Investigating the genetic diversity of bread wheat germplasm under drought stress

Introduction
Wheat ranks second among cereals in term of global production, but it has been affected by ongoing global climate changes. Drought stress is one of the important factors contributing to the reduction in wheat production. Developing drought-tolerant wheat varieties currently presents a major challenge for wheat breeders. Complex multigenic and multitrait genetic control, high genotype-environment interaction, low heritability, and challenges associated with high-throughput screening of plant traits and genes effective in drought tolerance due to the involvement of numerous traits and positive and negative correlations among them are some of the factors that have made wheat variety improvement for drought tolerance a challenging task. Selecting drought-tolerant genotypes as a cost-effective and biologically superior approach to increasing wheat production in low humidity areas has always been considered by breeders. Screening wheat genotypes to identify drought tolerance sources is essential and can be useful in selecting appropriate and desirable parents for implementing beneficial breeding programs. Local plant populations such as landraces are valuable resources for environmental stress tolerance, including drought stress. Local plant varieties are heterogeneous populations with local adaptation that provide genetic resources capable of meeting the needs imposed by new and emerging agricultural challenges under highly stressful conditions. The aim of this study was to identify wheat accessions tolerant to drought stress for use in future breeding programs.
Materials and methods
A total of 512 wheat accessions from the National Plant Gene Bank of Iran, along with the varieties Kavir, Roshan and Mahouti as controls, were evaluated for drought stress tolerance in the research field of Agricultural and Natural Resources Research Center of Yazd, Iran. The experiment was conducted in two separate augmented designs, one of which was considered to drought stress and the other to normal irrigation conditions. Drought stress was induced by limiting the irrigation cycle. Considering that out of the total studied genetic materials, 141 accessions survived, the evaluation of traits was perfoemed on both those surviving accessions and their corresponding counterparts in the normal experiment. For data analysis, descriptive statistics including minimum, maximum, average, and coefficient of variation were calculated. Correlation coefficients were estimated and tested to study the relationship between traits. Stepwise regression analysis was used to determine the role of each trait and to identify the important and influencing traits on five-spike grain weight. K-means cluster analysis method was used to separate the accessions. All statistical analyzes were performed using R and SPSS softwares.
Research findings
The results of descriptive statistics showed that the highest coefficient of variation under normal and drought stress conditions were attributed to five-spike grain weight (23.35%) and number of fertile tillers per plant (31.67%), respectively. Five-spike grain weight had the highest decrease in range (51.3%) under drought stress compared to normal conditions, and the highest decrease in mean was observed for number of fertile tillers per plant (59.3%) and five-spike grain weight (40.5%), respectively. A large number of accessions showed superiority over control varieties under both normal and drought stress conditions, so that 97 accessions had five-spike grain weight more than the control varieties. The accessions KC12856, KC12776, KC12783, KC12767, and KC12697 with values of 6.77, 6.75, 6.22, 5.94 and 5.86 g, respectively, had the highest five-spike grain weight under drought stress conditions. Number of spikelets per spike, number of florets per spikelet, 100-grain weight and number of grains per spike had significant and strong correlation coefficients at 1% probability level with five-spike grain weight under drought stress conditions. Based on the results of stepwise regression analysis under drought stress conditions, three traits including number of grains per spike, 100-grain weight and spike length were the most effective traits on five-spike grain weight and explained 90.6% of it’s variation. The studied genetic materials were grouped into five separated clusters using K-means cluster analysis.
Conclusion
The results of this experiment showed the valuable potential of the studied germplasm to drought stress tolerance. Evaluating the relationships among traits revealed the significance of number of grains per spikelet, number of spikelets per spike, number of florets per spikelet, and 100-grain weight under both normal and drought stress conditions. A considerable number of accessions were superior to the control varieties under both normal and drought stress conditions, indicating the richness and valuable potential of these genetic resources for drought tolerance related traits. Therefore, it is recommended to continue screening and evaluating the wheat collection of National Plant Gene Bank of Iran to identify drought stress-tolerant genetic resources. Based on the results of cluster analysis, the control cultivars with different characteristics were grouped in the separate clusters, and a distinguished cluster of the studied wheat accessions with superior traits was also formed. The diverse germplasm or superior accessions identified in this study can be used to develope the genetic base in crosses or as parents in breeding programs.