TIME SPAN OF PHYLOGENY'S INFLUENCE ON PHENOTYPES

Bivariate and multivariate correlations and regressions are widely used to make comparisons, functional interpretations, and predictions in biology and paleontology. Thirty years ago Joseph Felsenstein published an influential and widely cited study showing how phylogeny might influence quantitative comparisons in biology. Felsenstein dealt with the perceived problem of similarity due to common ancestry by introducing phylogenetically-constrained ‘independent contrasts,’ and then proposing that we study correlations between contrasts rather than correlations between species. Here I use Felsenstein’s ‘Brownian motion’ model of morphology diffusing through time to show that similarity due to common ancestry persists on short time scales, but is rarely important on time scales longer than about 10,000 generations. Ten thousand generations, or at most about 200,000 years, is trivial in the history of diversification of most taxa living today. Species that are similar over long intervals of time remain similar for reasons that are functional rather than phylogenetic.

Brownian motion is a random model, and the problem with Brownian motion as a model for evolution is that it is rarely scaled and calibrated realistically. Without scaling and calibration, Brownian motion is so general that it can be tailored to fit every case of interest – meaning random explanations can never be rejected. Realistic scaling requires that we recognize that natural selection and random drift take place on a generation-to-generation time scale, and realistic calibration involves rates on the order of 0.1 phenotypic standard deviations per generation on a generation-to-generation time scale. The reason similarity due to common ancestry is negligible in most instances is because functional constraints on morphology mean that observed disparity is small compared to the disparity expected for the time involved in an evolutionary diversification. This explains punctuated-equilibrium patterns too of course: most lineages that aren’t constrained functionally by their faunal or floral circumstances evolve rapidly beyond recognition while leaving little trace that they ever existed.