Identification of promising genotypes
Among the studied accessions, the trait days to first flowering ranged from 49.00 to 77.00 with an average of 58.15. Fifty seven genotypes were found significantly earlier to flowering. The genotype TNAU 82 was early in flowering (49 days), whereas TNAU 155 and TNAU 236 were recorded to be late in flowering (77 days). The trait days to fifty per cent flowering varied from 52.00 (IPS 113 and TNAU 82) to 80.00 (TNAU 236) with an average of 61.34 and 57 genotypes were significantly earlier in flowering. The plant height ranged from 36.01 cm to 86.35 cm with an average of 59.42 cm. About 55 genotypes were found significantly dwarf in nature. The genotype, APK 1 (86.35 cm) was the tallest and GPUK 3 (36.01 cm) was at the shortest. These traits can be considered as useful in the breeding programme for developing a short duration with non-lodging plant type.
Number of basal tillers per plant ranged from 11.00 (IPS 122 and TNAU 133) to 25.22 (TNAU 107) with an average of 17.36 and number of productive tillers per plant varied from 3.83 (DPS 95) to 10.73 (Sel 21) and 50 genotypes were found significant for productive tillers with an average of 6.46. The trait flag leaf blade length, IPS 123 recorded the maximum length (17.14 cm) and genotype APK 1 (1.62 cm) recorded the maximum flag leaf width. The trait peduncle length ranged from 3.78 (Sel 19) cm to 8.25 cm (TNAU 86). CO 3 (20.80 cm) registered maximum inflorescence length, TNAU 149 (9.3 cm) recorded the highest length of the longest raceme and the genotype RK 50 (11.6 cm) registered the highest thumb length. Maximum thousand grain weight was registered by APK 1 (4.91 g) with an average of 3.68 and the total of 49 genotypes recorded significant performance.
The genotypes IPS 123 and TNAU 162 witnessed the highest Zinc content (6.87 mg/100g) with an average of 3.61 mg/100g. Significance were observed for Zn content in 55 genotypes. For Iron content, TNAU 84 (25.03 mg/100g) recorded the highest value with fifty significant genotypes. The range of single plant yield realised in the present study ranged from 5.37 g to 31.37 g with an average of 15.19 g. Fifty one genotypes exhibited significant performance for single plant yield. Among them, the genotype Sel 21 registered as the high yielder (31.37 g).
Patil et al., (2019) observed wide range of variation for quantitative traits in finger millet accessions. The identified superior genotypes for 15 traits were given in Table 2.
Correlation coefficient analysis
Association of yield and other component traits helps plant breeders to focus on yield improvement in the desired direction. Among fifteen characters studied, the characters
viz., days to first flowering (r = 0.304), days to 50 per cent flowering (r = 0.305), plant height (r = 0.313), number of productive tillers (r = 0.482), peduncle length (r = 0.208), inflorescence length (r = 0.406), length of the longest raceme (r = 0.508) and thumb length (r = 0.278) were positively and significantly correlated with grain yield per plant (Table 3). Number of basal tillers (r = 0.162), thousand grain weight (r = 0.157), flag leaf length (r = 0.090), flag leaf width (r = 0.069), Fe content (r = 0.035) and Zn content (r = 0.014) expressed positive but non-significant association with grain yield. Similar results have also been reported earlier by
Vishnuprabha and Vanniarajan (2018) for Zn content in barnyard millet. Hence, it might be inferred that these traits could be considered as most important yield contributing traits in kodo millet. This is in accordance with the findings of
Plawani Panda (2015) who found that positive correlation of yield with days to first flowering, days to 50 per cent flowering, plant height, peduncle length and inflorescence length in barnyard millet;
Jadhav et al., (2015) for days to 50% flowering, plant height and productive tillers per plant in finger millet. While
Verma and Singh (1982) opined that plant height was positively correlated with grain yield in early and medium maturing genotypes in kodo millet.
Yadava and Jain (2006) indicated that plant height was significantly and positively correlated with grain yield in early and late maturing genotypes of kodo millet. In foxtail millet,
Pavithra (2015) registered positive correlation of yield with plant height, peduncle length and inflorescence length.
Prakash and Vanniarajan (2014) in proso millet and
Suryanarayana et al., (2014) in finger millet had similar findings in days to 50 per cent flowering and plant height;
Rameshwarkumar (2009) for peduncle length in little millet.
In terms of inter correlation among components studied, number of productive tillers revealed significant and positive association with thumb length, length of the longest raceme, thousand grain weight, inflorescence length and peduncle length. Peduncle length showed significant and positive association with inflorescence length, thumb length, length of the longest raceme, and thousand grain weight. Similar results were reported by
Plawani Panda (2015) for inflorescence length and thousand grain weight in barnyard millet;
Anantharaju and Ganesan (2005) for thousand grain weight in finger millet.
Inflorescence length showed positive and significant association with thumb length, length of the longest raceme and thousand grain weight. Similar results were reported by
Plawani Panda (2015) for thousand grain weight in barnyard millet. Length of the longest raceme showed a significant and positive association with thumb length and thousand grain weight. Thumb length showed positive association with thousand grain weight.
Examination of correlation among component characters revealed that strong associations are present among desirable component characters
viz., number of productive tillers, peduncle length, inflorescence length, length of the longest raceme and thumb length. Hence, selection criteria should consider all these characters for the improvement of grain yield. Undesirable association of some of the component characters might act as deterrent for the formulation of a comprehensive selection programme involving these traits. So, during selection programme, these factors might be considered with a caution.
Path coefficient analysis
Path coefficient analysis was undertaken to study the direct and indirect effects of the different traits on yield. The direct and indirect effects of fifteen characters on grain yield are presented in Table 4 and Fig 1. Path analysis revealed that inflorescence length (1.606) and plant height (1.179) had the highest positive direct effect on grain yield per plant which was followed by length of the longest raceme (0.617), flag leaf length (0.449) and number of productive tillers (0.370). Hence, direct selection for these traits would be rewarding for yield improvement, which will also reduce the undesirable effect of the component traits studied. The results were similar to the findings reported by
Plawani Panda (2015) for plant height and inflorescence length;
Prakash and Vanniarajan (2015) for plant height in barnyard millet;
Shalini et al., (2010) for plant height and number of productive tillers in proso millet.
Andualem and Tadesse (2011) and
Suryanarayana et al., (2014) for plant height in finger millet. It is known to contribute grain yield
via more number of grains per panicle which were in conformity with the findings of
Sonnad et al., (2008) in finger millet.
Regarding the indirect effect of component traits on grain yield, inflorescence length had high indirect effect through peduncle length (1.318), thumb length (1.094) and plant height (0.993). Whereas for plant height had high indirect effect through days to first flowering (0.984), days to fifty percent flowering (0.996) and thumb length (0.956). High and positive indirect effect of plant height through days to 50 per cent flowering was earlier reported by
Thakur and Saini (1995) and
Mishra (1996) in finger millet.
