Choice of Traits for Seed Yield Improvement based on Association Analysis Studies in Sesame (Sesamum indicum L.)

Sesame (Sesamum indicum L.) is one of the ancient cultivated oilseed crops of the world. Sesamum belongs to the Tubiflorae order and Pedaliaceae family (Shrikanth and Ghodake, 2022). It is also popular as til and gingelly and famously known as “Queen of Oilseeds”. Sesame is diploid (2n=26) and dicotyledonous. The genus Sesame has about 36 species (Esmaeil and Reza, 2019), of which Sesamum indicum L. is the most principal cultivated species. The plant yields rich edible oil due to the presence of powerful antioxidants Sesaminol and Sesamol. Sesame seeds are acknowledged as “the seeds of immortality”. Although the crop originated in Africa, India is considered to be the major Centre of genetic diversity (Maiti et al., 2018).
       
The global production of Sesame seeds was 6.2 million tonnes, led by Tanzania, India and Sudan (FAOSTAT, 2016). More than 6 million tons of sesame seeds have been produced under nearly 11 million ha categorizing sesame at the ninth rank among the major oil crops (FAOSTAT, 2017). The distribution of most of the species takes place in three regions viz., Africa, India and the Far East (Kehie et al., 2020). The composition of sesame concerning 100 g of seed is lipid contents- 48 g, carbohydrates-25.7 g, proteins-17 g, fibre-14 g and ash-6 g approximately. Sesame seeds contain 40 to 63 per cent oil rich in antioxidants and a substantial amount of Poly unsaturated fatty acids (Abate and Mekbib, 2015). Sesame seeds are rich in minerals such as calcium, phosphorous, magnesium and potassium in large quantities and similarly have vitamins such as Niacin, Thiamin, Riboflavin and vitamin B-6 (USDA Nutrient Database, 2015). Besides, it is used in the pharmaceutical and cosmetic industries as well (Kehie et al., 2020). Around 70 per cent of the World’s Sesame seed is processed into oil and meal. Sesame contains bactericidal and insecticidal activities and its antioxidants inhibit the absorption and production of cholesterol in the liver.
       
The understanding of the nature and magnitude of genetic variability is of immense value to the plant breeder for planning an efficient breeding programme to improve the yield potential of crops. Similarly, information on the genetic association of plant traits with seed yield has great significance to breeders in selecting desirable genotypes (Kehie et al., 2020). The phenotypic selection of parents based on their performance alone may not always be a dependable procedure since the phenotypic expression is highly influenced by environmental factors, which are non-heritable. It is thus, essential to select genotypes based on their genetic worth. Accordingly, correlation studies help in the selection of superior genotypes from diverse genetic populations (Jogdhande et al., 2017). In crop breeding, correlation coefficient measures the mutual relationship between various traits and determines the component character on which selection can be relied upon for genetic improvement of yield potential of crop plants (Kumar and Paul, 2016).However, in correlation studies, indirect associations become more complex and puzzling. Path analysis can avoid this complication by measuring the direct influence of one trait on the other as well as partitioning a given correlation coefficient into its components of direct and indirect effects (Manisha et al., 2018; Jogdhande et al., 2017). Path coefficient analysis is an effective means of analyzing direct and indirect causes of association and permits critical examination of specific traits that produce a given correlation. It provides information on the magnitude and direction of direct and indirect effects of the yield components (Chaudhary and Joshi, 2016).
               
However, the lack of information on the character association of yield and its contributing traits is believed to limit the genetic improvement of sesame in India. Hence, the present investigation was focused on gathering adequate information on the genetic association of yield and yield-related traits in sesame accessions collected from various areas of India.

Study area
 
The experimental study was performed at Student’s Farm, Department of Agriculture, Loyola Academy, Secunderabad during Kharif 2021.
 
Experimental material and design
 
The experimental material comprised of 60 genotypes (Table 1) seeded in a randomized full block design (RBD) with three replications. Each genotype was grown in a 5 m long row with 30 × 10 cm spacing. To ensure a successful yield, recommended agronomic practices were followed. Observations for thirteen quantitative characters like plant height (cm), number of branches per plant, seed yield per plant (g), 1000 seed weight (g), number of capsules per axil, capsule length (cm), number of capsules per plant, number of seeds per capsule, seed length (mm), seed width (mm), seed thickness (mm) and oil content were noted. Five randomly chosen plants in each entry of each replication were used for collecting the data.
 

 
Statistical analysis
 
Analysis of variance was carried out for the data using R software; to test for significant differences among the genotypes according to the standard statistical procedure described by Gomez and Gomez (1984).
       
Phenotypic and genotypic correlations between the quantitative traits were estimated using the method described by Miller et al., (1958). The correlation coefficient was analyzed on the tabulated data using META-R Version- 6.01 (Alvarado et al., 2017).
               
The proportion of direct and indirect contributions of various characteristics with seed yield was estimated through path coefficient analysis as suggested and elaborated by Dewey and Lu (1959).

Analysis of variance
 
Table 2 depicts the ANOVA findings for 13 features of 60 sesame genotypes. According to ANOVA, the mean sum of squares for the genotypes was extremely significant (P 0.05 and P 0.01) for all characters. This indicates that the material under research showed no significant changes in replication, proving that environmental error (genotype x environment) was less widespread. The material displayed considerable variation among genotypes.These findings demonstrated that substantial differences exist among genotypes for all parameters studied, which may provide breeders with a good chance to identify high-performing accessions for desired features to enhance crop breeding programmes. Shrikanth and Ghodake, 2022 and Esmaeil and Reza, 2019 discovered that the quantitative features of sesame genotypes differed significantly.
 

 
Correlation coefficient
 
The correlation coefficients between seed yield and yield contributing characters were worked out at the genotypic and phenotypic levels (Table 3).
 

 
Direct correlation
 
Positive significant correlation at genotypic and phenotypic levels was observed for 1000 seed weight (0.222, 0.170) with grain yield. A strong positive correlation between desirable traits like these is favourable to the plant breeder because it helps in the concurrent improvement of both the characters. Comparable findings were reported by Takele et al., (2021), Esmaeil and Reza (2019) and Shrikanth and Ghodake, (2022). A significant negative correlation was observed for days to 50% flowering (-0.262, -0.192) and seed yield. Similar findings were stated by Abate et al., (2018), Teklu et al., (2017) and Shrikanth and Ghodake, (2022).
 
Indirect correlation
 
Days to 50% flowering showed a positive and high significant correlation (genotypic and phenotypic) with capsules per plant (0.148, 0.137), while it was negatively associated with seed length (-0.203, -0.145) and seed width (-0.224, -0.205). Plant height exhibited positive and high significant correlation (genotypic and phenotypic) with number of branches (0.487, 0.446), capsule length (0.327, 0.282) and capsules per plant (0.420, 0.395). The number of branches executed a positive and high significant correlation (genotypic and phenotypic) with the number of capsules per axil (0.318, 0.301) and capsules per plant (0.702, 0.680) while negatively associated with seed thickness (-0.150, -0.129). 1000 seed weight exhibited positive and high significant correlation (genotypic and phenotypic) with number of capsules per axil (0.197, 0.165), capsule length (0.254, 0.201), seed length (0.418, 0.195), seed thickness (0.185, 0.111) and oil content (0.239, 0.216). The number of capsules per axil presented a positive and high significant correlation (genotypic and phenotypic) with capsules per plant (0.393, 0.389), but exhibited a negative association with seed width (-0.312, -0.273) and seed thickness (-0.169, -0.153). Capsule length presented positive and high significant correlation (genotypic and phenotypic) with number of seeds per capsule (0.434, 0.399), seed length (0.275, 0.197), seed width (0.299, 0.229) and seed thickness (0.221, 0.208). Capsules per plant displayed negative and high significant correlation (genotypic and phenotypic) with seed length (-0.187,-0.115), seed width (-0.282, -0.238) and seed thickness (-0.288, -0.267). The number of seeds per capsule displayed a positive and high significant correlation (genotypic and phenotypic) with seed length (0.183, 0.144), seed width (0.220, 0.195) and seed thickness (0.247, 0.199). Seed length exhibited a positive and high significant correlation (genotypic and phenotypic) with seed width (0.850, 0.577) and seed thickness (0.418, 0.308). Seed width presented a positive and high significant correlation (genotypic and phenotypic) with seed thickness (0.738, 0.605). Seed thickness displayed a positive and high significant correlation (genotypic and phenotypic) with oil content (0.212, 0.185). Shrikanth and Ghodake, (2022) reported a high, positive and significant association of days to 50% flowering, plant height, 1000 seed weight, number of capsules per axil, capsule length, capsule per plant, number of seeds per capsule, seed length and oil content. Esmaeil and Reza, (2019) and Kumar and Vivekanandan, (2019) also reported the same for plant height, 1000 seed weight, capsules per plant, seed length, seed width and seed thickness. Similar findings were reported by Teklu et al., 2017 for days to 50% flowering, plant height, number of branches, capsule length, capsules per plant, number of seeds per capsule and oil content.
       
If correlation bears a negative sign, it means that increase in the value of one character will lead to a decrease in the value of the second character and vice versa. Similarly, if correlation bears a positive sign, it means that increase in the value of one variable will lead to an increase in the second character. The magnitude of all genotypic correlations is higher than that of phenotypic correlations except for the number of branches and number of capsules per axil. It means that there is a strong association between these two characters genetically, but the phenotypic value is diminished by a significant interaction with the environment. Similar findings were reported by Abate et al., (2018).
 
Path coefficient
 
Path coefficient analysis splits the correlation coefficient into direct and indirect effects.
 
Genotypic path coefficient
 
In Table 4, the highest positive direct effects were exhibited by seed width (2.459) on the dependent character. High positive direct effects were expressed by 1000 seed weight (0.866) followed by the number of capsules per axil (0.660) on the dependent character. Moderate positive direct effects were exhibited by oil content (0.232) followed by the number of branches (0.226) on the dependent character. Low positive direct effects were expressed by days to 50% flowering (0.167) followed by the number of seeds per capsule on dependent traits. Shrikanth and Ghodake (2022) and Takele et al., (2021) reported high positive direct effects on grain yield. The highest negative direct effects were observed for seed length (-1.925) followed by seed thickness (-1.236) on dependent characters. High negative direct effects were observed for capsules per plant (-0.442) on dependent character. Low negative direct effects were observed for plant height (-0.122) followed by capsule length (-0.119) on dependent character. Similar findings have been reported by Teklu et al., (2017) and Kumar and Vivekanandan, (2019) for seed length, seed thickness, capsules per plant, plant height and capsule length.
 

 
Phenotypic path coefficient
 
In Table 5, low positive direct effects were identified for the number of branches (0.191), 1000 seed weight (0.174) and seed length (0.139) on the dependent character. Moderate negative direct effects were detected for days to 50% flowering (-0.222) and seed width (-0.265) on the dependent character. Similar findings were reported by Shrikanth and Ghodake, (2022); Esmaeil and Reza, (2019) and Teklu et al., (2017) for 1000 seed weight, capsule length, seed length, seed width and seed thickness.
 

               
The residual effect R=0.409 (P) and 0.343 (G) indicates that the component characters under study were responsible for about 96% and 97% of variability in seed yield per plant.

The mean sum of square (MSS) due to genotypes was highly significant (P<0.05 and P<0.01) for all the characters under study, indicating that there is a substantial amount of variability in the genotypes, according to the analysis of variance. A positive significant correlation at the genotypic level has been observed for 1000 seed weight (0.222) with grain yield. This type of positive correlation between desirable traits is constructive to the plant breeder as it helps in the concurrent improvement of both the characters. The highest positive direct effects were noted for 1000 seed weight (0.866) followed by the number of capsules per axil (0.660) on grain yield. Low positive direct effects were observed for the number of branches (0.191), 1000 seed weight (0.174) and seed length (0.139) on the dependent character. Moderate negative direct effects were detected for days to 50% flowering (-0.222) and seed width (-0.265) on the dependent character.

The authors are grateful to NBPGR, New Delhi and IIOR, Hyderabad for supplying the germplasm material.