Testing Correlations in Morphological and Molecular Evolution: a Meta-analysis Approach

  • Joel Heath

Student thesis: Master of Philosophy


Understanding the relationship between genomic and phenotypic evolution, and the factors that facilitate interactions between these processes, is of central importance in evolutionary biology. Mutations giving rise to beneficial phenotypes provide the foundation for descent with modification, and play a major role in evolutionary radiations. However, most genome-wide substitutions occur stochastically, whilst the tempo of phenotypic evolution is driven directly by selection, and it is unclear how these processes are linked. Deciphering the relationship between genomic and phenotypic rates of evolution will yield crucial insight into molecular processes driving adaptive evolution, and allow us to better understand the underlying forces structuring biodiversity. Here, correlations between molecular and morphological branch lengths were tested in a wide range of animal phylogenies reconstructed using Bayesian inference with datasets from 12 recently published total evidence studies. Correlations between rates of molecular and morphological evolution along branches were also tested in time calibrated phylogenies reconstructed using relaxed clock methods. Only very weak correlations were detected between molecular and morphological evolution in terms of both branch lengths and their rates, with the majority of groups studied showing no correlation at all. Analysing rates through time suggests that this apparent disconnect may stem from molecular evolution remaining unaffected by the intensity of selection that often drives accelerated morphological evolution during periods of heightened ecological opportunity. These results show that morphological characters often contain signal that is undetected in the molecular datasets most often used in taxonomic and phylogenetic research. This is likely because morphological variation in animals is predominantly driven by mutations occurring in non-coding regulatory sequences, rather than the molecular markers typically used to infer evolutionary relationships between taxa. Whilst the signal contained within molecular and morphological characters is conflicting, it may also be complimentary, and these dynamics should be considered in all studies aiming to test evolutionary hypotheses and reconstruct the tree of life.
Date of Award1 Aug 2020
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorRussell Garwood (Supervisor) & Robert Sansom (Supervisor)


  • Morphological
  • Evolution
  • Phylogenetics
  • Molecular
  • Correlation
  • Bayesian

Cite this