Trait-based insights of grain yield and major adaptive traits in wheat under terminal heat stress conditions

Introduction

Wheat (Triticum aestivum L.) ranks among the most vital cereal crops globally, cultivated on over 220 million hectares and playing a key role in ensuring global food security (Redhu et al., 2025) [39] . Global wheat production for the 2024/25 marketing year was projected around 800 million metric tons, reflecting a modest year-on-year rise (USDA 2024) [48] . With average global yields now approaching 3.54 t ha⁻¹, wheat remains central to cereal production (USDA 2024) [48] . Given its broad cultivation under both irrigated and rainfed systems, wheat ranks among the top cereals in many regions, frequently second to rice in irrigated systems and first under rainfed conditions (Qaseem et al., 2019) [36] . However, these global statistics mask regional vulnerability to climate change. Wheat yield and stability are increasingly threatened by rising temperatures, erratic rainfall, and more frequent heat extremes, especially in arid and semi-arid zones. Among various abiotic stresses, terminal heat stress, especially occurring during post-reproductive phase, is particularly damaging and represents a major constraint on wheat productivity (Redhu et al., 2025) [39] . Estimates suggest that in irrigated wheat-growing areas, up to ~ 40% may face episodic heat stress annually, contributing to global economic losses currently estimated in the multi-billion dollar range (Reynolds et al., 2007) [40] . Even a single heat event can cause significant losses if it occurs at critical growth stages (Asseng et al., 2011) [3] . Both anthesis and grain filling are two highly important stages as compared to prior developmental stages because there are few chances for recovery as compared to prior growth stages (Khadka et al., 2020) [25] . A rise of 1 °C in mean temperature during the reproductive period is projected to reduce wheat yields by around 6% globally, mostly via shortened grain filling (by 4.1 to 6.4%), reduced seed set, and compromised source-sink dynamics (Devi et al., 2024; Hassan et al., 2021) [10,18] . Heat stress has been shown to accelerate spikelet initiation and sporogenesis, ultimately resulting in spikelet sterility (Porter and Gawith 1999) [35] . Exposure to elevated temperatures (>20 °C) during the heading and anthesis stages promotes rapid spike enlargement but concomitantly decreases spikelet number (Semenov 2009) [42] . At the stage of floral initiation, heat stress exerts detrimental effects on microspore formation and pollen cell differentiation, frequently leading to complete male sterility (Freeha Anjum et al., 2008; Kaur and Behl 2010) [16,24] . In wheat, even a short-term (3-day) heat episode during flowering can cause structural abnormalities in florets (Hedhly et al., 2009). Furthermore, heat stress at 31/20 °C (day/night) has been reported to reduce grain size by disrupting endosperm cell structure and impairing the development of the aleurone layer (Dias et al., 2008) [12] . The combined effect is a contraction of yield potential and grain quality under heat stress episodes. Thus, developing wheat genotypes with enhanced tolerance to terminal heat stress is of high priority. One of the most reliable approach involves screening and selection of potential germplasm in diverse environments that can provide reliable data on grain yield performance under both environments i.e. stress-free and heat-stressed conditions (Langridge and Reynolds 2021) [27] . However, achieving heat tolerance solely through yield-based selection is highly challenging due to its complex heritability, which has limited the attained genetic gain (Redhu et al., 2024) [38] . A more effective strategy is to dissect component traits that confer adaptation under high temperature stress. Traits such as productive tiller number, biomass accrual, harvest index, kernel weight, and seed density are frequently implicated as yield determinants under stress (Dabi et al., 2016; Kumbhar and Larik 1983) [9,26] . To unravel the complex interplay among yield traits under stress, multivariate statistical methods (e.g. correlation, path analysis, multiple regression, and principal component analysis) provide powerful tools. These methods allow differentiation of direct and indirect effects, dimension reduction, and trait prioritization for breeding (Allard 1960; Bramel et al., 1984) [2,7] . In particular, biplot analyses based on vector angles (via cosine laws) help visualize associations and trade-offs among traits (Msundi et al., 2021) [32] . In this study, 200 recombinant inbred lines (RILs) developed from WH711 × WH1021 were assessed under timely and latesown environments. This study endeavours to discern an ideal model of wheat crop, capable of producing high-yielding genotypes under terminal heat stress. Our objectives were to (i) evaluate the impact of terminal heat stress on grain yield and its component traits, (ii) explore trait interrelationships under stress via multivariate approaches, (iii) identify key adaptive traits for breeding heat-tolerant genotypes, and (iv) selection of heattolerant genotypes.

Material and Methods

Experimental site and climate

The field trials were carried out in the 2018-19 Rabi season at CCSHAU farms, Hisar (29°09′ N, 75°43′ E; elevation 215.2 m above mean sea level). The soil is characterized as coarse sandy loam, slightly alkaline (pH ~7.6) with low organic matter content. The region experiences high solar radiation and episodic heat stress during late spring; full supplemental irrigation was provided as needed, ensuring non-water limitation. Meteorological data (maximum, minimum temperatures, rainfall, and sunshine hours) were recorded from a nearby weather station and summarized in Table S1 and Fig. 1.

Experimental Design and Sowing Treatments

To evaluate genotype performance under contrasting thermal regimes, two sowing dates, i.e., timely sowing (stress-free / yield potential environment) and late sowing (terminal heat stress environment) were used. Each genotype (line) was planted in an Alpha lattice design (ALD) with 5 incomplete blocks and 2 replications. Each plot comprised of 2 rows, each 2.5 m long spaced at 0.20 m (row-to-row), with a seeding rate of 100 kg ha⁻¹. Border rows were used to minimize edge effects. All agronomic practices, including fertilization, irrigation scheduling, weed management, and pest/disease control followed the recommended package for the Northwestern Plain Zone (NWPZ) to minimize non-genetic variation.

Plant materials

A population of 200 RILs developed from WH 711 × WH 1021 via the single-seed descent method was used (Table S2). WH 1021 is known for heat tolerance and high yield under late sowing, while WH 711 is agronomically superior under optimal, non-stress conditions but more heat susceptible. All entries (RILs and parents) were grown under both sowing conditions.

Measured traits and data collection

Phenological, morphological, and yield traits were systematically recorded for each recombinant inbred line (RIL). Phenological traits included days to anthesis (DA), days to heading (DH), and days to physiological maturity (DPM). For each RIL, heading was recorded when atleast 50% of plot spikes had emerged from the booting stage. The grain filling duration (GFD) was calculated as the interval between DPM and DA. Morphological and yield component traits measured were plant height (PH), number of productive tillers per meter (NPTM), number of spikelets per spike (NSS), spike length (SL), number of grains per spike (NGS), thousand-kernel weight (TKW), 100seed volume (HSV), and seed density (SD). Yield traits were assessed as grain yield per plot (GYP, g) and above-ground biological yield per plot (AGBYP, g). The harvest index (HI) was estimated as the proportion of grain yield to the total aboveground biomass (CM 1962). All phenological, morphological, and yield traits were measured randomly selected five plants from each entry, following standard agronomic protocols to ensure accuracy and comparability.

Statistical analysis

Mean values for each RIL under the two sowing environments were computed prior to analysis. A series of univariate and multivariate statistical approaches were applied. Pearson's correlation coefficients were estimated to assess pairwise trait associations (Al-Jibouri et al., 1958) [1] , and correlation matrices were visualized following Wei and Simko (Taiyun Wei 2021) [47] . Path coefficient analysis was conducted to partition the direct and indirect contribution of various traits to yield (Dewey and Lu 1959) [11] . Multiple linear regression was conducted to model yield as a function of component traits (Snedecor and Cochran 1967) [45] , while stepwise regression was used to select the most predictive subset of traits (Draper and Smith 1966) [13] . To explore multivariate patterns, principal component analysis (PCA) was employed to investigate multivariate patterns by reducing data dimensionality and examine covariance structures among traits (Kaiser 1958) [23] . In addition, cluster analysis was performed using Ward's method to group genotypes based on trait similarity. All statistical analyses were conducted using OPSTAT (Sheron et al., 1998) [43] , IBM SPSS Statistics v.27, and R software (packages: stats, factoextra, corrplot) (Kassambara and Mundt 2016) [23] .

Results

All the RILs and their parental lines were evaluated under two different environmental conditions, including a yield potential environment conducive to the expression of stress-free and heat heat-stressed environment where the environment had heat stress with an increase in temperature. Maximum, minimum temperature, and bright sunshine hours (in days) were presented in Table S1 and Fig. 1. An increasing trend of temperature was observed throughout the season, indicating the occurrence of heat stress during post-reproductive phase, especially in latesown trials. Additionally, the maximum number of sunshine hours was recorded in April (7.8 hours), which showed an increase in heat units, shortened the crop cycles, and consequently had adverse effects on grain yield.

Impact of terminal heat stress on grain yield and its component traits

Analysis of variance (ANOVA) for stress-free (a) and heatstressed (b) environments was shown in Table 1. The genotypes mean square (MS) was highly significant for all traits under both environments; however, most traits exhibited lower MS values under the heat-stressed environment, with exception of PH, ET, and NSS. On the other side, the mean square of variance due to blocks was found non-significant for all the traits under both environments except DA and GFD under heat-stressed environments. Table 2 presented the phenotypic variations and distributions of phenological and yield-related traits among RILs for. GYP and AGBYP were the most heat-sensitive traits among all the traits studied. The reduction score values of 37% and 30%, 21% and 23%, 34% and 27% were observed for WH 711 and WH 1021, respectively. GYP was recorded as 1050g and 660g for WH 711, 905g and 710g for WH 1021 under stress-free and heat-stressed environments, respectively. Likewise, GYP and AGBYP in RILs varied from 540-1220g and 217-900g; 1450-3500g and 900-2450g; 30.27-46.00% and 22.15-42.38% under stress-free and heat-stressed environments, respectively. On the contrary, HI for heat-tolerant parent (WH 1021) had a positive response under heat stress conditions with a negative (-2.11) reduction value. Among phenological traits, both parents and population showed an average reduction of 5, 7, and 18 days in terms of DH, DA, and DPM stages under a heat stress environment. Consequently, the highest reduction percent was recorded for grain filling duration with 27.50%, 21.95%, and 24.26% in WH 711, WH 1021, and RILs population, respectively. For spike-related traits, 12-20% of the reduction was recorded for WH711 and RILs, whereas 0.89% to 10% reduction was observed for the tolerant parent WH1021. NSS showed the lowest reduction among them. Substantial phenotypic variation was observed among RILs across both environments, with greater reductions (22.22%) in productive tillers comparatively (Table 2).

Trait associations and yield contributing patterns

The phenotypic correlation revealed that GYP exhibited a positively significant association with spike related traits including SL (0.191** and 0.332**), NSS(0.256** and 0.468**) and NGS(0.286** and 0.349**), yield component traits including AGBYP (0.802** and 0.808**), HI (0.395** and 0.732**) and other agronomic traits such as GFD (0.228** and 0.351**), NPTM (0.434** and 0.642**), TKW (0.230** and 0.286**) and SD (0.395** and 0.431**) under stress-free and heat-stressed environments, respectively ((Fig. 2 and Table S3). The AGBYP also exhibited a significant and positive relationship with NPTM (0.311** and 0.534**), SL (0.280**and 0.238**), NSS (0.232** and 0.307**), NGS (0.299** and 0.269**), and SD (0.208** and 0.332**) under both conditions. Furthermore, HSV showed a significantly positive association with thousand kernel weight (0.594** and 0.601**). On the other hand, other remaining traits, i.e., DH, DA, and DPM, showed negative correlation with GYP (-0.414**, -0.513**; -0.368**, -0.486**, and -0.388, -0.210** respectively) except GFD (0.288** and 0.351**). TKW also showed a negative association with DH (-0.519** and -0.376**) and DA (-0.431** and 0.389**). Path coefficient analysis revealed that the AGBYP contributed maximum (0.8814 and 0.8592) to GYP, followed by HI (0.5415 and 0.6257), TKW (0.2411 and 0.3413), HSV (0.1693), and NPTM (0.1512 and 0.1691) under both environments (Tables 3 and4). Although direct effects on GYP via DA (0.3858) and HSV (0.1693) were positive, the overall correlation was negative among them in the yield potential environment. Likewise, the direct effects on GYP via GFD (-0.9228), SL (-0.0492), NSS (-0.0213), and SD (-0.4276) were negative under heat heat-stressed environment, but they had a positive correlation with GYP. The traits, viz. NPTM (0.2097), SL (0.2800), NSS (0.2040), NGS (0.2925), TKW (0.1460) and SD (0.2267) contributed towards GYP through AGBYP under yield potential environment, whereas NPTM (0.8148), SL (0.6661), NSS (0.9546), NGS (0.9060), AGBYP (0.4220), HI (0.3033), TKW (0.5391), HSV (0.1833) SD (0.6137) contributed towards GYP via DA under heat stressed environment. Under stress-free conditions, GFD (0.1390) and SD (0.1805) contributed towards GYP via HI, whereas NPTM (0.3922, 0.2871), SL (0.2688, 0.2315), NSS (0.2901, 0.3117), NGS (0.2716, 0.2229), and TKW (0.1408, 0.2687) contributed towards GYP via AGBYP and HI. DH (0.2360), DA (0.5692), AGBYP (0.3159), and TKW (0.1708) also contributed positively towards GYP through GFD under heat heat-stressed environment.

Predictive modeling of grain yield

A simple regression model formula explained more than 99% (R 2 ) of the total variations of GYP for both environment, and the remaining less than 1% is probably due to the residual effects. The regression test showed that SD (104.51** and 63.735**), HI (22.46** and 17.113**), and AGBYP (0.349** and 0.309**) contributed significantly to GYP under both conditions, while HSV (37.39*) and NPTM (0.084*) were significant for both environments (Table 5). Multiple regression further predictive following equations for GYP:

Model for stress-free environment GYP: -823.7 -0.205DH -1.803DA + 1.154DPM -1.534GFD -0.063PH + 0.014NPTM + 0.714SL +0.076NSPS + 0.217NGPS + 0.349AGBYPP + 22.461HI -3.245TKW + 37.397HSV + 104.514SD Model for heat heat-stressed environment GYP: -623.76 + 0.455DH + 0.226DA -0.654DPM + 0.970GFD -0.059PH + 0.084NPTM + 0.387SL -0.543NSPS + 0.029NGPS + 0.309AGBYPP + 17.113HI -1.543TKW + 20.496HSV + 63.735SD Stepwise regression analysis revealed that AGBYP (62.74%), HI (36.59%), NGS (0.02%), SD (0.02%) and DA (0.2%) played a significant role in influencing GYP under yield potential environment whereas, four variables, including AGBYP (62.40%), harvest HI (36.80%), SD (0.08%), NPTM (0.02%) were included into the model in form of most important variables under heat stressed environment (Table 6). According to the results, more than 99% of the total variation in GYP was ascribed to these aforementioned variables under both conditions. The rest of the variables were removed from the models due to their low relative contributions. Moreover, the regression analysis predicted following equations for GYP:

Model for stress-free environment GYP: -803.384 + 0.3514AGBYP + 22.7522HI + 0.2672NGS +16.1568SD -0.3364DA Model for heat heat-stressed environment GYP: -563.434 + 0.3100AGBYP +17.01281HI + 24.0709SD + 0.0800NPTM

Principal component analysis of yield-related traits

GYP and its associated traits were mapped on principal component (PC) axes to delineate the proportion of variance explained by each trait (Fig. 3 a andb). Out of the 15 axes of differentiation, only 5 PCs were found with >1 eigen values and cumulatively contributed for more than 70% of the total variations of GYP under both conditions (Supplementary Table S4). The first PC was moderately correlated with GYP (0.37 and 0.40), GFD (0.35 and 0.28), AGBYP (0.24 and 0.31) and HI (0.22 and 0.28) under both conditions, while TKW (0.35) and NPTM (0.26) had significant contribution only under stress-free and heat stressed environments, respectively (Supplementary Table S5). The PC-2 was correlated with SD (0.39 and 0.36) and NPTM (0.31 and 0.26) under both conditions, whereas, GYP (0.32), AGBYP (0.29) and DA (0.22) showed contribution towards this component under stress-free environment. Rest of three PCs cumulatively accounted for more than 25% of total variation with the highest contributions from DPM, GFD, NSS, HI, HSV and SD in different components under both conditions. Acute angles were also observed among GYP, AGBYP, NPTM and SD which depicted positive association among traits and their positive contribution towards both components under both conditions.

Multivariate clustering of RILs based on adaptive trait profiles

On the basis on the relative magnitude of distance, parents and two hundred RILs were grouped into six clusters under both environments (Fig. 4a andb). The intracluster distance ranged from 709.22 (IV) to 1311.39 (II) and 587.11 (VI) to 900.81 (IV) under stress-free and heat-stressed environments and indicating less diversity within cluster, whereas the inter-cluster distance varied from 862.71 (between cluster I and IV) to 1935.37 (between cluster V and VI) and 765.36 (between cluster I and III) to 1412.78 (between cluster II and V) under stress-free and heat-stressed environments, respectively (Table S6 andS7). Cluster III comprised of largest number of RILs (75) followed by cluster IV (33 entries) and cluster II (31 RILs) under a stressfree environment, whereas cluster IV was recorded with the maximum number of RILs (66), followed by cluster VI (44 RILs) and cluster III (39 RILs) under a heat-stressed environment, respectively. Cluster V was comprised of 21 RILs and WH711 (agronomically superior parent) under a stress-free environment, and 9 RILs (15, 20, 33, 34, 59, 74, 100, 106, 117) with WH1021 (heat-tolerant parent) under a heat-stressed environment. This cluster can be categorized as heat-resilient high-yielding, supported by higher mean values for NPTM (138.45 and 131.85), NGS (59. 45 and 49.43), GYP (1042.52 g and 760.05 g), AGBYP (2790.91 g and 2005 g), and SD (1.51 g/cc and 1.50 g /cc), respectively (Table S8 andS9).

Discussion 4.1 Genetic variability and environmental effects

The evaluation of a large RIL population (WH711 × WH1021) under contrasting thermal regimes provided valuable insights into wheat yield performance and adaptive trait expression under terminal heat stress. The highly significant genotypic variance for all studied traits indicated the presence of ample genetic variability, which is essential for identifying promising heatresilient lines. Heat stress clearly harms wheat crop growth, development, total biomass, and potential grain yield (Zhang et al., 2018) [49] . Yield reductions under terminal heat stress can largely be attributed to pollen sterility, impaired assimilate partitioning to developing spikes, and shortened grain filling duration (Ji et [21,44,46] . The extent of trait reduction is influenced by both severity and duration of heat stress (Barnabás et al., 2008) [5] . The meteorological profile confirmed exposure to elevated temperature, extended sunshine hours, and low precipitation during the post-anthesis period, which accentuated early senescence and led to yield losses (Fig. 1). In our study, substantial reductions were recorded for GYP, AGBYP, and HI, with the magnitude of loss varying between the parents and the RILs. The heat-susceptible parent (WH711) exhibited greater reductions compared with WH1021, reinforcing the utility of tolerant parents in broadening the genetic base of heat-resilient wheat (Table 2). The RILs showed a wide spectrum of responses, validating their suitability for dissecting adaptive traits under stress. Terminal heat advanced phenological development (reduced DH, DA, DPM, and GFD) and compressed the grain-filling window, which in turn reduced grain number and weight (Table 2). These results corroborate mechanistic studies that attribute yield loss under reproductive heat to pollen sterility, impaired endosperm development, and shortened assimilate accumulation (Farooq et al., 2014; Zhang et al., 2018) [15,49] .

Trait associations: correlation and path analyses

Correlation analysis demonstrated that the AGBYP consistently exhibited the strongest positive association with grain yield in both environments (Table S3). This is expected, as biomass accumulation underpins assimilate availability for grain filling (Gaju et al., 2009) [17] . Other yield contributing traits, such as NPTM, HI, SL, and NGS, also correlated positively with yield (Fig. 2). These results were corroborated with earlier studies (Leilah and Al-Khateeb 2005) [28] . However, phenological traits such as DH, DA, and DPM generally showed negative associations with yield, likely reflecting an escape mechanism. whereby early-maturing genotypes partially avoid heat injury during grain filling. A negative association between DH and grain yield per meter has been reported earlier (Mansouri et al., 2018) [29] . GFD also had a significant positive correlation with GYP. Heat-tolerant wheat genotypes had been reported with 38% longer GFD as compared to heat-susceptible genotypes (Ihsan et al., 2016) [20] . Moreover, the grain filling rate is more imperative than the grain filling period (Borrill et al., 2015) [6] . The stressed plants may also have a higher grain filling rate, which, when combined with a shorter grain filling period, causes improper assimilate translocation and, as a result, lower yield production (Ihsan et al., 2016) [20] . Interestingly, plant height showed contrasting associations across environments (negative under optimal, positive under heat), implying selection must consider the G×E context (Moghaddam 1998) [31] . Under stress, slightly taller plants may sustain higher biomass and lower plant area index to minimize moisture demand and reduce moisture loss through transpiration (Khadka et al., 2020) [25] . The differential response of component traits with respect to grain yield could be attributed to environmental factors affecting plant growth (Asseng et al., 2002) [3] . Path analysis further highlighted the direct importance of AGBYP and HI on grain yield, corroborating earlier reports that emphasized biological yield and partitioning efficiency as critical determinants of productivity under stress (Leilah and Al-Khateeb 2005) [28] . Traits such as NPTM, NGS, TKW, and SD also exerted appreciable direct or indirect contributions, making them reliable selection indices. The positive relation between PH and GYP under a heat-stressed environment indicated its direct significant role in the determination of biological yield. Although the direct effects of GFD, SL, NSS, and SD on GYP were observed to be negative under stressed conditions but overall, they had a positive correlation with GYP, suggesting that these traits were strongly affected by associated traits. The residual effect was very low, indicating that the traits under study are sufficient for explaining the variability in grain yield.

Predictive models and key determinants of yield

Regression and stepwise analyses revealed that AGBYP, HI, and SD were the most consistent predictors of grain yield under both environments, while NPTM, HSV, and TKW showed environment-specific effects. These predictors cumulatively explained over 99% of yield variation, underscoring their robustness for breeding applications (Tables 5 and6). The overall results reflected the importance of AGBYP, HI, SD, and NPTM four variables under heat stress. These results reinforce the value of combining biomass accrual and partitioning efficiency with kernel traits in selection indices for heat resilience. Previous studies also indicated that the biological yield per unit, HI, thousand grain weight, and number of spikelets/m 2 are major contributors to variations in grain yield under a heat-stressed environment (Leilah and [28,32,41,46] .

PCA insights, trait integration, and RILs selection

PCA reduced dimensionality to five major components (>70% variance) (Table S4 andS5) and highlighted coherent trait groupings: biomass and HI components aligned closely with GYP, while other components captured phenology and kernel attributes. Traits, DPM, GFD, NSS, GYP, HI, TKW, and SD under a stress-free environment and PH, SL, GYP, HSV, and SD under heat heat-stressed environment are the most important variables observed in different principal components, indicating a wide scope of improvement for heat tolerance. Acute vector angles among GYP, AGBYP, NPTM, HI, and NSS. NGS and SD again confirmed concerted positive effects on yield (Fig. 3). Previous studies also listed grain filling duration, thousand kernel weight, grain yield, biological yield and harvest index as the most important traits for yield contribution (Leilah and Al-Khateeb 2005; Munjal 2017; Muñoz-Sanz et al., 2020) [28,33,34] . Multivariate clustering grouped both parental genotypes into different clusters under both environments, which evidently showed a significant source of genetic diversity and recombination phenomena in the segregating population. A total of six clusters were made with a lower intracluster distance and a higher intercluster distance, showing the potential opportunity of hybridization among diverse clusters to produce promising segregants for heat tolerance (Table S6). Nine RILs were grouped into cluster V along with WH1021, clearly exhibiting superior heat resilience and higher mean values for yield-related traits (GYP, AGBYP, NPTM, NGS, and SD) (Table S7-S9). Out of which, RIL 15, 20, and 100 were classified in the same cluster with an agronomically superior parent (WH 711) under a stress-free environment and a heat-tolerant parent (WH 1021) under a heat-stressed environment (Fig. 4). These results were matched with previous studies (Elahi et al., 2020; Mecha et al., 2017; Rajshree 2018) [14,30,37] .

Breeding implications and selection strategy

Collectively, results indicate that selection for (i) robust biomass production (AGBYP), (ii) efficient partitioning to grain (HI), and (iii) compact, heavy kernels (SD, TKW) will most effectively increase yield under terminal heat (Table S10). Number of productive tillers (NPTM) also emerges as an important secondary target under heat stress. We recommend a trait-based selection index combining AGBYP, HI and SD (with environment-specific weightings) rather than sole reliance on plot yield. The identified heat-resilient cluster and specific RILs (e.g., RILs 15, 20, 100) provide immediate breeding material for crossing and further validation in multi-location trials. -0.2929 0.3045 -0.0566 -0.009 0.0907 0.0356 -0.0175 -0.0063 -0.0023 0.2357 0.1081 0.0462 -0.0396 -0.0084 0.388 NPTM -0.7243 0.8148 -0.0645 -0.1327 -0.0218 0.1691 -0.0087 -0.0037 -0.0021 0.3922 0.2871 0.0174 0.1264 -0.2095 0.64 SPL -0.5572 0.6661 -0.0509 -0.1296 -0.0323 0.0123 -0.0492 -0.0095 -0.0024 0.2688 0.2315 0.1137 -0.0824 -0.0486 0.33 NS/S -0.753 0.9546 -0.0444 -0.2092 -0.0267 0.012 -0.0219 -0.0213 -0.0056 0.2901 0.3117 0.1019 0.0178 -0.1545 0.451 NG/S -0.7661 0.906 -0.06 -0.1849 -0.0124 0.0085 -0.0072 -0.0071 0.0167 0.2716 0.2229 0.0492 0.0627 -0.1498 0.35 AGBY -0.1695 0.422 -0.0376 -0.3393 -0.046 0.0396 -0.0154 -0.0072 -0.0053 0.8592 0.1834 0.0559 0.1027 -0.2123 0.83 HI -0.0644 0.3033 -0.0711 -0.2556 -0.0157 0.0317 -0.0182 -0.0106 -0.006 0.2518 0.6257 0.1466 -0.0142 -0.1712 0.732 TKW -0.1782 0.5391 0.0091 -0.4619 -0.0123 0.0035 -0.0164 -0.0064 -0.0024 0.1408 0.2687 0.3413 -0.3293 -0. 9, 10, 11, 12, 13, 16, 18, 19, 21, 23, 24, 26, 27, 29, 30, 31, 41, 43, 46, 50, 52, 53, 54, 65, 67, 68, 73, 79, 80, 81, 90, 91, 92, 95, 97, 98, 104, 105, 107, 112, 113, 115, 119, 121, 132, 142, 147, 154, 158, 159, 160, 161, 164, 172, 173, 176, 177, 181, 182, 184, 189, 190, 191, 196, 198, 199 V Stress-free 1279. 36 22 10, 14, 15, 16, 20, 55, 60, 70, 71, 73, 82, 84, 93, 95, 100, 110, 137, 164, 177, 195, 197

Acknowledgement

AcknowledgementThe authors sincerely acknowledge the Department of Genetics and Plant Breeding, as well as the field staff of the Wheat and Barley Section, CCS Haryana Agricultural University, Hisar, for their valuable support and assistance during the study.

References

1. Al-Jibouri HA, Miller P, Robinson H. Genotypic and environmental variances and covariances in an upland cotton cross of interspecific origin. Agronomy Journal. 1958;50:633-636.
2. Allard RW. Principles of plant breeding. New York: John Wiley and Sons, Inc.; 1960. p. 485.
3. Asseng S, Foster IAN, Turner NC. The impact of temperature variability on wheat yields. Global Change Biology. 2011;17:997-1012.
4. Asseng S, Turner NC, Ray JD, Keating BA. A simulation analysis that predicts the influence of physiological traits on the potential yield of wheat. European Journal of Agronomy. 2002;17:123-141.
5. Barnabás B, Jäger K, Fehér A. The effect of drought and heat stress on reproductive processes in cereals. Plant, Cell & Environment. 2008;31:11-38.
6. Borrill P, Fahy B, Smith AM, Uauy C. Wheat grain filling is limited by grain filling capacity rather than the duration of flag leaf photosynthesis: A case study using NAM RNAi plants. PLoS One. 2015;10:e0134947.
7. Bramel PJ, Hinz PN, Green DE, Shibles RM. Use of principal factor analysis in the study of three stem termination types of soybean. Euphytica. 1984;33:387-400.
8. CM D. In search of yield. J Aust Inst Agric Sci. 1962;28:171-178.
9. Dabi A, Mekbib F, Desalegn T. Estimation of genetic and phenotypic correlation coefficients and path analysis of yield and yield contributing traits of bread wheat (Triticum aestivum L.) genotypes. International Journal of Natural Resource Ecology and Management. 2016;1:145-154.
10. Devi S, Singh V, Kumar M, Yashveer S, Kumar R, Sharma S, Chawla R, Sapna Rani K, Redhu M. Assessing wheat genotype performance under heat, drought and combined stresses using multiple stress indices. Euphytica. 2024;220:164.
11. Dewey DR, Lu K. A correlation and path-coefficient analysis of components of crested wheatgrass seed production. Agronomy Journal. 1959;51:515-518.
12. Dias AS, Bagulho AS, Lidon FC. Ultrastructure and biochemical traits of bread and durum wheat grains under heat stress. Brazilian Journal of Plant Physiology. 2008;20:323-333.
13. Draper N, Smith H. Allied Regression Analysis. New York: Wiley; 1966.
14. Elahi T, Pandey S, Shukla R. Genetic variability among wheat genotypes based on agro-morphological traits under restricted irrigated conditions. Journal of Pharmacognosy and Phytochemistry. 2020;9:801-805.
15. Farooq M, Hussain M, Siddique KHM. Drought stress in wheat during flowering and grain-filling periods. Critical Reviews in Plant Sciences. 2014;33:331-349.
16. Anjum FA, Abdul Wahid AW, Farrukh Javed FJ, Muhammad Arshad MA. Influence of foliar applied thiourea on flag leaf gas exchange and yield parameters of bread wheat (Triticum aestivum) cultivars under salinity and heat stresses. 2008.
17. Gaju O, Reynolds MP, Sparkes DL, Foulkes MJ. Relationships between large-spike phenotype, grain number, and yield potential in spring wheat. Crop Science. 2009;49:961-973.
18. Hassan MU, Chattha MU, Khan I, Chattha MB, Barbanti L, Aamer M, Iqbal MM, Nawaz M, Mahmood A, Ali A. Heat stress in cultivated plants: Nature, impact, mechanisms, and mitigation strategies-A review. Plant Biosystems-An International Journal Dealing with All Aspects of Plant Biology. 2021;155:211-234.
19. Hedhly A, Hormaza JI, Herrero M. Global warming and sexual plant reproduction. Trends in Plant Science. 2009;14:30-36.
20. Ihsan MZ, El-Nakhlawy FS, Ismail SM, Fahad S. Wheat phenological development and growth studies as affected by drought and late season high temperature stress under arid environment. Frontiers in Plant Science. 2016;7:795.
21. Ji X, Shiran B, Wan J, Lewis DC, Jenkins CLD, Condon AG, Richards RA, Dolferus R. Importance of pre-anthesis anther sink strength for maintenance of grain number during reproductive stage water stress in wheat. Plant, Cell & Environment. 2010;33:926-942.
22. Kaiser HF. The varimax criterion for analytic rotation in factor analysis. Psychometrika. 1958;23:187-200.
23. Kassambara A, Mundt F. Factoextra: Extract and visualize the results of multivariate data analyses. CRAN: Contributed Packages. 2016.
24. Kaur V, Behl R. Grain yield in wheat as affected by short periods of high temperature, drought and their interaction during pre-and post-anthesis stages. Cereal Research Communications. 2010;38:514-520.
25. Khadka K, Earl HJ, Raizada MN, Navabi A. A physiomorphological trait-based approach for breeding drought tolerant wheat. Frontiers in Plant Science. 2020;11:715.
26. Kumbhar M, Larik A. Effect of plant density on the growth and yield potential of durum and bread wheat cultivars. 1983;32-35.
27. Langridge P, Reynolds M. Breeding for drought and heat tolerance in wheat. Theoretical and Applied Genetics. 2021;134:1753-1769.
28. Leilah AA, Al-Khateeb SA. Statistical analysis of wheat yield under drought conditions. Journal of Arid Environments. 2005;61:483-496.
29. Thell. Ssp. Turgidum conv. Durum (Desf.) Mackey] under south Mediterranean growth conditions: Stepwise and path analyses. International Journal of Agronomy. 2018;2018.
30. Mecha B, Alamerew S, Assefa A, Assefa E, Dutamo D. Genetic diversity based on multivariate analysis for yield and its contributing characters in bread wheat (Triticum aestivum L.) genotypes. Agri. Res. Tech. 2017;8:555748.
31. Moghaddam M. Genetic variation for and interrelationships among agronomic traits in landraces of bread wheat from southwestern Iran. J Genet Breed. 1998;52:73-81.
32. Msundi EA, Owuoche JO, Oyoo ME, Macharia G, Singh RP, Randhawa MS. Identification of bread wheat genotypes with superior grain yield and agronomic traits through evaluation under rust epiphytotic conditions in Kenya. Scientific Reports. 2021;11:1-11.
33. Munjal R. Genetic diversity in bread wheat for heat tolerance. Ekin Journal of Crop Breeding and Genetics. 2017;3:60-78.
34. Muñoz-Sanz JV, Zuriaga E, Cruz-García F, McClure B, Romero C. Self-(in)compatibility systems: Target traits for crop-production, plant breeding, and biotechnology. Frontiers in Plant Science. 2020;11:195.
35. Porter JR, Gawith M. Temperatures and the growth and development of wheat: A review. European Journal of Agronomy. 1999;10:23-36.
36. Qaseem MF, Qureshi R, Shaheen H. Effects of pre-anthesis drought, heat and their combination on the growth, yield and physiology of diverse wheat (Triticum aestivum L.) genotypes varying in sensitivity to heat and drought stress. Scientific Reports. 2019;9:1-12.
37. Rajshree SS. Assessment of genetic diversity in promising bread wheat (Triticum aestivum L.) genotypes. International Journal of Current Microbiology and Applied Sciences. 2018;7:2319-7706.
38. Redhu M, Singh V, Yashveer S, Sharma K, Kumar S. Unravelling the genetic basis of terminal heat tolerance and yield related traits in bread wheat (Triticum aestivum L.). Indian Journal of Genetics and Plant Breeding. 2024;84:523-531.
39. Redhu M, Singh V, Kumari A, Munjal R, Nimbal S, Niwas R, Yashveer S, Verma S, Langaya S, Devi S. Evaluation of physio-biochemical traits in bread wheat RILs for terminal heat stress. Cereal Research Communications. 2025:1-12.
40. Reynolds MP, Braun HJ, Pietragalla J, Ortiz R. Challenges to international wheat breeding. Euphytica. 2007;157:281285.
41. Saed-Moucheshi A, Pessarakli M, Heidari B. Comparing relationships among yield and its related traits in mycorrhizal and nonmycorrhizal inoculated wheat cultivars under different water regimes using multivariate statistics. International Journal of Agronomy. 2013;2013.
42. Semenov MA. Impacts of climate change on wheat in England and Wales. Journal of the Royal Society Interface. 2009;6:343-350.
43. Sheron O, Tonk D, Kaushik L, Hasija R, Pannu R. Statistical software package for agricultural research workers, OPSTAT. 1998.
44. Sinclair TR, Jamieson PD. Grain number, wheat yield, and bottling beer: An analysis. Field Crops Research. 2006;98:60-67.
45. Snedecor GW, Cochran WG. Statistical Methods. 7th ed. Ames, Iowa, USA: Iowa State University Press; 1967.
46. Sobhaninan N, Heidari B, Tahmasebi S, Dadkhodaie A, McIntyre CL. Response of quantitative and physiological traits to drought stress in the SeriM82/Babax wheat population. Euphytica. 2019;215:1-15.
47. Wei TY. R package 'corrplot': Visualization of a Correlation Matrix. 2021.
48. Zhang J, Zhang S, Cheng M, Jiang H, Zhang X, Peng C, Lu X, Zhang M, Jin J. Effect of drought on agronomic traits of rice and wheat: A meta-analysis. International Journal of Environmental Research and Public Health. 2018;15:839.