JWSS - Journal of Water and Soil Science

---

  The most desirable approach to improve characteristics such as yield with polygenic inheritance and genotype × environment interaction, is simultaneous selection using selection index based on related traits. This study was conducted Research Farm of Isfahan University of Technology, to compare selection indices for barley yield improvement under nitrogen stress and non stress conditions, using 49 F₁₃ recombinant inbred lines derived from a cross between Azumamugi and Kanto Nakate Gold. A split plot design with randomized complete block layout in three replications was used. In non-stress condition 100 kg nitrogen per hectare as urea was used at early reproductive growth, stem elongation and grain filling stages. In stress condition, 50 kg nitrogen per hectare was only used at early vegetative growth stage. Smith-Hazel and Pesek-Baker indices based on 4 traits (days to heading, harvest index, growth rate and nitrogen harvest index) and also direct and correlated response of these traits were calculated. The result of response to selection and correlated response indicated that the selection based on higher harvest index and nitrogen harvest index under both conditions screens high yielding lines. Harvest index had high weight in both indices under stress and non stress conditions. Under both conditions and for Smith-Hazel and Pesek-Baker indices, growth rate had negative correlation coefficients. Therefore, the selection based on these indices, separates lines with lower growth rate. In this study, Smith-Hazel index had higher efficiency.

---

To improve a complex character such as yield with low heritability, the use of indirect selection through other characters and a selection index based on different effective traits is recommended. This study was conducted to evaluate different selection methods by using 23 F2:4 wheat lines derived from the cross of Virmarin (susceptible cultivar) and Sardari (tolerant cultivar) at Research Farm of Isfahan University of Technology. A randomized complete block design with three replications was used in two irrigation treatments (i.e. irrigation after 70±3 and 120±3 mm evaporation from class A pan). For each irrigation treatment, different selection indices were constructed based on the number of spike per m², days to heading, number of grain per spike and 1000 – kernel weight. Also direct and correlated responses were calculated for these traits. The results of the response to selection and correlated response indicated that in the same conditions of this study, early heading cultivars with high number of spikes and low number of grain per spike with high 1000-kernel weight will have more yield potential. Number of spike per m² and 1000-kernel weight exhibited positive response in most of the indices and both irrigation treatments, while days to heading and number of grain per spike revealed negative responses. Thus selection based on these indices would result in early heading cultivars with high spike per m2, low number of grain per spike and high 1000–kernel weight. It was also found that Smith-Hazel and Brim-Williams indices had the highest efficiencies.