Application of Restricted Selection Indices for Grain Yield Improvement in Oats¹

Abstract

Genotypic correlations among grain yield, heading date, and height, calculated from a population of 1,200 F 9 ‐ derived oat ( Avena satlva L.) lines tested in six environments, showed that unrestricted selection for high grain yield would save undesirable lines with late heading dates and tall plant heights. We used restricted selection indices to examine the effect of restricting heading date and height on oat grain‐yield improvement. A restricted selection index for grain yield, with holding of heading date and height constant, was expected to be 57% as efficient for improving grain yield as unrestricted direct selection for grain yield; use of harvest index as a secondary trait in this restricted selection index increased the efficiency to 70%. More severe restrictions on heading date and height reduced the expected advance for grain yield, and very severe restrictions gave negative expected advance. Under some restrictions on heading date and height, indirect selection for grain yield through harvest index was expected to be almost as efficient as the use of a restricted selection index. With mild restrictions on heading date and height, much of the grain yield increase in oats may come from increased harvest index.