Expression and association of quantitative traits of wheat (Triticum aestivum L.) genotypes under different wheat growing environments

Abstract

Climate-resilient wheat genotype plays a crucial role in food and nutritional security of the world. The production of wheat is mainly limited by abiotic stresses such as heat stress and drought. To identify the most appropriate trait for the selection of high yielding genotypes under heat stress (HS) and heat drought (HD) environment, a field experiment was conducted in the western region of Nepal during the wheat growing season of 2021–22 to 2022–23 comprising 20 wheat genotypes under alpha lattice design. The additive main-effect multiplicative interaction (AMMI) and combined analysis of variance (ANOVA) showed environment had the most substantial effect on the expression of the quantitative traits studied with 24 and 48% yield reduction under HS and HD environments, as compared to irrigated. PCA extracted, days to booting (DTB), days to heading (DTH), days to anthesis (DTA), booting-anthesis duration (BtoA), spikelets per spike (NSPS) under irrigated; DTB, DTH, DTA, 10 spike weight (TSW), thousand kernel weight (TKW) under HS; and DTB, BtoA, spike length (SL), TSW under HD environment. Phenotypic correlation and path analysis revealed, to get high yield, selection can be done with DTB, DTH, DTA, booting-heading duration (BtoH), BtoA, plant height (Ph), spikes per meter square (NSPMS), SL, and TKW. Early booting and heading genotypes with longer spikes, more tillers, grains per spike, and seed weight benefits all environments. Extended booting-heading and booting-anthesis duration with shorter heading-anthesis duration are advantageous as well. Taller genotypes yield more under HS and HD environment whereas shorter genotypes are productive under irrigated.

Keywords:

• climate resilient
• heat stress
• drought
• earliness
• taller

1. Introduction

Hunger and malnutrition continue to be pressing global challenges, with significant implications for food security and human health. According to the FAO and United Nations, approximately 12.5% (1 billion) of the world’s population suffer from hunger each day, with 6.1% (488 million) suffering from extreme malnutrition, 5.77% (462 million) suffering from underweight problems, and 0.56% (45 million) from wasting. Furthermore, severe micronutrient deficiency affects around 17% (1.36 billion) of the global population (FAOSTAT, 2022). UN report shows, around 22% (149 million) of children below age five are stunted and 45% of child death below age five are associated with undernutrition in the world (FAOSTAT, 2022; Farcas et al., 2021; FAO, 2018). The Sustainable Development Goals (SDGs) established by the United Nations aim to address these issues by promoting food security, ending hunger, and improving human nutrition (Ferrari et al., 2020; Sachs et al., 2022). To achieve these goals, it is crucial to enhance the production and productivity of the most important cereal crop in the world, wheat (Shewry & Hey, 2015).

A steady trend of rising air temperatures from year to year leads to more extreme climate events. Drought is a steadily developing climatological phenomenon that covers entire countries, turning into an environmental problem not only of regional significance, but also of global scale. According to (IPCC, 2021), in 2020, the average annual temperature change in more than 150 countries was at least 1.0°C higher than the 1951–1980 average. A pronounced trend in temperature and precipitation changes, a decrease in moisture, and an increase in climate aridity invariably entail changes in ecosystems. Climate change will have an extremely adverse impact on agricultural yields, which will require increased adaptability, stability, diversification of crop production and adaptation to new changing requirements of agronomy, population nutrition and climate (Khan et al., 2021). Global climate change has recently greatly affected the productivity of many agricultural crops, including the productivity of an important grain crop—wheat worldwide, where it is cultivated on an area of 220 million hectares, with a gross production of 770 million tons (FAOSTAT, 2022).

Wheat (Triticum aestivum L.) plays a vital role in global food and nutritional security. Almost 35% of the world’s population consumes wheat as a staple food, and more than two-thirds of global wheat production is used for food, while one-fifth is used for livestock feed. Wheat is a vital source of food as it contains carbohydrates, fats, proteins, fiber, zinc, calcium, vitamin E etc. and contributes approximately 25% of the total calorie intake and 22% of total dietary protein (Bhandari et al., 2021; Shewry & Hey, 2015). Despite its significance, global wheat production still falls short of meeting the needs of the growing population (Djanaguiraman et al., 2020; Kamrani et al., 2017). In 2021/22, the global production of wheat reached 770 million metric tons (FAOSTAT, 2022), but it remains insufficient to feed the around 889 million people in the world (FAOSTAT, 2022). To meet the demand, the production of wheat should be increased by 15% (111.1 million metric tons). Furthermore, the demand is projected to increase by 25% by 2050, while climate-induced factors such as drought and heat stress are predicted to reduce wheat yield by 44–47% in South Asia (Lesk et al., 2016; Liu et al., 2016). As the leading crop, enhancing wheat production would play a vital role in addressing hunger and malnutrition. However, limited land availability and inadequate irrigation infrastructure hinder yield improvements through expansion. Despite a 14% increase in net cropping area from 1.36 billion ha to 1.5 billion ha, wheat production has had to accommodate a population rise of 200% from 3.5 billion to 7 billion between 1961 and 2011 (Bhanu, 2018; FAOSTAT, 2022).

The projected temperature increase of 0.92–1.07°C from 2016 to 2045 posed a significant threat to wheat production (WBG, 2022). Elevated temperature and low moisture stress induces physiological changes in the cell influencing source-sink transport in wheat (Zandalinas et al., 2018). Wheat has specific temperature requirements for growth, anthesis, and ripening, with optimal ranges of 16–22°C, 12–22°C, and 21–25°C, respectively (Akter & Rafiqul Islam, 2017a; Djanaguiraman et al., 2020). Temperatures exceeding 25°C causes heat stress on wheat, which negatively imparts wheat physiology and the biochemical process leading to poor wheat production (Akter & Rafiqul Islam, 2017b; Pan et al., 2018). Such temperature fluctuations, along with altered precipitation patterns and increased extreme weather events, had compounded the challenges of global wheat production (Djanaguiraman et al., 2020; Kamrani et al., 2017; WBG, 2022).

To mitigate the impact of climate change on wheat production and to address the challenges of hunger and malnutrition, it is crucial to develop climate-resilient wheat genotypes (Bhandari et al., 2021; Chaves et al., 2013; Shiferaw et al., 2011). Selection of climate-resilient genotypes should not just be focused on yield as yield is influenced by a variety of factors such as environments, genotype, genotype * environment interaction. Hence, climate-resilient genotypes should be selected by considering various yield-attributing parameters (Mahmudul et al., 2022). Addressing the challenge of low wheat productivity and increasing demand requires innovative strategies. Traditional methods of expanding agricultural land or improving irrigation infrastructure have limitations, making the breeding of stress-tolerant wheat genotypes a promising avenue for sustainable wheat production (Guzmán et al., 2017). The trait-based selection offers a promising approach to achieving this goal. Trait-based selection refers to the identification and selection of wheat genotypes with specific agronomic traits that contribute to higher yield potential and stress tolerance (Mwadzingeni et al., 2017; Tshikunde et al., 2019). Identification of the traits having direct and indirect effect on grain yield, breeders can develop improved wheat varieties that are better equipped to withstand heat stress and drought environments prevalent in the world (Crespo-Herrera et al., 2018; Joudi et al., 2014; Lopes et al., 2012).

In this study, we aim to evaluate the importance of trait-based selection in improving wheat yield under heat stress and heat drought environment. The findings would help the wheat improvement program of Nepal to develop climate-resilient wheat genotypes, thereby addressing the challenges of low productivity and the increasing demand for wheat. That could potentially help to eradicate the yield gap and combat the hunger and malnutrition (IPCC, 2021; Mwadzingeni et al., 2017; Tshikunde et al., 2019). Ultimately, our research aims to support the varietal improvement program in Nepal and contribute to global efforts in achieving food security and reducing hunger and malnutrition (IPCC, 2021; Mwadzingeni et al., 2017; Tshikunde et al., 2019).

2. Materials and methods

The field experiment was conducted in the agronomy farm of Institute of Agriculture and Animal Science (IAAS), Paklihawa Campus situated in the western region of Nepal at Bhairahawa, Rupandehi. The experiment comprises three wheat growing environments, irrigated (I), heat stress (HS), and heat drought (HD) laid during the wheat growing seasons of 2021–22 and 2022–23. The site of the experiment lies in the tropical Terai region of Nepal at the geographic location of 27°29’02“N and 83°27’17” E and an altitude of 104 meters above sea level. The experiment was conducted in loamy textured soil with pH 5.9 with sand, silt, and clay percentage of 31.3%, 48%, and 20.7%, respectively. The total Nitrogen (N), phosphorus (P₂O₅), and potassium (K₂O) content present in the soil was 0.07%, 13.53 kg ha⁻¹, and 160.8 kg ha⁻¹ whereas total organic matter, Boron (B), Sulphur (S), and Zinc (Zn) content were 2.13%, 0.18% ppm, 1.51 ppm, and 0.88 ppm, respectively.

Twenty elite wheat lines were used in the experiment as treatments 15 Nepal lines, three Bhairahawa lines, and two commercial checks, i.e., Bhrikuti and Gautam. The list of genetic materials used in the experiment is presented in (Table 1).

Table 1. Plant materials used in the experiment

S. N.	Genotypes*	Source	S. N.	Genotypes*	Source
1	Bhrikuti	CIMMYT, Mexico	11	NL 1376	CIMMYT, Mexico
2	BL 4407	Nepal	12	NL 1381	CIMMYT, Mexico
3	BL 4669	Nepal	13	NL1384	CIMMYT, Mexico
4	BL 4919	Nepal	14	NL 1386	CIMMYT, Mexico
5	Gautam	Nepal	15	NL 1387	CIMMYT, Mexico
6	NL 1179	CIMMYT, Mexico	16	NL 1404	CIMMYT, Mexico
7	NL 1346	CIMMYT, Mexico	17	NL 1412	CIMMYT, Mexico
8	NL1350	CIMMYT, Mexico	18	NL 1413	CIMMYT, Mexico
9	NL 1368	CIMMYT, Mexico	19	NL 1417	CIMMYT, Mexico
10	NL 1369	CIMMYT, Mexico	20	NL 1420	CIMMYT, Mexico

*The parentage of the genotype is confidential and is maintained by National Wheat Research Program (NWRP), Bhairahawa.

The field experiment was conducted using a serpentine alpha lattice design having two replications and five blocks for all tested environments. Each genotype was planted in a 10 m² plot with a plot dimension of 4m *2.5 m. The gap between blocks and replications was maintained at one meter. A row-to-row spacing of 25 cm was provided for a genotype in a plot whereas continuous sowing of plant to plant was done in an east-west direction. The seed rate was kept at 120 kg ha ⁻¹ and the fertilizer dose was provided at the rate of 100:60:40 NPK kg ha⁻¹. Fertilizers were applied by line placement method. Standard agronomic practices were conducted for wheat. Weeding was done manually two times at the jointing and booting stage. A full dose of irrigation was provided for irrigated and heat stress environment in six splits during pre-sowing, crown root initiation stage (CRI), jointing, booting, heading, and at the soft dough stage whereas, no artificial irrigation was provided for heat drought environment. The sowing time was maintained according to the environment. Sowing was done on 25th November for irrigated environments, and 25th December for heat stress and heat drought environment. The sowing of wheat under stress environments was late to coincide the reproductive and ripening stage of wheat with the terminal heat wave of March–April (Figure 1). Along with that, the temperature of the field during the reproductive and ripening stage of wheat sown under HS and HD environment was above 24 °C which created heat stress on wheat (Figure 1).

Figure 1. Maximum, minimum, and mean temperature along with 24h accumulated precipitation during the wheat growing season of 2021–22 (1a) and 2022–23 (1b).

A sampling of 10 random samples was done on each plot except from the border line. The phenological data of days to booting (DTB), days to heading (DTH), and days to anthesis (DTA) were taken when 50% of the population achieved their respective stage. Inter-phenological duration such as booting to heading duration (BtoH), booting to anthesis duration (BtoA), and heading to anthesis duration (HtoA) were collected. Plant height (Ph) was measured from the bottom to the top of the spike. Spike length (SL) was measured from the base of the spike to the top of the uppermost floret. The number of spikes per meter square, spikelets per spike (SPS), and grains per spike (GPS) were measured by counting them manually whereas 10 spike weights (TSW) and thousand kernel weights (TKW) were determined by weighing 10 spikes and 1000 grains on a weighing balance. The grain yield was determined by harvesting two quadrants of 1 m² plots, averaged, and converted to tons per hectare.

Characterization of the environments was done with daily maximum, minimum temperature and mean precipitation. The weather report of the experimental site was obtained from the Department of Hydrology and Meteorology (DTM), Bhairahawa (Figure 1).

The study employed additive main effects and multiplicative interaction (AMMI), combined analysis of variance (ANOVA), correlation, and path analysis to identify the most appropriate trait for abiotic stress tolerance in wheat.

The AMMI model ANOVA was conducted to assess the genotype-by-environment interactions. The AMMI model is represented by the formula.

$Yij=\mu +gi+ej+\sum \lambda ktk+\epsilon ij$ (Purchase et al., 2000)

where Yij is the observed value of the ith genotype in the jth environment, μ is the overall mean, gi is the ith genotype effect, ej is the jth environment effect, λk is the singular value of the kth principal component, tki is the score of the ith genotype on the kth principal component, and εij is the residual error.

A combined ANOVA was performed to quantify the effect of the tested wheat growing environment on the quantitative traits of wheat.

Correlation analysis was conducted to assess the degree of association between various quantitative traits and grain yield under tested environments. The Pearson correlation coefficient (r) was calculated using the formula,

$r\left(X,Y\right)=\frac{\sum \left[\left(Xi-\bar{X}\right)\left(Yi-\bar{Y}\right)\right]}{\left[\sqrt{\left(\sum {\left(Xi-\bar{X}\right)}^2\right)}\sqrt{\left(\sum {\left(Yi-\bar{Y}\right)}^2\right)}\right]}$

where Xi and Yi are the values of variables X and Y, respectively.

Path analysis was employed to determine the direct and indirect effects of different quantitative traits on grain yield. Principal Component Analysis (PCA) was conducted to identify the principal components and assess the contribution of different variables to the overall variability in the dataset. This involved calculating the eigenvalues and eigenvectors of the correlation matrix

Data entry and descriptive statistics were carried out in Microsoft Excel 2021. The combined analysis of variance (ANOVA), Pearson’s correlation coefficients, and principal component analysis (PCA) were conducted on IBM SPSS statistics V. 26. Additive main effect multiplicative interaction (AMMI), ANOVA of AMMI model and genotype (G) and genotype*environment (G*E) interaction was done from GEA-R Version 4.0 software provided by CIMMYT, Mexico, and the correlation plots were created using R-4.3.1.

3. Results and discussion

3.1. Effect of environment, genotype, and G*E interaction

The additive main effects and multiplicative interaction (AMMI) model analysis of variance (ANOVA) showed that the environment had a significant effect on the performance of all quantitative traits studied (p ≤ 0.05) (Table 2) i.e., DTB, DTH, DTA, BtoH, BtoA, HtoA, Ph, SL, NSPMS, NSPS, NGPS, TSW, TKW, and GY. The variation in DTB, DTH, DTA, BtoH, BtoA, Ph, SL, NSPM, NSPS, NGPS, TSW, and GY was mainly explained by environment whereas the majority of the variation on TKW and HtoA was mainly explained by genotype and genotype *environment interaction, respectively (Table 2). The environmental percentage variation explained on different traits evaluated in the experiment ranged from 31.99% for TKW to 94.29% for DTA. Whereas, the genetic percentage variation ranged from 3.2% for GY to 36.19% for TKW and the genetic*environmental percentage variation explained ranged from 2.18% for DTA to 31.85% for TKW (Table 2). The result implies environment (E), genotype (G), and G* E interaction have a substantial heterogenic effect across the traits evaluated. The result showed that DTB, DTH, DTA, BtoH, BtoA, Ph, SL, NSPM, NSPS, NGPS, TSW, and GY are mainly affected by environments, TKW by genetic variations among genotypes, and HtoA by genetic and environmental interactions. Hence, a trait performing well in one environment might not perform well in another environment. Understanding the effect of different wheat growing environments on quantitative traits is necessary for the trait-based selection of high-yielding stable genotypes. Along with the environmental effect, understanding genetic variation is necessary for effective breeding. The significant genotype explained percentage variation on different quantitative traits implies that there are considerable genetic variations across the genotypes evaluated (df = 19) for all traits except grain yield (Table 2). The genetic materials with high genetic variation provide a better chance of getting desirable traits that can be effective in heterosis breeding.

Table 2. Additive main effects and multiplicative interaction (AMMI) model analysis of variance (ANOVA) of quantitative traits studied

Traits	% Variation explained			Sum of square (SS)			Sum of Squares (SS) due to principal components						Residual
	Environment df= 5	Genotype df= 19	G*E df=38	Environment	Genotype	G*E	PC1 df= 23	PC2 df= 21	Pc3 df= 19	PC4 df= 17	PC5 df= 15	PC6 df= 13	df= 120
				df= 5	df= 39	df= 95
DTB	77.87***	14.74***	7.38***	5465.18	1034.82	517.98	297.49**	148.84**	35.53	21.5	14.6	0	194
DTH	90.36***	6.13***	3.50***	10461.34	709.75	405.58	258.25***	96***	25.03*	16.95	9.37	0	208.5
DTA	94.29***	3.53***	2.18***	12914.93	484	298.4	176.99***	51.63***	44.66***	22.12*	3	0	123
BtoH	71.29***	8.48***	20.22*	822.07	97.74	233.18	88.23***	77.05***	39.26*	16.87	11.75	0	214.5
BtoA	75.07***	11.93***	13.00**	1607.933	255.48	278.57	115.52***	67.88**	56.20*	26.66	12.29	0	202
HtoA	32.76***	26.95***	40.29*	134.82	110.91	165.76	65.81***	50.09**	22.24	19.33	8.28	0	149.5
Ph	88.06***	6.51***	5.42*	49686.42	3677.64	3057.77	1100.1***	1040.43***	503.83**	292.049	121.37	0	2599.6
SL	75.96***	11.18***	12.86***	382.27	56.28	64.7	28.99***	13.04***	10.199**	7.08*	5.4	0	43.04
NSPMS	64.08***	19.78***	16.14***	851418	262818.2	214535.1	90725.79***	53987.7***	34447.39*	23404.52	11970.67	0	148939.8
NSPS	76.16***	8.52***	15.32	1190.736	133.2024	239.55	123.72***	46.46*	36.11	24.66	8.61	0	300.25
NGPS	76.40***	8.40***	15.20*	17849.63	1962.36	3550.88	1415.21***	906.22*	558.8	453.45	217.19	0	3296.28
TSW	67.21***	10.47*	22.31	3031.87	472.53	1006.75	406.53**	441.29	155.08	126.48	77.36	0	1467.72
TKW	31.99***	36.16***	31.85**	1581.08	1786.83	1574.05	801.26***	386.54**	207.34	95.91	82099	0	1206.04
GY	87.20***	3.2	9.6	2.56E+08	9396528	28129337	14450454**	5640589	4910612	1777582	1350099	0	36705035
Range	(31.99–94.29)	(3.2–36.16)	(2.18–31.85)

Principal Component of AMMI (PC), Degree of Freedom (df), Sum of Squares (SS), Mean sum of Squares (MS). *, **, and *** denotes level of significance at 5%, 1%, and 0.1%, respectively.

3.2. Combined Analysis of Variance (ANOVA)

The combined analysis of variance (ANOVA) revealed that there was a significant reduction in the expression of yield and yield-attributing parameters of bread wheat under combined HS and combined HD environments as compared to combined irrigated environment (p ≤ 0.01). DTB, DTH, DTA, BtoH, BtoA, Ph, SL, NSPMS, NSPS, NGPS, TSW, TKW, and GY was reduced by (11.8, 12.3%), (14.5, 15.2%), (15.3, 16.2%), (42.8, 46.9%), (40, 44.4%), (34.3, 39.2%), (7.5, 16.3%), (4, 16.1%), (9.5, 23.7%), (12.3, 28.6%), (10.5, 13.1%), and (24.4, 48.7%) under combined HS and combined HD environments, respectively, as compared to the combined irrigated environment (Table 3). HD was the least suitable environment for the growth and yield performance of wheat.

Table 3. Combined analysis of variance (ANOVA) showing percentage reduction in yield and yield attributing parameters of bread wheat under HS and HD environments

Environments	DTB	DTH	DTA	BtoH	BtoA	HtoA	Ph (cm)	SL (cm)	NSPMS	NSPS	NGPS	TSW (g)	TKW (g)	GY (t ha^−1)
Irrigated	75.0±4.79^a	82.0± 82.05^a	85.5± 5.63^a	7.03±2.3^a	10.56± 3.0^a	3.54± 1.8^a	92.42± 9.9^a	10.05± 0.8^a	310.12± 75.8^a	17.57± 1.7^a	43.38± 8.1^a	23.88±4.4^a	40.40± 5.4^a	3.52±1.03^a
Heat stress	66.1± 2.43^b	70.1± 2.75^b	72.4±2.93^b	4.01±1.6^b	6.34±1.9^b	2.33±1.1^b	85.46± 10.2^b	9.65±1.1^a	272.81±84.1^b	16.62±1.9^b	39.27±7.8^b	20.92±3.2^b	36.15±4.5^b	2.66±0.89^b
Heat drought	65.8 ± 2.65^b	69.5±3.25^b	71.6±3.34^b	3.73±1.4^b	5.88 ±1.8^b	2.15±1.0^b	77.33±20.6^c	8.42±1.8^b	283.28±71.5^b	14.64±3.4^c	33.07±12.4^c	17.05±4.5^c	35.09±3.3^b	1.80±0.91^c
Mean	68.9±5.49	73.9± 7.02	76.5± 7.60	4.92±2.3	7.59±3.1	2.67±1.5	85.07± 15.7	9.37±1.5	288.74±78.6	16.27±2.7	38.57±10.5	20.62±5.0	37.22±5.0	2.66±1.17
CV	7.96	9.5	9.93	48.63	41.26	57.36	18.47	16.11	27.23	17.14	27.38	24.26	13.62	44.16
F-Value	***	***	***	***	***	***	***	***	***	***	***	***	***	***
% Reduction (vs. HS)	11.85	14.50	15.32	42.88	40.00	34.28	7.53	4.02	12.03	5.43	9.46	12.38	10.52	24.37
% Reduction (vs. HD)	12.28	15.25	16.24	46.98	44.38	39.22	16.33	16.17	8.65	16.71	23.75	28.60	13.14	48.72

The performance of wheat under combined heat stress and heat drought environment was poor as compared to irrigated environments. The poor performance of wheat under stress conditions were brought by a variety of factors. Moisture and heat stress inhibits the process of germination, growth, maturity, grain formation and yield (Akter & Rafiqul Islam, 2017a; Asseng et al., 2015; Mukherjee et al., 2019; Pask et al., 2014; Qaseem et al., 2019; Whittal et al., 2018). A drought environment retards the imbibition process of wheat seeds resulting in poor germination whereas heat stress induces tiller abortion and tiller mortality in developing tillers due to which the NSPMS under heat stress and heat drought environment are poor (Table 3). The optimum temperature for growth, spike initiation, booting, anthesis, and ripening of wheat is 16–20 °C 16–28 °C, 16–22 °C,12–22°C, and 21–25°C, respectively (Asseng et al., 2015; Mukherjee et al., 2019; Pask et al., 2014; Tack et al., 2015). The temperature of the field at the booting, heading, and anthesis stage were below 24°C under irrigated environment whereas it was above 24°C under heat stress and heat drought environment (Figure 1). Temperature above 18 °C during spike initiation, 22°C during booting-heading, and 24 °C during the anthesis-grain filling period is considered heat stress in wheat (Batool et al., 2019; Khan et al., 2020; Mathur et al., 2014). Temperature above optimum during growth accelerates the production of reactive oxygen species (ROS) (Lal et al., 2022; Posch et al., 2019). The biochemical changes due to high temperature induce the production of ethylene and abscisic acid (ABA) as well. Ethylene, ABA along with ROS induces senescence-related metabolic reaction (SRMR) causing oxidative damage to wheat (Christopher et al., 2018; Shirdelmoghanloo et al., 2016). As a result, the vegetative growth of wheat is accelerated, and wheat reaches to booting, heading, anthesis, ripening, and maturity stage earlier when sown under heat stress and heat drought environments (Akter & Rafiqul Islam, 2017a; Khan et al., 2020). Accelerated growth of wheat creates a shortage of time for tiller elongation, spike formation inside the stems, and photosynthate accumulation and as a consequence spike length (SL), numbers of spikelet per spike (NSPS), and spike weight (SW) reduces (Table 3). Temperature above 24 °C, during anthesis causes pollen abortion (Joshi et al., 2016), pollen mortality (Dwivedi et al., 2017a), and induces inefficient pollination on wheat as a result the number of fertile embryos decreases (Oyewole, 2016). The result is seen in NGPS with (9.46%, and 23.75%) reduction under heat stress and heat drought environment (Table 3). Due to a shorter graining filling period, and accelerated temperature, the seed gets shrunken, resulting in lower seed weight and lower 10 spike weight (Ni et al., 2017; Prasad & Djanaguiraman, 2014). The effect of heat stress and heat drought environments on seed weight was observed with (10.52%, and 13.14%) reduction in thousand-kernel weight (TKW) of wheat seeds under heat stress and heat drought environments, respectively (Table 3). The effect of wheat growing environments on yield attributing parameters is cumulated to grain yield (Figure 2) The yield of wheat is reduced by 25.37% and 48.72% under heat stress and heat drought environments, respectively (Table 3). Heat drought causes more severe damage on wheat growth and yield as the addition of moisture stress generally imbalances the cell water potential, the accent of sap, and the source-sink transport processes of wheat leading poor seed growth and yield (Shirdelmoghanloo et al., 2016).

Figure 2. Boxplot showing mean performance and the impact of heat stress and heat drought environment on days to booting (DTB), days to heading (DTH) and days to anthesis (DTA) (2a); booting to heading duration BtoH), booting to anthesis duration (BtoA), heading to anthesis duration (BtoA) (2b); plant height (Ph) (2c); spike length (SL); number of spikes per spikelets (NSPS), 10 spike weight (TSW) (2d); number of spikes per meter square (2e); number grains per spike, thousand kernel weight (TKW) (2f); and grain yield (GY)(2g).

The grain yield of wheat ranged from 4.61 t ha⁻¹ (BL 4919) to 3.09 t ha⁻¹ (NL 1179) under combined irrigated environment whereas from 3.04 t ha⁻¹ (BL 4919) to 1.8 t ha⁻¹ (NL 1387) under combined heat stress environment. Similarly, the wheat yield ranged from 2.23 t ha⁻¹ (Bhrikuti) to 1.52 t ha⁻¹ (NL 1179) under heat drought environment. The yield loss of wheat ranged from 6.2% (NL 1368) to 47.4% (NL 1387) under combined heat stress environments as compared to combined irrigated environment. Similarly, the yield loss ranged from 32% (Bhrikuti) to 56.3% (BL 4919) under combined heat drought environments as compared to combined irrigated environment (Table 4).

Table 4. Yield performance, percentage yield reduction, coefficient of variation (CV), and F-Value (df = 5) of elite wheat genotypes (df = 19) across tested environments (df = 5)

	2021 (t ha ^−1)			2022 (t ha ^−1)			Years combined yield (t ha ^−1)				% Reduction
Genotypes	I	HS	HD	I	HS	HD	I	HS	HD	Grand Mean Yield (t ha ^−1)	Vs HS	Vs HD	CV	F-Value
Bhrikuti	3.83^a	3.72^a	3.32^ab	2.72^b	1.76^a	1.13^a	3.28	2.74	2.23	2.75	16.3	32.0	37.17	**
BL 4407	3.96^a	3.02^b	2.55^b	2.66^b	2.24^b	0.94^c	3.31	2.63	1.75	2.52	20.5	47.3	36.63	**
BL 4669	4.61^a	3.52^b	2.22^c	2.19^c	1.72^d	1.15^e	3.40	2.62	1.68	2.57	23.0	50.5	46.81	***
BL 4919	6.25^a	3.79^b	3.08^c	2.96^c	2.29^d	0.94^e	4.61	3.04	2.01	3.22	34.1	56.3	52.79	***
Gautam	3.70^a	3.13^ab	2.20^ab	2.61^ab	2.29^ab	0.85^b	3.16	2.71	1.53	2.47	14.1	51.6	37.46	ns
NL 1179	3.85^a	3.20^a	1.88^ab	2.33^ab	2.26^ab	1.15^b	3.09	2.73	1.52	2.45	11.6	50.8	37.04	ns
NL 1346	5.04^a	3.48^b	2.86^bc	3.03^bc	2.18^cd	1.22^d	4.04	2.83	2.04	2.97	30.0	49.5	41.83	**
NL1350	4.18^a	3.57^ab	2.82^abc	3.05^abc	1.63^bc	1.13^c	3.62	2.60	1.98	2.73	28.1	45.3	40.53	*
NL 1368	4.39^a	4.26^a	2.41^b	2.01^b	1.75^b	0.98^b	3.21	3.01	1.70	2.64	6.2	47.1	48.91	**
NL 1369	4.39^a	3.43^b	2.36^c	2.79^bc	1.41^d	0.90^d	3.59	2.42	1.62	2.54	32.6	54.9	48.76	***
NL 1376	4.09^a	3.46^a	3.09^a	2.63^ab	1.45^b	1.10^b	3.36	2.46	2.10	2.64	27.0	37.7	41.68	*
NL 1381	4.43 ^a	3.31^b	2.29^cd	2.72^bc	1.67^de	1.05^e	3.58	2.49	1.67	2.58	30.3	53.2	44.87	**
NL1384	3.61^ab	3.32^ab	2.55^b	3.75^a	2.47^b	0.83^c	3.68	2.90	1.69	2.76	21.3	54.0	38.01	**
NL 1386	4.20^a	3.41^ab	2.62^ab	2.64^ab	1.48^ab	0.95^b	3.42	2.45	1.79	2.55	28.5	47.8	44.82	ns
NL 1387	4.15^a	2.18^b	2.51^b	2.63^b	1.39^bc	0.73^c	3.39	1.79	1.62	2.27	47.4	52.1	50.84	**
NL 1404	4.87^a	3.41^ab	2.58^b	2.55^b	1.95^bc	0.92^c	3.71	2.68	1.75	2.71	27.8	52.8	47.14	**
NL 1412	4.63^a	3.59^ab	2.75^bc	2.70^bc	2.32^c	1.01^d	3.67	2.96	1.88	2.84	19.4	48.7	40.81	**
NL 1413	4.07^a	3.27^ab	2.74^b	3.21^ab	1.77^c	1.06^c	3.64	2.52	1.90	2.69	30.9	47.7	39.57	**
NL 1417	3.80^a	3.72^ab	2.66^cd	3.18^bc	2.19^d	1.00^e	3.50	2.95	1.83	2.76	15.5	47.6	36.51	***
NL 1420	3.16^a	3.17^a	2.37^ab	3.01^a	2.21^b	1.19^c	3.09	2.69	1.78	2.52	13.0	42.3	29.17	**
Mean	4.26	3.40	2.59	2.77	1.92	1.01	3.52	2.66	1.80	2.66	24.4	48.7	40.66
Range	3.16–6.25	2.18–4.26	1.88–3.32	2.01–3.75	1.39–2.47	0.73–1.22	3.09–4.61	1.78–3.04	1.52–2.23	2.27–3.22	6.2–47.4	32–56.3	29.17–52.79

The coefficient of variation (CV) ranged from 29.17 for NL 1420 to 52.79 for BL 4919. Variation in CV signifies, the environment had a variable effect on the yield performance of wheat genotypes and there is substantial heterogeneity among the genotypes evaluated (Table 4).

3.3. Correlation and path coefficient analysis

The grain yield of wheat was found to have a significant positive correlation with BtoA and HtoA duration and a significant negative correlation with DTB and DTH under irrigated environments. Under HS environments, GY had a significant positive correlation with Ph, and no significant negative correlation was observed with other attributes studied. Whereas under HD environments, GY had a significant positive correlation with Ph, SL, TSW, and TKW, and no significant negative correlation was observed. SL and NGPS had a positive association with grain yield under all wheat growing environments whereas, DTB, DTH, and DTA had a negative association in all tested environments (Figure 3).

Figure 3. Combined phenotypic correlation across 14 quantitative traits of wheat genotypes under irrigated (3a), heat stress (3b), and heat drought (3c) environment.

Path analysis revealed that DTA, BtoA, SL NSPMS, NGPS, and TKW had a direct positive effect and DTB, DTH, BtoA, HtoA, Ph, NSPS, TSW had a direct negative effect on grain yield of wheat under irrigated environment. Under heat stress environment, DTB, DTA, BtoH, Ph, NSPMS, NSPS, NGPS, TSW, and TKW had a direct positive contribution to yield and DTH, BtoA, HtoA, and SL had a direct negative effect on grain yield of wheat. Under heat drought environment, DTH, BtoA, Ph, SL, NSPMS, NGPS, TSW, and TKW had a direct positive effect on the grain yield of wheat, and DTB, DTA, BtoH, HtoA had a direct negative effect on grain yield of wheat. NSPMS, NGPS, and TKW had a direct positive effect on yield under all environments whereas, HtoA duration had a direct negative in all environmental environments (Table 5).

Table 5. Combined path analysis among 14 quantitative traits studied under irrigated, heat stress, and heat drought environment under two years

Traits	ENV	DTB	DTH	DTA	BtoH	BtoA	HtoA	Ph	SL	NSPMS	NSPS	NGPS	TSW	TKW	GY
DTB	IR	−3.785	−7.272	7.968	−0.612	2.188	0.903	−0.086	0.003	0.158	−0.014	0.118	−0.003	−0.158	−0.591**
DTB	HS	2.326	−3.405	1.082	−0.951	0.602	0.025	−0.070	0.011	0.234	0.464	−0.002	−0.226	−0.129	−0.040
DTB	HD	−25.831	31.862	−8.710	18.958	−21.654	5.540	−0.284	−0.048	0.031	0.000	−0.016	−0.088	−0.132	−0.373
DTH	IR	−3.495	−7.875	8.012	−0.086	1.693	1.222	−0.112	0.004	0.180	−0.015	0.095	−0.004	−0.137	−0.518*
DTH	HS	2.164	−3.659	1.150	−0.572	0.442	0.029	−0.065	0.009	0.169	0.557	0.000	−0.176	−0.146	−0.100
DTH	HD	−24.047	34.226	−10.177	10.542	−15.340	4.898	−0.238	−0.040	0.011	0.000	−0.014	−0.058	−0.109	−0.347
DTA	IR	−3.464	−7.248	8.705	−0.316	1.296	0.598	−0.086	0.008	0.144	−0.015	0.076	−0.003	−0.111	−0.415
DTA	HS	1.864	−3.117	1.350	−0.515	0.086	−0.027	−0.068	0.008	0.107	0.454	−0.001	−0.169	−0.114	−0.144
DTA	HD	−19.711	30.515	−11.415	4.810	−5.416	1.362	−0.183	−0.030	−0.002	0.000	−0.005	−0.017	−0.098	−0.188
BtoH	IR	1.666	0.487	−1.976	1.390	−1.729	0.510	−0.037	0.000	0.010	0.002	−0.085	−0.001	0.090	0.326
BtoH	HS	−1.732	1.640	−0.544	1.277	−0.663	−0.009	0.052	−0.010	−0.262	−0.119	0.005	0.231	0.050	−0.084
BtoH	HD	18.026	−13.281	2.021	−27.168	24.561	−4.375	0.250	0.043	−0.057	0.000	0.012	0.109	0.120	0.261
BtoA	IR	2.953	4.754	−4.023	0.857	−2.805	−1.059	0.056	0.006	−0.125	0.009	−0.141	0.002	0.175	0.657**
BtoA	HS	−1.563	1.807	−0.129	0.945	−0.895	−0.075	0.032	−0.007	−0.257	−0.210	0.002	0.167	0.074	−0.112
BtoA	HD	20.234	−18.992	2.236	−24.138	27.645	−7.094	0.255	0.045	−0.049	0.000	0.019	0.117	0.107	0.385
HtoA	IR	1.880	5.293	−2.865	−0.390	−1.635	−1.818	0.106	0.007	−0.163	0.009	−0.084	0.003	0.121	0.464*
HtoA	HS	−0.564	1.018	0.359	0.109	−0.657	−0.103	−0.005	−0.001	−0.116	−0.191	−0.003	0.015	0.059	−0.081
HtoA	HD	17.355	−20.330	1.886	−14.414	23.786	−8.245	0.194	0.035	−0.027	0.000	0.022	0.096	0.064	0.422
Ph	IR	−1.139	−3.069	2.621	0.180	0.547	0.675	−0.287	0.123	0.110	−0.006	0.058	−0.005	0.054	−0.138
Ph	HS	−0.737	1.079	−0.417	0.301	−0.129	0.003	0.221	−0.019	0.013	−0.088	0.005	0.191	0.028	0.450*
Ph	HD	15.017	−16.693	4.269	−13.869	14.439	−3.280	0.489	0.050	−0.044	0.000	−0.001	0.101	0.186	0.664**
SL	IR	−0.062	−0.150	0.378	0.003	−0.087	−0.066	−0.182	0.193	−0.022	−0.007	0.099	−0.007	0.057	0.145
SL	HS	−0.849	1.119	−0.383	0.426	−0.229	−0.004	0.142	−0.029	−0.152	−0.260	0.004	0.319	0.116	0.220
SL	HD	13.108	−14.486	3.568	−12.239	12.974	−3.023	0.255	0.095	−0.070	0.000	0.019	0.128	0.216	0.546*
NSPMS	IR	−1.456	−3.446	3.041	0.033	0.854	0.718	−0.077	−0.011	0.412	−0.007	0.054	0.002	−0.102	0.017
NSPMS	HS	0.860	−0.979	0.228	−0.529	0.364	0.019	0.005	0.007	0.632	0.220	0.002	−0.296	−0.169	0.362
NSPMS	HD	−6.360	2.917	0.195	12.339	−10.780	1.780	−0.170	−0.053	0.125	0.000	−0.022	−0.158	−0.182	−0.369
NSPS	IR	−2.278	−5.009	5.379	−0.107	1.025	0.665	−0.074	0.057	0.121	−0.024	0.285	−0.007	−0.140	−0.106
NSPS	HS	1.229	−2.319	0.698	−0.173	0.214	0.022	−0.022	0.009	0.158	0.878	0.006	−0.168	−0.200	0.333
NSPS	HD	−4.211	6.464	−1.866	1.114	−2.491	0.997	−0.067	0.018	−0.041	0.000	0.013	0.089	0.069	0.087
NGPS	IR	−1.140	−1.902	1.693	−0.303	1.012	0.390	−0.043	0.049	0.057	−0.017	0.392	−0.007	−0.138	0.044
NGPS	HS	−0.335	−0.042	−0.141	0.475	−0.099	0.022	0.078	−0.009	0.072	0.388	0.014	0.113	−0.126	0.411
NGPS	HD	4.432	−5.250	0.645	−3.592	5.695	−1.926	−0.005	0.020	−0.030	0.000	0.093	0.095	−0.031	0.146
TSW	IR	−0.912	−2.273	2.199	0.068	0.387	0.398	−0.120	0.104	−0.079	−0.014	0.236	−0.012	−0.006	−0.024
TSW	HS	−1.149	1.409	−0.498	0.643	−0.327	−0.003	0.092	−0.020	−0.408	−0.322	0.003	0.458	0.180	0.058
TSW	HD	10.863	−9.449	0.907	−14.122	15.524	−3.793	0.237	0.058	−0.094	0.000	0.042	0.209	0.185	0.568**
TKW	IR	1.966	3.550	−3.184	0.412	−1.615	−0.720	−0.051	0.036	−0.137	0.011	−0.178	0.000	0.304	0.394
TKW	HS	−1.081	1.918	−0.553	0.228	−0.238	−0.022	0.022	−0.012	−0.385	−0.633	−0.006	0.297	0.278	−0.187
TKW	HD	11.754	−12.830	3.840	−11.223	10.139	−1.803	0.313	0.071	−0.078	0.000	−0.010	0.133	0.291	0.597**

Selection of genotypes that are earlier in booting, heading, and anthesis with longer spike length, NGPS, higher tillering capacity, and TKW would be beneficial for all environments. Whereas, the selection of genotype with longer booting to heading duration and booting to anthesis duration with shorter heading to anthesis duration would be beneficial across all tested environments. The taller genotype would give a higher yield under heat stress and heat drought environments whereas, the shorter genotype would be productive under irrigated environments. Late sowing has been a major reasons for terminal heat stress in wheat in south Asian countries including Nepal (Poudel et al., 2017; Shrestha et al., 2020). Generally, wheat sowing starts from 25^th November but due to climatic, geographic, input-related, and socioeconomic constraints, it is extended up to 25^th December in Nepalese context (Dhakal et al., 2020; Puri et al., 2020). The temperature of the field during booting, heading, and anthesis during early sowing stages (25^th November) is generally below 24 °C whereas, the temperature exceeds 24 °C during the booting, heading, and anthesis under late sown conditions (December 25th) under HS and HD environment. (Figure 1). This signifies the influence of heat stress on wheat due to late sowing. Since, wheat sowing in Nepal is closely associated with rice-wheat cropping pattern, the sowing time of wheat would get influenced by the growth cycle of rice varieties as well. Hence, it becomes crucial for breeding early maturing wheat cultivars (Asseng et al., 2015; Krupnik et al., 2021; Peña-Bautista et al., 2017). Hence, a genotype with earlier booting, heading and anthesis could significantly contribute to produce more under stress environments. Earliness in wheat is an avoidance mechanism against terminal heat stress (Mondal et al., 2013). The phenomena have been considered a key factor for stress tolerance and had extensively been used in climate-resilient wheat improvement programs (Chen et al., 2016a, 2016b; Gomez et al., 2014; Mondal et al., 2013, 2016; Pandey et al., 2015; Puri & Gautam, 2015; Yu et al., 2017).

Besides, earliness, plant breeders shall focus on attributes such as NSPMS, SL, NGPS, and TKW to produce more (Poudel et al., 2020; Puri et al., 2020). Genotypes with high tillering capacity are often associated with higher yield in wheat (Chavan et al., 2019; Dias de Oliveira et al., 2015). Wheat has a tillering ability to produce up to 500 tillers per meter square which can reduce up to 8–15% under HS and HD environments (Pandey et al., 2021; Tiwari et al., 2019). System of wheat intensification (SWI) also focuses on improving yield by improving the tillering ability of wheat (Rana et al., 2017). Since, under heat stress and heat drought environments, tiller suffers from tiller mortality and tiller abortion at rapid tiller growing stages, genotypes with a higher tillering ability and higher tiller survival rate could help to produce more under stress environments (Dwivedi et al., 2017b). Hence, heat stress and heat drought breeding shall focus on identification and selection of wheat genotypes having higher tillering capacity. Since, HS and HD environment severely affects SL (Zhang et al., 2020), GPS (Zhang et al., 2017), and TKW, it is crucial to consider these attributes as well. High wheat yields are often associated with longer spike lengths in wheat, increment in the spike length increases net spikelet per spike which upon successful pollination could produce higher number of grains per spike (Barber et al., 2015; Bheemanahalli et al., 2019; Mirosavljević et al., 2021). Spike itself performs photosynthesis and contributes around 9.8% − 39% to the yield of wheat (Zhang et al., 2020). Higher GPS on wheat causes competition among developing grains (Ullah et al., 2022) hence, selection along with TKW would enhance the success of wheat breeding (Mondal et al., 2013). reported, a genotype producing TKW of 34.5 under heat stress environment could produce a yield of 5 tons per hectare under irrigated environment.

Plant height has a crucial role in yield improvement program of wheat (Chairi et al., 2018; Chen et al., 2016a; Du et al., 2018). Plant height is associated with biomass accumulation, photosynthesis, and canopy temperature maintenance. Under irrigated environment, dwarf genotype resists lodging, which reduces yield by up to 40% (Berry & Spink, 2012; Berry et al., 2015; Chen et al., 2016a). Shorter height also improves source-sink transport by facilitating proper nutrition for developing grains. Hence, there is continuous work being done to incorporate dwarfish genes Rht1 (Rht-B1b), Rht2 (Rht-D1b), Rht1-D1c, and Rht8 to produce more under-irrigated environments (Chairi et al., 2018; Divashuk et al., 2013; Grover et al., 2018; Joudi et al., 2014; Lopes et al., 2013; S Kumar et al., 2013; Zhang et al., 2016). In contrast temperature plays a critical role under heat stress and heat drought environment. Maintaining a cooler canopy temperature is necessary to facilitate proper photosynthesis and adequate growth of grains. Taller plants have larger canopy coverage that reduces net canopy temperature as compared to dwarf genotypes. Larger biomass also increases net leaf area for photosynthesis thus producing more photosynthates to the grains (Acuña-Galindo et al., 2015; Balota et al., 2017; Cossani & Reynolds, 2015; Gao et al., 2017; Sharma et al., 2015). Hence, dwarf genotypes are promoted under irrigated environments whereas taller genotypes are promoted under heat stress and heat drought environments.

3.4. Principal Component Analysis (PCA)

Principal Component Analysis extracted five components for irrigated and heat stress environments whereas four components for heat drought environments. The first five principal components described 86.02%, and 79.84% of the total variation in yield under irrigated and heat-stress environments, respectively. Whereas the first four PC described 80.91% of the total variation in grain yield under heat drought conditions.

The majority of the variation on the data yield was governed by PC1 and PC2 under all tested environments. PC1 and PC2 explained a cumulative of 57.44%, 56.66%, and 61.95% of the total variation under irrigated, heat stress, and heat drought environment, respectively. Under Irrigated and heat stress environment, PC1 and PC2 showed negative and positive correlation with GY whereas PC1 and PC2 showed positive correlation with GY under HD environment (Table 6). Hence, a genotype with lower PC1 score and higher PC2 score would yield higher under irrigated and heat-stress environment. Whereas, a genotype with higher PC1 and PC2 score would yield more under heat-drought environment. Based on PCA biplot BL 4919 and NL 1350 were the highest yielding wheat genotype under irrigated environment. Whereas, NL 1384 and Bhrikuti were the most suitable wheat genotype under heat-stress and heat-drought environments.

Table 6. Correlation among PC1-PC5 with yield and yield attributing parameters of bread wheat under irrigated, heat stress and heat drought environment

Irrigated
	Eigenvalue	% Variance	Cumulative Variance	DTB	DTH	DTA	BtoH	BtoA	HtoA	Ph	SL	NSPMS	NSPS	NGPS	TSW	TKW	GY
PC 1	5.61	40.07	40.07	0.40	0.38	0.34	−0.13	−0.34	−0.28	0.17	0.06	0.19	0.31	0.22	0.17	−0.26	−0.24
PC 2	2.43	17.37	57.44	−0.11	−0.03	−0.03	0.23	0.20	0.01	0.38	0.51	−0.07	0.25	0.26	0.47	0.23	0.27
PC 3	1.65	11.78	69.21	0.00	0.23	0.10	0.54	0.15	−0.38	0.31	−0.06	0.31	−0.16	−0.42	−0.20	0.21	−0.04
PC 4	1.34	9.55	78.77	−0.14	−0.01	−0.09	0.33	0.17	−0.14	−0.33	−0.29	0.36	0.29	0.35	−0.06	−0.35	0.40
PC 5	1.01	7.25	86.02	0.16	0.05	0.31	−0.31	0.12	0.47	0.11	0.15	0.49	0.07	−0.11	−0.29	0.14	0.39
Heat Stress Environment
	Eigenvalue	% Variance	Cumulated Variance	DTB	DTH	DTA	BtoH	BtoA	HtoA	Ph	SL	NSPMS	NSPS	NGPS	TSW	TKW	GY
PC 1	5.37	38.33	38.33	−0.39	−0.36	−0.30	0.30	0.29	0.12	0.18	0.26	−0.25	−0.28	0.03	0.33	0.30	−0.01
PC 2	2.57	18.32	56.66	−0.04	0.01	−0.12	0.13	−0.07	−0.23	0.39	0.25	0.21	0.29	0.52	0.10	−0.28	0.45
PC 3	1.78	12.72	69.38	−0.03	0.13	0.39	0.32	0.53	0.46	−0.10	−0.07	−0.14	0.30	0.22	−0.02	−0.23	−0.09
PC 4	1.47	10.47	79.85	0.31	0.38	0.32	−0.07	−0.13	−0.12	0.13	0.39	−0.40	0.07	0.02	0.46	0.26	−0.12
PC 5	1.22	8.74	88.59	0.11	−0.06	0.23	−0.38	0.10	0.53	0.34	0.24	0.28	−0.16	−0.23	−0.07	0.13	0.37
Heat Drought Environment
	Eigenvalue	% Variance	Cumulated Variance	DTB	DTH	DTA	BtoH	BtoA	HtoA	Ph	SL	NSPMS	NSPS	NGPS	TSW	TKW	GY
PC 1	6.33	45.19	45.19	−0.34	−0.29	−0.20	0.30	0.33	0.28	0.30	0.30	−0.23	0.02	0.10	0.30	0.29	0.26
PC 2	2.35	16.77	61.95	0.30	0.38	0.38	−0.02	−0.09	−0.14	−0.03	0.17	−0.40	0.45	0.12	0.35	0.21	0.14
PC 3	1.49	10.63	72.58	−0.02	0.12	0.36	0.29	0.39	0.39	−0.23	−0.19	−0.01	−0.07	0.40	0.11	−0.41	−0.20
PC 4	1.17	8.36	80.93	0.11	0.28	0.39	0.27	0.21	0.09	0.28	−0.06	−0.01	−0.33	−0.62	−0.14	0.14	0.10

Based on the ranking of correlation of PC1 with morphological traits studied DTB, DTH, DTA, BtoH, and NSPS were extracted under irrigated environment whereas, DTB, DTH, TSW, TKW, and DTA under heat stress environment and DTB, BtoA, SL, and TSW under heat drought environment (Table 6).

PCA analysis suggested, trait-based selection should be focused on phenological stages especially, DTB, DTH, and DTA, besides that, inter-phenological interval also creates variation in yield of wheat. Along with that, selection based on spike-related parameters such as NSPS under irrigated, TSW under heat stress, and SL, TSW under heat drought environment can be considered (Figure 4).

Figure 4. Combined principal component analysis (PCA) under irrigated (4a), heat stress (4b), and heat drought (4c) environment.

4. Conclusion

Production of wheat in world is mainly limited by abiotic stresses such as heat stress and drought. The majority of wheat is grown in winter season in South Asian region where it suffers from terminal heat stress during the reproductive stage. Furthermore, uneven rainfall patterns and poor infrastructure development had also added drought stress to wheat. Hence, it is crucial to develop climate-resilient wheat genotypes for the food and nutritional security of the world. To identify the most appropriate for the selection of high-yielding wheat genotype, a field experiment was conducted in the wheat growing season of 2021–22 to 2022–23. The AMMI model ANOVA showed, the environment had the most substantial effect on the expression of the quantitative traits of wheat genotypes. The grain yield of wheat was reduced by an average of 24 and 48% under the HS and HD environment, respectively. To get a high-yielding genotype, selection should be done with DTB, DTH, BtoH, BtoA, Ph, NSPMS, SL, NGPS and TKW. Genotypes that are earlier in booting, heading, and anthesis with longer spike length, NGPS, higher tillering capacity, and TKW would be beneficial for all environments. Along with that, the genotype with longer booting to heading duration and booting to anthesis duration with shorter heading to anthesis duration would be beneficial across all tested environments. The taller genotype would give a higher yield under heat stress and heat drought environments whereas, the shorter genotype would be productive under irrigated environments.

Conflicts of interest

The authors declare they have no conflicts of interest.

Authors contribution

Radha Krishna Bhandari and Mukti Ram Poudel conceptualized the research. Radha Krishna Bhandari and Shivalal Nyaupane conducted the first year field experiment (2021-22) while Radha Krishna Bhandari conducted the second year field experiment (2022-23). Data collection, data analysis, visualization, manuscript preparation, review and editing were done by Radha Krishna Bhandari. The research was supervised by Mukti Ram Poudel. The final version of the manuscript has been approved by all the authors.

Disclosure statement

No potential conflict of interest was reported by the author(s).