The Stochastic Mode of Molecular Evolution: What Consequences for Systematic Investigations?

Abstract

Biochemical characters have been proposed as ideal sources of phylogenetic information. The homologous characters are easily identified, the relationships between such characters and the underlying genetic code is well understood, and, most importantly, biochemical characters are thought to evolve in a stochastic, clock-like manner. I present a modified version of the Zuckerkandl and Pauling (1965) model of stochastic evolution which is used to explore the consequences of a stochastic mode of evolution. The probability of observing shared character states is determined as a function of the evolutionary rate of a character, the time of independent ancestry for two sister taxa, the time of shared ancestry for the sister taxa independent of their next closest relative, and the number of functionally equivalent, equally probable character states. I found that, while many branching patterns can be reliably reconstructed using stochastically evolving characters, a large subset of theoretically possible phylogenies (as defined by the duration of shared and independent ancestry) would not be derived correctly. The model simulates a "best-case scenario" in which the rate of character evolution is constant over time. Violation of this assumption complicates phylogeny reconstruction and further limits the types of phylogenies that can be addressed with stochastically evolving characters. I discuss the implications of these findings for data analysis and experimental design.