Unraveling genetic variation and phenotypic plasticity for heat stress adaptation in diverse wheat germplasm

Introduction

Wheat (Triticum aestivum L.) is one of the world's most important cereal crops and contributes substantially to global food and nutritional security. Wheat is cultivated on more than 220 million hectares worldwide, with annual production exceeding 840 million tons, while India is the second-largest producer after China (USDA, 2025; FAOSTAT, 2024) [2] . Maintaining productivity under a warm climate is therefore a major priority for wheat improvement. High temperatures during reproductive and grain-filling stages speed up the development of crops, shorten the period of grain filling, lower the photosynthetic activity, and break the partitioning of assimilates, leading to reduced grain yield and grain weight (Asseng et al., 2015; Zhao et al., 2017) [3,4] . Global predictions have shown that for every 1 °C increase above the optimum, wheat yields decrease by about 6%, highlighting the need for heat-resilient cultivars (Asseng et al., 2015) [3] . The extent of damage caused by heat is genotype-dependent, allowing the possibility to exploit the genetic variation for adaptation. Grain yield is still the main breeding target, but its complex inheritance and sensitivity to environmental conditions may lower the efficiency of direct selection under stress. Phenological traits (booting, heading, flowering and maturity) influence the period of exposure to high temperature, while plant height and grain weight influence biomass and grain filling. These traits can serve as complementary selection criteria for heat adaptation (Qaseem et al., 2019; Langridge & Reynolds, 2021) [5,6] . Physiological traits can give more information on plant responses to heat stress. The Normalized Difference Vegetation Index (NDVI) gives an indirect measure of canopy greenness and crop status, while canopy temperature (CT) measures plant cooling capacity and the balance between transpiration and atmospheric demand. NDVI and CT have been validated by field phenotyping approaches which are useful to assess wheat responses to environmental stress and for selection in breeding populations (Araus et al., 2018) [7] . Quantitative genetic parameters are crucial to knowing how much of the observed phenotypic differences can be exploited by selection. Coefficients of variation for genotype and phenotype measure the extent of genetic and total phenotypic variation, respectively, while broad-sense heritability is a measure of the portion of phenotypic variation that is attributable to genetic differences. Heritability is an adjunct to genetic advance; it's the expected response to selection. These are beneficial in the conventional improvement especially if traits have high heritability with high genetic advance (Falconer & Mackay, 1996; Singh & Chaudhary, 1985) [8,10] . Phenotypic plasticity offers a complementary framework for assessing genotype-specific responses to contrasting environments. Plasticity measures the extent of change in a trait across environments and can capture coordinated responses across agronomic, developmental, and physiological traits, rather than focusing on yield loss alone. Variation in phenotypic plasticity among wheat genotypes can reveal contrasting capacities to adjust agronomic and physiological traits across environments, providing an additional criterion for assessing adaptation and stability (Elazab et al., 2025) [11] . Recent wheat studies have demonstrated the significance of phenotypic plasticity in understanding the variation in yield components, physiological traits, and environmental adaptation (Saieed et al., 2024; Elazab et al., 2025) [9,11] . Despite progress in wheat heat-stress research, relatively few studies have combined genetic variability, inheritance parameters, trait relationships, multivariate analysis, and multi-trait phenotypic plasticity within a diverse breeding panel. The present study therefore aimed to quantify genetic variability, heritability, and genetic advance for studies traits under Normal and Late-sown conditions; characterize relationships among key traits; identify stable heat-resilient wheat entries using an integrated multi-trait phenotypic performance approach.

Experimental material

The experimental material comprised 230 unique bread wheat entries, including 225 advanced breeding lines and five commercial checks. The experiment followed an alphalattice design with two replications; each divided into three incomplete blocks of 80 plots. Each block contained 75 breeding lines and the five checks, resulting in 225 unique breeding lines and 15 check plots (5 checks × 3 blocks) and 240 experimental units per replication. Details of the entries are provided in Supplementary Table S1.

Supplementary Fig 1:

Seasonal trajectories of weekly maximum (solid) and minimum (dashed) temperatures across Ludhiana (blue), Jabalpur (green), and Pusa (orange) during two consecutive Rabi cycles (2021-22 and 2022-23). The increase in temperature after March is indicative of differential accumulation of terminal heat stress during the crop reproductive phase.

Experimental layout and trait assessment

Each genotype was grown in a standardized six-row plot measuring 2.80 m in length with 20 cm row spacing. Recommended regional agronomic practices and uniform crop protection measures were followed. Days to booting (DTB), days to heading (DTHD), days to flowering (DAYSFL), and days to maturity (DAYSMT) were recorded when 50% of plants in a plot reached the respective stage. Plant height (PH), grain yield (GRYLD), thousand-grain weight (TGW), NDVI at anthesis (NDVIANTH), and canopy temperature at anthesis (CTANTH) were also recorded.

Statistical analysis

Data were analyzed for individual environments and across environments using linear mixed models fitted by restricted maximum likelihood (REML). were calculated as GCV = (√σ²G / X ̄) × 100 and PCV = (√σ²P / X ̄) × 100, respectively. Genetic advance as per cent of the mean (GAM) was worked out at 5% selection intensity with k = 2.06 following standard quantitativegenetic procedures (Falconer & Mackay, 1996) [8] Analyzes and plots were performed in R 4.5.2 using the packages lme4, lmerTest, FactoMineR, factoextra, corrplot, dplyr and related packages.

Results and Discussion

Anova

ANOVA showed significant genetic variation for all studies traits under both Normal and Late sown conditions (Tables 1 and2). The genotypic effects were highly significant (p < 0.001) for DTB, DTHD, DAYSFL, DAYSMT, PH, GRYLD, and TGW indicating considerable genetic diversity for phenological and yield traits. These results confirm the presence of exploitable genetic variation for traits associated with heat adaptation in the germplasm studied. The different thermal regimes resulted in evident phenotypic differences. Late sowing shortened crop development and maturity occurred up to 24 days earlier and reduced the grain yield as much as 47.9 per cent at Pusa. [12,5] .

Heritability and genetic advance

Genetic parameters varied among traits and environments (Tables 345). Under Normal sowing, GRYLD showed comparatively higher GAM at Pusa (12.47%) and Ludhiana (8.56%) than at Jabalpur (2.13%). Under Late sowing, GRYLD retained considerable genetic gain at Pusa (GAM = 11.11%), indicating exploitable genetic variation even under terminal heat stress. Broad-sense heritability for GRYLD was low under Late sowing at Ludhiana (H2 = 0.03) and Jabalpur (H2 = 0.05) but was moderate at Pusa (H2 = 0.12). Phenological traits showed low to moderate genetic advance while DAYSMT recorded particularly low GAM under stress (0.11% at Pusa and 0.19% at Ludhiana). Heritability and genetic advance together can be useful in guiding selection for traits where genetic control and expected gain were greater, direct selection is likely to be more effective; traits with lower estimates may be better used as complementary indicators rather than sole selection targets. Significance levels: *** = P < 0.001, ** = P < 0.01, * = P < 0.05, ns = non-significant. Significance levels: *** = P < 0.001, ** = P < 0.01, * = P < 0.05, ns = non-significant.

Multivariate variation and associations among traits

GRYLD was positively associated with TGW (r=0.55) indicating that grain weight retention contributed to yield performance in the evaluated germplasm like recent reports of positive relationships between yield and grain weight under heat stress (Bhandari et al., 2024) [13] . Similar studies associated with canopy temperature with wheat performance under heat stress (Yu et al., 2025) [16] , CTANTH was negatively correlated with GRYLD (r = -0.28) and TGW (r = -0.39), indicating that cooler canopies were associated with better yield and grain-weight maintenance. The phenological traits were highly correlated with each other, indicating a synchronization of crop development, and NDVI and CT were complementary indicators of canopy condition and thermal regulation. (Fig. 2). In addition, the multivariate structure of the evaluated traits was summarized using principal component analysis (PCA) (Fig. 3). PC1 and PC2 explained 56.9% and 23.6% of the total variation, respectively, and together accounted for 80.5% of GRYLD and TGW were the most important discriminators among genotypes, but phenological and physiological traits captured additional dimensions of variation. This is consistent with the present wheat research in which yield components along with phenological and physiological traits effectively phenotyped the genetic variation and identified the heat-adaptive genotypes (Khare et al., 2024) [14] . Combined PCA biplot displaying distinct phenotypic clustering of 240 elite wheat genotypes; the first two dimensions explained a robust 80.5% of the total phenotypic variance, while the 95% confidence ellipses statistically confirmed a highly significant spatial divergence between the optimal and heat-stressed environments.

Multi-Trait phenotypic plasticity

Phenotypic plasticity revealed substantial differences among traits under contrasting thermal conditions (Table 6). PI_PH had the highest variation (CV=84.97%) followed by PI_DTB (82.77%) and PI_DTHD (53.88%) indicating clear genotypic differences in developmental adjustment to late sown conditions. PI_DAYSMT (CV = 9.22%) and PI_CTANTH (10.94%) showed relatively narrow variation indicating more consistent responses across genotypes when compared. These trait-dependent differences in plasticity are in line with recent studies in wheat, where the magnitude and direction of phenotypic adjustment varies widely among traits and genotypes in different environments (Saieed et al., 2024; Elazab et al., 2025) [9,11] . The Overall PI was an integrated measure of multi-trait stability and ranked genotypes based on their overall response for agronomic, phenological and physiological traits (Table 7; Fig. 4). The Overall PI was lowest for G12072 (0.093), followed by G12031 (0.099), G12007 (0.102), G12037 (0.103), G11066 (0.106) and G10040 (0.106). These entries were relatively less plastic across multiple traits, showing more phenotypic buffering across different thermal conditions. Similar genotype-dependent differences in plasticity and yield stability have recently been demonstrated in bread wheat tested in different environments including late-sown heatstress conditions (Sangha et al., 2025) [15] . Trait-specific responses further emphasized different components of adaptation. G12070 had the lowest plasticity for GRYLD (PI_Yield = 0.148) with relatively stable yield expression, while G12031 had low CTANTH plasticity (PI_CTANTH = 0.206), suggesting the canopy-temperature response across thermal regimes was relatively consistent. The stability of TGW and NDVI among the selected entries further demonstrated the significance of preserving grain filling performance and canopy function under terminal heat stress. This is in line with recent evidence of plasticity of agronomic and physiological traits contributing to yield reliability and adaptation of wheat across stressful environments (Sangha et al., 2025; Elazab et al., 2025) [15,11] . Hierarchical clustering of standardized multi-trait plasticity indices further separated the germplasm into distinct response groups (Fig. 4). PI_DTB, PI_DTHD and PI_DAYSFL showed similar clustering patterns, consistent with coordinated plastic responses of developmental traits. Conversely, PI_NDVIANTH and PI_CTANTH showed relatively separated patterns, with a complementary variability of canopy greenness and thermal responses. Also, it has been proved that canopy temperature is an auxiliary indicator for wheat yield and grain performance under the field condition (Yu et al., 2025) [16] . Recent study for genetic analysis of wheat plasticity also reveals substantial genotype-dependent plasticity in flowering time and plant height further highlighting the genetic basis of developmental responses to environmental variation (Chen et al., 2025) [17] . The clustering pattern gave wide support for the Overall PI ranking with G12072, G12031, G12007, G12037 and G10040 being among the most stable entries in the ranking across the clustering analysis. These results together demonstrate that multi-trait plasticity can add to conventional trait means by identifying germplasm that exhibits relatively stable expression across contrasting thermal regimes, offering promising candidates for followup validation in heat-resilience breeding.

Conclusions

The study revealed substantial genetic variation for key agronomic, phenological, and physiological traits in diverse bread wheat germplasm evaluated under contrasting thermal regimes. Terminal heat stress accelerated crop development and substantially reduced grain yield, with pronounced effects on thousand-grain weight. Variation in heritability and genetic advance indicated differential potential for selection among traits, while trait associations and PCA demonstrated the complementary contributions of productivity, phenology, and canopy-related traits to heatresponse characterization. Multi-trait phenotypic plasticity further differentiated genotypes according to their stability across thermal conditions and identified G12072, G12031, G12007, G12037, and G10040 as the most promising stable entries. These genotypes provide valuable breeding material for further multi-environment validation and incorporation into wheat improvement programs aimed at enhancing yield stability and adaptation to terminal heat stress.

Acknowledgement

Acknowledgments: This research was supported by the Indian Council of Agricultural Research (ICAR), Government of India, and the ICAR-National Bureau of

References

1. United States Department of Agriculture, Foreign Agricultural Service. India: Wheat production, area and yield. Production, Supply and Distribution database. 2025.
2. Food and Agriculture Organization of the United Nations. FAOSTAT: Crops and livestock products database. FAO; 2024.
3. Asseng S, Ewert F, Martre P, Rötter RP, Lobell DB, Cammarano D, et al. Rising temperatures reduce global wheat production. Nature Climate Change. 2015;5(2):143-147. doi:10.1038/nclimate2470
4. Zhao C, Liu B, Piao S, Wang X, Lobell DB, Huang Y, et al. Temperature increase reduces global yields of major crops in four independent estimates. Proceedings of the National Academy of Sciences. 2017;114(35):9326-9331. doi:10.1073/pnas.1701762114
5. Qaseem MF, Qureshi R, Shaheen H, Shafqat N. Genome-wide association analyses for yield and yieldrelated traits in bread wheat (Triticum aestivum L.) under pre-anthesis combined heat and drought stress in field conditions. PLOS ONE. 2019;14(3). DOI:10.1371/journal.pone.0213407
6. Langridge P, Reynolds M. Breeding for drought and heat tolerance in wheat. Theoretical and Applied Genetics. 2021;134(6):1753-1769. doi:10.1007/s00122021-03795-1
7. Araus JL, Kefauver SC, Zaman-Allah M, Olsen MS, Cairns JE. Translating high-throughput phenotyping into genetic gain. Trends in Plant Science. 2018;23(5):451-466. doi:10.1016/j.tplants.2018.02.001
8. Falconer DS, Mackay TFC. Introduction to quantitative genetics. 4th ed. Longman; 1996.
9. Saieed MAU, Zhao Y, Chen K, Rahman S, Zhang J, Islam S, et al. Phenotypic plasticity of yield and yieldInternational Journal of Advanced Biochemistry Research https://www.biochemjournal.com related traits contributing to the wheat yield in a doubled haploid population. Plants. 2024;13(1):17. doi:10.3390/plants13010017
10. Singh RK, Chaudhary BD. Biometrical techniques in genetics and plant breeding. 2nd ed. Kalyani Publishers; 1985.
11. Elazab A, Castillo D, Matus I, Romero-Bravo S, Espinoza-Meza S, Sadras V, et al. The role of phenotypic plasticity of physiological and agronomic traits on the adaptation of spring wheat genotypes to Mediterranean environments. Field Crops Research. 2025;322:109740. doi:10.1016/j.fcr.2025.109740
12. Lobell DB, Sibley A, Ortiz-Monasterio JI. Extreme heat effects on wheat senescence in India. Nature Climate Change. 2012;2(3):186-189. doi:10.1038/nclimate1356
13. Bhandari R, Poudel MR. Genotype × environment interaction and selection parameters for high yielding wheat genotypes under irrigated and heat stress environment. Journal of Sustainable Agriculture and Environment. 2024;3. doi:10.1002/sae2.12098
14. Khare V, Shukla RS, Singh SK, Pandey S. Holistic approach to ascertain genetic variability and responsive trait selection for different heat regimes of central India via field screening of wheat recombinant inbred lines. Journal of Agronomy and Crop Science. 2024;210. doi:10.1111/jac.12674
15. Sangha JS, Wang W, Knox R, Ruan Y, Cuthbert RD, Isidro-Sánchez J, et al. Phenotypic plasticity of bread wheat contributes to yield reliability under heat and drought stress. PLOS ONE. 2025;20(3). doi:10.1371/journal.pone.0312122
16. Yu Y, Li C, Shen W, Yan L, Zheng X, Yao Z, et al. Correlation study between canopy temperature (CT) and wheat yield and quality based on infrared imaging camera. Plants. 2025;14(3):411. doi:10.3390/plants14030411
17. Chen Y, Dong H-B, Peng C-J, Du X-J, Li C-X, Han XL, et al. Phenotypic plasticity of flowering time and plant height related traits in wheat. BMC Plant Biology. 2025;25:636. doi:10.1186/s12870-025-06489-8