Re-examination of an introgression procedure of a polygenic resistance to powdery mildew from wild Prunus davidiana into peach in the light of molecular data and Monte-Carlo simulations.

摘要

A breeding programme has been undertaken to introgress a polygenic resistance to powdery mildew from P. davidiana clone P1908 into peach. The first step of the selection process was based on phenotypic data. Five interspecific hybrids between P1908 and Summergrand peach cultivar have been chosen for their good level of resistance to powdery mildew and each one was the starting point for backcross breeding schemes. The efficiency of this selection strategy was assessed in the light of molecular data (map and QTL detection) collected later on. Three alternative scenarios were compared by Monte-Carlo simulations: (1) random selection of F1 hybrids, (2) selection on phenotypic data, (3) selection on molecular markers. Simulations were performed in agreement with the genetic structure of the experimental population and the linkage mapping data available. For the selection of resistance loci, the selection with molecular markers appeared to be more efficient than the phenotypic selection, that itself was more efficient than random selection. Additional information about the genetic control of the resistance, collected after the selection step demonstrated the importance of preserving wild alleles over the whole genome. Because of this, the 3 scenarios were tested in reverse compared to their efficiency for the selection of resistance loci. When little is known about the QTLs controlling the character to be introgressed, which seems to be the case in fruit breeding, molecular markers could be used to monitor both the selection of target loci and the preservation of wild alleles on the rest of the genome.