EFFICIENCY OF MULTISTAGE MARKER-ASSISTED SELECTION IN THE IMPROVEMENTOF MULTIPLE QUANTITATIVE TRAITS

The application of marker-assisted selection (MAS) to breeding program mes depends on its relative cost and the expected economic return comp ared to conventional phenotypic selection. The relative efficiency of MAS can be increased through a two-stage selection scheme or through m arker-based, multiple-trait improvement. However, the effectiveness of these alternatives has not been quantified. In this study, we evaluat e the efficiency of MAS relative to conventional phenotypic selection and marker-only selection in multistage selection for the improvement of multiple traits. We further incorporate the costs of obtaining meas urements on phenotypic characters and marker loci into the objective f unction to evaluate the efficiency of MAS with respect to the gain per unit cost. Deterministic analyses indicate that excluding costs, mult iple-trait MAS can be used to increase the aggregate breeding values i n quantitative characters and is expected to be more effective than co nventional selection or single-trait MAS. Two-stage MAS has a slightly reduced gain because of culling in the first stage. If the objective function is to maximize the gain per unit cost, multiple-trait MAS is inferior to phenotypic selection in most of the selection schemes inve stigated when the cost ratio (r) of obtaining measurements on phenotyp ic characters to scoring marker loci is less than unity (r less than o r equal to 1.0) and the heritability (h(2)) is greater than 0.3. The e fficiency of MAS increases as r increases and h(2) decreases. For MAS to be more effective, it is necessary to decrease further the cost ass ociated with molecular marker assays.