Rates of molecular evolution: Phylogenetic issues and applications

The proper relationship between systematics practice and our understanding of evolution has been long debated. Systematists seek to avoid assumptions about evolutionary process in their methods, yet a growing body of evidence indicates that patterns in rates of evolution can be used to reduce effects of homoplasy. We review variable evolutionary rates for molecular characters in the context of constraints on mutation and fixation. Some constraints, like the genetic code for protein-coding genes, are consistent in the direction of their effects, whereas others, like population size and cladogenesis frequency, are historically variable within and among lineages. We review methods for assessing rate variability, and we estimate comparative absolute rates of change for five sets of mitochondrial DNAs in 12 vertebrates for application in phylogenetic analyses. Unequal weights for subsets of mitochondrial DNAs improved congruence with the most highly corroborated tree in many but not all cases. The largest data set (12,120 bases) yielded the same tree under all four weighting alternatives. This is consistent with the notion, echoing the law of large numbers, that as data sets increase in size, homoplasy will tend to cancel itself out. Even if this notion has validity, however, evolutionary biology will remain vital in systematics if we want to: match sets of taxa with characters likely to be historically informative (when data sets are not sufficiently large); avoid comparing characters with different histories due to reticulations, horizontal transfer, or lineage sorting; avoid assuming random distribution of homoplasy; be alerted to the possibility of long-branch attraction problems; and understand the cause of the hierarchy of taxa in nature as inheritance of genetic material and descent with modification.