Deviation from the line of least resistance does not exclude genetic constraints: a comment on Berner et al. (2010)

Berner et al. (2010) found that freshwater adaptation of three-spined sticklebacks had not followed the direction of maximal evolvability. Based on this, they suggested that ancestral variance structure has not appreciably biased adaptive diversification. We reanalyze their data to show that evolution has happened in directions of much larger than average evolvability, and we conclude that their data are consistent with an influence of ancestral variational constraints.