| Limited moisture availability due to low precipitation is a major constraint in winter wheat production throughout the Central Great Plains of the USA. Water stress at flowering or during grain filling is a characteristic of the environment of this region and other parts of the world where wheat is grown. Several experiments were conducted in the field and greenhouse both under well watered and water stress conditions. The objectives of this study were to (i) assess the usefulness of a chemical desiccant (NaClO3); (ii) evaluate water relations, canopy temperature depression (CTD), and chlorophyll content in selecting genotypes under drought stress and (iii) estimate the heritability of water potential (Ψ) and relative water content (RWC) by parent-offspring regression and parent-offspring correlations using F4 plants and their F5 progeny means.; The responses of nine cultivars to chemical desiccant, applied 10 days postanthesis, indicated that cultivars with heavier kernels generally suffered lower desiccation injuries (grain yield and kernel weight injury). There was no association between the drought stress and desiccant stress in the field; however, the chemical desiccant effects were more repeatable than drought because the fluctuations caused by natural environmental conditions were greatly reduced. The use of a chemical desiccant in the greenhouse is not recommended because the ranking of some cultivars (for grain yield and kernel weight injury) was different than in the field. The application of the chemical desiccation method to early generation bulk populations did not consistently differentiate among the crosses. The chemical desiccation injury was much higher in the F4 generation than F3 generation, possibly due to the deleterious effect of the desiccant on grain growth.; Cultivar differences were detected for Ψ, RWC, CTD, and chlorophyll content. Significant genetic variability was detected in the F4 and F5 generations among crosses for Ψ and RWC. The cross Sandy/TAM 107 consistently exhibited higher Ψ and RWC than Sandy/Yuma, and Sandy/Lamar crosses. Parent-offspring regressions and parent-offspring correlations were generally low for Ψ and RWC; however, Sandy/TAM 107 showed higher heritability both for Ψ (0.27–0.68) and RWC (0.45–0.59). In the unselected early-generation bulk populations, the crosses showed significant differences for RWC, while the crosses were the same for Ψ. Thus, the progeny selection method would be more beneficial than the bulk selection method for Ψ. The challenge is to use these traits for selection in actual breeding programs that will lead to superior performance of wheat genotypes in dry environments. |