Correlation and Path Coefficient Analysis of Yield and Yield-Attributing Traits in Proso Millet (Panicum miliaceum L.)

Introduction

Proso millet (PM) is popularly known as Broomcorn millet (China), Common millet (USA), Barri (India), Broomtail millet, Kashfi Millet, Red Millet, White Millet, Brown Millet, Chinese Millet, Kibi, Mijo (Spain), Panic Millet (France) and Gijang (Korea). It is a Summer annual grass commonly grown as a late-seeded Summer crop, capable of completing its life cycle in 60-100 days (Baltensperger, 2002) [1] . It is classified into five races viz., miliaceum, patentissimum, contractum, compactum and ovatum, based on differences in panicle morphology and shape. (Vetriventhan et al., 2019). In India, proso millet is grown across several states, including Tamil Nadu, Maharashtra, Karnataka, Andhra Pradesh, Bihar, Madhya Pradesh and Uttar Pradesh, covering about 0.41 lakh hectares with a production of 0.22 lakh tonnes of grain (Elangovan et al., 2023) [3] . Characterization of available germplasm for morpho-agronomic traits is essential to assess genetic diversity, broaden the genetic base and facilitate effective genetic improvement (Pucher et al., 2015) [9] . In proso millet breeding, correlation and path coefficient analyses are crucial for dissecting the relationships among yield and its component traits. Correlation analysis reveals the strength and direction of trait associations, while path analysis distinguishes direct and indirect effects, providing insight into causal relationships influencing grain yield. These tools aid breeders in identifying key traits with significant direct effects on yield, such as productive tillers, grain weight and flowering time, allowing for precise selection and effective genetic improvement. (Rajput et al., 2024) and (Mishra et al., 2022) [10,7] Hence, a study was undertaken to estimate the correlation and path coefficient among yield contributing traits in proso millet genotypes, providing valuable information for optimizing trait selection in plant breeding programs.

Materials and Methods

The study was carried out at RSJRS, Kolhapur, Maharashtra State during Summer 2024. Thirty-five genotypes of proso millet were obtained from AICRP on Small Millets, Zonal Agricultural Research Station, Shenda Park, Kolhapur, Maharashtra State. The experiment was laid out in randomized block design (RBD). The field was divided into three homogeneous replication blocks. Thirty-five genotypes were planted randomly in three replications. Each entry is represented by a single row of 4 m length spaced at 22.5 cm between the rows and 10 cm between the plants. Five random plants from each treatment in each replication were selected for recording observations. The association among the characters was assessed by estimating genotypic correlation coefficient using the procedure outlined by Singh and Chaudhary (1981) [13] . Path coefficient analysis was done according to the procedure suggested by Dewey and Lu (1959) [2] .

Results and Discussion

Correlation of grain yield per plant with its components

Correlation analysis reveals the interrelationships between quantitative traits, aiding in the selection of appropriate breeding strategies for crop enhancement. Genotypic correlation coefficients, estimated for all trait combinations represented in Table 1, reveal the direction and strength of genetic relationships. These insights help identify key traits for selection, thereby enhancing the efficiency and precision of crop improvement programs. Grain yield per plant had highly significant and positive correlation with fodder yield per plant (0.942), basal tiller number (0.869), panicle length (0.762), days to physiological maturity (0.740), days to 50 per cent flowering (0.725), plant height (0.648), flag leaf blade length (0.490) thousand grain weight (0.476), flag leaf blade width (0.371) and peduncle length (0.236), indicating that these traits contributed substantially to enhanced grain productivity. While, a highly significant and negative correlation was observed with protein content (-0.253), suggesting a possible inverse relationship between grain yield and grain protein percentage. The findings of the present study are in close agreement with those reported by Mishra et al., (2022) [7] , who observed a strong and significant positive correlation of grain yield with days to 50 per cent flowering, days to maturity, plant height and panicle length. Similarly, Santhoshkumar et al., (2023) [12] documented highly significant positive associations between grain yield per plant and traits such as basal tiller number, flag leaf blade dimensions and peduncle length. Furthermore, Hawlader (1991) and Kumar et al., (2022) [4,6] also reported a significant and positive relationship between grain yield and 1000-grain weight.

Path coefficient analysis

Path coefficient analysis enables the breakdown of correlation coefficients into direct and indirect effects, allowing for the identification of key contributing traits and a clearer understanding of each component's actual influence on yield. This method considers the cause andeffect relationships among variables by dividing the overall association into direct and indirect effects those mediated through other independent traits. The method helps breeders in formulating selection criteria by emphasizing characters with high direct effects on yield. When used alongside correlation analysis, path analysis offers deeper insights into the nature and strength of relationships between any two quantitative characters. The path coefficient analysis at the genotypic level represented in Table 2, revealed that fodder yield per plant exerted the highest positive direct effect (0.5548) on grain yield per plant, indicating its primary role in determining grain yield. It was followed by panicle length (0.2569), days to physiological maturity (0.1638), days to 50 per cent flowering (0.0807), basal tiller number (0.0749) and thousand grain weight (0.0624), all of which showed positive direct effects on grain yield. These traits can be considered as key selection criteria for yield improvement. In contrast, characters like peduncle length (-0.0923), protein content (-0.0691), flag leaf blade width (-0.0628), flag leaf blade length (-0.0072) and plant height (-0.0009) exhibited negative direct effects. The present study revealed that fodder yield per plant exerted the highest positive direct effect on grain yield, identifying it as a key driver of productivity in proso millet. Similar results were reported by Patil et al., (2024) [8] , who also highlighted fodder yield as the most influential contributor to grain yield. In addition, traits such as panicle length, basal tiller number, days to physiological maturity and 1000-grain weight also exhibited notable direct effects, suggesting their potential as reliable selection indices. These findings are consistent with those of Santhoshkumar et al., (2023) and Rajput et al., (2024) [12,10] , who found that traits such as number of grains per panicle, basal tiller number and days to maturity exhibited strong positive direct effects on grain yield per plant in proso millet. Salini et al., (2010) [11] reported similar results for 1000 grain weight. On the other hand, protein content showed a negative direct effect, indicating a possible tradeoff between yield and nutritional quality, as also reported by Kalinova and Moudry (2006). Traits like basal tiller number, panicle length, days to physiological maturity and days to 50 per cent flowering showed strong indirect effects on grain yield per plant primarily through fodder yield. Similar trends were observed by Santhoshkumar et al., (2023) [12] . Kumar et al., (2022) [6] found that traits like panicle length and flag leaf blade width exhibited significant positive indirect effects via thousand-grain weight, plant height and flag leaf sheath length.

Conclusions

The present study revealed that traits such as fodder yield per plant, basal tiller number, panicle length, days to 50 per cent flowering, days to physiological maturity, plant height and 1000-grain weight exhibited a highly significant and positive correlation with grain yield per plant. This suggests a direct linear relationship, implying that enhancing these traits could lead to an increase in grain yield. Conversely, protein content showed a highly significant and negative association with grain yield. Path analysis at the genotypic level showed that fodder yield per plant had the highest positive direct effect on grain yield, followed by panicle length, days to maturity, basal tillers, days to 50 per cent flowering and 1000 grain weight. Traits like flag leaf blade width, peduncle length, protein content, flag leaf blade length and plant height showed negative direct effects.

Acknowledgement

AcknowledgementThe Authors are thankful to AICRP on Small Millet, Zonal Agricultural Research Station, Shenda Park, Kolhapur, Maharashtra State for providing proso millet genotypes to conduct the experiment.

References

1. Baltensperger DD. Progress with proso, pearl and other millets. Trends New Crops New Uses. 2002;100:100-3.
2. Dewey DR, Lu K. A correlation and path-coefficient analysis of components of crested wheatgrass seed production. Agron J. 1959;51(9):515-8.
3. Elangovan M, Singode A, Venkatesh K, Amasiddha B, Karthik AS, Pandey S, et al. Agro-morphological characterization of proso millet germplasm for utilization. Int J Plant Soil Sci. 2023;35(16):381-92.
4. Hawlader MSH. Genetic variability and correlation study in proso millet (Panicum miliaceum L.) in Bangladesh. Ann Bangladesh Agric. 1991;1(2).
5. Kalinová J, Moudrý J. Content and quality of protein in proso millet (Panicum miliaceum L.) varieties. Plant Foods Hum Nutr. 2006;61(1):43-7.
6. Kumar GA, Vanniarajan C, Vetriventhan M, Sudhagar R, Saravanan S. Association and variability studies of quantitative traits in proso millet (Panicum miliaceum L.). Biol Forum Int J. 2022;14(3):208-14.
7. Mishra N, Tiwari S, Sapre S, Sahu VK, Tiwari PN. Morphological characterization and correlation assessment of quantitative and qualitative traits of proso millet (Panicum miliaceum L.) germplasm. Front Crop Improv. 2022;10(Special Issue III):1202-6.
8. Patil PB, Karad SR, Gajbhiye PN, Mote MS. Genetic diversity in proso millet (Panicum miliaceum L.). Biol Forum Int J. 2024;16(1):299-302.
9. Pucher A, Sy O, Angarawai II, Gondah J, Zangre R, Ouedraogo M, et al. Agro-morphological characterization of West and Central African pearl millet accessions. Crop Sci. 2015;55(2):737-48.
10. Rajput SG, Khound R, Santra DK. Phenotypic diversity for morpho-agronomic traits in the US proso millet International Journal of Agriculture and Food Science https://www.agriculturaljournals.com (Panicum miliaceum L.) core collection. Crop Sci. 2024;64(3):2599-611.
11. Santhoshkumar C, Vaithiyalingan M, Murugan E, Renuka R, Hemalatha G. Genetic variability, trait association and diversity study in proso millet (Panicum miliaceum L.). Electron J Plant Breed. 2023;14(3):1127-34.
12. Singh RK, Chaudhary BD. Biometrical methods in quantitative genetic analysis. New Delhi: Kalyani Publishers; 1981.
13. Vetriventhan M, Azevedo VC, Upadhyaya HD, Naresh D. Variability in the global proso millet (Panicum miliaceum L.) germplasm collection conserved at the ICRISAT genebank. Agriculture. 2019;9(5):112.