Effects of Two Yield Component Selection Procedures on Maize1

Abstract

Breeders of maize (Zea mays L.) have been concerned with relationships between yield and yield components and the effects of plant density on these relationships. We used three inbred lines (A251, A556, and A619) and backcross‐selection procedures for kernel depth, ear length, and ears per plant to develop six modified versions of each inbred: deep kernel (D), long ear (L), multiple ear (M), deep kernel and long ear (DL), deep kernel and multiple ear (DM), and long ear and multiple ear (LM). A normal (N) inbred plus its six modified versions were crossed with their counterparts in each of the two other inbred backgrounds to produce seven types of single‐cross hybrids: N, D, L, M, DL, DM, and LM. The 21 hybrids were evaluated at plant densities of 24,700, 49,400, and 74,100 plants/ha in 2 years at two locations per year. The objectives of this study were to compare the effects of backcross‐selection for one vs. two of the yield components and to estimate the relative effects of such backcross‐selection on response of hybrids to plant density.Expression of the selected yield component in single component types of hybrids usually exceeded that of their normal counterparts. Increase in expression of selected yield components in double component types occurred only for ears per plant at low plant density and ear length at higher plant densities. Yields of D, L, M, DL, DM, and LM hybrids were −4.2, +3.6, +6.6, −8.0, +0.4, and −2.2%, respectively, of yields of N hybrids when averaged over plant densities and genetic backgrounds. Modified hybrids, particularly double‐component types, did not respond well to medium and high densities. Within the highest yielding hybrid background (A619 ✕ A251), no modified hybrid significantly exceeded the yield of the normal hybrid at any density. Within the limits of our evaluation, backcross‐selection for yield components does not seem to be a useful procedure for improving high yielding hybrids grown at plant densities common in much of the Corn Belt. It might be a useful breeding procedure to improve yield in specific circumstances, e.g., in relatively low yielding hybrid backgrounds, especially at lower plant densities.