Assessment of Inter-Relation & Multivariate Analysis for Horticultural Traits in Garden Pea (Pisum sativum var. hortense)

Garden pea (Pisum sativum var. hortense L.) is the major food legume & ranks fourth in terms of production in the world (Vohra et al., 2015). It is a very protein rich and nutritious vegetable (Yadav et al., 1996) as its green seeds contain 7.2 % protein, vitamin A (139 I.U.), vitamin C (9 mg), phosphorus (139 mg), calcium (20 mg), carbohydrates (14.5 g) (Peter et al., 2012). Furthermore, it encompasses various health benefits by reducing colon cancer, coronary disease & LDL-cholesterol (Kour et al., 2020). In addition, pea pods radiating from mature, fresh green peas represent a copious source of dietary fiber, calcium & protein compared to broad bean & okra by-products (Kumari & Deka, 2021), as well as phenolic antioxidants (Hanan et al., 2020).

The most crucial objective of pea breeding is to enhance productivity, but pod yield is an intricate character influenced by polygenic traits & its expression depends on genetic factor, environment & their interaction (Burstin et al., 2015). Many studies have shown the effect of genetic factors, environment & their interaction on the pod yield of peas (Bocianowski et al., 2019). Thus, the improvement of crops depends upon the amount of variability present in the base material & effectiveness of selection (Kumari et al., 2012). The correlation provides insight into the relationships between characters & the influence of the environment. However, correlation studies do not provide an indication of the nature & extent of the involvement of traits in yield. This problem is overcome by path analysis, which divides the correlation coefficient into direct & indirect effects. Principal component analysis (PCA) helps in effective selection by identifying important traits & curtailing negligible, non-significant traits. This data gives proper weightage during the selection of traits, so that the enhancement of the desired trait can be achieved efficiently. Keeping the above facts in view & lacunas in the research in this aspect, we conducted this study to evaluate the interrelation and multivariate analysis of horticultural traits in garden peas.

Experiment Location

This research was conducted from 2022 to 2023 at the experimental station of the Agricultural Research Institute, Patna, Bihar. This place is characterized by a humid, sub-tropical climate with hot summers between March and June & cold winters between late October & early February.

Genetic Materials

Seven genotypes of garden peas (table 1) were collected from different sources & sown at two different sites on two dates, viz., 8 November 2022 (Early Sowing) & 10 December 2022 (Late Sowing). Before sowing, the field was prepared through 3 to 4 ploughings to achieve good soil tilth. Then, seed was sown at a row-to-row & plant-to-plant spacing of 30cm × 10cm. The standard package of practices as recommended by the Agricultural Research Institute, Patna, for vegetable crops was followed to raise the crop.

Table 1: List of genotypes

Experimental Design and Statistical Analysis

            All genotype was shown in a randomized block design with three replications. Data of eleven yield contributing characters viz., number of node to first flower (NFT), days to 50% flowering (DFF), days to first pod harvest (DPH), number of branch per plant (NBP), plant height in cm (PH), pod length in cm (PL), pod diameter in cm (PD), number of pod per plant (NPP), number of seed per pod (NSP), average pod weight in g (APW) & yield per plant in g (YPP) were taken by following the descriptors. The statistical analysis of pooled data was conducted to test for significant differences, as suggested by Panse & Sukhatme (1967). The correlation analysis, path analysis & PCA analysis were performed using RStudio software version 2023.03.1+446 & cluster analysis was conducted with the help of Past 4.03 software.

Correlation & path analysis of pea genotypes

Character association among eleven traits based on Pearson’s correlation analysis is given in Table 2 & figure 1. A non-significant correlation was found for the number of nodes to first flower, days to 50% flowering, days to first pod harvest, number of branches per plant, pod diameter & number of pods per plant with yield. Pod length (r=0.942; p≤0.01), number of seeds per pod (r=0.848; p≤0.05) & average pod weight (r=0.918; p≤0.01) had highly significant positive & strong (0:75 ≤r <1:00) relations with yield. Therefore, emphasis given to these traits during selection could be helpful for the genetic improvement of garden peas. Plant height (r=0.763; p≤0.05) had a significant negative & strong (0:75 ≤r <1:00) association with yield. Rahman et al. (2021) reported a positive correlation between yield & the number of seeds per pod. Kumar et al. (2015) also found a positive correlation between pod yield & pod length, as well as the number of seeds per pod, suggesting that these are significant traits contributing to pod yield. Adebisi et al. (2004) confirmed that the reflection of correlation values among the variables is crucial for selecting superior genotypes.

Table 2: Correlation analysis of eleven traits in garden pea genotypes

Legends: *P<0.05; ** P<0.01.

The path coefficient analyses (Table 3) provide a clear understanding of the nature of the association between the different traits in forming a proficient selection approach. The results revealed that the days to first pod harvest, number of branches per plant, number of pods per plant, number of seeds per pod & average pod weight had a positive direct effect on yield. However, the highest positive direct effect on yield was exhibited by average pod weight. Plant height showed a direct adverse effect on fruit yield per plant (-1.486), similar to its rYPP (-0.763), implying that a true relation exists between these traits. Moreover, plant height has a high indirect effect on yield per plant through pod length (2.281). Kumar et al. (2015) revealed the direct effects of the number of pods per plant & number of seeds per pod on pod yield & suggested that these traits can provide a reliable basis for improving pod yield in garden peas. In the same way, pod length had a positive indirect effect on yield through the number of seeds per pod (1.005), average pod weight (1.721) & plant height (1.293). Thus, correlation & path analysis confirms that the number of seeds per pod & average pod weight lead to an increase in yield in peas. Hence, selection based on the above-selected characters would be effective in increasing the yield of garden peas.

Figure 1: Correlation matrix of different characters in garden pea genotypes.

Table 3: Direct & indirect effects of eleven yield contributing characters in garden pea

Principal component analysis (PCA)

In the present research, the entire diversity was divided into seven groups, in which PC1 (6.58), PC2 (2.22) & PC3 (1.44) had more than 1 eigenvalue, representing the maximum variation (Table 4 & Figure 2). PCs with less than 1 eigenvalue were considered to have no practical significance (Tejaswini et al., 2018). Thus, it is inferred that vital features of the data correspond to the first three PCs, explaining 93.10% of the cumulative variation. Among all PCs, the highest was marked in PC1 (59.86%) while PC7 had the least proportion of variance, i.e., 0.00%. Ouafi et al., (2016) reported 3 PCs with more than 1 eigenvalue, accounting for a cumulative variation of 85.92%. Devi et al., (2021 & 2023) noted that 74.88% & 70.92% of the cumulative variation by principal components had more than one eigenvalue.

Table 4: Eigenvalues, variability % & cumulative variation % in garden pea

Figure 2: Scree plot showing the Eigenvalue (A) & percent proportion of variance (B)

Table 5: Principal loading factor of different traits in three principal factors

The factor loading of traits, as represented in Table 5 & fig 3(A), revealed that PC1 allowed for maximum positive loadings of traits such as pod length, pod diameter, number of seeds per pod, average pod weight & yield per plant, while other traits showed negative loadings. In PC2, plant height showed the maximum positive loading, while the other traits showed negative loading. The PC3 allowed maximum positive loading for the number of nodes to first flower & days to first pod harvest. The distinguished traits come collectively from diverse PCs & contribute to variability, having an affinity to remain interrelated (Sinha & Mishra, 2013). Thus, trait-based selection may be helpful in formulating a valuable selection approach for further pea improvement programs. Gixhari et al. (2014) reported that the number of seeds per pod & yield contributed positively to variability. Umar et al. (2014) found that pod length & pod diameter were related to the first principal component.

The biplot diagram (Figure 3B) shows the association among the traits & between genotype × traits. The vector length provides a measure of the contribution of traits to total variance; the longer the vector length, the greater the contribution of those traits. The traits days to first pod harvest, pod length & average pod weight showed the highest vector length, representing their involvement in the entire diversity. The genotypes closest to the vector trait are likely to be the best performers for particular traits. The genotype Punjab 89 performs better for the number of seeds per pod, while Azad Pea 3 performs best for pod diameter. The genotype 03L/22 is the best performing for plant height & 02L/22 performs better for the number of pods per plant.

Cluster analysis

The results (Table 6) revealed that the genotypes were grouped into three major clusters. Cluster I had the highest (57.14%) number of genotypes, viz., Kashi Mukti, Arkel, Azad Pea 3 & 01L/22, while cluster II had the lowest number (0.14%) of genotypes & cluster III contains two genotypes, viz., 02L/22 & 03L/22, which correspond to 28.57% of the total population. The genotypes of different origins show a random distribution in single clusters, indicating that the diversity is not associated with the geological origin (Devi et al., 2023). Such a relationship between geological allocation & diversity might be due to genetic construction, the history of selection, heterogeneity, or the closeness of traits (Sureja & Sharma, 2001). Cluster II with single genotypes specifies their unique identity. Our results align with those of Singh et al., (2013) & Arif et al., (2020).

Table 6: Cluster analysis of pea genotypes

Figure 3: Distribution of traits (A) & association between genotypes & traits (B) based on PC1 & PC2.

In the present investigation, it is found that pod length, number of seeds per pod & average pod weight had a highly significant positive & strong correlation with pod yield per plant. This suggests that these traits can serve as reliable selection criteria for enhancing pod yield in garden peas. The PCA analysis revealed three major PCs with around 93.10% of the cumulative variation. Further, it may be concluded that hybrids resulting from diverse crosses are thought to exhibit a large amount of variability & thus ample scope to identify or isolate transgressive segregants in the advanced generation.

Arif, U., Ahmed, M. J., Rabbani, M. A., & Arif, A. A. (2020). Assessment of genetic diversity in pea (Pisum sativum l.) Landraces based on physico-chemical & nutritive quality using cluster & principal component analysis. Pak. J. Bot., 52(2), 575-580. http://dx.doi.org/10.30848/PJB2020-2(2).

Aziz-ur-Rahman, Katoch, V., Rathour, R., Sharma, S., Rana, S. S., & Sharma, A. (2021). Studies on genetic parameters, correlation & path coefficient analysis in er2 introgressed garden pea genotypes. Legume Research- An International Journal, 44(6), 621-626. https://doi.org/10.18805/LR-4142.

Bocianowski, J., Ksiezak, J., & Nowosad, K. (2019). Genotype by environment interaction for seeds yield in pea (Pisum sativum L.) using additive main effects & multiplicative interaction model. Euphytica, 215,191. https://doi.org/10.1007/s10681-019-2515-1.

Burstin, J., Salloignon, P., Chabert-Martinello, M., Jean-Bernard Magnin-Robert, J., Siol, M., Jacquin, F., Chauveau, A., Pont, C., Gregoire Aubert, G., Delaitre, C., et al. (2015). Genetic diversity & trait genomic prediction in a pea diversity panel. BMCGenome, 16,105. https://doi.org/10.1186/s12864-015-1266-1.

Devi, J., Dubey, R. K., Mishra, G.P., Sagar, V., Verma, R. K., Singh, P. M., & Singh, J. (2021). Inheritance & stability studies of multi-flowering trait in vegetable pea (Pisum sativum L.), & its contribution in yield improvement. Sci Hortic., 287,110235. https://doi.org/10.1016/j.scienta.2018.09.048.

Devi, J., Dubey, R. K., Sagar, V., Verma, R. K., Singh, P. M., & Behera, T. K. (2023). Vegetable peas (Pisum sativum L.) diversity: An analysis of available elite germplasm resources with relevance to crop improvement. Spanish Journal of Agricultural Research, 21(2),1-13. https://doi.org/10.5424/sjar/2023212-19457.

Ghixari, B., Vrapi, H., & Hobdari, V. (2014). Morphological characterization of pea (Pisum sativum L) genotypes stored in Albanian genebank. Albanian J. Agric. Sci (special edition). https://doi:/10.13140/2.1.3931.5204.

Hanan, E., Rudra, S. G., Sagar, V. R., & Sharma, V. (2020). Utilization of peapod powder for formulation of instant pea soup powder. Journal of Food Processing & Preservation, 44(11), e14888. https://doi:/10.1111/jfpp.14888.

Kour, J., Nayik, G. A., U. L. Haq, R., Anand, N., Alam, M. S., Ramaiyan, B., & Banerjee, S. (2020). Pea: In Antioxidants in Vegetables & Nuts-Properties & Health Benefits (pp. 3–17). Singapore: Springer.

Kumar, R., Kumar, M., Dogra, R. K., & Bharat, N. K. (2015). Variability & character association studies in garden pea (Pisum sativum var. hortense L.) during winter season at mid hills of Himachal Pradesh. Legume Research, 38,164-168. https://doi:/10.5958/0976-0571.2015.00051.X.

Kumari, N., Srivastava, J. P., Singh, S. K., & Singh, I. P. (2012). Heritability & genetic advance in vegetable pea (Pisum sativum L.). Annals of Agricultural Research New Series, 33(4),244-246. https://epubs.icar.org.in/index.php/AAR/article/view/42713.

Kumari, T., & Deka, S. C. (2021). Potential health benefits of garden pea seeds & pods: A review. Legume Science, 3(2),11-13. doi.org:10.1002/leg3.82. https://doi.org/10.1002/leg3.82.

Ouafi, L., Alane, F., Rahal-Bouziane, H., & Abdelguerfi, A. (2016). Agro-morphological diversity within field pea (Pisum sativum L.) genotypes. African Journal of Agricultural Research, 11(40),4039-4047. https://doi.org/10.5897/AJAR2016.11454.

Peter, K. V. & Hazara, P. (2012). Hand Book of Vegetables. Stadium Press LLC. pp-481.

Singh, S. R., Lal, S., Ahmed, N., Srivastava, K. K., Kumar, D., Jan, N., Amin, A. & Malik, A. R. (2013). Determination of genetic diversity in strawberry using the principal component analysis & single linkage cluster analysis. Afri. J. Biot., 12(24),3774-3782. https://doi.org/10.5897/AJB2012.3003.

Sinha, K. A., & Mishra, P. K. (2013). Agro-morphological characterization & morphology based genetic diversity analysis of landraces of rice variety (Oryza sativa L.) of Bankura district of west Bengal. International Journal of Current Research, 5(10),2764–69.

Sureja, A. K., & Sharma, R. (2001). Genetic divergence in garden pea (Pisum sativum L.). Veg Sci, 28,63-64.

Tassoni, A., Tedeschi, T., Zurlini, C., Cigognini, I. M., Petrusan, J. I., RodríguezÓ, & Signori, F. (2020). State-of-the-art production chains for peas, beans & chickpeas-valorization of agro-industrial residues & applications of derived extracts. Molecules, 25(6),1383. https://doi.org/10.3390/molecules25061383.

Tejaswini, K. L. Y., Manukonda, S., Ravi Kumar, B. N. V. S. R., Ramana Rao, P. V., Lal Ahamed, M., & Krishnam Raju, S. (2018). Application of principal component analysis for rice F₅ families characterization & evaluation. Emer Life Sci Res, 4(1), 72-84. doi.org/10.31783/elsr.2018.417284. https://doi.org/10.31783/elsr.2018.417284.

Umar, H. M. I., Ur-rehman, S., Bilal, M., Atif, S., Naqvi, H., Manzoor, S. A., Ghafoor, A., Khalid, M., Iqbal, M. T., Qayyoum, A., Ahmad, F., Irshad, M. A. (2014). Evaluation of genetic diversity in pea (Pisum sativum) based on morpho-agronomic characteristics for yield & yield associated traits. J. Biodivers. Environ. Sci. 4(5),323-328.

Vohra, K., Dureja, H., & Garg, V. (2015). An insight of pulses: From food to cancer treatment. J Pharmacogn Nat Prod., 1,108. https://doi.org/10.4172/2472-0992.1000108.

Yadav, L.M., Sharma, P.P., & Maurya, K.R. (1996). Effect of nitrogen and phosphorous on growth and yield of some varieties of pea. Madras Agric. J., 83(3), 142-147. https://doi.org/10.29321/MAJ.10.A00988.