Identification of Iranian bread wheat with superior performance under well-watered and rain-fed conditions using linear selection indices.

backgroundWheat (Triticum aestivum L.) is one of the most important staple crops worldwide, yet its productivity is severely constrained by drought stress, which is becoming increasingly frequent under climate change. Genetic improvement through multi-trait selection indices provides an efficient strategy for simultaneously improving yield and adaptive traits while minimizing undesirable trade-offs. The present study evaluated the efficiency of seven selection indices—the linear phenotypic selection index (LPSI), base linear phenotypic selection index (BLPSI), restricted linear phenotypic selection index (RLPSI), predetermined proportional gain linear phenotypic selection index (PPG-LPSI), eigen selection index method (ESIM), restricted eigen selection index method (RESIM), and linear molecular selection index (LMSI)—for identifying superior bread wheat genotypes under well-watered and rain-fed (drought) environments. A panel of 266 Iranian bread wheat genotypes (180 landraces and 86 cultivars), representing winter, facultative, and spring growth habits, was assessed across nine environments (five well-watered and four rain-fed) during 2016–2020 in a randomized complete block design with two replications. Eight agronomic traits including grain yield (GY), plant height (PH), days to flowering (DF), days to maturity (DM), thousand kernel weight (TKW), spike area (SA), grain number per spike (GN), and spike weight (SW) were recorded.

resultsAnalysis of variance revealed highly significant genotypic effects for all traits, with higher heritability estimates for phenological traits (DF, DM, PH) compared with grain yield. Strong positive correlations were observed between GY and GN, SW, and TKW, whereas negative associations were detected with DF, DM, and PH, particularly under drought conditions. Direct and correlated selection responses showed reduced efficiency under drought compared with well-watered environments, highlighting the importance of multi-trait selection strategies. Among the evaluated indices, LPSI, BLPSI, and LMSI consistently demonstrated higher efficiency, achieving positive and balanced genetic gains across traits in both conditions. Notably, LMSI, by integrating genomic information from 16,060 molecular markers obtained through genotyping-by-sequencing, outperformed purely phenotypic indices under stress environments by providing higher selection accuracy and genetic gain. Using a selection intensity of 5%, the indices identified a subset of superior genotypes. Under well-watered conditions, genotypes such as Shanghai, Khazar1, Naz, Moghan1, 621908, and 626358 were consistently superior, while under drought stress, Alborz, Koohdasht, Shiroodi, Atrak, Falat, and 623318 were identified as top-performing.

conclusionsMulti-trait selection indices, particularly genomic-based LMSI, effectively enhanced selection accuracy and genetic gain. Spring-type genotypes were predominantly selected under both well-watered and rain-fed conditions, highlighting the role of growth-habit diversity. These results provide guidance for breeding programs targeting drought resilience and yield stability in wheat.