Combining ability and Heterosis analysis for grain yield and yield associated traits in Pea (Pisum sativum L.) 

Pea (Pisum sativum L.) is an important legume grown as a garden and field crop throughout the temperate region of the world; it is also grown as a cool season crop. It ranks third in production among the grain legumes after soybeans and beans in the world. Pea is valued primarily for the nutritional quality having high protein as 20-30%, vitamin A -5%, beta carotene 4% and  sufficient carbohydrates. The growth and development of peas are determined by the interaction of genetic factors, the environment and agricultural practices (Acikgoz et al., 2009).
       
Breeding for superior varieties requires selection of parents capable of transmitting their desirable qualities. A rational approach for breeding is to select parents based on their combining ability rather than visual observations of their traits. The concept of combining ability analysis has significant practical importance in plant breeding (Tyagi and Srivastava, 2001) The general combining ability reveals the presence of additive type of gene action, whereas specific combining ability reveals the presence of non-additive type of gene action. India is the second largest producer of vegetables, next only to China, in the world with a production of 40 million tonnes from four million hectares of land area. In spite of that, this seemingly high level of production can provide only 208 grams of vegetables per capita (Sharma, 2003), as against the suggested dietary intake of 275g and 250 g per capita per day for adult male and female, respectively for undertaking moderate work (Swaminathan, 2002).

Fifteen genotypes of pea including Pear polo, KSP-110,L-116, DS-10, GS-10, G-10, PB-89, Magic-100, AP-3, VRPE-24, Jyoti, PB-01, Rachana, Sonali-10 and Shilpa-10, out of these Pear polo, KSP-110, L-116, DS-10, GS-10, G-10, PB-89, Magic-100, AP-3, VRPE-24, Jyoti, PB-01 as a line and  Rachana, Sonali-10, Shilpa-10 as a tester were carried out at the Experimental Research Farm of School of Agriculture, Lovely Professional University, Phagwara, Punjab during rabi season (2016). The experiment was conducted in line x tester mating design. Seeds of F₁s hybrids and parental lines were sown in a randomized block design with three replications on 12 November 2016.  Each plot consisted of 36 F₁ and 15 parent plants on three 1.5 m apart and 50 cm rows. Plants spacing were 10 cm. The soil of experimental site is sandy loam. Geographically, Punjab is situated at 31°.15 N latitude and 75°.42 E longitudes in North Gangetic plain in eastern part of Punjab. The temperature ranges between 22 and 46°C in the summer. Winters in Punjab experience very large diurnal variations with warm days and downright cold nights. The average annual rainfall is 1100 mm. Quantitative data were collected on five plants in each plot. Days to 75% flowering, plant height, number of primary branches, pods/plant, pod length (cm), pod density, number of seeds per pod, days to maturity, 100-seed and yield/plant were recorded. Regarding the statistical analysis data recorded on parents and the F₁ hybrids were analyzed together as suggested by Singh and Chaudhary (1979). The combining ability analysis was done by Kempthorne (1957). Heterosis over mid parent or the better parent of the observed characters was calculated according to Kempthorne (1957).

The analysis of variance for all entries including parents and hybrids for 10 characters revealed that treatment variances for all the 10 characters were highly significant (Table 1). This indicates the presence of variability among treatments. Variability is the most important characteristic feature of any population. Estimation of genetic variability is an important prerequisite for realizing response to selection as the progress in breeding depend upon its amount, nature and magnitude. The breeder should have the capability of distinguishing the genetic and non-genetic components of variation occurring in a population. In the present investigation, out of 10 traits studied all traits viz., days to 75% flowering, number of primary branches per plant, plant height, number of pod per plant, pod length, seed per pod, pod density, days to maturity, 100 seed weight and seed yield per plant showed high magnitude of variability in terms of hybrids and seven traits viz., plant height, number of pod per plant, seed per pod, pod density, days to maturity, 100 seed weight and seed yield per plant showed high magnitude of variability in terms of parents vs hybrids. Similar findings were reported by Ceyhan et al., (2008) in pea, Kanagarasu et al., (2013) in maize, Motamedi et al., (2014), and Ahmed et al., (2015) in corn and pea. Parents Vs F₁ hybrid performance is probably the most basic comparison in quantitative inheritance and the degree of hybrid vigor provides the preliminary idea about the probable gene action involved in determining the particular character and it is the  simplest and the easiest measure of genetic diversity of these parents. Heterosis is the superiority of F₁ over its parents. Traditionally the deviation of F₁  performance from mean of the parents is considered as the estimate of heterosis. However, only the superiority of hybrid over the superior parent is of practical value. Therefore only the heterosis over superior parent is discussed. The heterosis for the most of the characters was observed in both the directions (i.e. positive and negative). The magnitude of overall heterosis ranged from 26.86 per cent for pod per plant to -56.57 percent for yield per plant. Positive heterosis was observed for the maximum crosses under the characters number of pods per plant.  The maximum positive heterosis for the character number of pod per plant was recorded for the cross DS-10 x Shilpa-10 (26.86). In the previous studies, Mehmet et al., (2008) reported high estimates of heterosis for seed yield and moderate estimate of heterosis for number of pod per plant in pea which was contradictory with the present findings. Negative heterosis is desirable in pea for days to 75% flowering, plant height and days to maturity as these would help in the development of early maturing cultivars (Table 4). Negative heterosis was observed for all the 36 cross combinations for the character days to maturity.  The value of heterobeltiosis heterosis ranged from -0.67 to -17.46% with the relative heterosis of -9.98%. The maximum heterosis was recorded for the cross DS-10 x Rachana, (-17.46) (Table 4). Likewise, similar finding had also been reported by Ceyhan et al., (2003) in pea and Motamedi et al., (2014). The value of negative heterosis for days to 75% flowering ranged from -34.02% to -21.95%, with the heterobeltiosis and relative heterosis (Table 4). The maximum heterobeltiosis was recorded for the hybrid Vrpe24 x Sonali-10 (-34.02%). Similar result was also reported by Alam et al., (2008).
 

 



       
Combining ability studies are useful not only in analyzing the genetic architecture of the character under study but also in ranking the parental lines on the basis of their performances in the crosses. The information thus obtained helps in designing suitable breeding procedure for genetic amelioration of the crop and selection of suitable parents which when crossed will give rise to more desirable segregates. Griffings (1956) presented a model to show that variance for gca involves additive and additive x additive gene actions. Variance for sca on the other hand, depends on the dominance and epistasis components of genetic variation. In autogamous crops, gca component is very important since only this component could be fixed in further selection. Normally sca could not contribute tangibly in improvement of self-pollinated crop except where commercial exploitation of heterosis is feasible. However, if the crosses showing high sca effects involved either the parents or at least one parent possessing high gca effects. They could be exploited for practical breeding. Analysis of variance for combining ability revealed that presence of both additive and non-additive gene actions for the characters studied, which was indicated by the significance of both the GCA and SCA variances. For days to 75% flowering, plant height, number of pod per plant, pod length, seed per pod, pod density, 100 seed weight and seed yield per plant the SCA variance were highly significant and also ratio of GCA/SCA was less than unity indicating the involvement of non-additive gene action (Table 2). The variances due to the SCA were higher in magnitude than GCA for all the traits. These results are encouraging from the view of heterosis exploitation. In the previous studies Singh et al., (2012) and  Motamedi et al., (2014) revealed that estimates of SCA variances were higher than GCA variances for all the traits studied, thus indicating predominance of non-additive gene action for these traits. The combining ability analysis revealed significant variances for GCA and SCA for the traits plant height and 100 seed weight. Similar, justification was also reported by Enrique Luis Cointry et al., (2013) in which significant variances for GCA and SCA for plant height was obtained. The GCA and SCA were recorded for most of the traits viz, seed yield, plant height, pod length, number of pod per plant and pod density to be significant (Table 3). These findings of the present investigations was also supported by Ercan ceyhan et al., (2004). The study revealed significant difference in GCA and SCA effects for the traits viz., days to 75% flowering, number of primary branches per plant, pod per plant, days to maturity and seed yield per plant. These findings were supported by the previous work of Singh et al., (2001) and Dixit et al., (2003). Four parental lines viz. PB-01, Jyoti, PB-89 and DS-10 had good combining ability or were good combiners for seed yield as they expressed highly significant and positive gca effect at 1% level. These results were in agreement with the findings of Uddin et al., (2006) and Espinosa and Ligarreto (2005) who revealed significant differences for GCA and SCA indicating the presence of additive as well as non-additive gene effects for controlling the traits. The five lines, pear polo, DS-10, G-10, AP-3 and VRPE-24 are the good combiners for the trait plant height in pea. In case of pod length two lines, Jyoti and PB-01 are the good combiners.
 

 

The present investigation has provided some useful information regarding magnitude of heterosis, gene action and combining ability effects as well as performance of parents and F₁s for seed yield per plant and its related components. The parent Jyoti was found to be the good general combiner for number of primary branches per plant, plant height, number of pod per plant, pod length, seed per pod, days to maturity and seed yield per plant. The parental line PB-89 was found to manifest the best combining ability for the traits like number of primary branches, number of pod per plant, pod length pod density, days to maturity and seed yield per plant. Whereas, the parent VRPE-24 reported good general combining ability for days to 75% flowering, primary branches per plant, plant height, pod length, number of seed per pod.  The hybrid AP-3 X Sonali-10 was found to be the good specific combiner for plant height, pod length, pod density and 100 seed weight. The hybrid VRpe-24 X Sonali-10 was found to have the good specific combining ability for the traits like pod length, pod density, 100 seed weight.