Genetic Analysis of Pigeonpea [Cajanus cajan (L.) Millsp.] Hybrids for Yield and Yield Attributes

Pigeonpea [Cajanus cajan (L.) Millsp.] commonly known as ‘Redgram’ or ‘tur’ or ‘arhar’ is a legume crop with diploid chromosome number, 2n=22, belongs to the tribe phaseoleae, sub tribe cajaninae of the sub family papilionoidae, family fabaceae. In India pigeonpea is the second most important pulse crop after chickpea and is cultivated in an area of 5.33 m ha with a production of 4.87 Mt and an average productivity of 913 kg/ha. (Indiastat, 2016-2017). It is consumed in the form of split pulse (dal) and seeds are rich sources of protein (22.3%), fat (1.7%), carbohydrate (57.6%), iron, iodine and other essential aminoacids such as lycine, cystine and arginine. A number of varieties have been released and the area also increased but the yield has remained unacceptably low because of several biotic and abiotic constraints. Hence, hybrid breeding is expected to break the yield barrier and increase the yield. Genetic variability is of greatest interest to the plant breeder as it plays a vital role in framing successful breeding programme.  Heritability of a metric character is a parameter of particular significance to the breeder as it measures the degree of resemblance between the parents and the off-springs and its magnitude indicates the heritability with which a hybrid can be identified by its phenotypic expression. While genetic advance aids in exercising the necessary selection pressure. Correlation enables to identify the characters or combination of characters which might be useful as indicator of high yield by way of evaluating relative influence of various characters on yield and among themselves as well. The original concept of correlation was given by Galton (1889) and later elaborated by Fisher (1918). Genetic diversity is the basis for any crop improvement programme. Mahalanobis D² analysis is a useful tool to assess the genetic divergence among population.

The experimental material comprised of forty-eight hybrids and two checks evaluated in a randomized block design with two replications at Agricultural Research Station, Tandur, PJTSAU, Telangana, India during Kharif 2018.  Each entry was sown in four rows of 4m length with a row to row and a plant to plant distance of 120 cm and 20 cm respectively. The recommended agronomic and plant protection measures were followed in order to raise a normal crop. Observations on nine different quantitative characters viz., days to 50% flowering, days to maturity, plant height, number of branches per plant, number of pods per plant, number of seeds per pod, pod length, test weight and seed yield were recorded on five randomly selected competitive plants except for days to 50% flowering and days to maturity where data was recorded on plot basis. Mean values from the five randomly selected plants from each of the two replications were averaged and expressed as the mean of the respective character and considered for statistical analysis using Windostat version 9.2. Phenotypic and genotypic coefficients of variation (PCV and GCV) were computed according to Burton (1952). Heritability (h²) in the broad sense was calculated according to the formula given by Allard (1960) for all characters. From the heritability estimates, the genetic advance was estimated by the following formula given by Burton (1952). Correlation coefficients were calculated at genotypic and phenotypic level using the formulae suggested by Falconer (1964). A measure of group distance based on multiple characters was given by Mahalanobis (1936) using D² statistic. The data collected on different characters was analyzed using Mahalanobis’ D² analysis to determine the genetic divergence among the hybrids.

Analysis of variance for the experiment involving a set of 50 entries for all the characters revealed highly significant differences among the hybrids for all the characters indicating the presence of sufficient variability. The results are presented in Table 1. The results pertaining to genetic parameters viz., phenotypic coefficient of variation (PCV), genotypic coefficient of variation (GCV), broad sense heritability (h²_bs) and genetic advance as percent of mean (GAM) for all the nine characters are furnished in Table 2 indicated that there is almost perfect relation between PCV and GCV of each character. This could be seen from the highest magnitude of both PCV (31.56%) and GCV (30.56%) for seed yield and number of pods per plant (23.23-20.89%) suggesting that these characters were under the influence of genetic control. So the characters can be relied upon and simple selection can be practiced for further improvement. Similar results of high PCV and GCV for seed yield were earlier reported by Pandey et al., (2015) and Satapathy et al., (2019) and for number of pods per plant by Pushpavalli et al., (2018). Number of branches per plant, number of seeds per pod, pod length and test weight exhibited moderate magnitudes of both PCV and GCV, respectively. The phenotypic coefficient of variation was higher than genotypic coefficient of variation (GCV) for all the characters under study but the GCV was greater than the variation produced by the environment for all the characters. It signifies the role played by the environment in the expression of these characters. The results are in agreement with findings of Bhadru et al., (2011), Yerimani et al., (2013) and Satapathy et al., (2019). All the characters recorded high estimates of heritability. Broad sense heritability estimates ranged from 61.80 per cent (Number of seeds per pod) to 99.43 per cent (Yield). All the characters recorded high estimates of heritability indicating that they were least influenced by the environmental effects, however selection for improvement of such characters may not be useful, because broad sense heritability is based on total genetic variance which includes additive, dominant and epistatic variances. Thus, the heritability values along with estimates of genetic advance would be more reliable than heritability alone (Johnson et al., 1967). High heritability coupled with high genetic advance was observed for number of branches per plant, number of pods per plant, number of seeds per pod, pod length, test weight and yield. Thus, these traits are predominantly under the control of additive gene action and hence these characters can be improved by pedigree method of breeding. Similar results for high heritability coupled with high genetic advance for various traits have also been reported earlier for number of pods per plant, test weight and seed yield (Magar, 2003, Ram et al., 2016; Pushpavalli et al., 2017; Reddy et al., 2019).
 

 

        
Character association studies (Table 3) revealed that number of pods per plant (0.9282 G, 0.8174P) followed by number of branches per plant (0.6008 G, 0.4792 P), number of seeds per pod (0.5954 G, 0.4651 P), pod length (0.4027 G, 0.2960 P) and plant height (0.3149 G, 0.2736 P) manifested significant and positive correlation with yield at both phenotypic and genotypic levels. (G and P indicates genotypic and phenotypic correlations, respectively).
 

        
Thus, it can be inferred that selection based on any one of these traits either alone or in combination, will result in identifying high yielding strains. Highly significant and positive correlation between grain yield and other traits as obtained in the present investigation have also reported for number of branches per plant and number of pods per plant (Kumar et al., 2014); for plant height, pod length, number of seeds per pod, number of branches per plant and number of pods per plant (Baldaniya et al., 2018); Significant positive correlation of number of pods per plant and test weight with seed yield were reported by Devi et al., (2012), Mittal et al., (2010), Pandey et al., (2015), Sharma et al., (2012) and Pushpavalli et al., (2018).
        
Mahalanobis’s Euclidean squared distances grouped the material into nine clusters. The cluster composition is given in the Table 4. The clustering pattern was depicted by ward’s minimum variance dendrogram (Fig 1) Among the clusters, cluster II was the largest comprising of 14 hybrids followed by cluster I with 13 hybrids, cluster VI with 9 hybrids, cluster IV with 6 hybrids, cluster IX with 4 hybrids and the remaining clusters are solitary (III, V, VII and VIII). The average intra and inter cluster D² values are presented in Table 5. Most of the intra clusters were closely related and clustervalues ranged from 0.00 (cluster III, V, VII and VIII) to 6.73 (cluster IX). From the inter cluster D² values, it could be observed that the highest divergence was between cluster IV and cluster VIII (12.67) whereas the lowest divergence was noticed between cluster VII and cluster VIII (4.16). Greater the distance between the two clusters, wider the genetic diversity among the entries of these clusters. Cluster mean values (Table 6) for yield contributing characters indicated that considerable differences existed for all the traits. Highest cluster mean values were noticed for Days to 50% flowering in cluster III (129.50), days to maturity in cluster III (174.50), plant height in cluster VIII (199.50), number of branches per plant in cluster VII (21.50), number of pods per plant in cluster VI (234.44), number of seeds per pod in cluster VIII (4.00), pod length in cluster IX (5.13), test weight in cluster IX (14.13) and seed yield in cluster VIII (2425.50). The data on inter cluster distances and per se performance of hybrids were used to select genetically diverse and agronomically superior entries.
 

 

 

 

       
The characters contributing to greater divergence should be given more importance for effective selection. The contribution of different characters towards the genetic diversity is presented in Table 7. Seed yield (ranked first 413 times out of 1225 total numbers of combinations) contributed 33.71 per cent to the divergence of hybrids. This was followed by days to 50% flowering (30.53%) by 374 times, test weight (12.16%) by 149 times, plant height (9.22%) by 113 times, pod length (5.14%) by 63 times, number of seeds per pod (1.55%) by 19 times and days to maturity (0.65%) by 8 times contributed less towards divergence. The characters contributing maximum to the divergence were given greater emphasis for further selection in pigeonpea improvement. The results revealed that seed yield and days to 50% flowering have contributed more towards divergence, so the direct selection for these traits would be helpful as evident from the number of times these traits appeared or ranked first during contribution to diversity.
 

The characters viz., number of branches per plant, number of pods per plant, number of seeds per pod, pod length, test weight and seed yield exhibited high heritability estimates coupled with high genetic advance which suggested that these traits were amenable for further improvement by following simple selection methods. Character association revealed significant positive association of single plant yield with plant height number of branches per plant, number of pods per plant, number of seeds per pod and pod length. Based on cluster mean values the hybrids of cluster VIII exhibited highest per se performance for seed yield hence these are identified as potential parents for developing high yielding hybrids.

The authors are very much thankful to Agricultural Research Station, Tandur, Professor Jayashankar Telangana State Agricultural University, Hyderabad for providing the facilities, funds and the material for carrying out the investigation.