Evaluating triticale genotypes based on drought resistance indices

Introduction
Triticale is a relatively new man-made cereal crop created by crossing wheat (Triticum spp.) and rye (Secale cereale) to combine the high grain quality and productivity of wheat with the stress tolerance and adaptability of rye. This crop has received attention worldwide due to its high protein content, tolerance to a wide range of environmental stresses, and its ability to thrive in marginal soils. Triticale genotypes are generally more resilient than wheat under drought, salinity, and other abiotic stresses. However, despite its potential, the large-scale cultivation of triticale in many countries, including Iran, remains limited, partly due to a lack of comprehensive knowledge regarding its performance under different environmental conditions. Since drought is one of the most critical limiting factors for cereal production in arid and semi-arid regions, identifying genotypes with enhanced drought tolerance is of paramount importance. The application of drought tolerance indices, which combine yield performance in stressed and non-stressed conditions, is a widely accepted approach to screen for stable and adaptable genotypes. Nevertheless, relying on a single index may be misleading, as each index captures only a specific dimension of tolerance. Therefore, integrative multivariate approaches such as principal component analysis (PCA), biplot, and heatmap clustering can offer a more comprehensive understanding and accurate discrimination of tolerant and high-yielding genotypes.
Materials and methods
To investigate the effect of different irrigation regimes and to identify triticale genotypes with superior drought tolerance, a two-year field experiment was conducted using a split-plot design in randomized complete blocks with three replications. The study involved four irrigation treatments: full irrigation (control), irrigation withholding at the flowering stage, at the milky stage, and at the doughy stage of grain development. Nine improved triticale genotypes, mostly derived from CIMMYT lines and recently developed in national breeding programs, were evaluated. Grain yield was measured for each plot and subsequently used to calculate a range of drought tolerance indices, including tolerance index (TOL), stress tolerance index (STI), geometric mean productivity (GMP), mean productivity (MP), yield index (YI), yield stability index (YSI), stress susceptibility index (SSI), and stress tolerance score (STS). Combined and simple analyses of variance (ANOVA) were performed using SAS software, while multivariate statistical methods such as PCA-based biplot and heatmap clustering were applied using R software to visualize genotype–index relationships and to classify genotypes into tolerant and susceptible groups.
Results and discussion
The combined ANOVA across two years revealed significant effects of genotype, irrigation treatment, and their interaction, while year effects were not significant, confirming the consistency of results. Grain yield decreased significantly under water deficit treatments compared to full irrigation, with the strongest reduction observed when irrigation was withheld at the flowering and milky stages. Among the nine genotypes, Sanabad (G1), Paj (G2), and ET-85-04 (G4) consistently exhibited the highest mean yields across normal and stressed conditions, while ET-85-04 and ET-83-20 showed the lowest variability across environments, indicating strong stability. Biplot analysis demonstrated that combining multiple indices provided clearer separation of tolerant genotypes compared to individual indices. Notably, STI, GMP, and MP were strongly correlated with grain yield under both stressed and non-stressed conditions, confirming their reliability as selection criteria. Heatmap clustering further confirmed that genotypes ET-85-04, Sanabad, and Paj grouped together with favorable stress tolerance indices, suggesting their superior adaptability. In contrast, some genotypes, such as ET-92-18 and ET-92-15, showed weaker associations with tolerance indices and were classified as less suitable for drought-prone areas. The use of multivariate methods was proven to be a powerful approach, integrating diverse tolerance indices to provide a more holistic assessment. These findings highlight that genotypes with high yield potential and stable performance across environments are ideal candidates for breeding programs aimed at improving drought tolerance in triticale.
Conclusion
In conclusion, the results of this two-year field evaluation emphasize that relying on a single drought tolerance index is insufficient for accurate identification of tolerant genotypes. Instead, combining multiple indices through multivariate statistical methods such as PCA-based biplot and heatmap clustering offers a superior strategy to distinguish stable and drought-tolerant lines. Based on both yield performance and tolerance indices, ET-85-04 was identified as the most promising genotype under both stress and non-stress conditions. Alongside Sanabad and Paj, this genotype demonstrated a balanced combination of high productivity, drought tolerance, and stability, making it a strong candidate for release as a new triticale cultivar. These results suggest that triticale can be effectively incorporated into cropping systems in drought-prone regions, potentially offering a reliable alternative to wheat, particularly in areas where water resources are limited. The findings also provide valuable insights for future breeding programs, reinforcing the necessity of integrative statistical approaches in screening and selecting stress-tolerant genotypes.