Assessment of Durum Wheat Genotypes by Drought Tolerance Indices under Drought and Normal Condition

Wheat is the most important crop in the world and it is cultivating in about 220.19 million hectares around the world. Wheat demand is increasing with the continuous increase in human population and it is expected to reach up to 40% in 2030. Mottaleb et al., (2023). However, the world’s wheat consumption is expected to expand beyond production raising concerns about future food security. In India, during 2018-19 rabi season, wheat was cultivated in 29.55 mha constituting 24.35 per cent of the total crop acreage and production in 2018-19 has made a landmark achievement by producing 101.20 mt by registering another record in average national productivity i.e. 3424 kg/ha. (Anonymous, 2019). Drought (water deficits) and heat (high temperatures) stress are the major abiotic constraints, under the current and climate change scenario in future.
               
Reforming drought tolerance in wheat species is one of the most important solutions to fight the drought. It makes identifying drought tolerant genotypes difficult. Use of high yielding genotypes having drought tolerance is an efficient approach to lessen its damaging effects. With declining resources of water and escalating intensity of drought, loss of yield is a dangerous alarm in these regions. However, attaining drought tolerance exclusively yield depended is complex due to its intricate heritability. Likewise, choosing genotypes having tolerant genes is a tricky task (Mitra, 2001). Alternatively, some statistical parameters as well as drought tolerance indices could be employed to compare the changes in grain yield in normal and drought conditions for the assortment of genotypes of high yields and drought tolerance. They have been widely used for screening drought-tolerant genotypes in durum wheat (Patel et al., 2019). On the other side, performance increase in water shortage condition requires identifying drought-tolerant genotypes and management affairs to maximize accessible water. Thus, improvement of drought tolerance in crops is a major objective of most crop breeding programs, particularly in arid and semi-arid areas of the world. To characterize, detect and evaluate the response of the population and differentiate drought resistance genotypes, several selection indices have been suggested based on a mathematical relationship between irrigated conditions and drought conditions (Sadeghzadeh Ahari et al., 2009). Thus, the aim of this study was to evaluate the efficiency of several drought tolerant indices for screening and identification of drought tolerant durum wheat genotypes.

Plant material and experimental setup
 
Twenty durum wheat (Triticum durum Desf.) genotypes were used in the study based on their reputed differences in yield performance under irrigated and non-irrigated conditions (Table 1). Experiments were conducted at the research farm, Wheat Research Station, Vijapur (latitude 23.350 N, longitude 072.550 E and altitude 126 m above sea level) in rabi seasons of 2017-18 to 2018-19 (two cropping years). The annual average temperature of the study area was 28.6^oC and 12.6^oC (Fig 1) with maximum and minimum temperatures, respectively, with the soil type classified as sandy loam soil with a pH of 7.74. During both the crop seasons no rainfall was received, so experiment was conducted in truly natural drought conditions. Each genotype was sown in four rows 2.5-meter in length and 20 cm apart, with a seed rate of 150 kg/ha in a randomized complete block design replicated twice.  Standard agronomic practices were adopted to raise a good crop.




 
Estimation of drought tolerant indices
 
Drought tolerance indices were calculated using following formulas:

  (Fischer and Maurer, 1978)

  (Fernandez, 1992)

  (Rosielle and Hamblin, 1981)

  (Fernandez, 1992)

  (Rosielle and Hambling, 1981)



  (Gavuzzi et al., 1997; Lin et al., 1986)

  (Bouslama and Schapaugh, 1984)

  
Where,
Ysi and Ypi = Yields of a given genotype under stress and optimum condition, respectively.
Ys and Yp = Average yield of all genotypes under stress and optimal conditions, respectively.
 

 
Statistical analysis
 
Analysis of variance was performed for each index according to Steel and Torrie (1980) by the computer program MSTATC software. Principal component analysis (PCA) was used to classify the screening methods as well as the genotypes. Correlation was also used to identify tolerant and high-yielding genotypes using IRRI Star software, based on principal component analysis.

Pooled analysis of variance
       
The results of pooled analysis of variance in normal and drought stress conditions and stress tolerance indices showed significant differences among the genotypes for Ys, Yp and stress tolerance indices (Table 2). These results indicate presence of high genetic variation among genotypes, which could be a useful source for the selection of drought-tolerant genotypes. Kumar et al., (2021) and Mumtaz and Khan (2023) also reported significant differences between studied genotypes of wheat .


 
Mean values of drought tolerance indices
 
To study suitable stress tolerance indices for the selection of genotypes under drought stress condition, the yield of genotypes under both normal and stress conditions were recorded for calculating different sensitivity and tolerance indices (Table 3). Studies revealed that in wheat, the effect of drought stress is more pronounced during the reproductive stage (Nezhadahmadi et al., 2013). In our study, 11 (55%) accessions from the study panel revealed a GY-SSI value of <1, indicating drought tolerance, whereas 09 (45%) showed a GY-SSI higher than 1, implying that these genotypes are drought susceptible. This suggests that in this study, drought stress was moderate but enough to facilitate the selection of drought-tolerant accessions. Moderate drought stress was reported as recommended to select drought-tolerant wheat lines (Ali and El-Sadek, 2016). In the non-stress condition (Yp) the highest values for grain yield was recorded for genotype GW 1339 (8254.29 kg ha^-1) followed by GW 2017-877(7839.58 kg ha^-1) and GW 2017-864 (7512.50 kg ha^-1), while genotype GW 2017-870 (5158.33 kg ha^-1) followed by GW 2015-699(5462.50 kg ha^-1) and GW 2017-874(5564.58 kg ha^-1) had the lower grain yield in that order. On the other hand, under drought stress condition (Ys) the highest grain yield was observed by genotype GW 1339 (2970.83 kg ha^-1) followed by GW 2017-878(2889.58 kg ha^-1) and HI 8737(2845.83 kg ha^-1). Genotypes GW 2017-866(1881.25 kg ha^-1) followed by GW 2017-854 (1900.00 kg ha^-1) and GW 1349(1947.92 kg ha^-1) recorded lower grain yield.


       
Genotypes GW 2017-878(2854.25), GW 2017-870 (3020.83) and GW 2015-699(3216.67) were found more tolerant based on TOL, in which lower values of TOL identified tolerant genotypes (Table 3). GW 2017-878(2889.58 kg ha^-1) had the second-highest grain yield in drought stress condition. Therefore, it is obvious that TOL index has been relatively successful in stress conditions to select the genotypes which had the high grain yield and lead the selection towards more efficient and tolerant genotypes.
       
The stress intensity (SI) value was 0.65 which provides an opportunity to evaluate durum wheat genotypes under severe stress conditions. Based on the SSI, the genotypes GW 2017-878(0.75), GW 2017-870 (0.85) and HI 8737(0.88) were identified as drought tolerance genotypes, while the genotypes GW 2017-877(1.13) and GW 2017-854(1.11) displayed the higher values of SSI (Table 3). Ghafari (2008) stated that genotype evaluation through SSI, categorizes experimental materials according to tolerance and stress sensitivity. Through this index, tolerant and sensitive genotypes can be specified without regarding their performance potential.
       
According to the MP, the genotypes GW 1339 (5612.56), GW 2017-864(5134.38) and HI 8737(5016.15) were found drought tolerance genotypes and the genotypes GW 2017-870(3647.92), GW 2015-699(3854.17) and GW 2017-874 (3859.38) were identified as drought susceptible ones in stressed condition (Table 3). MP is based on the arithmetic means and therefore, it may have an upward bias due to a relatively larger difference between Ypi and Ysi-. Generally higher MP value is an indicator of genotypes with higher yield potential. So, the MP index leads the selection towards more efficient genotypes in both stress and non-stress conditions. Among the tested genotypes, genotype GW 1339 (2970.83 kg ha^-1, 8254.29 kg ha^-1) had the highest grain yield in stress and non-stress condition. Verma and Singh (2023) also reported GW 1339 as high yielding genotype in central zone of India. The results of this study corresponded to the results of Shirinzadeh et al., (2009).
       
GMP values recorded were highest in genotype GW 1339 (4922.03) followed by GW 2017-864(4530.32) and HI 8737(4434.88). In terms of harmonic mean (HARM), the genotypes GW 1339(4329.06), HI 8737(4013.79) and GW 2017-864(4005.17) were identified as drought tolerant genotypes, while the other remained remaining genotypes showed lower values of HM (Table 3). The results of both GMP and HM indices were completely similar. It seems that this similarity is due to the nature of their calculating formulas and so it is logical to use one of them in future studies. Results for MP, GMP and HARM were following the findings of Mirzaei et al., (2014).
       
The higher values for stress tolerance index (STI) were observed for genotypes GW 1339(0.576), GW 2017-864(0.481) and HI 8737(0.457) in that order and genotypes GW 2017-870 (0.240) and GW 2017-866(0.254) had the lower stress tolerance index (STI). The high values of STI in these genotypes indicated high drought tolerance and high potential yield. Mevlüt and Sait (2011) reported relatively similar ranks for the genotypes observed by GMP and MP parameters as well as STI, which suggests that these three parameters are similar for screening drought-tolerant genotypes.
       
In case of the yield index (YI), the genotypes GW 1339 (1.28), HI 8737(1.23) and GW 2017-878(1.23) were recognized as more drought tolerant genotypes (Table 3).

The higher value for YSI was observed in genotypes GW 2017-878(0.52), GW 2017-870 (0.44) and GW 2015-699(0.42) (Table 3). The genotypes with high YSI are expected to have high yield under stressed and low yield under non-stressed conditions (Mohammadi et al., 2010).  Genotype GW 2017-878(2889.58 kg ha^-1) had a high yield in stress condition, but it didn’t have high yield in normal condition.
 
Correlation analysis
 
Albeit selection based on a combination of indices may provide a more useful criterion for improving drought tolerance, however, correlation analysis between grain yield and drought tolerance indices can be a good criterion for screening the best genotypes and indices. Thus, a suitable index must be significantly correlated with grain yield under both the conditions (Mitra, 2001). A positive and non-significant correlation between (Yp) and (Ys) indicates that the yield under stress condition (Ys) has no significant correlation with the yield under non-stress environment (Yp). This indicates high stress intensity. Similar results were earlier reported by (Alefsi David et al., 2020).
       
A positive correlation between Yp and SSI (r=0.53), Yp and TOL (r=0.90) and a negative correlation between Ys (drought stress) and SSI (r= - 0.59), Ys (drought stress) and TOL (r= -0.08) (Table 4) suggest that selection based on SSI and TOL will result in reduced yield under irrigated conditions. Especially, negative correlation between SSI and Ys was expected because genotypes that suffer less yield loss from irrigated to drought conditions also tend to have high yield in stressful environments. SSI identified some genotypes such as GW 2017-878, GW 2017-870 and GW 2015-699 as stress-resistant though they did not have outstanding yield performance in stress primarily because of their low potential yield (Table 3). On the other hand, the correlation between Yp and SSI was negligible (r= 0.53). The Ys was significantly correlated (P<0.01) with all indices except TOL, whereas Yp was highly significantly correlated with all indices except YI. Highly correlated indices with both the Ys and Yp are most appropriate for identifying stress tolerant cultivars.


       
Both Yp and Ys were significant and positively correlated (P<0.05) and (P<0.01) with, STI (r=0.77 and 0.84), HM (r=0.57 and 0.96), GMP (r=0.78 and 0. 84) and MP (r=0.94 and 0.63) (Table 4). Positive significant correlation between GMP and MP and TOL in both drought and normal conditions shows that their effects are stronger than those of SSI and DI. These observed relationships are in consistence with numerous studies. Indices which had high correlation with grain yield in both stressed and non-stressed conditions have been selected as the best ones, because these were able to separate and identify genotypes with high grain yield in both conditions. With respect to the results of correlation coefficients of different indices and grain yield in two drought stress and normal irrigation conditions, genotypes which had higher values of these indices were more effective in identifying high high-yielding lines under drought stress as well as non-stress conditions. Many studies reported positive relationships between Ys and the most popular and widely used indices MP, GMP, STI, SSI, TOL (Farshadfar et al., 2012; Naghavi et al., 2013). Karimizadeh and Mohammadi (2011) were in opined that MP, GMP, STI, HARM and YI indices are preferred for practical usage. The observed relations were in consistence with those reported by Talebi et al., (2009) and Mohammadi et al., (2010) in durum wheat. 
       
In drought stress condition, YSI had a positive and significant correlation with grain yield(r=0.597). While, it had a negative and significant correlation with grain yield in normal condition (r= - 0.532). So, it cannot be a proper index for selecting the genotypes which have a high yield in normal irrigation and drought stress conditions. (Sio-Se Mardeh, 2006). Genotype GW 2017-878(2889.58 kg ha^-1) had high yield in stress condition, but it didn’t have a high yield in normal condition. The genotypes with high YSI are expected to have high yield under stressed and low yield under non-stressed conditions (Mohammadi et al., 2010).
 
Principal components analysis
 
In further evaluation of relations between genotypes and drought tolerance indices, principal components analysis was performed. Table 5 shows latent roots and special vector of under-study genotypes for the first two components, the most variations between data expressed by two components (99.00%). The first vector showed 61% of variations and showed that indices GMP, STI, HARM, MP, YI, Ys and Yp in the formation of this component has the highest positive coefficient, since high values of these indices were optimal and considering the positive relation of the first component with these indices, if we selected the top level, the genotypes were selected which had high and stable yield in different environments (drought stress, non-stress). So, this component was identified as drought tolerant component (Farshadfar et al., 2001). The second component had 38% of these variations (Fig 2). This component has high and positive correlation with YSI, Ys and YI. Tahmidul et al., (2023) and Sheeba and Mohan (2025) obtained similar results in first principal component analysis of drought tolerance in wheat and rice crop.

In the present investigation, significant differences among the grain yield under stress, non-stress condition and stress indices were observed suggesting presence of genetic variation among genotypes for drought. Based on indices and association results STI, HARM, GMP and MP indices were the best predictors of yield under water-stressed and non-stress environments.  Based on indices selection, principal component analysis GW 1339, GW 2017-864 and HI 8737were the high-yielding and drought tolerant genotypes among the 20 lines evaluated. These lines could be a good source of germplasm for deriving drought tolerant lines. Principle component analysis showed that most variance among data is justified by first two components with 99% of the total changes.
 
Disclaimers
 
The views and conclusions expressed in this article are solely those of the authors and do not necessarily represent the views of their affiliated institutions. The authors are responsible for the accuracy and completeness of the information provided, but do not accept any liability for any direct or indirect losses resulting from the use of this content.

The authors declare that there are no conflicts of interest regarding the publication of this article. No funding or sponsorship influenced the design of the study, data collection, analysis, decision to publish, or preparation of the manuscript. Informed consent. All animal procedures for experiments were approved by the Committee of Experimental Animal care and handling techniques were approved by the University of Animal Care Committee.