Identification of Superior Barley Genotypes Using Selection Index of Ideal Genotype (SIIG)

The main objective of the study was to evaluate and select the superior barley genotypes based on grain yield and some pheno-morphological traits using a new proposed selection index (SIIG). For this purpose, one-hundred-eight pure and four local cultivars (Norouz, Auxin, Nobahar, and WB-97-11) were evaluated as reference genotypes in four warm regions of Iran, including Ahvaz, Darab, Zabol, and Gonbad, during the 2020–2021 cropping seasons. The results of REML analysis showed that the heritability of all traits (except plant height) was higher in Gonbad than in other environments, while the lowest values were estimated in Ahvaz and Zabol environments. In addition, among the measured traits, the thousand kernel weight and grain filling period showed the highest and lowest values of heritability (0.83 and 0.01, respectively). The results showed that the seed yield of genotypes 1, 108, 3, 86, 5, 87, 19, 16, 15, 56, and 18 was higher than the four reference genotypes, and, on the other hand, the SIIG index of these genotypes was greater than or equal to 0.60. Based on the SIIG discriminator index, 4, 8, 31, and 28 genotypes with values greater than or equal to 0.60 were identified as superior for Darab, Ahvaz, Zabol, and Gonbad environments, respectively. As a conclusion, our results revealed that the SIIG index has ideal potential to identify genotypes with high yield and desirable traits. Therefore, the use of this index can be beneficial in screening better genotypes in the early stages of any breeding program for any crop.


Introduction
Barley (Hordeum vulgare L.) has been recognized as one of the most compatible crops with production in many different regions of the world [1,2]. In addition, this cereal crop ranks fourth in the world in terms of economic importance after wheat, rice, and corn [3]. The results of REML analysis using the BLUP statistic for the studied genotypes are shown in Table 2. Accordingly, the highest heritability of genotypes was for TKW (0.83) and the lowest for GFP (0.01) and YLD (0.28). The genotype × environment interaction variance for DHE, DMA, GFP, and YLD traits was higher than the genotypic variance of these traits, while the opposite was true for PLH and TKW traits. Graphic phenotypic variation for the measured traits is shown Figure 1. The results indicated that the highest values of DMA, DHE, and PLH were recorded in the Moghan environment, while the lowest values were recorded in the Ahvaz environment. The lowest values of YLD and TKW were observed in Zabol and the highest values in the Darab environment. In addition, the Ahvaz and Darab environments showed the lowest and highest GFP, respectively. environment. In addition, the Ahvaz and Darab environments showed the lowest and highest GFP, respectively.
Based on the results in Table 3, the highest grain yield was recorded for the Zabol (4805 kg ha −1 ) and Darab (4768 kg ha −1 ) environments, respectively. The Ahvaz and Zabol environments also showed the lowest and highest grain yield, respectively, and the difference between them was 3396 kg ha −1 , which was higher than the average grain yield obtained in Moghan (3056 kg ha −1 ). The SIIG selection index was used to select the best genotypes for each test environment and all warm environments in terms of grain yield and other measured traits (Tables  4 and 5). According to the results of the SIIG index, genotypes 37, 107, 38, 71, 105, 104, 99, and 63 with the highest SIIG value (between 0.601 and 0.719) were identified as the best genotypes in the Ahvaz environment. The average grain yield and SIIG values of identified genotypes were higher than those of the reference genotypes (Table 3). In the Zabol environment, genotypes 54, 56, 1, 5, 18, 96, 4, 26, 108, and 16 were considered superior genotypes, along with the Nowruz cultivar with the highest SIIG index value (0.703-0.773). The range of the SIIG index for selected genotypes in the Darab environment was 0.659 and 0.611, and some genotypes, such as 86, 87, 1, and 108, were considered ideal genotypes compared to others. In addition, the Nowruz cultivar with a yield of 6957 kg ha −1 and a high SIIG value (0.611) was identified as the best reference genotype in this environment (Table 3). Genotypes 105, 80, 84, 36, 3, 99, and 107 with the highest SIIG index value (0.706-0.706) were the best genotypes in the Gonbad environment, respectively. The SIIG index values and grain yield of these genotypes were higher than those of all control genotypes (Table 3). Based on the average data from the four test environments, the identified superior genotypes were 86, 108, 3,1,87,105,99,80,4,18,5,109,97,15,82,56,16, and 23. The SIIG values for these genotypes ranged from 0.608 to 0.726 (Table 4). Based on the results in Table 3, the highest grain yield was recorded for the Zabol (4805 kg ha −1 ) and Darab (4768 kg ha −1 ) environments, respectively. The Ahvaz and Zabol environments also showed the lowest and highest grain yield, respectively, and the difference between them was 3396 kg ha −1 , which was higher than the average grain yield obtained in Moghan (3056 kg ha −1 ).
The SIIG selection index was used to select the best genotypes for each test environment and all warm environments in terms of grain yield and other measured traits (Tables 4 and 5). According to the results of the SIIG index, genotypes 37, 107, 38, 71, 105, 104, 99, and 63 with the highest SIIG value (between 0.601 and 0.719) were identified as the best genotypes in the Ahvaz environment. The average grain yield and SIIG values of identified genotypes were higher than those of the reference genotypes (Table 3). In the Zabol environment, genotypes 54, 56, 1, 5, 18, 96, 4, 26, 108, and 16 were considered superior genotypes, along with the Nowruz cultivar with the highest SIIG index value (0.703-0.773). The range of the SIIG index for selected genotypes in the Darab environment was 0.659 and 0.611, and some genotypes, such as 86, 87, 1, and 108, were considered ideal genotypes compared to others. In addition, the Nowruz cultivar with a yield of 6957 kg ha −1 and a high SIIG value (0.611) was identified as the best reference genotype in this environment (Table 3). Genotypes 105, 80, 84, 36, 3, 99, and 107 with the highest SIIG index value (0.706-0.706) were the best genotypes in the Gonbad environment, respectively. The SIIG index values and grain yield of these genotypes were higher than those of all control genotypes (  The results of the correlation of the SIIG index with the measured morpho-phenological traits are shown in Figure 2. Under Ahvaz conditions, the SIIG index showed a positive and significant correlation with YLD (0.95 **), while it negatively and significantly correlated with DMA (−0.67 **) and DHE (−0.53 **) (Figure 2A). In the Zabol environment, the SIIG index showed a significant positive relationship with YLD (0.95 **), DMA (0.28 **), PLH (0.23 *), and GFP (0.22 *) ( Figure 2C). Based on the data obtained in the Darab environment, associations between SIIG index and YLD (0.92 **), DMA (0.50 **), and GFP (0.49 **) were positive and significant ( Figure 2D).
In addition, a significant positive association was observed between the SIIG index with YLD (0.91**) and TKW (0.74**) in the Gonbad environment ( Figure 2B). Based on the data, the selection index showed a positive and significant correlation with TKW and DMA traits ( Figure 2E).

Discussion
The advantages of the REML method over classical methods are that it is highly efficient in augmentation designs and reduces the number of negative estimates of genetic parameters due to problems, especially inadequacy of experimental design, that arise in classical methods [21]. In the present study, our results showed that the heritability of most traits was higher in the Moghan environments compared to other test environments ( Table 2). Despite the short flowering and maturation periods in the Darab environment compared to Gonbad and Zabol, the grain yield in the Darab environment was higher than in the other two environments (Figure 1), indicating ideal barley production conditions in Darab compared to the Gonbad and Zabol environments. This result can be appreciated by breeders and farmers to optimize barley yield in low-yielding environments in Iran's tropical climate. Therefore, additional studies (such as stability tests) of selected genotypes can be used to develop high-yield genotypes or improve grain yield stability and productivity through appropriate breeding strategies.
As the results showed, groups 1 (genotypes with SIIG values greater than or equal to 0.7) and 2 (genotypes with SIIG values greater than or equal to 0.6 and less than 0.7) were superior genotypes based on the SIIG index (Table 5). Although all traits will eventually be reflected in grain yield, selection based on various traits can be effective in improving the process of breeding programs. One of the advantages of using the SIIG index is that all traits are considered and their effects are shared by genotype [8]. In other words, in this index, different traits will directly participate in the selection of genotypes [12]. Since irrigated crops are not usually exposed to water stress, dwarfism may be a key trait for lodging [22]. On the other hand, in breeding for drought stress tolerance, early maturity has an important role in improving grain yield [23][24][25]. Of course, it should be noted that early maturity will be useful when the grain filling period is not limited and the plant has the necessary opportunity to complete this period to prevent shrinkage and loss of grain weight [26]. Since the genotypes in the present study were tested in the warm environments of Iran, dwarfism, early maturity, and increasing the length of the grain-filling period were taken into account as selection criteria when choosing desirable genotypes. For other traits, such as TKW and YLD, the observation of high values was considered in the selection of desirable genotypes.
The different response of genotypes in the studied environments is due to the genotype × environment interaction, so, using the SIIG index, efforts were undertaken to introduce the best genotypes into additional experiments based on different traits in each environment. The results of correlating the SIIG index with various traits showed that, in all environments, there was a significant correlation between grain yield and SIIG index, indicating the effectiveness of the SIIG index in selecting high-yield genotypes. Traits with high variability will contribute more to the numerical value of the SIIG index [12]. Moreover, in each environment, the SIIG index introduced leading genotypes and showed their distance from other genotypes. The SIIG index is a selective model and is used to select the most ideal genotypes. In other words, using the SIIG index, different traits can be turned into a single index, and the selection of superior genotypes can be undertaken more reliably and accurately [20]. Another feature of the SIIG index is the integration of traits with different units [8].
The results of grouping genotypes based on SIIG index in Ahvaz showed that, as the SIIG value was lowered, the YLD and GFP values decreased and the DMA and DHE values increased, while no major changes were observed in TKW and PLH. Therefore, the use of the SIIG index led to the selection of genotypes with high yield and early maturity in Ahvaz. The results of the SIIG index in Darab showed that, with the reduction in SIIG, the amount of YLD and GFP increased, while the amount of DMA and DHE decreased. Therefore, in Darab, selection using the SIIG index led to the selection of high-yield but late-maturity genotypes. In Gonbad, as the SIIG value decreased, the amount of YLD, TKW, and GFP decreased, but not much change was observed in other traits. In Zabol, the results showed that, with decreasing SIIG value, YLD decreased, but no significant changes were observed in other traits. In general, the results of grouping genotypes in terms of the traits studied using the SIIG index in all environments showed that, as the SIIG value increased, YLD increased, but other traits increased or decreased due to correlation with YLD. Therefore, group 1 genotypes (0.70 ≤ SIIG) in any environment are superior genotypes that can outperform other genotypes in terms of YLD and other traits.

Genetic Materials and Setup Experiments
In this study, 108 pure barley genotypes were used along with 4 check varieties (cv. Nooruz, cv. Auxin, cv. Nobahar, and WB-97-11). This set of barley genotypes is derived from hybridization between local Iranian cultivars and international genetic materials obtained from ICARDA's national barley breeding programs in SPII, Karaj, Iran. Nooruz, Auxin, and Nobahar are new and improved cultivars with high yield potential and excellent adaptability in different regions of warm Iran. Therefore, they were chosen as a reference for evaluating new genotypes. The experiment was performed at the following four locations: Gonbad (37 • 15 00" N 55 • 10 02" E), Zabol (31 • 01 43" N 61 • 30 04" E), Darab (28 • 45 07" N 54 • 32 40" E), and Ahvaz (31 • 19 13" N 48 • 40 09" E) during the 2020-2021 growing seasons. The meteorological characteristics of each environment are shown in Table 6.
The studied genotypes were planted in six lines along 6 m at a distance of 15 cm from each other on December 6. Seed consumption was determined by 300 seeds per square meter and thousand kernel weight for each genotype. Seeds were sown using an experimental plot planter (Wintersteiger, Ried, Austria). The fertilizer composition was 32 kg ha −1 nitrogen (twice), and di-ammonium phosphate and potassium sulfate were 100 and 50 kg ha −1 , respectively (before planting). After the removal of perimeter plants, all experimental plots were harvested with an experimental grain harvester (Wintersteiger, Ried, Austria). The traits studied were the number of days to heading (DHE), days to maturity (DMA), plant height (PLH), thousand kernel weight (TKW), and grain yield (YLD). Table 6. Monthly meteorological data in 2020-2021 cropping seasons in the warm environments of Iran.

Darab Ahvaz
Temperature ( • C) Rainfall (mm)  Depending on the number of genotypes and the different traits measured, the data matrix was formed as the following equation (matrix D).

Converting the Primary Data Matrix (Matrix D) to a Normal Matrix (Matrix R)
The following relation is used to normalize the row data (without unifying the data): At this stage, according to the type of trait and the researcher's opinion for each trait, the best (ideal) and the weakest (non-ideal) genotypes were selected. For example, in terms of grain yield, the maximum yield of a given genotype was considered the ideal value and the non-ideal value the lowest. In the case of days to maturity (DMA), minimum values are favorable. The distance coefficients from ideal genotypes (d i + ) and weak genotypes (d i were estimated based on the following relations. where r ij is the normalized value of ith genotype (i = 1, 2, . . . , n) in terms of jth trait (j = 1, 2, . . . , m). r j + and r j − indicated the normalized values of ideal genotypes and weak genotypes for each jth trait, respectively.

Calculating the Ideal Genotype Selection Index (SIIG)
In the last step, Equation (6) shows the ideal genotype selection index for genotype: The SIIG i value varies from 0 to 1, and genotypes with SIIG ≈ 1are selected as superior genotypes in terms of grain yield and other measured traits. All R scripts to calculate this index are shown in Appendix A.

Statistical Data Analysis
The experiment data were subjected to the calculation of analysis of variance based on the REML model using ACBD-R software [27]. Correlation analyses and heatmaps were performed using package 'metan' [28].

Conclusions
Overall, the SIIG index results in Darab, Ahvaz, Zabol, and Gonbad showed that genotypes 4, 8, 31, and 28 with SIIG values greater than or equal to 0.600 (0.60 ≤ SIIG) can be identified as superior genotypes for grain yield and other phonological traits. Since grain yield and related traits are inherited quantitatively, it is necessary to consider all yield-related traits to identify ideal genotypes in any breeding programs. Our results showed that the SIIG index has ideal potential to identify high-yielding genotypes with desirable traits. Therefore, the use of this index can be beneficial in screening the superior genotypes in the early steps of any breeding program for any crop.