Physiological and antioxidant responses of synthetic hexaploid wheat germplasm under drought

Abstract

Background

As a result of the world population and climate change impact increases (especially in arid environments), there is a critical need for high-yield, drought-tolerant wheat. Synthetic hexaploid wheat derived lines (SHW-DL), were created artificially by crossing different durum wheat cultivars (AABB) with accessions of Aegilops tauschii (DD), a beneficial source of new genes for common bread wheat (Triticum aestivum L). Here, we studied the response of a panel of 91 SHW-DL for drought tolerance based on physiological, antioxidant enzyme activities, and drought tolerance indices.

Results

A wide range of variation and high values of heritability observed for grain yield, physiological and antioxidant traits indicating that the SHW-DL panel constitutes a valuable gene source for drought tolerance improvement of wheat. Despite decreases in grain yield (YLD), leaf area index (LAI), and relative water content (RWC) an increase in the content of malondialdehyde (MDA) was observed. Moreover, drought streass increased the antioxidant enzyme activities of ascorbate peroxidase (APX), catalase (CAT) and peroxidase (POD), and also photosynthetic pigments, proline (Pro), and MDA content. With higher values of grain yield, physiological and biochemical traits such as photosynthetic pigments, and RWC, and lower content of MDA, and peroxidase (H₂O₂) activity, SHW-DL performed better compared to common wheat lines under water stress conditions.

Conclusions

Different responses to water stress within the germplasm and between synthetic and common wheat suggest that selection for adaptive and suitable genotypes is possible for drought tolerance in synthetic wheat germplasm. Genotypes 54, 98, 102, 105, 122, 124, 143, 159, 196, and 198 were identified to be directly used in breeding programs or indirectly by crossing them with other wheat germplasm collections.

Background

Common wheat (Triticum aestivum L.) is the most important cereal crop and the main food source worldwide. Three different diploid genomes include T. urartu (A genome donor), Aegilops speltoides (B genome donor), and Ae. tauschii (D genome donor) constitute the base for bread wheat [23]. During domestication and modern improvement, bread wheat has gone through a genetic bottleneck resulting in a narrow genetic variation [13]. To increase the genetic diversity of cultivated wheat, plant breeders suggested producing synthetic hexaploid wheat (SHW) (Fig. 1). Synthetic hexaploid wheat (SHW) (2n = 6x = 42, AABBDD) is the consequence of interspecific cross between durum wheat (2n = 4x = 28, AABB, T. turgidum L.) and goat grass (2n = 2x = 14, DD, Ae. tauschii Coss.) which can boost the genetic variation of common wheat [23, 28]. Using the hybridization of these primary SHW lines to a variety of elite hexaploid wheat lines, synthetic-derived lines with different levels of genetic introgression from the ancestors were produced [23]. Due to biotic and abiotic stress resistance [6, 38], higher rates of photosynthesis [16], and a significant increase in grain yield [17], SHW-derived lines (SHW-DL) might be suitable for drought and heat-stressed environments. Becker et al. [6] showed that water extraction from deeper depths in drying soil increased in SHW-DL and the productivity was higher compared to winter wheat using root morphological traits. Also, it is shown that SHW-DL has better agronomic performance and higher antioxidant activities [45], indicating the tremendous genetic potential of synthetic wheat for improving yield and yield stability under drought and heat-stressed conditions.

Fig. 1

Development of synthetic hexaploid wheat (AABBDD) in comparison with the emulating evolution of the hexaploid wheat (AABBDD)

Drought is a complex phenomenon that negatively causes a range of impacts, especially in agriculture under climate change. Drought can cause losses in performance and growth in different stages of wheat with the greatest impact during anthesis [10]. Drought evokes oxidative stress by producing toxic reactive oxygen species (ROS) such as hydrogen peroxide (H₂O₂), superoxide radical (O₂⁻), singlet oxygen (¹O₂), and hydroxyl radical (HO) [26]. Active oxygen species can disrupt normal cellular metabolism and damage lipids, proteins, chlorophyll, and DNA [49]. To alleviate the drought impacts, plants have adapted mechanisms including drought tolerance, escape, and avoidance. Plant cells try to scavenge ROS and their effects by activation of antioxidant enzymes such as catalase (CAT), superoxide dismutase (SOD), guaiacol peroxidase (POD), and ascorbate peroxidase (APX) as a defense system [25, 49]. A positive relationship between drought tolerance and antioxidant enzyme activities has been shown [26]. Non-enzymatic antioxidants such as carotenoids, glutathione, ascorbic acid, and proline may also play a role [33]. Furthermore, the intercellular concentration of malondialdehyde (MDA) is synthesized due to the degradation of polyunsaturated lipids by ROS, indicating the extent of oxidative stress due to increasing ROS [41]. Under drought stress conditions, plant cells accumulate free amino acids, sugars, and proline to adjust osmotic pressure and facilite the restoration of cell turgor [33].

Plant breeders practically are looking for genotypes with high yield potential as well as less yield reduction under drought. The efficiency of selecting superior genotypes may be improved by applying appropriate drought and susceptibility indices [19]. Effective indices to screen drought-tolerant genotypes include stress susceptibility index (SSI), relative stress index (RSI), tolerance index (TOL), mean productivity (MP), yield stability index (YSI), harmonic mean (HM), geometric mean productivity (GMP), stress tolerance index (STI), yield index (YI), and combination of significant indices (CSI) [9, 11, 19,20,21,22, 34, 40, 42]. The combination of best-identified tolerance and susceptibility indices seems to be the most effective method for screening superior lines [42].

Because of limited genetic diversity within wheat germplasm and the prospect of increased drought events, the development of new varieties adapted to stress environments is inevitable. To mitigate the problem, many researchers advocate using SHW-DL because it represents a wider, more comprehensive genetic base from tetraploid and diploid relatives of wheat [7, 38]. Previous studies have mainly focused on SHW-DL aspects related to biotic stresses and less abiotic ones. Therefore, exploiting their genetic potential under abiotic stresses such as drought is needed to facilitate their utilization in cultivar development, especially in arid and semiarid regions. Information about the response of SHW-DL to drought stress in arid climates in terms of productivity, biochemical, physiological, and antioxidative properties, and its comparison with common wheat is limited. The objectives of this study were to evaluate the genetic diversity of a large panel of SHW-DL under drought stress based on biochemical, physiological, and antioxidative enzymes and identify promising lines for drought tolerance based on drought and susceptibility indices and physiological attributes.

Results

Analysis of variation and variation estimates for traits

Combined analysis of variance showed significant (P < 0.05) effects of the moisture regimes for the studied traits and genotypes (except chl a, chla/chlb and LAI) (Table S1). There were significant differences between the two experimental years except for CAT, protein, Chl b, and Pro. The interaction of genotype × year was significant for all the traits excluding Pro, RWC, and LAI. The range, genotypic coefficients of variation (GCV), and heritability estimations (h²) are presented in Table 1. The GCV ranged from 4.56 (RWC) to 49.06 (POX) for normal water conditions and 3.83 (chla/chlb) to 57.71 (POX) for the water stress conditions. The range of heritability was 18.38 (for CAT) to 86.08 (for POX) for normal water conditions and 16.68 (chla/chlb) to 89.11 (POX) for the water stress conditions. Antioxidant enzymes (CAT, APX, and POX), protein, H₂O₂, and YLD had higher heritability under water stress conditions (Table 1). The heritability of grain yield was moderate (32.24) which was lower than the h² of physiologic traits (except RWC and LAI).

Table 1 Range, genotypic coefficient of variation (GCV) and broad sense heritability (h²) of SHW-DL under two conditions

Effects of water stress

A wide range of variations was observed for grain yield from 684 g m^− 2 to 992 g m^− 2 under water stress and normal conditions, respectively. Grain yield was decreased by 64% and 71% under water stress in the two experimental years, respectively. In addition, the activities of antioxidant enzymes (CAT, APX, and POX) and pigment content (including Chl a, Chl b, Tch, and CAR) increased significantly due to water stress (Table 2). Moreover, water deficit increased the content of MDA, H₂O₂, and PRO, whereas the content of protein was unchanged (Table 2).

Table 2 Mean of physiological traits of SHW-DL evaluated under two irrigation levels during two years

Comparison of SHW and common wheat and screening of genotypes

Significant differences were observed between SHW-DL and common wheat lines for most studied traits (Table S1). The results showed that the activities of APX, POX, and H₂O₂ in common wheat genotypes were higher than SHW-DL in water stress conditions, whereas the content of CAT, Chl b, Tch, RWC, and YLD were higher in SHW-DL (Table 3). There was no difference between SHW-DL and common wheat genotypes for the content of protein, Chl a, CAR, MDA, LAI, and Pro under water stress conditions. Under normal conditions, RWC in SHW-DL was higher than common wheat, whereas, for most of the other traits, no significant differences were observed between these two groups.

Table 3 The comparison between SHW-DL and common wheat using physiological traits, YLD and LAI

Principal component analysis based on physiologic traits, grain yield, and drought tolerance indices was performed (Fig. 2a and b). The PC graph revealed that the first two components explained 59.54% and 64.93% of variation at normal and water stress conditions, respectively. The graphs revealed that pigments content (Chla, Chlb, CAR, Tch) were grouped together and positively associated with PC1 under both moisture conditions. Moreover, drought tolerance indices (STI, CSI, and YSI) were grouped together and positively associated with PC2 under normal moisture conditions. Under water stress conditions, drought tolerance indices (STI, CSI, and YSI) along with CAT and PRO were positively associated with PC2. As PC1 had a higher correlation with pigment content (including Chl a, Chl b, Tch, CAR), it may be considered a “chlorophyll” factor (Fig. 2a). The second PC2 had a positive correlation with yield and stress tolerance indices, CAT and Pro, and was named “yield potential and drought tolerance” factor (Fig. 2a). Selection based on high PC2 and moderate PC1 may result in superior genotypes. Therefore, genotypes 54, 82, 98, 102, 105, 122, 124, 143, 159, 161, 168, 196, 198, and 199 may be selected for drought tolerance and high productivity under water stress conditions. Genotypes 54, 58, 80, 98, 102, 105, 122, 124, 154, 159, and 196 may be suggested for cultivation under normal water conditions. Across both water environments and based on multivariate and univariate analysis, genotypes 54, 98, 102, 105, 122, 124, 143, 159, 196, and 198 were introduced with the highest productivity and drought tolerance. Amongst these high-yielding genotypes, the highest increase in pigment contents and lowest in MDA content belonged to genotype 54, and the highest increase in Pro, and chla/chlb was found in genotype 198. Moreover, the protein content of genotype 102 was high, whereas the content of CAT and APX was high in genotype 159 (Table S2).

Fig. 2

Principal component analysis of traits measured in the synthetic hexaploid wheats derived lines under normal (a) and water stress (b) conditions. Horizontal and vertical axes are the first and second principal components, respectively. CAT (µmoles of H₂O₂ decomposed min^− 1 mg^− 1 protein)Catalase, APX (µmole of monodehydroascorbate formed min^− 1 mg^− 1 protein)Ascorbate peroxidase, POX (increase in absorbance min^− 1 mg^− 1 protein) Peroxidase, Prot. (mg/ml) Protein, Chl a (mg/g) Chlorophyll a content, Chl b (mg/g) Chlorophyll b content, CAR (mg/g) Carotenoid content, Tchl (mg/g) Total Chlorophyll content, YLD (g/m²) grain yield, Chl a/Chl b ratio of chlorophyll a to chlorophyll b, H₂O₂ (µm/g) Hydrogen peroxide, MDA (µm/g) Malondialdehyde content, PRO (µm/g) Proline content, RWC (%) relative water content, LAI Leaf area content (mm²), STI stress tolerance index, YSI yield stability index, CSI Combination of significant indices.The Number of genotypes is according to the first column (Genotype code) in table S1

Interrelationships of traits

Significant positive relationships were observed amongst pigment contents (Chl a, Chl b, CAR, Tch) (r = 0.53–0.97) under both normal and water stress conditions (Table 4). Grain yield was positively correlated with MDA under normal conditions (r = 0.38) while a negative correlation was observed under water stress conditions (r=-0.18). The relationship between POX and APX was positive and significant under normal water treatment (r = 0.38). Also, there was a positive correlation between protein and H₂O₂ in normal water conditions (r = 0.57). The relationships between drought tolerance indices with YLD were positive and significant under both water conditions (r = 0.59–0.91). A positive correlation was observed between H₂O₂, and Pro under both water conditions (r = 0.29–0.30) (Table 4). There was a negative correlation between drought tolerance indices (CSI) and MDA (r=-0.24).

Table 4 Correlation coefficients among different traits, STI and YSI of 99 SHW population evaluated under normal (above diagonal) and water-stressed (below diagonal) environments

Discussion

Due to a genetic bottleneck within wheat germplasm during its evolution and unexpectedly increased drought severity in the near future, there is a need to develop new wheat varieties. To broaden the genetic diversity of bread wheat pedigrees, SHW-DL is suggested as a bridge. Previous studies have mainly focused on SHW-DL aspects related to biotic stresses. Also, several studies were performed to evaluate the different mechanisms of drought stress on various kinds of wheat. Still, little is known about the tolerance reaction of SHW-DL in dry regions. Therefore, exploiting their genetic potential under abiotic stresses such as drought is needed to facilitate their utilization in variety development. In this regard, 99 wheat genotypes, including 91 SHW-DL and eight varieties of common wheat, as control, were evaluated under two water environments. Based on physiological, biochemical, antioxidant enzyme activities and drought tolerance indices, the response of synthetic and common wheat genotypes was assessed, and the superior SHW-DL genotypes were identified.

The physiological traits investigated on the SHW-DL significantly varied among genotypes in two water environments during two experimental years. High genotypic coefficient of variation (GCV), as the suitable measurement of genetic diversity, was observed for all measured traits, especially for POX (53.38), APX (39.53), MDA (38.64), and H₂O₂ (34.47). In line with Blum [10] and Ebrahimiyan et al. [18], the genotypic variation related to most traits, such as yield, decreased under the water stress environment compared to the normal condition. In addition, broad-sense heritability (h²) of most traits showed wide genetic variation. Dabholkar [15] defined heritability estimates between 5 and 10% as low, 10–30% as medium, and > 30% as high. In the current research, the heritability of most variables was classed as high, indicating that this panel of SHW-DL could be used as an excellent resource in wheat breeding programs. Saed-Moucheshi et al. [43] reported high heritability for SOD, MDA, H₂O₂, APX, YLD, and Pro in triticale under water stress conditions. A high heritability for CAT was also reported under saline environments in wheat [8].

According to our results, water stress treatment reduced grain yield by 31% compared to the control condition. Pradhan et al. [39] also reported a 36% reduction in wheat production due to water stress. The results of this study indicated that water stress led to initiating ROS scavenging processes through enhancing antioxidant enzyme activities (CAT, POX, APX) and the content of non-enzymatic antioxidants (such as proline) to regulate the intracellular level of H₂O₂. ROS increase resulted in decreased YLD, LAI, and RWC. Under water stress conditions, reduced yield is attributed to lower water potential, stomatal conductance, transpiration, photosynthetic rates, and dry matter buildup [1, 31]. In the absence of water, plants maintain osmotic adjustment by maintaining a high water content. In response to drought stress, the accumulation of proline in plants can protect enzymes as an osmolyte and increase membrane stability [36]. Our results showed that genotypes with high yield potential and drought tolerance had higher content of proline compared to sensitive ones, which agrees with previous research [34]. Moreover, previous research reported that most of the pigment content, such as chlorophyll, carotenoid, and chlorophyll stability index, decreased in the early period of drought stress but increased in later periods [47]. Our results showed that the pigment content increased in water stress conditions when leaves were sampled after ten days of applying water stress. Tang et al. [46] reported that SHW-derived varieties had higher chlorophyll contents in leaves and better performance under adverse environments than non-synthetic-derived varieties at most times after anthesis. They also reported that the increase in the contents of Chl a, Chl b, and Chl a + Chl b with the aggravation of drought stress protects the photosynthetic system and reduces the damage. In addition, it was reported that the decrease in pigment content might be due to the decrease in RWC [47]. The later increases may be attributable to higher moisture at the inner surface as a result of leaf rolling which leads to boost resistance to stress [32]. Furthermore, MDA has been proven to be an essential biomarker of oxidative stress [27]. Our results indicated that minimum MDA and highest pigment content were recorded in genotype 54, which was a drought-tolerant genotype. Different responses to water stress within the germplasm and between synthetic and common wheat suggest that selection for adaptive and suitable genotypes is possible for drought tolerance in this germplasm.

The results showed thatSHW-DL performed better under water stress compared rather than common wheat, especially in terms of YLD, Tch, and RWC. The higher RWC and chlorophyll content may have played a role in SHW-DL having a higher YLD. Our findings were consistent with previous research that revealed synthetic hexaploid wheats performed better compared to conventional bread wheats using different morphological and physiological traits under control and osmotic stresses [48]. Synthetic wheat can relocate metabolites quicker from leaves and stems to developing grain under drought stress [48]. Our results indicated that SHW-DL had more Tch and Chl b under water stress. Pradhan et al. [39] showed that the decrease in grain yield of synthetic wheat (25%) was lower than bread wheat (47%) which may be due to the amount of leaf chlorophyll under drought. The lower rate of MDA increase in synthetic genotypes also indicates that SHW-DL has improved antioxidant systems to cope with droughts.

Based on the correlation and PCA analysis, three tolerant indices of STI, YSI, and CSI positively correlated with yield in both moisture conditions. STI and YSI can help distinguish genotypes with different drought tolerance and stability levels, respectively. Moreover, YSI had a significant positive relationship with RWC, indicating that selection based on higher RWC may lead to higher stable genotypes under drought, which agrees with Ebrahimiyan et al. [18]. Also, a negative association between the drought tolerance index (CSI) and MDA might be used to select drought tolerance indirectly. This study used a combination of significant indices (CSI) as a criterion to identify drought-tolerant lines, like that in the previous study [42]. Utilizing physiological traits, productivity, and stress tolerance indices, drought-tolerant genotypes were identified. In this regard, genotypes 54, 98, 102, 105, 122, 124, 143, 159, 196, and 198 revealed higher rank. These synthetic hexaploid wheat-derived lines with higher values of enzyme activities and pigment content can be introduced as genotypes with performing better under water stress conditions in wheat breeding programs.

Conclusions

High genetic variation in terms of antioxidant enzyme activities, the content of non-enzymatic antioxidants, productivity, and drought tolerance indicated that the studied SHW-DL population constitutes functional genetic variability for productivity and drought tolerance. This potential is helpful for wheat breeding to face global environmental changes, especially in arid environments. The SHW-DL performed better in comparison with common wheat lines due to seed yield under water stress which indicated that SHW-DL has recognizable effects to enhance genetic diversity and adaptive potential of modern hexaploid wheat. The high productivity potential of SHW-DL genotypes under drought was correlated with improved physiological traits such as higher maximum photosynthetic pigments. Also, the minimum MDA and highest pigment content were recorded in genotypes with the highest drought tolerance. Genotypes with the highest drought tolerance (54, 98, 102, 105, 122, 124, 143, 159, 196, and 198) were identified to be directly used in breeding programs or indirectly by crossing them with drought-sensitive and high-yielding germplasm. These findings have significant implications in breeding drought-tolerant wheat varieties using SHW as genetic resources to cope with the genetic bottleneck.

Methods

Plant material and experimental site

In the present study, 99 wheat accessions including 91 SHW-DL and eight bread wheat cultivars were evaluated under two different moisture environments of normal and water stress. SHW-DL was provided from the gene bank at CIMMYT. The bread wheat cultivars, including AAC Scotia, Carbery, Norwell, and Sable from Canada and Pishtaz, Qhods, Kavir, and Roshan from Iran, were used as control (Table S3). Each SHW-DL was developed by crossing an Ae. tauschii accession with a modern tetraploid wheat genotype. The initial crosses used 19 different Ae. tauschii and 13 different tetraploid accessions, resulting in 23 primary SHWs (see Table S4). In the panel of synthetic lines, at least one cross has been made with elite hexaploid wheat, resulting in different degrees of synthetic genetic material, ranging from 2nd -degree (primary SHW crossed to one common hexaploid line) to 5th -degree synthetic (primary synthetic crossed to four common hexaploid lines).

The experiments were performed during two winter cropping seasons (October to June of 2018–2019 and 2019–2020) at the Research Farm of Isfahan University of Technology in Najaf Abad, Iran (32 ̊ 30′ N, 51 ̊ 20′ E). A simple lattice square design (11 × 11) was used in the field and some extra genotypes were used as filler. Using long-term climatic data, the station’s mean annual temperature and precipitation were 14.5° C and 140 mm, respectively. The trend of temperature and humidity of Najaf Abad during the growing season (Oct-Jun) of 2018–2019 and 2019–2020 was depicted in Fig. 3. There is silty clay loam soil at the site with a pH 8.3. It is calcareous, containing 390 g kg^–1 ca. carbonate equivalent, 4.0 g kg^–1 organic carbon, and 0.77 g kg^–1 total nitrogen. There were no saline or sodic elements in the soil. In Table S5, soil analysis is shown for two growing seasons (Oct-Jun) 2018–2019 and 2019–2020.

Fig. 3

The trend of temperature and humidity during the growing season of wheat (Oct-Jun) 2018–2019 and 2019–2020

Growth conditions and irrigation treatments

Planting rows were 2m long and 20 cm apart in each plot, resulting in a plant density of 300 plants per m². During both consecutive years, four replications of the experiments were divided into two water regimes of normal and stressed conditions (Figure S1). Water stress was applied at the flowering stage. To apply water stress, irrigation was administered in stress treatment when 85–90% of water at field capacity (FC) had been depleted from the root zone. A 40–45% FC irrigation was maintained until maturity for normal water treatment [3]. Soil moisture was measured by standard gravimetric methods [14] at depths of 0–30 and 30–60 cm using the auger. Based on the following Eq. (1), the irrigation depth was determined:

$$\:I\:=\:\left[\right(FC-\theta\:)/100]\:D\times\:B$$

(1)

where I is the irrigation depth (cm), FC is the soil gravimetric moisture percent at field capacity, θ is the soil gravimetric moisture percent at irrigating time, D is the root zone depth (60 cm), and B is the soil bulk density at root zone (1.4 g cm^− 3). A basin irrigation system was used to apply the water. Using a volumetric counter, we measured the volume of water applied for each treatment. Based on the following Eq. (2), the irrigation efficiency (Ea) was assumed to be 75% when determining the depth of irrigation (Ig):

$$\:Ig\hspace{0.17em}=\hspace{0.17em}I\:\times\:\:100/\:Ea$$

(2)

Trait measurements

Immediately after ten days of applying water stress, RWC and Leaf area index (LAI) were performed. For the measurement of physiological traits, including proline content (Pro), antioxidant enzyme activities, MDA, H₂O₂, chlorophyll, and carotenoid content, flag leaves of five plants from each genotype were harvested and stored at ultra-low temperature (-80^◦ C) for later use.

Measurement of biochemical traits

Leaf area (m²) per ground area (m^–2) was recorded as LAI. Leaf relative water content (RWC) was measured 10–12 days after water stress treatment following the Barrs and Weatherley [4] method according to the recommended Eq. (3):

$$\:RWC\:\left(\%\right)\:=\:(FW-\:DW)\,\times\,100\:/\:(TW-\:DW)$$

(3)

FW, DW, and TW represent the fresh weight, dry weight, and turgid weight, respectively.

Measurement of antioxidant enzyme activities

Based on Bradford [12], taking the absorbance ratio at 590 nm to 450 nm, protein content in extracts was determined spectrophotometrically using bovine serum albumin (BSA) as a standard. Under the ice-cold condition, antioxidant enzyme activities were measured using 500 mg of powdered leaf tissue and extraction buffer containing 1% polyvinyl pyrrolidone and 0.5% Triton X-100 in 100 mM potassium phosphate buffer (pH 7.0). The extract was centrifuged at 15,000 rpm for 20 min at 4 ̊C and the supernatant was used to assay the antioxidant enzymes. For ascorbate peroxidase activity (APX) (µmole of monodehydroascorbate formed min⁻¹ mg⁻¹ protein), Nakano and Asada (35) were followed spectrophotometrically by a decrease in the absorbance at 265 nm in 2 min. The assay buffer contained 2950 µl potassium phosphate buffer (50 mM, pH 7.0) containing 0.5 mM H₂O₂, 5 mM ascorbate) and 50 µl enzyme extract. APX activity was defined using the extinction coefficient 2.8 mM⁻¹ cm⁻¹. To assay peroxidase (POD) activity (increase in absorbance min⁻¹ mg⁻¹ protein), 50 µl enzyme extract was added to 2950 µl potassium phosphate buffer (50 mM, pH 7.5) containing 9 mM guaiacol and 10 mM H₂O₂. The formation of tetraguaiacol was monitored for 2 min as the absorbance at 470 nm increased. Using the extinction coefficient of 26.61 mM⁻¹ cm⁻¹, the POD activity was computed. Catalase activity (CAT) (µmoles of H₂O₂ decomposed min⁻¹ mg⁻¹ protein) was quantified by consumption of H₂O₂ and monitored spectrophotometrically at 240 nm for 2 min (2). The assay buffer contained 50 mM potassium phosphate buffer (pH 7.0), 50 µl enzyme extract, and 15 mM H₂O₂.

Measurement of other physiological traits

Leaf proline content (Pro) was measured following the Bates et al. [5] method. Leaves (200 mg) were ground in 10 ml of 3% aqueous sulfosalicylic acid, and the extract was filtered. Then, 2 ml of the extract was added to the test tube containing 2 ml of each ninhydrin reagent and glacial acetic acid. The mixture was boiled for one hour in a water bath at 100 ̊ C. The mixture was cooled on ice, and 4 ml of toluene was added and mixed. The toluene phase was separated, and its absorbance was measured at 520 nm using a HITACHI U-1800 (Japan) spectrophotometer against a toluene blank.

Leaf chlorophyll and carotenoid contents were assayed spectrophotometrically following the Lichtenthaler and Buschmann [29] method. First, a leaf sample of 0.1 g was powdered and homogenized in 10 ml of acetone (80%) in the dark. Samples were centrifuged at 5000 × g for 15 min (5810R, Eppendorf Refrigerated Centrifuge, Germany), and the absorption of the extracted solution was recorded at 662 and 645 nm (for Chl a and Chl b) and 470 nm (for carotenoids) against a blank (i.e. 80% acetone) (U-1800 UV/VIS, Hitachi, Japan). Finally, the results were expressed as milligrams of pigment per gram of leaf fresh weight [4, 5, 6, 7].

$$\eqalign{{\text{Chl a }}\left( {{\text{mg/g Fw}}} \right){\text{ = }} & {\text{ }}\left[ {{\text{12}}{\text{.21 }}\left( {{\text{AB}}{{\text{S}}_{{\text{663}}}}} \right){\text{ - 2}}{\text{.81 }}\left( {{\text{AB}}{{\text{S}}_{{\text{646}}}}} \right)} \right]{\text{ }} \cr & {{ \times }}\left[ {\left( {{\text{10 ml Acetone}}} \right){\text{ / 1000}}} \right] \cr}$$

(4)

$$\eqalign{{\text{Chl b }}\left( {{\text{mg/g Fw}}} \right){\text{ = }} & {\text{[20}}{\text{.13 (AB}}{{\text{S}}_{{\text{646}}}}{\text{) - 5}}{\text{.03 (AB}}{{\text{S}}_{{\text{663}}}}{\text{)]}}\; \cr & {{ \times }}\;\left[ {\left( {{\text{10 ml Acetone}}} \right){\text{ / 1000}}} \right] \cr}$$

(5)

$$\eqalign{{\text{CAR }}\left( {{\text{mg/g Fw}}} \right){\text{ = }} & {\text{[[[1000 (AB}}{{\text{S}}_{{\text{470}}}}{\text{) - 3}}{\text{.27 Chl a -- 104Chl b}}\left] {{\text{ / 227}}} \right] \cr & {\text{/ Weight }}\left( {\text{g}} \right)\left] {{{ \times }}} \right[\left( {{\text{10 ml Acetone}}} \right){\text{ / 1000]}} \cr}$$

(6)

$$\:Tch\:(mg/l)\hspace{0.17em}=\hspace{0.17em}Chl\:a\hspace{0.17em}+\hspace{0.17em}Chl\:b\:$$

(7)

Lipid peroxidation level was measured in terms of malondialdehyde (MDA) content following Heath and Packer [24] and H₂O₂ content according to Loreto and Velikova [30] methods. Homogenized fresh leaf (approximately 0.2 g) in cold TCA (5 mL of 0.1%) was centrifuged at 12,000 × g for 15 min at 4 ̊ C and then, 0.5 mL of the supernatant was added to 0.5 mL of 10 mM potassium phosphate buffer (pH 7.0) and 1 mL of 1 M KI. The absorbance of the mixture was measured at 390 nm by a UV-visible spectrophotometer. The H₂O₂ content was determined using a standard curve prepared with known concentrations of H₂O₂ and expressed as mmol g⁻¹fresh weight.

Estimation of drought tolerance indices

Stress tolerance index (STI), yield stability index (YSI), and combination of significant indices (CSI) were determined, suggested by Fernandez (1992), using the following formulas (8, 9 and 10):

$$\:STI\:=\:\left(Yp\:*\:Ys\right)/\left(Xp\right){^2}\:$$

(8)

$$\:YSI\hspace{0.17em}=\hspace{0.17em}Ys\:/\:Yp\:$$

(9)

$$\eqalign{{\text{CS}}{{\text{I}}_{\text{i}}}{\text{ = 1/2 }} & {\text{[}}\left( {{{\text{r}}_{{\text{YP}}{\text{.MP}}}}{{ \times M}}{{\text{P}}_{\text{i}}}} \right){\text{ + }}\left( {{{\text{r}}_{{\text{YP}}{\text{.GMP}}}}{{ \times GM}}{{\text{P}}_{\text{i}}}} \right){\text{ + }}\left( {{{\text{r}}_{{\text{YP}}{\text{.HM}}}}{{ \times H}}{{\text{M}}_{\text{i}}}} \right)\cr{\text{ + }} & \left( {{{\text{r}}_{{\text{YP}}{\text{.STI}}}}{{ \times ST}}{{\text{I}}_{\text{i}}}} \right){\text{ + }}\left( {{{\text{r}}_{{\text{YS}}{\text{.MP}}}}{{ \times M}}{{\text{P}}_{\text{i}}}} \right){\text{ + }}\left( {{{\text{r}}_{{\text{YS}}{\text{.GMP}}}}{{ \times GM}}{{\text{P}}_{\text{i}}}} \right)\cr{\text{ + }} & \left( {{{\text{r}}_{{\text{YS}}{\text{.HM}}}}{{ \times H}}{{\text{M}}_{\text{i}}}} \right){\text{ + }}\left( {{{\text{r}}_{{\text{YS}}{\text{.STI}}}}{{ \times ST}}{{\text{I}}_{\text{i}}}} \right) \cr}$$

(10)

where Ys is the grain yield of a test genotype under water stress conditions; Yp is the grain yield of a test genotype under normal conditions, and Xp is the mean yield of test genotypes under normal conditions.

Statistical analyses

Prior to performing the ANOVA analysis, the Smirnov-Kolmogorov normality test was performed to verify the normal distribution of the residues using SAS 9.440 software [44]. Normal distribution was achieved without transforming data. The relative efficiency of the experimental designs of Lattice design was less than randomized complete block design (RCBD). Therefore, phenotypic and physiologic data were analyzed separately following the RCBD procedure of SAS 9.440 software. To partition the total variation into components based on the expected mean square in SAS 9.440 software, an analysis of variance (ANOVA) was conducted. If the F value was significant in ANOVA, the least significant difference (LSD) value was calculated at the 5% probability level. Pearson’s correlation coefficients (r) among the physiologic and phenotypic traits were calculated separately for the normal and water stress treatments using PROC CORR in SAS software. Based on a correlation matrix constructed over all replicate data, principal component analyses (PCA) were performed to reduce data dimensions and simplify genotype selection. To show the relationships among studied genotypes based on recorded traits, PCA bi-plots were plotted separately for normal and water stress conditions. The following Eq. (11) was used to calculate the genetic coefficient of variation (GCV):

$$\:GCV\:=\:(g\:/\mu\:)\:*\:100\:$$

(11)

where σ_g and µ are the square root of the genotypic variance, and phenotypic mean, respectively [37].

The expected mean squares of ANOVA were used to estimate heritability [12, 20]:

$$\text{h}^{2}=\sigma^{2}_{\text{g}}/(\sigma^{2}_{\text{g}}+\sigma^{2}_{\text{e}}/\text{r})$$

(12)

where σ²_g, and σ²_e are the variance components of genetic and error, respectively.