Selection for environmental variability of body size in mice.

Abstract

1. An experiment with mice is described which was intended to show whether the degree of sensitivity to environmental influences can be changed by selection of extreme phenotypic deviants for a metric character. 2. Body weight was taken as a metric character and opposite extremes (large × small) were selected as parents in the “extreme” line. Individuals closest to the mean were selected as parents in the “central” line. The coefficient of variation of body weight was followed through 13 generations of this selection. 3. No change of variability was detected in the extreme line. The males, but not the females, of the central line showed a small decline of variability. 4. The two lines diverged in body weight in both sexes in the second half of the experiment, the extreme line being the heavier. The cause of this difference is not known. 5. The conclusion to which the results point is that selection of extreme phenotypic deviants did not materially change the sensitivity to environmental influences. An increase in environmental variance from this cause is therefore unlikely to be an important consequence of unidirectional selection.