Correlation and path analysis studies in F3 generation of cowpea [Vigna unguiculata (L.) Walp.]

Introduction

Cowpea (Vigna unguiculata (L.) Walp.) is an important pulse crop grown widely in India and other tropical regions for its nutritional value, adaptability, and role in sustainable agriculture. It is a rich source of protein, minerals, and dietary fiber making it an essential component in the diets of rural and urban populations. In addition to its nutritional benefits, cowpea improves soil fertility through nitrogen fixation and is well suited for cultivation in drought-prone and low-input environments. In regions like the Konkan, cowpea is commonly cultivated during the rabi season in rice fallows, offering a profitable option to farmers. However, the genetic improvement of cowpea has been relatively limited due to the use of narrow genetic base and lack of systematic selection for complex traits such as seed yield. Seed yield is influenced by several interrelated plant characters, which makes its direct improvement a challenging task. To improve seed yield effectively, it is important to understand the relationships between yield and its contributing traits. Correlation analysis helps in identifying the strength and direction of association among traits, while path coefficient analysis provides a clearer picture by separating these associations into direct and indirect effects. This approach helps in identifying the most important traits that directly contribute to yield and can be used for effective selection in early segregating generations. The present investigation was undertaken in the F 3 generation of cowpea to assess the association between seed yield and related quantitative traits at both genotypic and phenotypic levels and to determine the direct and indirect contributions of these traits through path coefficient analysis. The study aims to identify key characters that can be used as selection criteria for the development of high-yielding cowpea genotypes.

Ratnagiri, Maharashtra. The experimental material for the present study comprised 72 F 3 generation lines of cowpea, developed at the Department of Agricultural Botany, College of Agriculture, Dapoli. These 72 F 3 lines, along with five check varieties, were evaluated using an Augmented Block Design (ABD) comprising four blocks. Each block contained 18 lines and all five checks, which were repeated across the blocks for comparison. Each entry was sown in two rows, with a spacing of 30 cm between rows and 30 cm between plants. Each row contained 15 plants, making a total of 30 plants per line.

Cultural Practices

The crop was grown under uniform agronomic conditions. Land preparation was done through standard tillage practices to ensure a fine tilth. Fertilizers were applied at the rate of 25 kg nitrogen and 50 kg phosphorus (P₂O₅) per hectare. Half the nitrogen dose was applied at sowing, and the remaining half was applied 30 days after sowing. Recommended plant protection measures and irrigation were provided as required to ensure healthy crop growth.

Statistical Analysis

 Genotypic and phenotypic correlation coefficients were calculated to estimate the degree of association between seed yield per plant and other quantitative traits using the method suggested by Johnson et al. (1955) [8] .  Path coefficient analysis was performed following the method of Dewey and Lu (1959) [2] to partition the correlation coefficients into direct and indirect effects, with seed yield per plant considered as the dependent variable. The residual effect was also computed to account for the unexplained variation in seed yield.  For interpreting the magnitude of direct and indirect effects from path analysis, the classification proposed by Lenka and Mishra (1973) was followed:  Negligible: 0.00 -0.09  Low: 0.10 -0.19  Moderate: 0.20 -0.29  High: 0.30 -0.99  Very High: >1.00

Results and Discussion

Correlation analysis Phenotypic Correlation Coefficient

Seed yield per plant showed highly significant positive correlation with test weight (0.346), number of pods per cluster (0.387), number of seeds per pod (0.444), number of branches per plant (0.453), number of clusters per plant (0.625), number of pods per plant (0.670) and harvest index (0.958). It had significant positive correlation with pod length (0.273). It showed non-significant positive correlation with days to maturity (0.039), iron content (0.045), days to 50% flowering (0.051) and days to first flowering (0.060). Seed yield per plant showed nonsignificant negative correlation with plant height (-0.155) and protein content (-0.059). The similar results were observed by Sogalad et al. (2022) [16] and Gawande et al. (2024) [6] for number of clusters per plant, number of pods per plant, number of pods per cluster, number of seeds per pod, pod length, test weight and harvest index. Number of branches per plant exhibited highly significant positive correlation with number of pods per cluster (0.296), number of seeds per pod (0.300), seed yield per plant (0.453), number of clusters per plant (0.489), harvest index (0.492) and number of pods per plant (0.571). Number of clusters per plant showed highly significant positive correlation with harvest index (0.593), seed yield per plant (0.625) and number of pods per plant (0.822). Number of pods per cluster showed highly significant positive correlation with Seed yield per plant (0.387), harvest index (0.398) and number of pods per plant (0.415). Number of pods per plant showed highly significant positive correlation with harvest index (0.642) and seed yield per plant (0.670). Harvest index exhibited highly significant positive correlation with seed yield per plant (0.958). Protein content had non-significant negative correlation with iron content (-0.063) and seed yield per plant (-0.059). Iron content had non-significant positive correlation with seed yield per plant (0.045).

Genotypic Correlation Coefficient

Seed yield per plant exhibited highly significant positive correlation with pod length (0.383), test weight (0.432), number of seeds per pod (0.553), number of pods per cluster (0.565), number of branches per plant (0.620), number of clusters per plant (0.715), number of pods per plant (0.901) and harvest index (0.996). It had non-significant positive correlation with iron content (0.031), days to 50 per cent flowering (0.085), days to first flowering (0.093), days to maturity (0.101) and non-significant negative correlation with protein content (-0.076) and plant height (-0.197). The similar results were observed by Sogalad et al. (2022) [16] and Gawande et al. (2024) [6] for number of clusters per plant, number of pods per plant, number of pods per cluster, number of seeds per pod, pod length, test weight and harvest index. Number of branches per plant exhibited highly significant positive correlation with number of seeds per pod (0.346), pod length (0.356), number of pods per cluster (0.445), number of clusters per plant (0.522), harvest index (0.599), seed yield per plant (0.620) and number of pods per plant (0.828). Number of clusters per plant showed highly significant positive correlation with pod length (0.328), harvest index (0.672), seed yield per plant (0.715) and number of pods per plant (0.975). Number of pods per cluster showed highly significant positive correlation with number of seeds per pod (0.313), test weight (0.332), number of pods per plant (0.369), harvest index (0.544) and seed yield per plant (0.565). Number of pods per plant showed highly significant positive correlation with harvest index (0.806), seed yield per plant (0.901) Harvest index exhibited highly significant positive correlation with seed yield per plant (0.996). Protein content had non-significant negative correlation with seed yield per plant (-0.076). Iron content had non-significant positive correlation with seed yield per plant (0.031).

Path coefficient analysis Phenotypic Path Coefficient Analysis

At the phenotypic level, harvest index exerted the highest positive direct effect on seed yield per plant, with a value of 0.9195, indicating that this trait contributed substantially and directly to yield. Number of clusters per plant and number of pods per plant also showed positive direct effects, with values of 0.0894 and 0.0457 respectively, although their contributions were relatively lower. Similar results were reported by Bhagwati et al. (2017) [1] , Patil et al.

(2021) [15] and Gawande et al. (2024) [6] . Protein content (0.0184), test weight (0.0146), days to first flowering (0.0143), days to maturity (0.0322), and number of pods per cluster (0.0143) recorded negligible but positive direct effects on seed yield per plant. The similar results were obtained by Jogdhande et al. (2017) [7] , Patil et al. (2021) [15] and Pareet et al. (2022) [14] . Negative direct effects were observed for days to 50% flowering (-0.0540), number of branches per plant (-0.0513), pod length (-0.0323), number of seeds per pod (-0.0388), plant height (-0.03), and iron content (-0.0023) suggesting that these traits had minimal direct contributions to seed yield at the phenotypic level. The result was in conformation with reports of Dinesh et al. (2017) [3] and Sogalad et al. (2022) [16] . Overall, harvest index emerged as the most influential trait contributing directly to yield, while most other traits contributed through indirect effects.

Genotypic Path Coefficient Analysis

At the genotypic level, harvest index recorded the highest positive direct effect on seed yield per plant, with a value of 0.6245, followed by days to maturity (0.4649) and number of clusters per plant (0.3288) on seed yield per plant. Bhagwati et al. (2017) [1] , Pareet et al. (2022) [14] and Gawande et al. (2024) [6] observed similar results in cowpea for positive direct effect on seed yield per plant. Days to 50 per cent flowering (0.133), number of branches per plant (0.1322) and protein content (0.143) had low positive direct effect. Similar results were reported by Jogdhande et al. (2017) [7] and Lonare et al. (2024). Number of pods per cluster (0.0291), number of pods per plant (0.0244), pod length (0.0461), number of seeds per pod (0.0212) and test weight (0.076) had negligible positive direct effect on seed yield per plant. Patil et al. (2021) [15] , Pareet et al. (2022) [14] and Gawande et al. (2024) [6] observed similar results for seed yield per plant. Plant height (-0.1272) and iron content (-0.1873) had low negative direct effect on seed yield per plant. Similar results were reported by Jogdhande et al. (2017) [7] for plant height on seed yield per plant. Days to first flowering had high negative direct effect (-0.6976). Pareet et al. (2022) [14] observed similar negative direct effect of days to first flowering on seed yield per plant. This analysis indicates that days to maturity, harvest index, and number of clusters per plant had the strongest direct contributions to seed yield per plant at the genotypic level, and should be considered as important selection criteria in cowpea improvement programs. 0.0143 0.0113 0.0103 0.0018 0.003 0.0007 0.003 -0.0012 -0.0006 -0.0005 -0.0007 0.0006 0.0013 0.0002 D50 -0.0427 -0.054 -0.0452 -0.0088 -0.0102 -0.0015 -0.0091 -0.0007 0.0002 0.0018 0.0017 -0.0029 -0.006 -0.0056 DM 0.0232 0.0269 0.0322 0.0047 0.004 0.001 0.0031 -0.0001 0.0009 0.001 -0.0001 0.0014 0.0015 0.0063 NBPP -0.0065 -0.0083 -0.0075 -0.0513 -0.0251 -0.0152 -0.0293 0.0007 -0.0081 -0.0154 -0.0047 -0.0253 0.0084 -0.0052 NCPP 0.0187 0.0169 0.0111 0.0437 0.0894 0.0106 0.0734 0.0059 0.015 0.0208 0.0104 0.053 -0.0063 0.0189 NPPC 0.0008 0.0005 0.0005 0.0048 0.0019 0.0163 0.0067 -0.0009 0.003 0.0033 0.0035 0.0065 0.0002 0.0007 NPPP 0.0097 0.0077 0.0044 0.0261 0.0375 0.0189 0.0457 -0.0009 0.0072 0.0086 0.0052 0.0293 -0.0023 0.0068 PH 0.0025 -0.0004 0.0001 0.0004 -0.002 0.0017 0.0006 -0.03 0.0036 0.0053 0.0064 0.0043 -0.0021 -0.004 PL 0.0013 0.0001 -0.0009 -0.0051 -0.0054 -0.0059 -0.0051 0.0038 -0.0323 -0.0192 -0.0045 -0.0107 -0.0035 -0.0022 NSPP 0.0013 0.0013 -0.0012 -0.0116 -0.009 -0.0079 -0.0073 0.0068 -0.023 -0.0388 -0.0019 -0.02 -0.0041 0.0009 TW -0.0008 -0.0005 0.0001 0.0013 0.0017 0.0032 0.0017 -0.0031 0.002 0.0007 0.0146 0.005 -0.0013 -0.0028 HI 0.0368 0.0494 0.0388 0.4526 0.5451 0.3658 0.5905 -0.1332 0.3051 0.4733 0.3133 0.9195 -0.0649 0.0549 PC 0.0017 0.0021 0.0009 -0.003 -0.0013 0.0002 -0.0009 0.0013 0.002 0.002 -0.0017 -0.0013 0.0184 -0.0012 IC -0.0003 -0.0024 -0.0045 -0.0023 -0.0049 -0.001 -0.0034 -0.0031 -0.0016 0.0005 0.0044 -0.0014 0.0014 -0.

Conclusion

From this study, it can be concluded that seed yield per plant was positively and significantly associated with number of branches per plant, clusters per plant, pods per cluster, pods per plant, pod length, seeds per pod, test weight, and harvest index at both genotypic and phenotypic levels. These traits play a crucial role in yield determination and can be considered important for selection. Path analysis revealed that harvest index, number of clusters per plant, and pods per plant had high positive direct effects at the phenotypic level. At the genotypic level, harvest index, days to maturity and clusters per plant showed high direct effects on seed yield. Therefore, these traits should be prioritized as key selection criteria for improving seed yield in cowpea breeding programs. The results also suggest that selection based on both correlation and path analysis can enhance breeding efficiency. The strong genetic associations observed in this study indicate that early generation selection for these traits can lead to substantial genetic improvement. Overall, these findings provide a useful basis for identifying superior genotypes and designing effective yield improvement strategies in cowpea.

References

1. Bhagwati B, Sharma PP, Deva RM. Correlation coefficient analysis for various quantitative traits in cowpea (Vigna unguiculata (L.) Walp.) genotypes under different environments. J Pharmacogn Phytochem. 2017;6:1994-2001.
2. Dewey DR, Lu K. A correlation and path-coefficient analysis of components of crested wheatgrass seed production. Agron J. 1959;51(9):515-518.
3. Dinesh HB, Viswanatha KP, Lohithaswa HC, Pavan R, Singh P. Variability, correlation and path analysis studies in F3 generation of cowpea (Vigna unguiculata (L.) Walp.). Int J Curr Microbiol Appl Sci. 2017;6(9):1420-1428.
4. Federer W. Augmented designs with one-way elimination of heterogeneity. Biometrics. 1961;17:447473.
5. Gaiwal KB, Toprope VN, Patil SS, Chaudhari BD, Dalvi VT. Genetic variability, heritability and genetic advance studies in F₂ and F₃ populations of cowpea [Vigna unguiculata (L.) Walp.]. Pharma Innov J. 2022;11(12):1853-1856.
6. Gawande Y, Rathore NR, Sinha A. Genetic variability, character association and path analysis studies in cowpea (Vigna unguiculata (L.) Walp.) genotype. Int J Curr Microbiol Appl Sci. 2024;13(12):250-257.
7. Walp.) genotypes. Int J Curr Microbiol Appl Sci. 2017;6(6):3305-3313.
8. Johnson HW, Robinson HF, Comstock RE. Estimates of genetic and environmental variability in soybean. Agron J. 1955;47:314-318.
9. Khan H, Viswanatha KP, Sowmya HC. Study of genetic variability parameters in cowpea (Vigna unguiculata (L.) Walp.) germplasm lines. Bioscan. 2015;10(2):747-750.
10. Lenka D, Mishra B. Path coefficient analysis of yield in rice varieties. Indian J Agric Sci. 1973;43:376-379.
11. Lokesh GY, Niranjana M. Correlation and path analysis studies in F2 population of cowpea (Vigna unguiculata (L.) Walp.). Int J Pure Appl Biosci. 2018;6(1):279-283.
12. Lonare KM, Pethe UB, Kunkerkar RL, Dhopavkar RV, Kadam JJ, Palshetkar MG, et al. Genetic variability, heritability and genetic advance studies in F₂ generation of cowpea (Vigna unguiculata (L.) Walp.). Int J Adv Biochem. 2024;8(10):1091-1095.
13. Pareet SB, Pethe UP, Palshetkar MG, Rathod RR, Dhopavkar RR. Correlation and path analysis studies in red cowpea. Agriways. 2022;10(1):1-5.
14. Patil SE, Pethe UB, Mahadik SG, Dalvi VV, Joshi MS. Correlation and path analysis study in F₃ generation of cowpea [Vigna unguiculata (L.) Walp.] genotypes. Pharma Innov J. 2021;11(12):3749-3754.
15. Sogalad MS, Deshpande SK, Kavyashree NM. Genetic variability, correlation and path analysis studies in grain cowpea (Vigna unguiculata (L.) Walp. subsp. unguiculata). Pharma Innov J. 2022;11(12):3749-3754.