2016) and Parveen et al., (2018); for spike length observation of Srivastava et al., (2012), Barot et al., (2013) Mandal and Madhuri (2016), Kumar et al., (2017) and Farooq et al., (2019); observation of Saxena and Rawat (2011) and Padhar et al. (2013) for peduncle length; Khan et al., (2007), Barot et al., (2013), Mandal and Madhuri (2016), Parveen et al., (2018) and Patel et al., (2018) for productive tillers/plant; for spikelets per spike with the observation of Saxena and Rawat (2011), Abro et al., (2016) and Kumar et al., (2017); for flag leaf area with the observation of Ahmed et al., (2017), Parveen et al., (2018), Sattar et al., (2018) and Farooq et al., (2019); for thousand grain weight Hei et al., (2016); Rani et al., (2017), Rajput and Kandalkar (2018) and Patel et al., (2018); by Çifcil and Yağdı1 (2010) and Ahmed et al., (2017) find similar observations for grains per spike; for grain weight per spike with the experiment of Çifcil and Yağdı1 (2010), Singh et al., (2013) and Kumar et al., (2017); by Khan et al., (2007), Singh et al., (2013), Padhar et al., (2013) and Ishaq et al., (2018) for biological yield per plant; Saxena and Rawat (2011), Barot et al., (2013), Ahmed et al., (2017) Parveen et al., (2018) and Rajput and Kandalkar (2018) for grain yield per plant and for harvest index by Padhar et al., (2013), Barot et al., (2013) and Devi et al., (2013).
The understanding of inheritance for various traits and identification of superior parentes are important pre-requisite for an effective breeding programme. The gca and sca effects are good measure of additive and non-additive gene action, respectively. This information about the nature of gene action in irrigated and water stress condition of the genotypes can help to determine an appropriate breeding strategy for yield improvement in wheat for water limited environment.