Recurrent Parent Recovery in a Backcross

Backcrossing solves a specific problem: an established variety is excellent except for one thing, and a donor somewhere has the gene that fixes it. Cross to the donor once, then cross the offspring repeatedly back to the good variety, selecting each time for the trait you are importing.

Each backcross halves the donor's remaining contribution. The first cross gives 50% recurrent parent, BC1 gives 75%, BC2 87.5%, and BC4 reaches 96.875%. Most programmes stop between BC4 and BC6, at which point the product is the original variety in all visible respects plus the new gene.

The expectation hides the problem that dominates real backcrossing. The figure is a genome-wide average, but the donor segment carrying the target gene is deliberately retained every generation, and everything physically linked to it comes along. That linkage drag can span tens of centimorgans and drag deleterious donor alleles into an otherwise recovered variety — a classic case being disease resistance genes introgressed from wild relatives that arrive with a yield penalty nobody selected for.

Marker-assisted backcrossing exists to attack exactly this. By genotyping for markers flanking the target, a breeder can select the rare individual whose donor segment has been trimmed by recombination on both sides, cutting drag in two or three generations rather than waiting for chance. Selecting simultaneously for recurrent parent markers across the rest of the genome — background selection — can reach BC6-equivalent recovery by BC2, which is worth several years.