The present investigation was carried out with 34 diverse wheat genotypes (Table 1), of which 18 were landraces collected across the Bundelkhand region (Uttar Pradesh); seven commercial durum cultivars were obtained from Indian Institute of Agricultural Research (IARI), Regional Research Station, Indore (MP), which are recommended for cultivation in Central Zone (CZ); seven durum cultivars were obtained from Punjab Agricultural University (PAU), Ludhiana (Punjab) representing Northwest Plane Zone (NWPZ), and two bread wheat cultivars were from IIWBR Karnal (Haryana). The landraces were collected from six districts (Banda, Mahoba, Chitrakoot, Jhansi, Hamirpur, and Lalitpur) of the Bundelkhand region, conserved by farmers for traditional cultivation. The local wheat collections possessed good adaptability to rainfed farming and are preferred by farmers of this region. The experiments were conducted for two consecutive crop seasons 2020–2021 and 2021–2022 at two locations (Table 2):

a. Crop Research Farm, Banda University of Agriculture and Technology Banda, UP (25.5269°N latitude and 80.3418°E longitude)

b. Experimental Farm, Krishi Vigyan Kendra, Jhansi (25.5347° N latitude, 78.5742° E longitude)

Bundelkhand region in Uttar Pradesh state of India is known for extreme heat and drought stress, which severely affect the growth, development, and yield of wheat crop. This region frequently experiences extreme temperature, usually 50°C during summer and 5°C in the winter season. The summer season (April–June) characterized by strong heat wave and dusty, gusty, and dry wind blowing over the entire region, which often lead to fatal heat strokes. The average rainfall of this region is 800–900 mm per annum; however, most of rainwater is lost as runoff due to undulated topography. During winter season (November–March), few rain showers are received occasionally, which supplies inadequate moisture to crops. The whole Bundelkhand region is known for predictable drought and heat stress during the anthesis to grain filling stage of winter season crops.

The experiments were conducted under three environments, viz., the optimum environment (OE), heat stress environment (HSE), and the heat–drought environment (HDSE) (Table 2). The timely sown (OE) and late sown experiments (HSE) were given irrigation (flood irrigation) at the required field capacity throughout the crop period, whereas in the late sown restricted irrigation, no irrigation was given during the reproductive phase (HDSE). The genotypes were planted in a two-row plot of 2 m length with two replicates following Alpha lattice design for each condition. The distance of 5 m was kept between the trials to check the moisture percolations in the drought experiment. The sowing was done with the modified precision numeric seed planter. The sowing and harvesting dates for each environment for both the years is given in Table 2. The crop was supplemented with 120 kg/ha nitrogen and 60 kg/ha phosphorus. Total N was split into three doses of 40 kg/ha each and applied at the three-leaf stage, tillering, and again at heading stage. The recommended agronomic practices of weeding and crop protection were applied to raise healthy crop plants.

The data on yield and yield-contributing traits were recorded for each genotype in all three environments during 2020–2021 and 2021–2022.

The Banda and Jhansi locations are typically represented by hot and dry environment of Central Zone (CZ) in India. The weather data were recorded at the Meteorological Observatory Unit (MOU) of the University at Banda and KVK, Jhansi for two years (Figure 1). The maximum temperature was recorded more than 35°C at the grain filling stage during 2020–2021 and 2021–2022 at the Banda and Jhansi locations. At Banda location, the cropping season during 2020–2021 was extended due to 10 mm rainfall in the month of March, while during 2021–2022, the temperature in the last week of February increased rapidly exposing the crop to terminal heat. During 2020–2021, maximum precipitation (958 mm) was received from June to October during 2020–2021, while 929 mm of rainfall was recorded in the monsoon season during 2021–2022. During the month of February 2021, very less rainfall of 10.57 mm was received in the month of February 2021, whereas rainfall of 104 mm was received at early vegetative stage of the crop in the first week of January 2022. During 2022, the rainfall was received only in the early vegetative stage, and the crop was exposed to drought stress during reproductive phase. The OE trials were irrigated at both locations to avoid the effect of heat and drought stress. The relative humidity (RH) at the grain filling stage during 2020–2021 was high due to rain in the month of March, while RH was low during 2021–2022 at the time of grain filling.

Combined analysis of variance using linear mixed model (LML) indicated significant effects of environment, location, year, and their interaction on most of the measured traits (Table 3A). The genotypes exhibited highly significant variation for all traits except TKW and HI. The pooled analysis of variance indicated significant differences in all interactions; thus, a separate analysis of variance was performed for each location. The individual location ANOVA analysis also indicated significant differences between genotypes, and Gen × Env interaction was also significant for most of the traits (Tables 3B, C).

Phenotypic trait variation was observed across 34 genotypes under OE, HSE, and HDSE, indicating considerable genetic variation in performance across environments. Some genotypes exhibited higher reduction, while other genotypes managed to grow and remain relatively healthy. The performance of genotypes under combined heat–drought stress was severely affected compared to heat stress alone at Banda and Jhansi. The mean of 11 agro-morphological traits of the wheat lines under optimal and stress conditions is presented in Table 4 and Figure 2. The highest heritability was reported for PH (98%) followed by TKW (95%), whereas least was reported for NSS (45%).

Heat stress significantly affected the performance of all the traits. At Banda location, highest reduction was reported in NT (24%) followed by SY (20%), DM (18%), and BY (16%) as compared to optimum environment, whereas least reduction of 4% was reported in PH and NSS (Table 5). Similarly at Jhansi, highest reduction was reported in SY (44%) followed by (34%), and DM (22%) as compared to optimum environment, whereas least reduction of<5% was reported in PH and NSS (Table 5). The maximum yield reduction of 48% was recorded in genotype PDW-233 at Banda, whereas yield reduction of 66% was reported in Local-5 at Jhansi location. The genotype Local-4 recorded minimum SY reduction of 5.6% at Banda, while a minimum 15% SY reduction was recorded in the HI 8627 genotype at Jhansi.

Combined heat–drought stress also affected agro-morphological traits across years at both locations. At Banda, highest reduction was recorded for BY (37%) followed by SY (36%) and NT (27%), whereas the least affected trait was NSS, which showed a reduction of 8% under combined heat–drought stress. Spike yield (59%) showed the highest reduction at Jhansi followed by (57%) and DM (25%), whereas NSS recorded the minimum reduction of 7% (Table 5). The PH and NSS traits were the least affected traits under heat–drought stress at both locations. Highest SY was reported for genotype Local-17 (985 g) at Banda, whereas the genotype PDW-274 recorded with a maximum SY of 490 g at Jhansi. Genotype Local-3 and HI8627 recorded minimum SY reduction of 19% and 35% at Banda and Jhansi, respectively. Under HDSE, DFF significantly reduced from 88 days to 74 days. Genotypes Local-17 and Local-13 were earliest in heading at Banda, whereas genotypes Local-3, Local-9, HI-8777, Local-13, Local-14, and DBW 187 were earliest in heading at the Jhansi.

Among all genotypes, Local-17 and HI 8802 performed well under HSE and HDSE with higher yield at the Banda location. At Jhansi, genotypes DBW 187 and HI 8777 performed well under HSE with higher yield, while under HDSE genotype, PDW 274 and HI 8777 were better performing genotypes.

Stress indices (STI, SSI, MP, TOL, and MSTI) were estimated based on seed yield under stress environments (HSE and HDSE) and OE conditions. Based on the STI and SSI, the genotypes were grouped into tolerant and sensitive to heat and combined drought-heat stress conditions (Supplementary Tables S1–S4).

Based on STI, five genotypes local-17 (2.16) and PDW 274 (1.88), HI-8802 (1.83), HI-8713 (1.54), and Local-5 (1.32) performed better under heat stress environment at Banda (Table 7). Similarly, five genotypes DBW 187 (1.77), HI-8777 (1.69), Local-4 (1.53), Raj 4120 (1.48), and PDW 274 (1.40) were tolerant at Jhansi (Table 7). Thus, these genotypes are considered as heat stress tolerant. Based on SSI, genotypes PDW 233 (2.05) and Local-16 (1.96) were identified as heat susceptible at Banda (Supplementary Table S1), whereas at Jhansi, the maximum value of SSI was exhibited in Local-5 (1.34) and PDW 233 (1.33) (Supplementary Table S3); thus, these genotypes are considered as heat stress susceptible. The maximum value of TOL was recorded in PDW 233 (437) and Local-16 (365) in the HSE at Banda (Supplementary Table S1), whereas genotypes Local-5 (522) and PDW 233 (516) had a maximum value of TOL under HSE at Jhansi (Supplementary Table S3).

Under HSDE, highest values for STI were recorded in genotypes Local-17, PDW 274, HI-8802, HI-8713, and HI-8759 at Banda (Table 8). At Jhansi, genotypes PDW 274, DBW 187, HI-8777, Raj 4120, and Local-4 had the highest value of STI under HDSE. Hence, these genotypes can be considered as heat–drought stress tolerant. Based on SSI, the higher value was recorded in genotypes PDW 233 and PDW 291 at Banda (Supplementary Table S2), whereas at Jhansi, genotypes PDW 233 and DWL 5023 recorded the highest values (Supplementary Table S4). The maximum values of TOL recorded were in genotypes PDW 233 and PDW 291 at Banda, whereas genotypes PDW 233 and Local-5 recorded maximum values at Jhansi.

In summary, genotypes Local-17 and PDW 274 were selected to be heat and combined heat and drought tolerant along with yield at Banda. Similarly, genotypes DBW 187, HI 8777, and PDW 274 performed better at Jhansi under both environments compared to the rest of the genotypes.

The weather conditions of the Bundelkhand region are hot and dry during the reproductive phase for the cool season crops like wheat. The temperature and rainfall pattern are consistently favoring experiments focused on screening germplasm for abiotic stresses, more specifically heat and drought stress. The meteorological data clearly demonstrate that more than 95% rainfall was received from June to September in both cropping seasons at Banda. Similarly, at the Jhansi location during monsoon season, the major part of the total rainfall was received. Therefore, at both locations, weather conditions were ideal for challenging plants to stress conditions (Figure 1). During crop season, very less amount of precipitation was received in the reproductive phase, while crop water requirement in the vegetative phase was maintained by irrigation. The temperature in HSE and HDSE experiments was >35°C during the anthesis period, and it reached up to 40°C during the grain filling stage, although experiments under optimum environments were raised under non-stressed conditions.

Our result showed that combined stress penalty on seed yield at Jhansi was more severe (59% under HDSE as compared to 44% under HSE). The maximum yield penalty (36%) was recorded in HDSE, whereas yield penalty of 20% was estimated under heat stress at Banda location (Table 4). The maximum yield for each genotype was harvested at Banda due to the long duration of crop compared to Jhansi. The climatic conditions particularly rainfall and temperature patterns and recorded yield penalty due to stress environments at both locations reflect that Bundelkhand is an appropriate location for terminal heat and drought stress evaluation.

The combined stress is more detrimental on plant phenological traits than heat stress alone at both the locations (Table 5). Several reports have shown that the combination of heat and drought stress is common in dry and semi-dry regions worldwide and reported extensive yield losses. The results of the present study are in agreement with the study of Yashavanthakumar et al. (2021), which reported the confounding effect of heat and drought resulting in yield loss (56%), followed by drought (41.1%) and least affected by heat stress alone (4.8%) over the control environment. Results of the present study also showed that heat and heat–drought stress adversely affected all measured trait significantly, which is also reported in other wheat crop experiments (Hussain et al., 2021; Yashavanthakumar et al., 2021) and Durum wheat (Aberkane et al., 2021; Ayed et al., 2021; Laddomada et al., 2021). Drought stress at pre- and post-anthesis stage significantly reduced grain yield (Y), spike length (SL), number of grains spikes-1 (NGS) and thousand kernel weight (TKW), while kernel abortion (KA) was increased (Shokat et al., 2020). Our study is in agreement with Qaseem et al. (2019), who reported that the combined heat and drought stress is a more devastating factor to plant development as compared to any stress alone. However, there was less impact of stress conditions on PH and NSS compared to optimum environments. Although, both traits are highly stable and were less affected by the environmental factors. This finding is contrary to previous studies, which have suggested that stress has a significant impact on plant height reduction (Sattar et al., 2020). In conclusion, the genotypes Local-17, PDW 274, HI 8802, and HI 8713 have the highest yield potential under OE at the Banda location, whereas DBW 187, HI 8777, Local-4, and PDW 274 were superior at Jhansi location.

The association among all measured agro-morphological traits with seed yield was studied using pooled data over environments under OE, HSE, and HSDE (Tables 6A, B). The NGS and TKW directly positively associated with seed yield across the normal and stress environments, indicating that the genotypes were able to maintain both the traits under stress conditions. The selection of the genotypes under stress conditions is most reliable taking NGS and TKW as targeted traits (Dodig et al., 2012; Sareen et al., 2014). There was a significantly negative correlation of PH (−0.44*) and NT (−0.61) with SY under HSE, whereas SY had a significantly negative correlation with DFF (−0.53**) and DM (−0.61**). The traits DFF and DM indicated that heat and drought escape mechanisms as short duration genotype are preferred for drought and high-temperature tolerance (Reynolds et al., 2007; Punia et al., 2011; Mondal et al., 2013; Gupta et al., 2020). The days to maturity and early heading escapes the stress during the critical reproductive phase and increases productivity before the onset of heat and drought stress (Blum, 2010). NGS, NSS, and TKW are the main traits contributing towards yield under stress environments, whereas BY is a key trait to be considered under combined heat and drought stress.

The regression analysis indicated that NGS is the key trait that is contributing towards grain yield under stress environments (Figure 3). High temperature reduces grain weight (Wardlaw et al., 1980) due to reduction in both duration and rate of grain filling and high respiration rate (Tashiro and Wardlaw, 1990); however, yield reduction is primarily due to reduction in grain weight (Guttieri et al., 2001). Tyagi et al. (2003) also reported that grain weight per spike could be used as a measure of heat tolerance. Similarly in the present study, TKW is found to be significantly contributing towards SY under heat stress.

In the present study, analysis of variance revealed the significant difference among 34 wheat genotypes for all measured traits. The significant variation in genotype, environment, year, and location indicated variable performance of genotypes across environments. The genotype × environment interaction for SY and its associated trait shows the possibilities of selecting genotypes with less yield penalty under stress environments. The significant effect of different environments on the performance of genotypes was reported by Mohammadi et al. (2010) and Talebi et al. (2010) in bread wheat.

Stress selection indices STI, SSI, MP, TOL and MSTI, are widely used for the identification of tolerant genotypes under stress environments. The selection of heat- and drought-stress tolerant genotypes was done using stress tolerance index in wheat was adopted in several studies (Puri et al., 2015; Aberkane et al., 2021a; Poudel et al., 2021). The STI is a more appropriate index to identify tolerant genotypes in the panel of germplasm. Earlier many researchers used STI to screen germplasm in chickpeas (Erdemci, 2018; Shabani et al., 2018). The genotype having higher values of STI, MP, and MSTI indicates their superior performance under both normal and stress environments. Additionally, the genotypes selected based on STI will have higher stress tolerance and grain yield, as it involves the stress tolerance intensity and GMP.

MP favors higher yield potential and lower stress tolerance, and the selections based on MP generally increase the average performance of genotypes in both stress and non-stress environments and fails to distinguish between stress-tolerant and high-yielding genotypes.

Larger values of TOL represent more sensitivity to stress; thus, smaller values are favored for selecting tolerant genotypes. The smaller the values of SSI and TOL, the greater the value of tolerance but fails to distinguish between genotypes performing better under both environments and performing better only in under stress. Selection based on TOL and SSI favors those genotypes having lower yield potential under normal and high yield potential under stress environments. Thus, TOL also fails to distinguish between tolerant and high-yielding genotypes.

The STI is used for the identification of wheat genotypes that produce higher yields under non-stressed and stressed environments. Genotypes having higher values of STI are tolerant and have high yield potential, as it involves the stress tolerance intensity and GMP. Thus, STI is the best one to distinguish between genotypes performing superior/poor either under stress or non-stress environments or under both environments.

In our screening experiments under HSE and HDSE at Banda, genotypes Local-17, PDW 274, HI 8802, and HI 8713 have the highest values of STI, MP, and MSTI; therefore, these are considered as tolerant to heat and combined heat–drought stress (Table 7). Similarly at Jhansi, DBW 187, HI 8777, Local-4, Raj 4120, and PDW 274 showed the highest value for STI, MP, MSI, and MSTI under HSE and HDSE. Our result is in support of the finding of Kamrani et al. (2017) for the selection of higher-yielding and heat tolerant genotypes based on MP and STI. The genotype having a higher value of TOL and SSI indicates more reduction in yield under HSE and HDSE, thus indicating stress susceptibility (Tables 7, 8) (Nouri et al., 2011; Poudel et al., 2021). The smaller the values of SSI and TOL, the greater the value of tolerance but fails to distinguish between genotypes performing better under both environments and performing better only in under stress. The selection based on TOL and SSI favors those genotypes having lower yield potential under normal and stress environments. The susceptibility indices TOL and SSI were less effective than tolerance indices STI, MP, and MSTI to select suitable lines for drought stress (Kamrani et al., 2017).

In our study, genotype PDW 274 showed higher values of stress indices STI, MP, and MSTI and ranked in the top 5 genotypes across the environment at both locations (Tables 6, 7). The other genotypes PDW 233 and PDW 291 showed the highest stress susceptibility index (SSI) across the environments at both locations. The reliability of STI for the selection of stress-tolerant genotypes with high seed yield under normal and stressed conditions further confirms the consistency of this index in selecting for high productivity under stress conditions (Fernandez, 1992).