Estimating Rate and Time in Molecular Phylogenies: Beyond the Molecular Clock?

Abstract

The study of rates of character evolution has been a cornerstone of evolutionary biology since the pioneering work of Simpson (1944). It has occupied a similar position in molecular evolutionary studies since Zuckerkandl and Pauling’s (1962, 1965) proposal of the molecular clock. There is a fascinating contrast between these two works, however. Simpson used information about time, from the fossil record, to draw inferences about rates and modes of evolution. His main conclusion was that such rates are highly variable. Although also using information from fossils, Zuckerkandl and Pauling came to just the opposite conclusion about rates of protein evolution. They then argued that if proteins evolved at a roughly constant rate, a study of rates and modes of evolution could be used to say something about timing of events in evolutionary history. Both these ideas about the tempo of character evolution have achieved nearly the status of null hypotheses in their respective disciplines. Although Simpson clearly inferred that some morphological rates have been nearly linear, or “clock-like” over at least moderate periods of time (e.g., Simpson, 1944, pp. 203-204), few paleontologists or morphologists give credence to the notion of morphological clocks. And although there is indisputable evidence that many genes and proteins do not evolve at a constant rate through time (Britten, 1986; Avise, 1994), molecular rate constancy continues to be viewed as a reasonable model even across vast reaches of the tree of life (Wray et al., 1996).