Molecular Branch Lengths Fit Trait Evolution Better than Does Time
Year: 2019
Type: dataset
Abstract: Phylogenetic comparative methods (PCMs) have become a sophisticated suite of techniques for analyzing trait evolution. Powerful PCMs incorporate branch length estimates. Phylogenetic trees with branches scaled to time are called chronograms, and those to substitution are called phylograms. Given identical topologies, chronograms and phylograms may have branch lengths proportional to one another, as expected from the molecular clock hypothesis. However, this hypothesis is often violated. Unlike phylograms, chronograms, coupled with the Brownian motion model, assume that a trait evolves at a constant rate over time and across species. This assumption is unrealistic. Researchers have developed methods to allow variable rates of trait evolution, but why add complexity if phylograms are sufficient at explaining trait variation? We assembled published pairs of chronograms and phylograms with identical topologies (n = 100 jawed vertebrate species) and six phenotypic trait datasets. Phylograms better fit the evolution of the six traits after accounting for model complexity than do chronograms. Also, the variable-rates method we used rescales (modifies) phylogram branch lengths less than chronograms, indicating a better match to trait data. This work will improve comparative studies by refining how researchers employ phylogenies in fields ranging from ecology and biomedicine to linguistics and anthropology.
Source: Figshare
Language: English
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W4394111056