The minimum number of genes contributing to quantitative variation between and within populations

Abstract

A procedure is outlined for estimating the minimum number of freely segregating genetic factors, nE, contributing to the difference in a quantitative character between two populations that have diverged by artificial or natural selection. If certain simple criteria are satisfied approximately on an appropriate scale of measurement, nE can be estimated by comparing the phenotypic means and variances in the two parental populations and in their F1 and F2 hybrids (and backcrosses). This generalizes the method of Wright to genetically heterogeneous (or wild) parental populations, as well as inbred lines. Standard errors of the estimates are derived for large samples. The minimum number of genes involved in producing a large difference between populations in a quantitative trait is typically estimated to be about 5 or 10, with occasional values up to 20. This strongly supports the neo-Darwinian theory that large evolutionary changes usually occur by the accumulation of multiple genetic factors with relatively small effects.