DISCUSSION
In this study, we investigated how the selection response depends on and shapes the genetic architecture of a multicomponent sex pheromone signal. Through truncation selection, we gradually increased and decreased the relative amount of acetates across 10 generations. As we hypothesized, we found that the acetates responded readily to selection, diverging three-quarters of a standard deviation away from the control line in 10 generations in both the high and low lines. However, in contrast to our hypothesis that selection would also result in indirect responses in the other traits, we found that the response was limited to the acetates only. In addition, we found that levels of genetic variance were constant through time, opposite to the expected erosion of genetic variance in response to selection. Lastly, in line with our expectations, the genetic covariance structure changed in a way that likely facilitated a selection response in the pheromone. We discuss each of these results further below.