“Artificial selection to increase the phenotypic variance in gmax fails”

Posted on

Jacqueline L. Sztepanacz and Mark W. Blows

The level of standing genetic variance does not reflect the strength of stabilizing selection on traits

Many traits are subject to stabilizing selection that acts to reduce or maintain levels of genetic variance and the status quo. Each generation genetic variation is increased by mutation and removed by selection, such that the resulting genetic variation, known as the standing genetic variance, should reflect the outcome of these two competing processes. As a consequence, some theories predict that the magnitude of standing genetic variance in traits may reflect the strength of stabilizing selection that acts on them. Traits with high standing genetic variance may be subject to weaker stabilizing selection than traits with low standing genetic variance. This pattern is, of course, correlative. More direct evidence for the strength of stabilizing selection may be gained by manipulating traits to increase their variance, where the extent to which a trait increases in variance indicates the strength of stabilizing selection acting on it.

Working with a population of the Australian fruit-fly Drosophila serrata, Sztepanacz and Blows applied artificial selection to increase the variance in two traits, where the strength of stabilizing selection acting on each was predicted to differ by an order of magnitude. Contrary to their expectations the trait predicted to be under strong stabilizing selection increased in phenotypic variance to a large extent, by more than a third of a phenotypic standard deviation. In contrast, the trait predicted to be under weak stabilizing selection showed a completely unexpected response, decreasing in variance. While their results overall provided evidence for stabilizing selection acting in this population, the pattern was not consistent with predictions based on the levels of estimated mutational variance and standing genetic variance in these traits. The reasons why remain unresolved, but the authors explore several possibilities in the paper. Read the Article