Evaluation of Genetic Variation for Drought Tolerance and Determination of the Best Selection Criteria in Safflower Genotypes (Carthamus tinctorius L.)

Abstract

 
Abstract
In order to evaluate genetic variation and drought tolerance of safflower cultivars (Carthamus tinctorius L.), an experiment was conducted using fifteen cultivars in a randomized complete block design with three replications under drought and normal conditions during 2014-2015 farming season. Drought tolerance indices, such as tolerance (TOL), stress tolerance index (STI), stress susceptibility index (SSI), mean productivity (MP) and geometric mean productivity (GMP) were calculated to distinguish cultivars having the best seed yield as well as drought tolerance. The correlation coefficients illustrated that STI and GMP were the best and efficient selection criteria to distinguish drought tolerant and high-yielding cultivars. Significant and positive correlation was found between yield in both stress and normal conditions with GMP, MP and STI. Principal component analysis (PCA) showed that first and second PC accounted for 97.1% of the total variation. Biplot graphical display represented that lines 2, 11, 14 and 15 were highly adapted to the both normal, stress conditions, and classified them in high-yielding and drought tolerant groups, while genotypes numbered as 10, 12 and 13 were potential and stable under normal. condition. Based on data analysis, cultivars numbered as 1, 5, 6 and 9 had lowest yield under both moisture regimes, lines 3, 4, 7 and 8 showed high-yielding under stress regimes. Cluster analysis ordered the genotypes into six groups with 5, 3, 2, 2, 2 and 1 genotypes, respectively. In conclusion, present investigation revealed that drought conditions induced reduction of yield of some cultivars, while others were tolerant to drought stress. Hence, breeders can select drought tolerant safflower lines based on the GMP and STI indices.
   
 

Keywords


 

 

 

       
 

Original Research

 
   

             Research on Crop Ecophysiology                                 Vol.11/2  , Issue 2 (2016), Pages: 58 - 67

 

 

 
 

 

 

 

 

 


Evaluation of Genetic Variation for Drought Tolerance and Determination of the Best Selection Criteria in Safflower Genotypes (Carthamus tinctorius L.)

 

 

 

 

 

Ahmad Reza Golparvar1*, Kourosh Keighobadi2, Mohammad Mehdi Gheisari3, Amin Hadipanah4

 

1-Department of Agronomy and Plant Breeding, College of Agriculture, Isfahan

 

(Khorasgan) Branch, Islamic Azad University ,Isfahan, Iran

 

2-Plant Improvement and Seed Production Research Center, Isfahan (Khorasgan) Branch, Islamic Azad University ,Isfahan, Iran

 

3-Toxicology Research Center, Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan, Iran

 

4-Department of Horticultural, Science and Research Branch, Islamic Azad University, Tehran, Iran

 

 

 

 

 

* Corresponding author email: dragolparvar@gmail.com

 

Received: 5 March 2016                                                                          Accepted:  20 April 2016

 

 

 

Abstract

 

In order to evaluate genetic variation and drought tolerance of safflower cultivars (Carthamus tinctorius L.), an experiment was conducted using fifteen cultivars in a randomized complete block design with three replications under drought and normal conditions during 2014-2015 farming season. Drought tolerance indices, such as tolerance (TOL), stress tolerance index (STI), stress susceptibility index (SSI), mean productivity (MP) and geometric mean productivity (GMP) were calculated to distinguish cultivars having the bestseed yield as well asdrought tolerance. The correlation coefficients illustrated that STI and GMP were the best and efficient selection criteria to distinguish drought tolerant and high-yielding cultivars. Significant and positive correlation was found between yield in both stress and normal conditions with GMP, MP and STI. Principal component analysis (PCA) showed that first and second PC accounted for 97.1% of the total variation. Biplot graphical display represented that lines 2, 11, 14 and 15 were highly adapted to the both normal, stress conditions, and classified them in high-yielding and drought tolerant groups, while genotypes numbered as 10, 12 and 13 were potential and stable under normal. condition. Based on data analysis, cultivars numbered as 1, 5, 6 and 9 had lowest yield under both moisture regimes, lines 3, 4, 7 and 8 showed high-yielding under stress regimes. Cluster analysis ordered the genotypes into six groups with 5, 3, 2, 2, 2 and 1 genotypes, respectively. In conclusion, present investigation revealed that drought conditions induced reduction of yield of some cultivars, while others were tolerant to drought stress. Hence, breeders can select drought tolerant safflower lines based on the GMP and STI indices.

 

   

 

Key words: Safflower, Drought tolerance indices, Genetic improvement, Biplot, Cluster analysis.

 

 

 

Introduction

 

 

 

Safflower (Carthamus tinctorius L.) is one of the seed oil crops grown in Iran. It is one of the plants, which have a high conformity to various conditions such as resistance to drought, and it is appropriated to be grown in arid and semi-arid areas. Due to the growing request for edible oils, improvement of oilseed crop is very important (Safavi et al., 2013; Rameshknia et al., 2013). In normal condition plants are subjected to various stresses factors with harmful influence on growth and crop production (Roudbari et al., 2012). Drought as an environmental stress is the limitation that induces a highly negative effect on yield (Khalili et al., 2014). Drought tolerant is an important characteristic for increasing and enhancing crop efficiency in dry regions (Guo et al., 2009). Recognition of the important yield component is very efficient in genetic programs of these traits via indirect selection (Golparvar, 2011). The identified genes from wild plant species provide a mean for sustaining genetic imorovment in plant cultivated in dry regions. The cultivated lines tolerated less drought stress than wild plants and fluctuating water stress levels caused meaningful more declines in the seed yield of cropping genotypes as compared with wild genotypes (Majidi et al., 2011). The inheritance of agronomic traits was studied in safflower under drought stress condition. In order to improve seed yield and seed yield of safflower under drought regimes, obtained outcomes could be suitable for designing of breeding programs (Mirzahashemi et al., 2014). Detection of the drought tolerance genes in barley (Hordeum Vulgare L.) will facilitate the molecular mechanisms conception of drought tolerance, and also facilitate the genetic breeding of barley via marker-assisted selection or transformation of genes. These results showed that new understanding into further comprehension of drought tolerance procedures in barley plants could be provided (Guo et al., 2009). Various plants reacte to drought stress differently. Drought condition induced varied molecular and physiological responses such as changes in gene expression in plants (Savitri et al., 2013). Shiranirad et al. )2011) announced that drought is a common obstacle seriously influencing rapeseed (Brassica napus L.) production, mostly in arid region in the world. They observed that MP, GMP and YI parameters were the best for screening high seed yield genotypes under stress conditions. Farshadfar et al. (2013) cleared that grain yield of bread wheat in normal and stress regimes were significantly and negatively correlated with SSI. Their findings indicated that some indices such as RDI, ATI, SNPI and DI can be used as the most favorable indicators for identifying drought tolerant genotypes. Cluster analysis of mentioned investigation classified the cultivars into three groups including; tolerant, susceptible and semi tolerant or semi-sensitive to drought regimes. In order to assess of drought tolerance indices under different environmental conditions for screening of Turkish oat (Avena sativa L.), fourteen landraces and cultivars were used. The experiments were applied both under rain-fed and irrigated regimes for three cropping seasons. Correlation coefficient matrix showed that the drought parameters were significantly inter-correlated with each other and can be classified into four groups. Their results demonstrated that the STI, GMP, MP, YI and HM indices under dry and irrigated conditions can be suggested to screen drought tolerant cultivars with high-yielding potential (Akcura et al., 2011).

 

   

 

Materials and methods

 

 

 

A filed trial was carried out during 2014-2015 at the Isfahan (Khorasgan) Branch, Islamic Azad University, research station (50˚ 44΄ N, 32˚ 40΄ W and altitude of 1517 m above mean sea level). The study location is characterized by arid climate with an annual average rainfall of 120 mm, and the annual mean maximum and minimum temperatures of25 °C and 1 °C, respectively. Soil type of the study site was silty loam and soil pH was 7.7 to 8. Generally, there is no precipitation during safflower growth cycle in this region.

 

Fifteen spring safflower cultivars including U.S.10, Kuseh landrace, Nebraska-10, Gila, S149, Bushehr landrace, Shiraz landrace, Arak-2811, Kerman landrace, Isfahan landrace, C111, Lordegan landrace, S3110, A.C.Sterling and Semnan landrace were planted at first  of March 2012. The plots comprising of three rows were 3 m long and 0.5 m apart. Interplants distance within rows was 5 cm, hence, seedling density was 400/000 plants ha-1. The experiment was watered at planting and flowering stages. Irrigation regimes were started at emerging of seedling. Two irrigation programs were considered in this study: IR1, irrigation after 75 mm cumulative evaporation from class A evaporation pan (CE) during the whole growth cycle as optimum irrigation treatment. IR2, irrigation after 150 mm cumulative evaporation from class A evaporation pan (CE) during the whole growth cycle as stress treatment. Various drought tolerance indices were evaluated (Table 1).

 

 

 

Table 1. Drought tolerance indices to calculate the reaction of safflower cultivars to stress

 

Code

Drought tolerance indices

Equation*

References

1

Stress Susceptibility Index (SSI)

SSI =

Fischer and Maurer, 1978

2

Geometric Mean Productivity (GMP)

GMP=

Fernández et al., 1992

3

Stress Tolerance Index (STI)

STI =

Fernández et al., 1992

4

Mean Productivity (MP)

MP =

Rosielle and Hambling, 1981

5

Tolerance Index (TI)

TOL=

Rosielle and Hambling, 1981

 

* Ys and Yp are seed yield in stress and normal conditions, respectively.

 

The research was conducted in two independent randomized complete block design (as stress and non-stress conditions) with three replications in each experiment. Analysis of variance and Duncan’s multiple range test was utilized for mean comparisons were applied using proc GLM procedure of SAS software (version 9.2, SAS institute Inc., NC, USA). Correlation analysis, principal component analysis (PCA) and biplot graphical display were done by using STATGRAPHICS PLUS software while cluster analysis based on Ward’s method was carried out by SAS9.2 software.

 

   

 

Results and discussion

 

 

 

Yield ranged from 3534 kgha-1 (cultivar S149) to 1366 kgha-1 (cultivar Arak-2811) in non stress treatment (Yp) (Table 2). The values of yield under stress conditions varied from 1326 to 1966 kgha-1 and the Semnan and Lordegan landraces had lower seed yields Ys. Gila, U.S.10 and Nebraska-10 showed higher yield (1966, 1946 and 1934 kgha-1, respectively) in Ys non-stress. Mean yields under and stress conditions, were 2253 and 1657 kg ha-1, respectively revealing a reduction of 27% compared to normal irrigation conditions (data not shown).

 

 

 

Table 2. Average values of drought tolerance indices in safflower cultivars

 

Code

genotype

Yp

Ys

TOL

MP

SSI

GMP

STI

1

Esfahan landrace

1714

1472

242

1593

 0.52

 1588.39

  0.49

2

Kuseh landrace

2392

1740

652

2066

  1.01

 2040.11

  0.81

3

Arak-2811

1366

1686

-320

1526

      -0.87

 1517.58

  0.45

4

Nebraska-10

1838

1934

-96

1886

      -0.19

 1885.38

  0.69

5

Semnan landrace

1566

1326

 240

 1446

 0.57

 1441.01

  0.4

6

Lordegan landrace

1494

1326

 168

 1410

 0.42

  1407.49

  0.39

7

Bushehr landrace

1496

1846

       -350

1671

      -0.87

  1661.81

  0.54

8

Shiraz landrace

2146

1680

466

1913

 0.81

 1898.75

  0.7

9

Kerman landrace

2234

1360

874

1797

  1.46

 1743.05

  0.59

10

A.C.Sterling

2600

1446

 1154

2023

  1.66

 1938.96

  0.74

11

S3110

3146

1820

 1326

2483

  1.57

 2392.84

  1.12

12

C111

3080

1466

 1614

2273

  1.96

 2124.91

  0.88

13

S149

3534

1754

 1780

2644

  1.88

 2489.7

  1.22

14

U.S.10

2794

1946

848

2370

  1.13

 2331.76

  1.07

15

Gila

2406

1966

440

2186

  0.68

 2174.9

  0.93

 

 

 

The data indicated that drought stress could significantly reduce yield. The genotypes S149 and S3110 showed properseed yield under both moisture regimes (Table 2). The values of mean productivity (MP) varied from 1410 kg ha-1 (Lordegan landrace) to 2644 kg ha-1 (line S149) and the genotypes S149, S3110, U.S.10, C111, Gila, Kuseh landrace, A.C.Sterling and Esfahan landrace were the most productive (1955 kg ha-1). Based on geometric mean productivity (GMP), yield varied from  1407.5 kg ha-1 (Lordegan landraces) to 2490 kg ha-1 (line S149), proposing that the genotypes 13, 11, 14, 15, 12, 2 and 10 were the most productive. TOL index varied from -350 to 1780 kg ha-1. Lower or negative TOL indices show tolerance to irrigation stress. Hence, Esfahan, Semnan and Lordegan landraces, Nebraska10, A.C.Sterling, S3110, C111 and S149 were more tolerant (297 kg ha-1). Stability tolerance index (STI) ranged from  0.39 (Lordegan landrace) to  1.22 (S149). The value of stress susceptibility index (SSI) varied from -0.87 (lines Arak and Bushehr landraces) to 1.96 (line C111). To detecti the most desirable drought tolerance measures, correlation coefficient between yields under non-stress and stress conditions, and other quantitative indices of drought tolerance were estimated (Table 3). The outcomes indicated that the indices GMP, MP, STI and SSI were very similar for selection as Yp. This was supported by the high correlations among Yp and SSI (r= 0.84), TOL (r= 0.94), MP (r=0.95), GMP (r= 0.92) and STI (r=0.93). Correlation analysis demonstrated that the indices GMP and STI were similar for selection as Ys. Correlations between yields under stress regimes and GMP (r= 0.58) and STI (r=  0.57) confirmed this conclusion. The indices SSI, TOL and MP illustrated the lowest correlation with Ys (Table 3). Results of Safavi et al. (2013), investigations indicated that significant positive correlation was observed between grain yield in the drought regimes (Ys) with indicator stress tolerance index (STI), harmonic mean (HAR) and geometric mean productivity (GMP) and therefore these indices were suitable criteria for screening stress tolerant cultivars. Majidi et al. (2011), believed that GMP, STI and HM are superior criteria for identifying high yield genotypes under drought and normal regimes. The present results verified significant and positive correlation amongst Yp and Ys with GMP and STI; so these indices may be better predictors of Yp and Ys than MP, SSI and TOL indices. Our findings are in coincident with study of Rameshknia et al. (2013) who believed STI and GMP indices were the best parameters for identification and screening of genotypes under normal and stress regimes in breeding programs. Safavi et al. (2013), also stated that tolerant index (TOL) and mean productivity (MP) can be regarded as desirable indices for detecting drought tolerant genotypes. Khalili et al. (2014) announced STI, MP, GMP and YI indices were the most appropriate criteria in safflower breeding plans and they revealed that these indices were used for screening high-yielding cultivars under both normal and stress conditions. In assessment of genetic properties of drought tolerance indices of durum wheat, the parameters such as; MP, GMP and STI had high positive genetic correlations with each other as well as with grain yield under stress regime (Ys) and normal condition (Yp). Hence, through these indices it is possible to select high-yielding cultivars in either conditions (Hussain Ali, 2015).

 

 

 

     

 

Table3. Pearson’s correlation coefficients among drought tolerance indices

 

index

Yp

Ys

SSI

TOL

MP

GMP

STI

Yp

1

 

 

 

 

 

 

Ys

  0.249ns

1

 

 

 

 

 

SSI

  0.849**

      -0.231ns

1

 

 

 

 

TOL

  0.94**

      -0.095ns

  0.955**

1

 

 

 

MP

  0956**

  0.521*

  0.678*

  0.799**

1

 

 

GMP

  0.927**

  0.587*

  0.634*

  0.746**

  0.994**

1

 

STI

  0.93**

  0.574*

  0.635*

  0.754**

0.993**

  0.996**

1

 

For abbreviations, see Table1.

 

 *, ** significant at 0.05 and 0.01 probability levels; ns: not significant.

 

 

 

SSI values changed from -0.69 – 1.54, which were significantly and positively correlated with yield under non-stress and TOL index and negatively correlated with Ys. MP is the mean production under both moisture regimes, and was highly correlated with Yp and TOL indices. Results of Rameshknia et al. (2013) assessment illustrated that STI, GMP and MP indices could screen tolerant and sensitive genotypes under both environmental conditions, and mentioned indices that could be for selection of tolerant cultivars of spring safflower. TOL varied from -350- 1780 kg ha-1. A positive correlation between TOL and Yp (yield under non-stress conditions) and a negative correlation between TOL and yield under water stress (Ys) offered that selection based on TOL indices resulted in reduced yield under optimum irrigation regime. Hussain ali (2015) revealed that the genetic correlation of TOL and SSI indices with yield under stress conditions were high and negative, while correlation coefficient between TOL index and Yp was high and positive. Their findings cleared that selection can be based on TOL index to improve drought tolerance in durum wheat. Our correlations coefficient matrix illustrated that both GMP and STI indices were correlated with yield under both conditions. Moreover, a suitable index must be significantly correlated with yield in any of the two moisture regimes and show a low coefficient of variation. Therefore, these indices can be used to determine drought resistance cultivars with high yield in both moisture regimes. Selection based on a combination of indices may be more useful for improving drought resistance of safflower, but correlation coefficients are helpful for determining the degree of overall linear association between any two attributes (Safavi et al., 2013). Hence, a better approach than a correlation analysis such as
biplot analysis is required to identify supreme cultivars for both moisture regimes. For further assessment of relation among drought tolerance indices, principle component analysis was applied. Accordingly, PC, two components accounted for 97.1% of the total variation (Table 4). The results of the principle component analysis of safflower cultivars indicated that the first PC accounted for 77.3% of the total variation, while the second PC justified 19.81% of the remaining variation (Table 4). Also reported that the first component with more than 68% of total variation is able to separate high-yielding and seed yield cultivars from other cultivars.

 

 

 

Table4. Principal component analysis for drought tolerance indices in safflower cultivars

 

 

Indices

 

 

PC1

PC2

 

 

GMP

0.416

0.198

 

 

MP

0.424

0.121

 

 

SSI

0.346

-0.442

 

 

STI

0.416

0.191

 

 

TOL

0.385

-0.361

 

 

Yp

0.42

-0.112

 

 

Ys

0.167

0.76

 

 

Total Variation %

77.3

97.1

 

 

Variation %

77.3

19.81

 

 

For abbreviations, see Table1.

 

 

 

 

 A biplot diagram from the first and second factor components is shown in Figure1. The biplot is divided into four classes named A, B, C and D based on the two first principle components. Lines which were located in zone A (2, 14, 15 and 11) demonstrated high yield under both  moisture conditions. Hence, these cultivars can be used as tolerant varieties into following breeding procedures for selection of drought tolerant and high-yielding cultivars under stress regime. Lines 13, 10 and 12 which were placed in region B, had suitable potential under both moisture conditions Safavi et al., (2013) believed that some indices such as STI, GMP, HAR and MP were more able to screening drought tolerant varieties and based on correlations between mentioned indices Yp and ys vectors (the angle between the vectors) in the bi plot graph, STI was the favorable index for identifying drought tolerant cultivars in safflower.

 

 

 

 

Figure1. Biplot graphical display of 15 safflower varieties and 7 drought indices (See Tables1and2 for abbreviations and genotype cods).

 

 On the other hand, genotypes that were located in zone D (genotypes 1, 5, 6 and 9) had the lowest yields under stress and normal conditions. Genotype number C area 3, 4, 7 and 8 were located in C area and had low and high-yield under normal and stress regimes, respectively. Accordingly, genotypes of the area A was classified therfore high-yielding and drought resistance groups. Majidi et al. (2011) indicated that wild genotypes had a low yield but their seed yield was stable when the environment changed. As these landraces make a favorable genetic source for transferring drought tolerant genes to other genotypes. Cluster analysis of drought tolerance indices classified the mentioned 7 indices into three groups with 4, 2 and 1 indices, respectively (Figure 2). Group 1 consisted of indices with high positive values for first principle components (GMP, STI, MP and Yp indices). These results were verified by the biplot graph analysis which could locate genotypes 15, 14 and 11 with high GMP, STI and MP values into group A. Group 2 included indices with negative SSI and TOL values in second principle components. Ys index (yield under stress conditions) with lowest and positive high values for first and second principle components was located into group 3 (Figure 2).

 

 

 

 

Figure2. Dendrogram from cluster analysis of drought tolerance indices based on WARD’s method

 

 Cluster analysis based on yield under both moisture regimes and drought tolerance indices classified the cultivars into six groups with 5, 3, 2, 2, 2 and 1 genotype, respectively (Figure 3). Group 3 included genotypes with high Ys, Yp, MP and GMP values, and is considered as a drought tolerant group with high-yielding under normal and stress conditions. Genotypes 14 and 15 (Gila) with high drought resistance and high GMP values were located in the same group. Roudbari et al. (2012) concluded that Gila genotype is more suitable genotype for drought stressed conditions. Grain yield, as a gross selection criterion for drought tolerance, is a complex characteristic that is defined by several metabolic, biochemical and physiological plant operations.

 

 

 

 

Figure3.  Dendrogram from WARD cluster analysis of safflower cultivars based on drought tolerance indices (See Tables2 for abbreviations and genotype codes).

 

Group 4 which included genotypes 9 and 10 with low seed yield under drought regime. Genotypes 1, 3, 5, 6 and 7 were classified into group 5. These lines showed lower drought tolerance than the genotypes of group 4. The last group consisted of line 12 that had the lowest and high yield under drought and normal conditions respectively, and classified into susceptible group.

 

,In conclusion the results of present study, showed that moisture regimes had a clear impact on yield of safflower genotypes, so that drought conditions could decline yield up to 1657
 kg ha-1. This reduction is 27% compared to the normal treatment. Gila and S149 had higher yields in during stress and normal conditions, respectively. Genotypes Semnan and Lordegan landraces had the lowest seed yield under both moisture conditions. According to the results, GMP and STI were correlated with Yp and Ys, so they were determined as the best drought tolerance indices to select drought tolerant safflower cultivars. Selection based on these indices may be useful for determining a genotype with good seed yield under both stress and normal regimes. We can suggest that the genotypes Gila and U.S.10 can be recommended as candidate cultivars for drought resistance in arid area.

 

 

 

 

 

References

 

Akcura M, F. Partigoc and Y. Kaya (2011): Evaluating of drought stress tolerance based on selection indices in Turkish bread wheat landraces. The Journal of Animal and Plant Sciences, 21(4): 700-709.

 

Farshadfar E, MM Poursiahbidi and SM. Safavi (2013): Assessment of drought tolerance in land races of bread wheat based on resistance / tolerance indices. International journal of Advanced Biological and Biomedical Research., 1(2): 143-158.

 

Fernandez GCJ, (1992): Effective selection criteria for assessing plant stress tolerance. In: Proceedings of the International Symposium on Adaptation of Vegetables and Other Food Crops in Temperature and Water Stress (Kuo CG, ed.). Publication, Tainan, Taiwan. 1992.

 

Fischer RA and R. Maurer (1978): Drought resistance in spring wheat cultivars. I. Grain yield responses. Australia Journal of Agricultural Research, 29: 897-912.

 

Golparvar AR, (2011): Assessment of Relationship between seed and oil yield with agronomic traits in spring safflower cultivars under drought stress condition. Journal of Research in Agricultural Science., 7(2): 109-113.

 

Guo P,  M. Baum, S. Grando, S. Ceccarelli, G. Bai, R. LI, M. Koreff, R.K. Varshney, A. Garner and J. Valkoun (2009): Differentially expressed genes between drought-tolerant and drought-sensitive barley genotypes in response to drought stress during the reproductive stage. Journal of Experimental Botany., 60(12): 3531-3544.

 

Hussani Ali I, (2015): Genetic Properties of drought tolerance indices in durum wheat. Jordan Journal of Agricultural Sciences., 11: 15-19.

 

Khalili, M., A.L.  Pour-Aboughadareh  and M.R. Naghavi (2014):  Evaluation of drought tolerance in safflower genotypes based on drought tolerance indices. Notulae Botanicae Horti Agrobotanici, 42(1): 214-218.

 

Majidi MM., V. Tavakilo, A. Mirlohi and M.R. Sabzian (2011): Wild safflower species (Carthamus oxyacanthus Bieb.): A possible source of drought tolerance for arid environments. Australian journal of crop science, 5(8): 1055-1063.

 

Mirzahashemi M, P. Golkar and G. Mohammadi-Nejad (2013): Gene effects for agronomic traits in safflower (Carthamus tinctorius L.) under drought stress. Ethno-Pharmaceutical Products Journal, 1(1): 23-28.  

 

Rameshkina Y, B. Tahmasebpoor and E. Sabbaghzadeh (2013): Evaluation the quantitative indices of drought tolerance in spring safflower. Bulletin of Environment, Pharmacology and Life Sciences, 2 (8): 4-12.

 

Rauf, S, (2008): Breeding sunflower (Helianthus annuus L.) for drought tolerance. Communications in Biometry and Crop Science. 2008. Vol. 3, No. 1, pp. 29–44.

 

Rosielle AA and J. Hamblene (1981): Theoretical aspects of selection for yield in stress and non-stress environments. Crop Science, 21: 943-946.

 

Roudbari Z, J. Saba and F. Shekari (2012): Use of physiological parameters as tools to screen drought tolerant safflower genotypes. International Research Journal of Applied and Basic Sciences, 3 (12): 2374-2380.

 

Safavi SM, SS. Pourdad and SA. Safavi (2013): Evaluation of drought tolerance in safflower (Carthamus tinctorius L.) under non-stress and drought stress conditions. International journal of Advanced Biological and Biomedical Research., 1(9): 1086-1093.

 

Savitri, ES, N. Basuki, N. Aini and EL. Arumingtias (2013): Identification and characterization drought tolerance of gene LEA-D11 soybean (glycine max L. Merr) based on PCR-sequencing. American Journal of Molecular Biology, 3: 32-37.

 

Shiranirad, AH and A. Abbasian (2011): Evaluation of drought tolerance in rapeseed genotypes under non-stress and drought stress conditions. Notulae Botanicae Horti Agrobotanici, 39(2): 164-171.

 

 Shraf A, M. Abd El-Mohsen, M.A. Abd El-Shafi, E.M.S. Gheith and H. Suleiman (2015): Using different statistical procedures for evaluating drought tolerance indices of bread wheat genotypes. Advance in Agriculture and Biology, 4 (1): 19-30.