Forty generations of bidirectional selection for mating frequency in male Japanese quail.

A bidirectional replicated selection experiment for high (H1 and H2) or low (L1 and L2) cumulative number of complete matings (CNCM) in male Japanese quail was conducted for 40 generations. In the S32 generation, a subline was taken from each selected line and selection was relaxed. In the randombred control line (C), CNCM and unselected traits changed significantly over generations. Means of the selected lines were adjusted each generation for deviations from the control means. After 40 generations of selection, there was a 21-fold difference in CNCM (59.4 vs 2.8) between Lines H1 and L1. Whereas means increased and variation decreased in the high lines, means decreased and variation increased in the low lines. Regressions of mean CNCM on generation of Lines H1, H2, L1, and L2 were 1.15 +/- 0.08, 0.61 +/- 0.08, -0.26 +/- 0.04, and -0.34 +/- 0.03, respectively. Although responses to selection were observed throughout the 40 generations in Line H1, the low lines appeared to have reached a limit to selection after the S30 generation. Relaxed lines provided supporting evidence for this conclusion. Mean CNCM decreased in the relaxed high lines to that of the control, whereas the low relaxed lines remained at the same level as their corresponding selected lines. Line H2 went into extinction in the 37th generation as a result of reduced fitness. Realized heritabilities of CNCM were 0.09, 0.07, 0.06, and -0.15 in Lines H1, H2, L1, and L2, respectively. As correlated responses to the selection, male quail in the high lines were heavier, exhibited greater relative aggressiveness, and had larger cloacal glands than those of the control and low lines.