Introduction: It is crucial  to introduce and develop new high-yielding dryland wheat cultivars with resistance to biotic and abiotic stresses, given the extent of wheat cultivation in tropical regions and the occurrence of drought in recent years. On-farm experimentation provides new insights and is a suitable method for informing farmers about novel technologies, such as new crop varieties. Using the SIIG index and cluster analysis, the current study aimed to identify the superior wheat genotypes for grain yield and other agronomic traits in tropical drylands and introduce them to farmers.
Material and Methods: In current study, 33 varieties/ pure lines selected from the DARI and ICARDA wheat breeding programs were cultivated on farmers' fields of Lorestan province using a randomized complete block design (RCBD) with two replications during 2017-18 cropping season. The genotypes evaluated consisted of 12 cultivars and 21 pure lines of bread wheat and durum wheat, respectively. The pedigree of all genotypes is presented in table 1. The following nine characteristics were recorded: days to heading (DH), plant height (PH), Days to maturity (DM), spikes per square meter (SSM), number of grains per spike (NGPS), 1000 kernel weight (TKW), biological yield (BY), Grain yield (GY), and harvest index (HI). Using ANOVA, all investigated characteristics were analyzed. Genotypes means were compared using the least significance difference (LSD) at 5% and 1% probability level. Correlation analysis was performed using the Pearson method. Selection index of ideal genotype (SIIG) was utilized to select the genotypes with the highest yield and agronomic traits. In addition, cluster analysis using the WARD method and principal component analysis were employed to categorize genotypes. To analyze the data, MSTATC, IBM SPSS Statistics ver. 22 and Excel were utilized.
Results and Discussion:  Paya and Kabir varieties and G17, G18, and G31 lines had the highest average grain yields, with yields of 5,280, 3,985, 4,210, 4,003, and 3,950 kg/ha, respectively. The SIIG index indicated that Kabir and Paya varieties, and G31, G18, G29, G17 and G16 lines with a high SIIG value (0.822, 0.720, 0.791, 0.779, 0.767, 0.736 and 0.745, respectively) were superior genotypes, whereas G13 and G17 lines with a low SIIG (0.582 and 0.576, respectively) were the weakest genotypes for the majority of traits evaluated in the current study. Grain yield exhibited a significant and positive correlation with biological yield (BY), number of grains per spike (NGPS), and harvest index, according to correlation analysis (HI). Days to heading and maturity correlated negatively with grain yield. Therefore, BY, NGPS, HI, and early maturity may be suitable indicators for selecting high-performing genotypes in wheat breeding programs under rainfed conditions. Strong and positive correlation was observed between the SIIG index and grain yield. This issue reflected a larger proportion of grain yield in the SIIG index. Cluster analysis classified genotypes into three primary classes. Class 1 consisted of maturity date genotypes with high yield, Class 2 consisted of early maturity genotypes with low yield, and Class 3 consisted of early maturity genotypes with low yield. In the current study, first -class genotypes with high yield and earliness were deemed promising genotypes for planting in tropical regions under rainfed conditions or as parents for improving yield and desirable agronomic traits in wheat breeding programs.
Conclusion: The best genotypes, as determined by the SIIG index and cluster analysis, are the Kabir and Paya varieties and the G31, G18, G29, G17, and G16 lines, which are suitable for planting in drylands and further breeding programs. Significant and positive correlations between grain yield and biological yield (BY), number of grains per spike (NGPS), and harvest index (HI) indicated that these traits could be regarded as suitable indicators for enhancing grain yield in wheat under rainfed conditions. In the current study, genotypes with higher grain yield and desirable agronomic traits had the highest SIIG index. SIIG index could therefore be utilized as a suitable method for identifying the best genotypes on various crops.