Estimation of genetic parameter
The estimates of genotypic coefficient of variation (GCV) and phenotypic coefficient of variation (PCV) of yield and yield attributing characters are presented in (Table 2). In the present investigation, high genotypic coefficient of variation and phenotypic coefficient of variation was observed for ear height and lowest was observed for DM. Heritability estimate differed for different traitsranged from (99.15%) - (44.37%). Expected genetic advance as percent of mean was highest for ear height followed by kernel length. Traits with high heritability and high genetic advance as percent of mean was observed in days to 50% pollen shed, days to 50% silk, plant height, ear height ,tassel length, branches per tassel, leaves per plant, kernels per row, ear leaf length, leaf area, kernel length and grain yield. It indicates that these traits are governed largely by genes with additive gene action. Simple selection such as mass selection may improve such traits. However, testing genotypes across locations and years will give us a clear picture. Decision on choice of appropriate breeding methods will be possible if germplasm
s are tested across environments.
Correlation coefficient
Correlation study helps to determine the dependence of the traits on some relatively independent traits which are less affected by environments. The relationship of a trait with yield and other component characters could also be useful while selecting plants from a genetically diverse population and in choice of parents for a hybridization programme. Since the grain yield depends upon many yield contributing characters, it becomes essential to study the association of each such character to the yield. Character with high heritability, relatively simple inheritance and with desirable association to yield is considered for selecting plants and such a selection effects correlated response in the progeny. Further, path coefficient analysis helps to estimate the cause and effect of various yield components and finally gives a clear picture on high direct and indirect effects of the independent traits on the dependent variable such as yield.
The mean values of twenty-three quantitative traits in forty entries comprising of thirty five germplasms and five check hybrid varieties of maize were analyzed for elucidating the simple correlation coefficient between them (Table 3).
Grain yield per plant had positive and highly significant correlation with traits namely ear length, ear diameter, kernel rows per ear, kernels per row, kernel length, kernel width and 100 kernel weight. Days to 50% pollen-shed and days to 50% silk showed highly significant and negative correlation with the grain yield (Table 3). Similar finding was reported by
Pavan et al., (2011). They opined that grain yield had positive significant genetic correlation with ear length, ear circumference, number of kernel rows/ear, number of kernels/row and 100-grain weight.
El-Shouny et al (2005) also reported positive and significant with ear diameter, ear length, number of kernels per row, 100-kernel weight, number of rows per ear with grain yield per plant. The cause of correlation can be genetic or environmental. Genetic cause may be attributed to pleiotropism or linkage or both.
Path analysis
Grain yield is an ultimate product of interaction among its component traits under the various influences of environment. Path analysis provides an effective means of partitioning direct and indirect causes of association. It measures the relative importance of each component traits towards grain yield. Results of path analysis were presented on Table 4.
Days to 50% silk, ear length, ear diameter, plant height, branches per tassel, kernels per row, 100 kernel weight, kernel length and kernel width had showed positive direct effect on grain yield. Highest positive direct effect was recorded for days to 50% silk. Hence, these traits can be considered as the important for selection in a maize breeding program for grain yield improvement. This finding was in agreement with the report of
Mogesse (2021) that on ear length, 1000-kernel weight and number of kernel rows per ear contributed directly to grain yield.
Yahaya (2021) also reported plant height followed by 1000 grain weight gave positive direct effects on grain yield.
Lal et al., (2022) also found that kernels per row gave positive direct effect on grain yield.
Trait days to 50% silk had recorded high or very high positive direct effect on grain yield along with high heritability. If we select plants with lateness with respect to days to 50% silk there will be corresponding increase in grain yield, too. Trait days to 50% pollen shed had exhibited high negative direct effect on grain yield along with high heritability. Therefore, selection of plants with fewer days to 50% pollen shed may yield high yielding plants. This finding was in agreement with the report of
Sumalini, (2015).
Shukla, (2017) also found that days to male flower initiation showed negative direct effect on grain yield.
Principal component analysis
The characters which showed significant to highly significant correlation with grain yield per plant were studied for principal component analysis. In the present study, PCA created four principal components with an Eigenvalue of more than 1.0, accounting for 67.74% of the overall variation present in the maize germplasm (Table 5). The first principal component showed the highest variability, 28.87%, with Eigenvalue of 6.64, while the fourth component was the lowest contributor. In PC1, the highest value was observed for days to 50% silk (10.82) which was followed by days to 50% pollen shed (10.15), ear diameter (8.84) and ear length (8.30). The second component,
i.
e. PC2, contributed 18.34% to the total variability with Eigenvalue of 4.22. The third component (PC3) contributed 11.98% to the total variability with an Eigenvalue of 2.76. The character grain yield showed the highest contribution to the variance so far as the PC3 was concerned. Similarly, the fourth component contributed 8.55% to the total variability with an Eigenvalue of 1.97. The characters, namely, kernels per row (12.48), number of rows per ear (8.38), 100 kernel weights (11.25) and kernel length (8.84), showed the highest value of PC4 only. PCA identified few prominent traits that play an important role in classifying the variation existing in the germplasm. Therefore, emphasis should be given on the traits for selection of individuals from a heterogenous population during the crop improvement programme.
Al-Naggar et al. (2020) performed PCA analysis among the nineteen maize genotypes and reported two main components contributing 57.91% of the total variability. In India, thirty maize accessions were studied by
Suryanarayana et al. (2017) for PCA and cluster analysis. They found 85.31% of the total variance along with greater than one Eigenvalue.
