QTL detection and marker-assisted composite line development

Efficiency of marker-assisted selection (MAS) in a composite line from a cross between inbred lines was evaluated for a trait with heritability 0.1 or 0.25 using simulations. Only significant marker effects identified in a QTL scan in the initial cross were used for MAS. Three selection criteria were investigated: BM = sum of BLUE of marker effects + BLUP of polygenic BV; MR = sum of BLUE of marker effects; and M = number of favorable marker alleles. Reponses to selection over 10 generations, starting from the F2, were compared with standard BLUP selection on EBV derived from phenotype. All MAS strategies outperformed BLUP selection in initial generations by rapidly increasing the frequency of detected QTL. Criterion BM outperformed other criteria even when phenotypes were available only in the F2. With phenotypes available in all generations, MR and M tended to be comparable. When phenotypes and marker data were available only in the F2, MR gave the lowest genetic gain. The efficiency of MAS depended on the stringency of the threshold used for QTL detection. With phenotype and marker data in all generations, responses increased when the threshold became less stringent (α = 0.05 versus 0.01) but differences in response were limited. Benefits of MAS over standard BLUP were lower with high heritability and when a greater proportion of favorable QTL alleles originated from one of the lines for models that did not include polygenic effects and large marker intervals. Responses decreased initially and increased in later generations when marker intervals were small.