1 Introduction

Chickpea is one of the most important pulse crops of India, playing a crucial role in agricultural production by enriching the soil through biological nitrogen fixation. Chickpea has special significance predominantly in the vegetarian diet as it plays an important role in human nutrition as a rich source of protein for large population sectors in the developing world and is considered a healthy food in many developed countries. Hybridization followed by selection to isolate desirable recombinant from the superior crosses is one of the important breeding methods for breaking yield barriers and combining different traits in one variety. The ability of parents to combine well depends upon complex interaction among genes, which cannot be predicted from yield performance and adaptability of the parents (Allard, 1960). The information on nature and magnitude of genetic components of variation for yield and its component traits for a crop is essential to adopt an effective breeding programme for its improvement. Genetic studies on this legume lag far behind its economic importance. Although, crop improvement efforts have improved the adaptation of chickpea to warmer conditions in the subtropics and many varieties have been developed over years, but no substantial gain is realized in the enhancement of genetic yield potential of the crop and production and productivity remained almost static over years. So a comprehensive study of genetic architecture of quantitative characters of economic value is essential to improve the yield potential of this crop. On the basis of meager information available on gene effects controlling the yield and its component traits in chickpea, the present study was undertaken to understand the nature and magnitude of genetic components in divergent parents and the extent of heterosis for seed yield and its component characters. Thus, main objective of this investigation was to estimate the extent of heterosis for seed yield and its component characters and to isolate better crosses.

2 Materials and Methods

The five generations of six crosses were developed by using seven divergent genotypes of chickpea (Cicer arietinum L) viz., JG16 x ICC 96029, JG74 x ICC 96029, JG130 x ICC 96029, JGG 1 x ICC 96029, Narsinghpur Bold x ICC 96029, JG 130 x JG 322. F₁^, F₂ and F₃ generations were grown along with their parents.  The experiment was laid out at Seed Breeding Farm, Department of Plant Breeding & Genetics, JNKVV, and Jabalpur during 2012-13 in randomized complete block design with 3 replications. Randomization was carried out in each replication among five populations viz, P_1, P_2, F_1, F₂ and F₃ of six crosses. Row length was 4 m and row to row distance 30 cm and plant to plant spacing 10 cm. Standard agronomic practices and plant protection methods were recorded as and when required. The observations were recorded on quantitative traits on randomly selected competitive plants from each of the parents, F₁, F₂ and F₃ populations per replication. As the F₃ population was included in the experiment, the data were first subjected to D scaling test of Mather (1949) and Hayman and Mather (1955) to detect the presence of epistasis, then the five parameter model of Hayman (1958) was fitted to the data to obtain estimates of (m), (d), (h), (i), (l) heterosis and inbreeding depression were worked out simultaneously.

3 Results and Discussion

The estimates of simple scaling test (Table 1) were found significant for plant height, biological yield, harvest index and seed yield per plant in JG 74 x ICC 96029; significant value for biological yield and seed yield per plant was recorded in JG 130 x ICC 96029. In JG130 x JG322 significant value was obtained for primary branches, number of seeds per plant, biological yield and seed yield/plant. It indicates the preponderance of non - allelic interaction in the expression of these characters.

In chickpea only additive gene effects and additive x additive type of interaction effects are more useful for exploration in breeding programme through appropriate selection schemes. The information on mean degree of dominance will also give an idea about the possibility of improvement of a particular character through selection in early or later generation. Results of present investigation revealed that both additive (d) and additive x additive type (i) of gene action has greater contribution in the expression of majority of characters in different crossesbut there estimates were found negative and low in magnitude, as compared to the dominance (h) and dominance x dominance (l) interaction in almost all the cases.

Significant negative estimates of both additive (d) and additive x additive type (i) of gene action was found for days to 50% flowering, days to pod initiation, days to maturity, plant height and number of primary branches in the crosses JG 16 x ICC 96029, JG 74 x ICC 96029, JG 130 x ICC 96029, JGG 1 x ICC 96029 and Narsinghpur Bold x ICC 96029 (Table 2; Table 3; Table 4; Table 5; Table 6; and Table 7). The cross JG130 X JG 322 had predominance of negative additive effects only for plant height, number of primary branches and number of seeds per plant. Similar results were drawn by Gupta (2005) and suggested that top consideration should be given to these traits. This type of gene action is fixable; hence selection would be effective for these traits. Bulk / pedigree method of selection can be effective means of improvement in this crop.

Significant positive dominance effect (h) was observed for days to 50% flowering, days to maturity, effective pods per plant, number of seeds per plant and biological yield per plant in JG 16 x ICC 96029. Significant negative estimate was recorded for plant height and positive estimates were found for effective pods per plant and number of seeds per plant in JG 74 x ICC 96029. JG 130 x ICC 96029 and JGG1 x ICC 96029 exhibited significant positive dominance effect for number of seeds per plant, biological yield per plant and seed yield per plant. In JG 130 x JG 322, it was significant positive for number of primary branches and biological yield per plant, whereas it was negative in case of days to pod initiation. These findings are in confirmation with those of Girase and Deshmukh (2000) and Gupta (2005).In most of the crosses, non-fixable gene effects were higher in magnitude than the fixable gene actions indicating role of non-additive gene effects. Recurrent selection viz. diallel selective mating or bi-parental mating in early segregating generation might prove to be useful approach for tangible improvement of chickpea.

Dominance X dominance (l) gene effects also played significant positive role in JG 16 x ICC 96029 for almost all the characters except days to maturity, pod filling percentage and harvest index. It was found significant negative for plant height and number of primary branches in JG 74 X ICC 96029 and for plant height and seed yield per plant in cross JG 130 x ICC 96029. Positive (l) gene effects were also noted for days to 50% flowering in JG 130 x ICC 96029, for days to pod initiation and number of primary branches in cross JGG 1 x ICC 96029, for number of seeds per plant in Narsinghpur Bold x ICC 96029 and for days to 50% flowering, days to pod initiation and pod filling percentage in cross JG130 x JG 322. This type of epistatic interaction is non- fixable, so heterosis breeding is effective. Similar results were also reported by Pankaj et al. (2002), Katiyar (2003) and Gupta (2005). Since, characters showing dominance or dominance x dominance gene interaction are not fixable that is why selection will not be much effective for the traits showing non-additive gene action. Therefore, breeding method such as recurrent selection or bi-parental mating may hasten the role of genetic improvement for these traits.

In the crosses JG 16 x ICC 96029, JG 74 x ICC 96029 and JGG 1 x ICC 96029 both duplicate and complementary type of gene action played equally important role in the expression of characters. Majority of characters in crosses JG 130 x ICC 96029, Narsinghpur Bold x ICC 96029 and JG 130 x JG 322 exhibited duplicate epistatic interaction. The superiority of these crosses may be due to duplicate gene interaction, which can be exploited in the subsequent generation. In such crosses non additive gene effects played a predominant role in association with additive component and the biparental breeding method can be used to exploit both the components.

Over dominance played an important role in the expression of characters in crosses JG16 x ICC 96029, JG 130 x ICC 96029 and JG 130 x JG 322. The crosses JG 74 x ICC 96029 and JGG 1 x ICC 96029 exhibited equal magnitude of partial dominance and over dominance whereas, major role of partial dominance was found in Narsinghpur Bold x ICC 96029. Hence, it could be concluded from this study that the genetic improvement of seed yield in these populations may be possible by adopting reciprocal recurrent selection followed by isolation of high yielding genotypes in later generations.

Exploitation of heterosis is one of the important breeding options for breaking yield barriers. Moreover, the estimates of heterosis and inbreeding depression together provide information about the type of gene action involved in the expression of various quantitative characters. If there is a high heterosis followed by inbreeding depression, it indicates the presence of non-additive gene action. Whereas,  if the performance is same in F₁ and F₂ generations, it reveals the presence of additive gene action. The yield co ntributing traits revealing high additive gene action are likely to show high heritability and can be improved through simple selection methods of breeding. The traits governed by dominance gene action are likely to be less heritable and may be best utilized through the use as F₁ hybrids, in case of heterosis breeding. Estimation of heterosis and inbreeding depression also give a preliminary idea about the genetic control. The crosses with positive heterosis and negative inbreeding depression will definitely be useful for breeders in the development of high yielding lines.

In cross JG 16 x ICC 96029 significant positive heterosis (Table 8) was observed for days to 50% flowering, days to pod initiation, days to maturity, primary branches, effective pods per plant, number of seeds per plant and seed yield per plant. The cross JG 74 x ICC 96029 and  JG 130 x ICC 96029 exhibited significant positive heterosis for days to 50% flowering, days to pod initiation and days to maturity, whereas in Narsinghpur Bold x ICC 96029 it was noted for days to pod initiation, days to maturity and number of seeds per plant. Significant negative heterosis was recorded for plant height and number of primary branches in cross JG74 X ICC 96029 and for plant height in cross JG 130 X JG 322. This study confirms the earlier findings for heterosis over better parent Sarode et al. (2000), Hegde et al. (2002) and Gupta et al. (2003). Negative heterosis for phenological characters viz, days to 50% flowering, days to pod initiation, days to maturity and plant height while positive heterosis for other important traits would help to screen cross combinations for high yield.

Inbreeding depression values varied considerably for different characters for all six crosses with respect to sign and magnitude. None of the crosses had very high magnitude of significant heterosis coupled with negative inbreeding depression. JG 74 x ICC 96029 had very high magnitude of significant negative heterosis coupled with negative inbreeding depression for plant height and number of primary branches. While, high magnitude of significant positive heterosis coupled with positive inbreeding depression was observed for days to 50% flowering, days to pod initiation, number of primary branches, effective pods per plant, number of seeds per plant and seed yield per plant in cross JG 16 x ICC 96029, for days to pod initiation and effective pods per plant in cross JGG 1 x ICC 96029 and for number of seeds per plant in cross Narsinghpur Bold x ICC 96029.

The significant positive heterosis value for seed yield per plant was recorded in JG 16 x ICC 96029, it is mainly due to heterosis through major yield components like days to 50% flowering, days to pod initiation, days to maturity, primary branches, effective pods per plant and number of seeds per plant. The improvement of seed yield in this cross may be possible by practicing selection for these characters by growing large F₂ populations. Very low magnitude of heterosis estimates were observed for yield and its components. Hence, it may not be possible to fix desirable genes in early generation of selection. So selection in the late generations may be advantageous for the improvement of seed yield.

References

Allard R.W., 1960, Principles of Plant Breeding, John Wiley and Sons. Inc., New York and London

Girase V.S., and Deshmukh R.B., 2000, Gene action for yield and its components in chickpea, Indian J Genet., 60(2): 185-189

Gupta S.K., Singh S., and Kaur A., 2003, Heterosis for seed yield and its component traits in desi x desi and desi x kabuli crosses of chickpea (Cicer arietinum L.), Crop Improvement, 30: 203-207

Hayman B.I., 1958, The separation of epistatic from additive and dominance variation in generation means, Heredity, 12:371–390

https://doi.org/10.1038/hdy.1958.36

Hayman B.I., and Mather K., 1955, The description of genie interaction in continuous variation, Biometrics, 11, 69–82

https://doi.org/10.2307/3001481

Hedge V.S., Yadav S.S., and Kumar J., 2002, Heterosis in short duration chickpea (Cicer arietinum L.), Crop improvement, 29(1): 94-99

Katiyar M., 2003, Genetic analysis of yield and its component traits in Kabuli chickpea, Indian J. Pulse Res., 16: 92-94

Mather K., 1949, Biometrical Genetics, Dovers publications Inc., New York, pp.382

Sarode N.D., Deshmukh R.B., Patil J.V., Manjare M.R., and Mhase L.B., 2000, Heterosis and combining ability in chickpea (Cicer arietinum L.), Legume Research, 23:3, 206-209

Singh S.P., Singh K.P., Singh B., and Singh G., 2002, Studies of heteobeltosis and inbreeding in chickpea, Indian J. Pulses Res., 15: 90-92