Abstract
This paper is a review of experiments, performed in our laboratory during the past 20 years, designed to analyse the significance of different components of random variability in quantitative traits in laboratory rats and mice. Reduction of genetic variability by using inbred strains and reduction of environmental variability by highly standardized husbandry in laboratory animals did not remarkably reduce the range of random variability in quantitative biological traits. Neither did a tremendous increase of the environmental variability (i.e., living in a natural setting) increase it. Therefore, the postnatal environment cannot be that important as the source of random variability.
Utilizing methods established in twin research, only 20–30% of the range of the body weight in inbred mice were directly estimated to be of environmental origin. The remaining 70–80% were due to a third component creating biological random variability, in addition to the genetic and environmental influences. This third component is effective at or before fertilization and may originate from ooplasmic differences. It is the most important component of the phenotypic random variability, fixing its range and dominating the genetic and the environmental component.
The Gaussian distribution of the body weights observed, even in inbred animals, seems to be an arrangement supporting natural selection rather than the consequence of heterogeneous environmental influences. In a group of inbred rats, the males with the highest chance of parenting the next generation were gathered in the central classes of the distribution of the body weight.