The distribution of genetic founder contributions within and among selection and control lines of an experiment with laying hens.

Pedigree analysis was used to monitor genetic dynamics in retrospect in two identically selected lines (S1, S2) and a control (CL) of laying hens, originating from a common base. The pedigree of the last generation animals had nine generations of ancestors the members of the first known generation of ancestors were taken as founders. In each generation and in all lines one son was randomly chosen from each of 20 sires. Selection of females was randomly one daughter from each of 80 dams in CL and the best 80 according to a complex criterion in S1 and S2. The experiment included seven generations of selection. After three generations 51.6, 51.5 and 95.5% of the female founders of the base generation remained in S1, S2 and CL, respectively. The corresponding figures for the founder males were 67.7, 65.1 and 96.2%. These fractions did not change much in the following generations. On average 88.5, 87.1 and 78.8% of the remaining founders of the last generation were present in the pedigree of a last generation animal in S1, S2 and CL, respectively. The average genetic contribution of a founder male to the genome of the last generation was 1.22, 1.25 and 1.04% in S1, S2 and CL, respectively. The corresponding female contributions were 0.64, 0.60 and 0.49%, respectively. The most important founders of S1 and S2, which contributed 90% of the genome of their own line, contributed 80% of the genome of the other line and 73% of the one of CL in both cases. The most important founders of CL contributed 82% to the genomes of S1 and S2. The numbers of founder genome equivalents, were 9.8, 10.3 and 12.2 for S1, S2 and CL in the last generation and indicate a considerable loss of founder alleles in all lines. Effective population sizes calculated from average rate of inbreeding obtained from pedigrees were 91.1, 98.3 and 122.5 for S1, S2 and CL, respectively. Using the squared founder contributions for this purpose resulted in effective sizes of 0.966, 0.941 and 0.959 of the former values for S1, S2 and CL, respectively. Mating only animals with four different grand sires reduced inbreeding compared with random mating over the duration of the experiment. The effectiveness of this scheme was, however, strongly reduced in the last generation.