| Abstrakt: |
Four lines of mice were created from F2's of highly inbred lines. One of the lines was maintained as a control line; in the other three lines an initial nine generations of mass selection entirely on rate of gain between 4 and 11 weeks of age was accompanied by marked increase in rate of gain, moderate increases in feed intake and, therefore, marked increases in gross efficiency. Selection was modified at Generation 9; during the next 12 generations Line 1 was selected for efficiency of feed use, Line 2 was selected for feed intake, and Line 3 was selected for rate of gain as in the previous generations. After the selection scheme was modified, rate of gain continued to increase in all three selected lines with the most rapid gain occurring in fact in Line which was selected for efficiency; regression on generation number for Generations 10 to 21 were, respectively .75±.113, .40±.100 am .38±.115, the lowest being the line selected for rate of gain. Feed intake continued to increase in all lines; the regressions on generation being 1.99 ±. 936, 3.11 ± .708 and 1.50 ± .918 grams. Gross efficiency, or grams gain per gram feed, likewise, improved; regressions being .0031 ± .00067, .0006 ± .00057 and .0013± .00052 g/g per generation, respectively. Realized heritability values for growth rate were .23 ± .020 (1–21) over 21 generations in Line 3, 0.29 ± .034 over the first nine generation in Line 2, and 0.22 ± .067 over the first nine generations in Line 1. Realized h2over Generations 10 through 21 for feed efficiency in Line 1 and for feed intake in Line 2 were .17 ± .042 and .20 ±.057, respectively. Heritabilities calculated from paternal half-sib correlations were in agreement with the realized values, although intrageneration estimates varied considerably with a standard error frequently greater than 1. Pooled over generations, genetic correlations among the three characters rate of gain (G), food intake (F) and efficiency (E), were as follows: rA(G.F)= 0.88 ± .09; rA(GE)= 0.91± .04; rA(EF)= 0.52 ± .33. These values appear to be reasonable estimates of the genetic relationships existing, as evidenced by the striking correlated response obtained during selection. |
