RFLP-based estimates of parental contribution to F-2- and BC1-derived maize inbreds

Selection and genetic drift during inbreeding may cause differences between the actual and expected proportions of the genome derived by an inbred from each of its parents. We used 70 RFLP loci to determine the frequency and magnitude of deviations from the expected parental contribution among F-2- and BC1-derived maize (Zea mays L.) inbreds. Assuming inbreds i and j were the parents of inbred k, the parental contribution of i to k was estimated as p = (S-ik - S-ij)/(1 - S-ij), where S-ik and S-ij were the average proportions, across the ten linkage groups in maize, of RFLP loci with alleles common to the inbreds in subscript. Bootstrap confidence intervals (CIs) were obtained for p by re-sampling RFLP similarity for each linkage group. Among 62 F-2-derived inbreds, 13 had estimates of p that deviated significantly from the expected value of 0.5. One F-2-derived inbred obtained p = 0.801 of its genome from a parent. Among 34 BC1-derived inbreds, eight had estimates of p that deviated significantly from the expected contribution of 0.75 from the recurrent parent. Two inbreds, both from the same BC1 population, had an estimated p greater than or equal to 0.94. The results suggested that selection during backcrossing generally favored the recurrent parent over the donor parent. Among the inbreds with significant deviations from the expected p, the width of 95% CIs with 70 RFLP loci was > 0.20. Inbreds selfed from the same F-2 or BC1 population varied in p, indicating that coefficients of co-ancestry calculated from pedigree records may give erroneous estimates of genetic relationship.