Bayesian phylogenetic estimation of fossil ages

Area of Science:

• Evolutionary Biology
• Paleontology
• Biostatistics

Background:

• Integrating morphological fossil evidence and molecular sequences for divergence dating is now possible using Bayesian probability.
• The fossilized birth-death tree prior and Lewis-Mk model enable estimation of divergence times and phylogenetic relationships for fossil and extant taxa.

Purpose of the Study:

• To investigate the internal consistency of Bayesian models for divergence dating.
• To assess the accuracy and precision of phylogenetic estimates for fossil ages using empirical datasets.

Main Methods:

• Utilized a Bayesian statistical framework combining morphological and molecular data.
• Applied the fossilized birth-death tree prior and Lewis-Mk model.
• Analyzed two well-characterized datasets of fossil and extant species (penguins and canids).
• Estimated the age of each fossil within the datasets.

Main Results:

• Phylogenetic estimates of fossil ages demonstrated high accuracy, with credible intervals rarely excluding true ages.
• Median relative errors were 5.7% and 13.2% for the two datasets.
• Median relative standard errors (RSD) were 9.2% and 7.2%, indicating good precision.
• Estimated fossil ages were, on average, less than 2 million years from their geological strata midpoint ages.

Conclusions:

• The Bayesian statistical model is an adequate fit for geological and morphological data.
• The framework accurately models the evolution of discrete morphological traits in fossil and extant taxa.
• This approach has potential applications in dating temporally unconstrained fossils and testing evolutionary hypotheses.