On the analysis of evolutionary change along single branches in a phylogeny.

Comparative biologists are sometimes interested in estimating the evolutionary rate along single branches in a phylogeny. I evaluate two methods by which the evolutionary rate along single branches can be compared with the evolutionary rate throughout the rest of the tree. The first is McPeek's contrasts method, and the second is a likelihood method presented independently in two recently published studies. Although the latter method was developed primarily for the comparison of rates among clades, the approach is equally suited for the analysis of evolutionary rate along single or isolated branches. I find that Type I error is acceptable in both methods but that power and parameter estimation are relatively poor in McPeek's method as it is typically applied.