Some practical considerations for using RFLP markers to aid in selection during inbreeding of maize

If molecular markers are to be routinely used in maize (Zea mays L.) breeding for selection of quantitative trait loci (QTL), then consistent marker-trait associations across breeding populations are needed, as are efficient methods for weighting information from different markers. Given 15 restriction fragment length polymorphism (RFLP) markers associated with grain yield in testcrosses of 220 [BS11(FR)C7 X FRMo17] F2 individuals to FRB73, separate weighting schemes were attempted in order to maximize the frequency of favorable marker genotypes associated with increased grain yield in selected F2 individuals and F2:S4 lines. The following principles were apparent: (1) Differential weighting among markers, in addition to weighting individual marker genotypes on the basis of associated mean effects, should be emphasized when using markers to select in breeding populations. This is due to limited population sizes that can readily be handled. (2) Relatively few markers may need to be used to screen segregating populations (e.g., F2) of limited size for loci affecting complex traits, such as combining ability for grain yield, assuming prior knowledge of marker-QTL associations. Markers given greatest weight (largest estimates of associated effects) will determine most selections. (3) When marker-based selection is among individuals at higher levels of inbreeding (e.g., S4) within selected families, more markers need to be used in screening because those associated with relatively small effects have an increased chance of affecting selection. These results suggest a qualitative approach for utilizing RFLP markers to aid in selection of complex traits in commercial hybrid maize breeding programs. Commercial research programs produce thousands of crosses each year aimed at inbred line development. Discovery of molecular markers with consistent QTL associations across breeding populations and close QTL linkages would allow for rapid screening of new F2 populations at a few key markers. Early elimination of individuals with undesirable genotypes would reduce the extent of hybrid performance testing necessary during later stages of inbreeding.