Exploiting genetic variability for simultaneous improvement of yield and blast resistance in pearl millet

Introduction

Pearl millet [Pennisetum glaucum (L.) R. Br.] is the fourth most important cereal crop in India after rice, wheat and maize, and remains a critical source of food, fodder and nutritional security for farmers of the semi-arid and arid tropics, where its tolerance to drought, heat and poor soils gives it an agronomic advantage over other cereals. Genetic improvement of the crop for grain and fodder yield depends on the presence of adequate genetic variability in the breeding population and on the extent to which that variability is heritable and transmissible to the next generation. Phenotypic expression of a quantitative trait is governed jointly by genotype and environment, and the observed phenotypic variance can be partitioned into genotypic and environmental components. The genotypic and phenotypic coefficients of variation (GCV and PCV) provide a standardised measure of the extent of variability present for a trait, while heritability in the broad sense indicates the proportion of that variability which is genetically transmissible. However, heritability estimated in isolation can be misleading, since it does not distinguish additive from non-additive gene action; heritability estimates are therefore most informative for predicting breeding progress when interpreted together with genetic advance, as originally proposed by Johnson et al. (1955) [3] . A trait combining high heritability with high genetic advance as percent of mean is generally governed by additive gene effects and responds well to simple mass or pedigree selection, whereas high heritability accompanied by low genetic advance points to the influence of non-additive gene action or strong genotype-environment interaction, calling for alternative breeding approaches such as heterosis breeding or recurrent selection. Disease pressure adds a further dimension to selection strategy in pearl millet, since blast caused by Pyricularia grisea (Cooke) Sacc. has, in recent years, become a significant constraint to both grain and fodder productivity of the crop (Thakur et al., 2011) [6] . Where yield evaluation and disease screening are carried out on the same breeding material under natural epiphytotic conditions, the genetic parameters for yield traits can be interpreted alongside disease reaction to identify traits that could serve as reliable indirect selection criteria for combined yield improvement and disease tolerance. With this background, the present investigation was undertaken to estimate the components of genetic variability, heritability and genetic advance for nine morpho-phenological and yield-related traits in a set of pearl millet advanced breeding lines and checks, and to examine their relevance to blast-resistance breeding based on a companion natural-condition blast screening trial on the same material.

Materials and Methods

The experimental material comprised the same set of 100 advanced pearl millet breeding lines (ABLB-1 to ABLB-100) and seven check cultivars (ICMB-94555, ICMB-98222, PMBSN-24, PMBSN-26, PMBSN-27, VPMV-9 and ICMV-221) evaluated for blast reaction in the companion study, laid out in an Augmented Block Design during Kharif 2025 at the Regional Agricultural Research Station, Vijayapura, College of Agriculture, Vijayapura, with a spacing of 45 cm × 15 cm and a plot size of two rows of 3 m each. Disease incidence was recorded using the formula given by Mayee and Datar (1986) [4] .

Per cent disease index (PDI) =

Sum of individual disease ratings × 100 No. of leaves observed × maximum disease grade Foliar blast severity rating scale (0-9) Observations were recorded on five randomly selected, competitive plants per plot for nine morpho-phenological and yield characters: panicle length (PL, cm), panicle girth/width (PW, cm), number of productive tillers per plant (NPT), test weight (TW, g per 1000 seeds), plant height (PH, cm), dry fodder yield (FY, kg ha⁻¹), stay-green percent (SG), days to 50 percent flowering (DFF) and grain yield (GY, kg ha⁻¹). Analysis of variance was carried out following the augmented block design procedure to test the significance of differences among genotypes, checks, and the genotypesversus-checks contrast for each character. Genotypic and phenotypic coefficients of variation were computed following Burton (1952) [2] and categorised as low (<10%), moderate (10-20%) or high (>20%). Broad-sense heritability was estimated as the ratio of genotypic to phenotypic variance following Burton and Devane (1953) [1] , and genetic advance at 5 percent selection intensity, along with genetic advance as percent of mean, was estimated following Johnson et al. (1955) [3] . Standard statistical procedures followed Panse and Sukhatme (1954) [5] .

Results and Discussion

Analysis of variance

Mean sum of squares from the analysis of variance (Table 1) revealed highly significant (P<0.01) differences among genotypes (ignoring blocks) for all nine characters studied, confirming the presence of substantial genetic variability in the material. The genotypes-versus-checks contrast was also highly significant for every trait, indicating that the advanced breeding lines as a group differed meaningfully from the released check cultivars. Similarly, the mean sum of squares among checks alone was highly significant for all traits, reflecting the wide genetic base already present among the seven check cultivars used as standards. Error mean squares were comparatively low relative to the genotype and check mean squares for most traits, and coefficients of variation ranged from 1.59% (days to 50 percent flowering) to 13.50% (number of productive tillers per plant), indicating an acceptable level of experimental precision.

Genetic variability, heritability and genetic advance

Estimates of mean, range, genotypic and phenotypic coefficients of variation, broad-sense heritability, genetic advance and genetic advance as percent of mean for the nine characters are presented in Table 2. For every character, the phenotypic coefficient of variation (PCV) was numerically higher than the corresponding genotypic coefficient of variation (GCV), as expected, since phenotypic variance includes an environmental component in addition to the genotypic component. The difference between PCV and GCV was narrow for grain yield (11.84% vs. 11.57%), number of productive tillers (34.66% vs. 31.83%) and staygreen percent (10.53% vs. 9.58%), indicating relatively limited environmental influence on the expression of these traits, whereas the gap was comparatively wider for plant height (16.08% vs. 10.25%) and dry fodder yield (11.93% vs. 7.53%), pointing to greater sensitivity of these two traits to environmental fluctuation (Sharma et al., 2018) [7] Number of productive tillers per plant recorded the highest GCV and PCV among the traits studied (31.83% and 34.66%, respectively), classifying it as highly variable, followed by grain yield and plant height, both in the moderate range. Days to 50 percent flowering showed the lowest GCV (2.79%) and PCV (3.81%), reflecting the comparatively narrow range of flowering time (45-52 days) in the material and limited scope for improvement of this trait through direct selection alone. Broad-sense heritability was highest for grain yield (92.50%) and stay-green percent (91.13% [3] . Stay-green percent likewise combined very high heritability (91.13%) with moderately high genetic advance (18.86%), suggesting good potential for improving fodder quality retention through direct selection.

In contrast, plant height and dry fodder yield showed moderate-to-low heritability coupled with low-to-moderate genetic advance, and days to 50 percent flowering combined moderate heritability (53.54%) with the lowest genetic advance over mean (4.21%) recorded among the nine traits. Such combinations point to a greater relative contribution of non-additive gene action and genotype-environment interaction in the expression of these traits, indicating that direct selection is likely to be less effective and that breeding strategies exploiting non-additive variance, such as heterosis breeding, may be more rewarding for improving plant height and fodder yield in this material.

Per se performance of genotypes and checks

Per se performance of the 100 breeding lines for the nine morpho-phenological characters is presented in Table 2, and that of the seven check cultivars in Table 3

Phenotypic Correlation Analysis of Disease Incidence with Grain Yield and Yield Attributes

Phenotypic correlation analysis (Table 5) revealed that disease incidence was significantly and positively correlated with plant height at 45 days (r = 0.23*) and 85 days (r = 0.29**), indicating that taller plants tended to show higher disease incidence. Grain yield showed significant positive associations with fodder yield (r = 0.39**) and plant height at 85 days (r = 0.25**). Further, plant height at 85 days was significantly correlated with fodder yield (r = 0.47**), panicle length (r = 0.53**), panicle girth (r = 0.37**) and productive tillers (r = 0.26**). Overall, fodder yield and plant height at 85 days were the most important traits positively associated with grain yield in the evaluated genotypes.

Overall, these results support a breeding strategy in which the advanced lines identified as Resistant or Moderately Resistant in the companion blast-screening trial (ABLB-10, ABLB-9, ABLB-12, ABLB-16, ABLB-17, ABLB-43, ABLB-58, ABLB-78 and ABLB-83) are prioritised as parents in hybridisation programmes aimed at combining high heritability-high genetic advance yield traits with durable blast resistance, followed by multi-location, multiseason evaluation to confirm the stability of both the yield advantage and the disease reaction before their release as cultivars or use as breeding parents.

Conclusion

Significant genotypic variability was recorded for all nine morpho-phenological and yield traits studied in the pearl millet breeding material. Grain yield, number of productive tillers per plant and test weight combined high heritability with high genetic advance over mean, indicating predominantly additive gene action and good scope for direct phenotypic selection, whereas plant height, dry fodder yield and days to 50 percent flowering were more strongly influenced by non-additive and environmental effects. Considered alongside a companion natural-condition blast screening trial on the same material, in which grain yield was significantly associated with plant height and significantly depressed by disease severity, the results suggest that selection for grain yield and its high-heritability component traits under natural blast pressure offers a practical route to simultaneously advancing yield potential and blast tolerance in pearl millet improvement programmes.

References

1. Burton GW, Devane EH. Estimating heritability in tall fescue (Festuca arundinacea) from replicated clonal material. Agronomy Journal. 1953;45(10):478-481.
2. Burton GW. Quantitative inheritance in grasses. Proceedings of the 6th International Grassland Congress. 1952;1:277-283.
3. Johnson HW, Robinson HF, Comstock RE. Estimates of genetic and environmental variability in soybeans. Agronomy Journal. 1955;47(7):314-318.
4. Mayee CD, Datar VV. Phytopathometry. Technical Bulletin No. 1. Parbhani: Marathwada Agricultural University; 1986.
5. Panse VG, Sukhatme PV. Statistical methods for agricultural workers. New Delhi: Indian Council of Agricultural Research; 1954.
6. Thakur RP, Sharma R, Rao VP. Screening techniques for pearl millet diseases. Information Bulletin No. 89. Patancheru, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics; 2011. p. 15.
7. Sharma B, Chugh LK, Sheoran RK, Singh VK, Sood M. Study on genetic variability, heritability and correlation in pearl millet germplasm. Journal of Pharmacognosy and Phytochemistry. 2018;7(6):19831987.