INTRODUCTION

Among the cereals, barley is the firstborn domesticated food-cereal across the globe (Bendada et al., 2021); is one of a few crops grown successfully in semi-arid areas, where rainfall varies significantly with year. Algeria with its topographical and bioclimatic characteristics which make it possible to show a diversity of landscapes and cropping systems, cereal growing is the predominant speculation of agriculture. It extends over an annual area of ​​about 3.6 million hectares compared to the useful agricultural area (UAA) (MADR, 2012). Yield stability is an important criterion for the development of cultivars intended for environments with variable rainfall. Many methods of stability analysis are proposed in the literature such as parametric and non-parametric stability indices (Benmahammad et al., 2010; Rose et al., 2008). Barley (Hordeum vulgare L.) is predominant to be the most drought tolerant of the small grain cereal crops (barley, durum wheat and bread wheat) and is a major crop in Middle East and North Africa countries, because it is the predominant crop lower 300 mm of annual rainfall. In Mediterranean areas barley is primarily grown as animal feed and both grain and straw yields are used (Fatma et al.,2018). Drought stress decreases grain yield of barley genotypes through negative influence on yield components i.e. No. of plants/unit area, No. of spikes and grains per plant or unit area and 1000-grain weight, which are determined at different stages of plant development (Haddadin, 2015; Al-Ajlouni et al., 2016). The importance of G × E interactions in national cultivar evaluation and breeding programs have been demonstrated in almost all major crops, including wheat genotypes (Frih et al., 2021). Multi Environment Trials are important in plant breeding and agronomy for studying yield stability and predicting yield performance of genotypes across environments. Genotype x environment interactions (GEI) complicates the identification of superior genotypes (Sankar et al., 2021) but their interpretation can be facilitated by the use of several statistical modelling methods. GEI occurs when the genotypes respond differently across environments, and it is considered one of the main factors limiting progress in breeding and, hence, in agricultural production. The first and most common approach is parametric, which relies on distributional assumptions about genotypic, environmental and GxE effects. The second major approach is the non-parametric or analytical clustering approach, which relates environments and phenotypes relative to biotic and abiotic environmental factors without making specific modelling assumptions. Several parametric methods including univariate and multivariate ones have been developed to assess the stability and adaptability of varieties. The parametric approach continent many indices such as the regression coefficient (bi; Finlay and Wilkinson, 1963), variance of deviations from the regression (S²_di; Eberhart and Russell, 1966), Wricke’s ecovalence stability index (Wi²; Wricke, 1962), Shukla’s stability variance (σi²; Shukla, 1972), environmental coefficient of variance (CVi; Francis and Kannenberg, 1978) and the yield stability index (YSi; Kang, 1991). The second group of analytical methods includes non-parametric methods such as Nassar and Huhn’s statistics (S⁽¹⁾, S⁽²⁾; Nassar and Huhn, 1987), Huhn’s equation (S⁽³⁾ and S⁽⁶⁾; Huhn, 1990), Thennarasu’s statistics (NP⁽ⁱ⁾; Thennarasu, 1995). Non-parametric statistics are a feasible alternative to parametric statistics because their performance is based on ranked data (Nassar and Huhn, 1987) and no assumptions are needed about the distribution and homogeneity of the variance of the errors. Because each method has its own merits and weaknesses, most breeding programs now incorporate both parametric and non-parametric methods for the selection of stable genotypes. The aim of this study is to select adapted and stable barley genotypes based on the some parametric and non-parametric methods.

MATERIALS AND METHODS

2.1.  Plant material and field conditions

Field experiment was conducted during the 2017-2020 cropping seasons at the experimental field of ITGC, Setif, Algeria (5°20’E, 36°8’N, 958 m above mean sea level). The statistical design employed was based on a complete randomized block design (CRBD) with three replications. Sixteen Barley (Hordeum vulgare L.) genotypes were used in this study. The seeds were sown using an experimental drill in 1.2 m×2.5 m plots consisting of 6 rows with a 20 cm row space and the seeding rate is about 250 seeds per m².

2.2.  Statistical analysis

2.2.1.  Parametric measures

The regression coefficient (bi) is the response of the genotype to the environmental index that is derived from the average performance of all genotypes in each environment (Finlay and Wilkinson 1963). If bi does not significantly differ from 1, then the genotype is adapted to all environments. In addition to the regression coefficient, variance of deviations from the regression (S²_di) has been suggested as one of the most-used parameters for the selection of stable genotypes. Genotypes with an S²_di=0 would be most stable, while an S²_di>0 would indicate lower stability across all environments. Hence, genotypes with lower values are the most desirable (Eberhart and Russell 1966). Wricke (1962) proposed the concept of ecovalence as the contribution of each genotype to the GEI sum of squares. The ecovalence (Wi) of the i^th genotype is its interaction with the environments, squared and summed across environments. Thus, genotypes with low values have smaller deviations from the mean across environments and are more stable. Finally, the coefficient of variation is suggested by Francis and Kannenberg (1978) as a stability statistic through the combination of the coefficient of variation, mean yield, and environmental variance. Genotypes with low CVi, low environmental variance (EV), and high mean yield are considered to be the most desirable.

2.2.2. Non-Parametric measures

Huhn (1990) and Nassar and Huhn (1987) suggested four non-parametric statistics. We use during this study two parameters: (1) S⁽¹⁾, the mean of the absolute rank differences of a genotype over all tested environments; (2) S⁽⁶⁾, the sum of squares of rank for each genotype relative to the mean of ranks. The lowest value for each of these statistics reveals high stability for a certain genotype. In addition, four NP (1–4) statistics are a set of alternative non-parametric stability statistics defined by Thennarasu (1995). We use just tow parameters (NP⁽²⁾ and NP⁽⁴⁾). These parameters are based on the ranks of adjusted means of the genotypes in each environment. Low values of these statistics reflect high stability. The data were analyzed by the using of the online software (STABILITYSOFT) developed by Pour-Aboughadareh et al. (2019).  

RESULTS AND DISCUSSION

3.1. Parametric measures

The Table 1, showed that the values of regression coefficient (bi) varied from 1.685 for the advanced line G5 (2 row spike type) to 0.172 for the local landrace Saida 183 (6 row spike type). This variation in regression coefficients indicates that genotypes had different responses to environmental changes. Based on the definition described by Pour-Aboughadareh et al. (2019) the genotypes with low values (b_i<1) are very suitable to low-yielding environments, but the contrary for the genotypes with high values (b_i>1).

Table 1: Parametric, non-parametric stability index and mean grain yield (q ha-1) for the barley genotypes tested under semi-arid conditions

The local landrace Saida 183 and Tichedrette are very suitable to growing under the poor condition or just under rainfall conditions. In addition, the graphical distribution (Figure 1) between the regression coefficient and the mean grain yield of tested genotypes proved that the adapted and stable genotypes with high mean grain yield under these conditions are G2, Fouarra and G12.

Figure 1: The relationship between the regression coefficients and mean grain yield (q ha-1) for Barley genotypes tested

The advanced line G1, G3, G5, G6, G10 and G13 are greater specificity of adaptability to high-yielding environments (Irrigated conditions). According to Megahed et al. (2018) genotypes with regression coefficient greater than unity would be adapted to more favorable environments. The values of deviation from regression (S²_di) classified the genotype G4 (2 row spike type) as the most desirable genotypes, but with mean grain yield (39.8) lowest than the general mean of grain yield (42.26). The combination between the S²_di and the mean grain yield of tested genotypes proved that the advanced line G5, G6 and G13 have lowest values of S²_di and highest mean grain yield (>general mean of grain yield). Genotypes with high mean yield, a regression coefficient equal to the unity (bi = 1) and small deviations from regression (S²_di=0) are considered stable (Eberhart and Russell, 1966). The graphical distribution (Figure 2) between the Wricke’s ecovalence stability index (W_i²) and the mean grain yield of tested genotypes proved that the adapted and stable genotypes with high mean grain yield under these conditions are G13, G6, G10 and Fouarra. The genotype Saida 183 displayed high ecovalence and is classified as unstable genotype with lowest mean grain yield (35.02). The lowest value of W_i² is registered by G9, this one have bi equal to the unity (0.932), lowest values of S²d_i (0.062) and mean grain yield equal 41.11 (General mean of GY=42.26). In contrary, based on the environmental coefficient of variance (CVi) the genotype Saida183 is very stable, but with lowest mean grain yield (35.02). Many studies confirmed the efficiency of using like these parametric index to select adapted and stable barley genotypes (Ramla et al., 2016; Verma et al., 2019) and stable durum wheat genotypes (Guendouz and Hafsi, 2017). 

Figure 2: The relationship between the Wricke’s ecovalence stability index (Wi²) and mean grain yield (q ha-1) for Barley genotypes tested

3.2.  Non-parametric measures

Accordingly, Si⁽¹⁾ and Si⁽⁶⁾ of the tested genotypes (Table 1) showed that the advanced line G6 had the lowest values; therefore, this genotypes were regarded as the most stable genotypes according to Si⁽¹⁾ and Si⁽⁶⁾ with highest grain yield 49.92 q ha^-1. In addition, the graphical distribution (Figure 3) between the Nassar and Huhn’s non-parametric index (Si⁽¹⁾) and the mean grain yield of tested genotypes showed that the adapted and stable genotypes with high mean grain yield under these conditions are G6, Fouarra, G10, G13 and G2.

Figure 3: The relationship between the Nassar and Huhn’s non-parametric index (Si(1)) and mean grain yield (q ha-1) for Barley genotypes tested

In contrary, the advanced line G1 had the highest values of  Si⁽¹⁾  and Si⁽⁶⁾ and high mean grain yield (42.32) over general grain yield equal 42.26 q ha^-1. Our results are in according with the research of Khalili and Pour-Aboughadareh (2016), which proved that the indices of Nassar and Huhn’s are very suitable to select stable and adapted barley genotypes. Based on the non-parametric index developed by Thennarasu (1995), the advanced line G6 is the more stable genotype over all genotypes tested with lowest values for the both index tested (NP2 and NP4) and highest mean grain yield (49.92 q ha^-1). In addition, the graphical classification based on the distribution (Figure 4) between the Thennarasu’s non-parametric index and the mean grain yield of tested genotypes showed that the adapted and stable genotypes are G6, Fouarra, G10, G13, G2 and G12. Many researchers suggested that the used non-parametric measures cited below in the selection of stable durum wheat and barley genotypes (Guendouz and Hafsi, 2017; Hannachi et al., 2019; Khalili and Pour-Aboughadareh, 2016) are very suitable under arid and semi-arid conditions.

Figure 4: The relationship between the Thennarasu’s non-parametric index (NP4) and mean grain yield (q ha-1) for Barley genotypes tested

3.3.  Association among stability parameters and grain yield

The results of Spearman’s coefficient of rank correlations between mean grain yield and the different parametric and non-parametric stability measures are shown in Table 1. The mean grain yield correlated significantly and negatively with all non-parametric indices tested. Many studies registered like this significant correlation in Barley (Khalili and Pour-Aboughadareh, 2016). Thus, selection of stable genotypes based on these stability parameters may not enable barley breeders to identify genotypes that are both high-yielding and stable. A study of durum wheat genotypes using the same stability parameters (Kilic et al., 2010) also identified below-average-yielding genotypes as the most stable and the highest-yielding genotypes as more unstable. In addition, no significant correlations are registered between the mean grain yield and the parametric indices (Table 1). As illustrated in the Table 1, significant correlation registered between the different parametric and non-parametric indices. Many studies revealed that S⁽¹⁾ and S⁽⁶⁾ were positively and significantly correlated with each other and with NP⁽²⁾ and NP⁽⁴⁾ (Pour-Aboughadareh et al., 2019). During this study significant and positive correlation registered between S⁽¹⁾ and S⁽⁶⁾ (r = 0.86^***) and among S⁽⁶⁾ and NP⁽⁴⁾ (r=0.99^***), Kilic (2012) reported that this significant positive correlation between these stability parameters suggests that these parameters would play similar roles to select adapted and stable genotypes. The Wricke’s ecovalence stability index (W_i²) registered positive and significant correlation with all non-parametric indices tested; these results indicate that these parameters plays similar roles in the selection of stable barley genotypes. 

3.4.  Classification based on principal component analysis

Principal component (PC) analysis based on the rank correlation matrix was performed and presented in Figure 5.

Figure 5: Biplot of IPC1 (F1) and IPC2 (F2) of the rank correlation matrix of the stability parameters with grain yield and Barley genotypes tested

The results proved that the first and second principal components of the rank correlation accounted for 57.16% and 18.76% of the variation, respectively, making a total of 75.92% of the original variance among the stability parameters, many studies have been reported like these results in durum wheat (Kilic et al., 2010) and barley (Mut et al., 2010). The static and dynamic yield stability concepts describe the differential response of genotypes to variable environments (Becker and Leon, 1988). Based on the PC analysis the parametric indices bi and CVi are associated with dynamic stability but other indices are associated with static stability. In addition, the principal component analysis classified the genotypes Fouarra and G2 in dynamic stability group with highest grain yield.  The high yield performance of released genotypes is one of the most important targets of breeders; therefore, they prefer a dynamic concept of stability because this concept of stability means that a genotype would show high response to different levels of inputs such as fertilizer, temperature and humidity. 

CONCLUSION

The classification based on the mean grain yield describe that the genotypes G6, Fouarra and G2 had the highest grain yield. The selection of adapted and stable genotypes based on Parametric and Non-parametric indices of tested genotypes proved that the adapted and stable genotypes with high mean grain yield under these conditions are G2, G6, Fouarra, G10, G13, G2, G12 and G13. The Principal component analysis classified the genotype Fouarra and G2 in dynamic stability group with highest grain yield.

Reference

Al-Ajlouni, Z.I., Al-Abdallat, A.M., AlGhzawi, A.A., Ayad, J.Y., Abu Elenein, J.M., Al-Quraan, N.A., Stephen Baenziger, P., 2016. Impact of pre-anthesis water deficit on yield and yield components in barley (Hordeum vulgare L.) plants grown under controlled conditions. Agronomy 6(33), 1-14.

Becker, H.C., Leon, J., 1988. Stability analysis in plant breeding. Plant Breeding 101, 1–23.

Bendada, H., Guendouz, A., Benniou, R., Louahdi, N., 2021. The effect of spike row type on the grain yield and grain filling parameters in barley (Hordeum vulgare L.) genotypes under semi-arid conditions. Agricultural Science Digest  DOI: 10.18805/ag.D-311.

Benmahammed, A., Nouar, H., Haddad, L., Laala, Z., Oulmi, A., Bouzerzour, H. 2010. Analyse de la stabilite des performances de  rendement  du  blé  dur  (Triticum durum  Desf.)  sous  conditions  semi-arides.  Biotechnologien , Agronomie , Societe et Environnement 1(14), 177-186.

Eberhart, S.A.T., Russell, W.A.N., 1966. Stability parameters for comparing varieties. Crop Science6, 36–40.

Fatma M.A.M., El-Khawaga, A.A., Ali, M.M.A., Hassan, A.I.A., 2018. Stability analysis of barley genotypes under different water stress levels.Zagazig Journal of Agricultural Research 45(5), 1521–1545.

Finlay, K.W., Wilkinson, G.N., 1963. Adaptation in a plant breeding programme. Australian Journal of Agricultural Research 14, 742–754.

Francis, T.R., Kannenberg, L.W., 1978. Yield stability studies in short-season maize: I. A descriptive method for grouping genotypes. Canadian Journal of Plant Science 58, 1029–1034.

Frih, B., Oulmi,  A., Guendouz, A., 2021. Study of the drought tolerance of certain of durum wheat (Triticum durum Desf.) genotypes growing under semi-arid conditions in Algeria. International Journal of Bio-resource and Stress Management 12(2), 137–141.

Guendouz, A., Hafsi, M., 2017. Comparison of parametric and non-parametric methods for selecting stable and adapted durum wheat cultivars under semi-arid conditions. Jordan Journal of Agricultural Sciences 13(3), 655–662. 

Haddadin, M.F., 2015. Assessment of drought tolerant barley varieties under water stress. International Journal of Agriculture and Forestry 5(2), 131–137.

Hannachi, A., Fellahi, Z.E.A., Bouzerzour, H., 2019. Analyse de l’adaptabilité et la stabilité de quelques variétés de blé dur (Triticum durum Desf.) aux conditions sud mediterraneennes. Revue Agriculture 10(2), 56–67.

Huhn, M., 1990. Non-parametric measures of phenotypic stability. Part 1: Theory. Euphytica 47, 189–194.

Kang, M.S., 1991.  Modified rank-sum method for selecting high yielding, stable crop genotypes. Cereal Research Communication 19, 361–364.

Khalili, M., Pour-Aboughadareh, A., 2016. Parametric and non-parametric measures for evaluation yield stability and adaptability in barley doubled haploid lines. Journal of Agricultural Science and Technology 18, 789-803.

Kilic, H., Mevlut, A., Husnu, A., 2010. Assessment of parametric and nonparametric methods for selecting stable and adapted durum wheat genotypes in multi-environments. Notulae Botanicae Horti Agrobotanici Cluj-Napoca 38, 271–279.

Kilic, H., 2012. Assessment of parametric and nonparametric methods for selecting stable and adapted spring bread wheat genotypes in multi-environment. Journal of Animal and Plant Sciences 22, 390–398.

MADR, Ministere  de  l’Agriculture  et  du  Développement  Rural, 2012. Statistiques  agricoles,  superficies  et  productions, Direction des Statistiques Agricoles et des Enquêtes Economiques, Série B. Avable on http://madrp.gov.dz/ . Accessed on December 15^th , 2021.

Megahed, F.M.A., El-Khawaga, A.A, Ali, M.M.A., Hassan, A.I.A., 2018. Stability analysis of barley genotypes under different water stress levels. Zagazig Journal of Agricultural Research 45(5), 1521–1545. 

Mut, Z., Gulumser, A., Sirat, A., 2010. Comparison of stability statistics for yield in barley (Hordeum vulgare L.). African Journal of Biotechnology 9, 1610–1618.

Nassar, R., Huhn, M., 1987. Studies on estimation of phenotypic stability: Tests of significance for non-parametric measures of phenotypic stability. Biometrics 43, 45–53.

Pour-Aboughadareh, A.M., Yousefian, H., Moradkhani, P., Poczai., Siddique, K.H.M., 2019. STABILITYSOFT: A new online program to calculate parametric and non-parametric stability statistics for crop traits. Applications in Plant Sciences 7(1), e1211.

Ramla, D., Yakhou, M.S., Bilek, N., Hamou, M., Hannachi, A., Aissat, A., Mekliche-Hanifi, L., 2016. Grain yield stability analysis of barley doubled haploid lines in algerian semi-arid zones. Asian Journal of Crop Science 8, 43–51.

Rose,  L.W.,  Das,  M.K.,  Taliaferro,  C.M.A., 2008.  Comparison  of  dry  matter  yield  stability  assessment  methods  for  small numbers of genotypes of Bermuda grass. Euphytica, 164, 19–25.

Shukla, G.K., 1972. Some statistical aspects of partitioning genotype-environmental components of variability. Heredity 29, 237–245.

Sankar, S.M., Singh, S.P., Prakash, G., Satyavathi, C.T., Soumya, S.L., Yadav, Y., Sharma, L.D., Rao, A.R., Singh, N., Srivastava, R.K., 2021. Deciphering genotype-by-environment interaction for target environmental delineation and identification of stable resistant sources against foliar blast disease of pearl millet. Frontiers in Plant Science 12, 656158.

Thennarasu, K., 1995. On certain non-parametric procedures for studying genotype- environment interactions and yield stability. PhD thesis, PJ School, Indian Agricultural Research Institute, New Delhi, India.

Verma, A., Kumar, V., Kharab, A., Singh, G., 2019. Quantification of G×E interaction for feed barley genotypes by parametric and non-parametric measures. Bangladesh Journal of Botany 48(1), 33-42.

Wang, Y., Ren, X., Sun, D., Sun, G., 2015. Origin of worldwide cultivated barley revealed by NAM-1 gene and grain protein content. Frontiers in Plant Science 6, 803.

Wricke, G., 1962. Ubereine Methode zur Erfassung der okologischen Streubreite in Feldversuchen. Zeitschrift für Pflanzenzuchtung 47, 92–96.