Introduction: Millet species are among the earliest plants to be domesticated and have historically served as staple foods in Central and Western Asia (particularly in China, India, and Russia), Europe, and parts of Africa. Millets belong to the tribe Paniceae within the grass family (Poaceae). Finger millet (Eleusine coracana L.) is one of the economically important species. This species belongs to the tribe Chlorideae and is cultivated mainly in parts of India and Africa, where it is used both as grain and fodder. Compared to other millet species, finger millet requires moderate climatic conditions and adequate rainfall. The tribe Paniceae, on the other hand, is considered one of the largest tribes in the Poaceae family, comprising 71 genera and approximately 1400 species distributed mainly in tropical and temperate regions. Panicum is one of the largest genera, including about 400 species that are mostly distributed in subtropical and temperate zones. Most recent studies have focused on understanding the correlations between grain yield and its contributing components to determine which traits should be prioritized when selecting superior cultivars. Therefore, this study was conducted in Birjand to develop a preliminary strategy for identifying high-performing cultivars.
Materials and Methods: In this experiment, various morphological, phenological, and quantitative traits were measured, including plant height, stem diameter, number of leaves, number of tillers, number of fertile tillers, 1000-seed weight, leaf-to-stem ratio, fresh and dry weight of stems and leaves, fresh and dry forage yield, grain yield, reaction to rust disease, and days to flowering. To assess resistance to rust disease, a scale from 1 to 5 was used, where 1 indicated susceptibility and 5 indicated high tolerance to the disease. For determining grain yield, panicles from the two middle rows were harvested after removing 0.5 meters from each end of the plot at physiological maturity. After threshing and cleaning, the grain yield was calculated. After collecting the data, statistical analysis was performed using the MSTAT-C software. To compare the means, Duncan's multiple range test was employed at the 5% probability level. Simple correlation coefficients between traits were calculated based on the results of the first and second years using SPSS software. Environmental and genetic variance components were also estimated based on the expected mean squares. Principal component analysis was conducted using SPSS software to reduce data dimensionality and interpret the existing variation among genotypes.
Results and Discussion: The results of the combined analysis of variance over three experimental years showed that all studied cultivars significantly differed (p ≤ 0.01) for all measured traits, indicating substantial genetic variability among them. The genotype × year interaction was statistically significant (p ≤ 0.01) for grain yield and dry forage yield, and significant at the 5% level for fresh forage yield, while no significant interaction was observed for other traits. Mean comparisons using Duncan’s multiple range test revealed that cultivars KCM19, KCM6, KCM5, and KCM1 produced the highest number of tillers, whereas KCM8 had the lowest. In terms of fertile tiller count, KCM2 and KCM5 had the highest, while KCM14 had the lowest. Regarding leaf number, KCM1, KCM5, and KCM6 had the highest values, while KCM8 and KCM20 had the lowest. For plant height, KCM3, KCM6, and KCM10 showed the maximum heights with averages of 87.4, 87.2, and 88.9 cm respectively, while KCM12 had the minimum height at 61.9 cm. In terms of stem diameter, KCM2 had the maximum value (4.5 mm), while KCM1 had the smallest (3.2 mm).
Conclusion: Combined ANOVA of morphological and quantitative traits indicated significant differences among cultivars in terms of yield performance. Among the tested cultivars, KCM1, KCM8, KCM14, KCM18 were identified as the most promising. Correlation analysis between grain yield and other traits showed that dry forage yield (r = 0.59^**), number of leaves (r = 0.32^*), plant height (r = 0.65^**), panicle length (r = 0.51^**), and number of tillers (r = 0.34^**) had positive and significant correlations with grain yield. Genetic variation coefficients were higher for fresh and dry forage yield and leaf number compared to other traits, indicating considerable diversity among the studied lines. Given the genetic diversity among genotypes, principal component analysis (PCA) was carried out to determine the role and contribution of each trait. PCA explained the existing variability through five main components. Based on the eigenvalues and eigenvectors within each component, these findings can be effectively utilized in breeding programs to exploit trait relationships and improve selection efficiency.