Field versus controlled environmental experiments to evaluate the heat stress response of barley (Hordeum vulgare L.)

The complexity of heat stress hinders both the exploration of the genetic basis of stress response and breeding of genotypes with increased stress tolerance. Our main goal was to analyze and compare the possibilities of evaluating heat stress responses of barley cultivars in field sowing and controlled environmental experiments. For this purpose, a four-year field-sown experiment was carried out at one location in a panel of 190 winter and facultative barleys. In parallel, a subset of 28 cultivars were included into controlled environmental tests, where their reactions were determined to single heat stress treatment applied at heading and to combined heat stresses applied at first node appearance and then at heading. Based on the grain-yield related parameters, seven distinct clusters of the cultivars could be established with specific reaction patterns across the years. There was one year with close to optimal weather conditions and one year, when heat stress occurred during flowering and grain setting, making it possible to evaluate the heat stress responses of the 190 barley genotypes. In the heat stress prone 2022 year, the general trends were a strong reduction in the reproductive tiller number and a slight reduction in the fertility. In several groups, these negative effects were compensated with significant increases in grain number per ears and with strong increases in the average grain weight. Under controlled conditions, heat stress significantly reduced most of the grain-yield related traits. Among the more tolerant genotypes, two basic response types could be distinguished. One group was able to better preserve the grain number and weight in the main ear under heat stress, while the other was more able to allocate resources into the side tillers during the recovery period. In the combined heat stress, the average trait values were similar to those in the single stress or even lower, and there was no general priming effect clearly detectable. In the case of the 28 genotypes, there were significant correlations between the stress-induced changes in grain-yield related traits measured under field and under controlled conditions, underlining the possibility of combining the information originating from the two different environments.