Probabilistic methods surpass parsimony when assessing clade support in phylogenetic analyses of discrete morphological data

Abstract

Fossil taxa are critical to inferences of historical diversity and the origins of modern biodiversity, but realizing their evolutionary significance is contingent on restoring fossil species to their correct position within the tree of life. For most fossil species, morphology is the only source of data for phylogenetic inference; this has traditionally been analysed using parsimony, the predominance of which is currently challenged by the development of probabilistic models that achieve greater phylogenetic accuracy. Here, based on simulated and empirical datasets, we explore the relative efficacy of competing phylogenetic methods in terms of clade support. We characterize clade support using bootstrapping for parsimony and Maximum Likelihood, and intrinsic Bayesian posterior probabilities, collapsing branches that exhibit less than 50% support. Ignoring node support, Bayesian inference is the most accurate method in estimating the tree used to simulate the data. After assessing clade support, Bayesian and Maximum Likelihood exhibit comparable levels of accuracy, and parsimony remains the least accurate method. However, Maximum Likelihood is less precise than Bayesian phylogeny estimation, and Bayesian inference recaptures more correct nodes with higher support compared to all other methods, including Maximum Likelihood. We assess the effects of these findings on empirical phylogenies. Our results indicate probabilistic methods should be favoured over parsimony.

Archosauromorph extinction selectivity during the Triassic–Jurassic mass extinction

“…; O'Reilly et al . ). The accuracy of these maximum likelihood trees can be improved by accounting for uncertainty using bootstrapping and subsequent collapsing of poorly‐supported branches into polytomies (Brown et al .…”

“…; O'Reilly et al . ). The use of bootstrapping could be incorporated into similar comparisons between inferred trees and informal composite trees in the future, but at present the use of bootstrapping with this method would cause problems of pseudoreplication, as the input trees are assumed to be independent, but this is not the case with data from bootstrapping as they are based on partially overlapping data.…”

Constraining uncertainty in the timescale of angiosperm evolution and the veracity of a Cretaceous Terrestrial Revolution

“…Furthermore, perceptions of ‘key characters’ have invariably been formulated within the increasingly out‐moded parsimony‐based phylogenetic framework (Wright & Hillis, ; O'Reilly et al ., , ; Puttick et al ., ) used to infer both seed plant relationships and the phylogenetic distribution of characters. Symptomatically, much of the controversy over seed plant relationships is rooted in the false precision of parsimony‐based phylogenetic analyses of morphological characters (O'Reilly et al ., , ; Puttick et al ., ). At the least, the hypotheses of character evolution used to discriminate stem‐ and crown‐angiosperm fossil taxa should be reviewed within a probabilistic framework that can better accommodate the uncertainty associated with such inference.…”

Comparative evaluation of maximum parsimony and Bayesian phylogenetic reconstruction using empirical morphological data

“…Therefore, at least for the data set analysed by the authors, priors and MCMC proposals were not an issue. Although both probabilistic methods are comparable in building trees from morphology (O'Reilly et al ., ), differently from BI, the formal analysis of phylogenetic information under ML was studied more thoroughly (Massingham & Goldman, ; Geuten et al ., ; Mauro et al ., ). If the BI‐ML equivalence demonstrated by Brown et al .…”

“…Researchers were thus prompted to investigate the performance of Markovian models in estimating phylogenetic relationships relying on morphological data when compared to the widely used parsimony approach. Previous simulation studies have shown that the Bayesian model‐based inference provides more accurate phylogenetic trees than the parsimony algorithms across a range of different conditions (Wright & Hillis, ; O'Reilly et al ., , ). However, such accuracy comes at the cost of poor precision, as the recovered Bayesian consensus trees are largely unresolved (Brown et al ., ; O'Reilly et al ., ).…”