Parsimony, not Bayesian analysis, recovers more stratigraphically congruent phylogenetic trees

Area of Science:

• Evolutionary biology
• Paleontology
• Systematics

Background:

• Accurate phylogenetic trees are crucial for reconstructing evolutionary histories.
• Probabilistic phylogenetic analyses (e.g., Bayesian) are often suggested to outperform traditional parsimony techniques for morphological data.
• Discrepancies exist between simulation findings and empirical data regarding method accuracy.

Purpose of the Study:

• To empirically compare the accuracy of Bayesian and parsimony phylogenetic analyses using morphological data.
• To evaluate phylogenetic tree congruence with stratigraphic age ranges as an independent benchmark.
• To investigate the reasons for discrepancies between simulation and empirical study results.

Main Methods:

• Analysis of 167 independent morphological data matrices from fossil tetrapods.
• Construction of phylogenetic trees using both Bayesian and parsimony methods.
• Assessment of stratigraphic congruence between reconstructed phylogenies and known taxon age ranges.

Main Results:

• Bayesian phylogenetic trees showed significantly lower stratigraphic congruence compared to parsimony trees.
• Empirical data analysis indicates parsimony methods are more accurate for morphological data phylogenetic reconstruction.
• Stratigraphic data serve as a reliable independent benchmark for evaluating phylogenetic accuracy.

Conclusions:

• Parsimony analysis appears more accurate than Bayesian analysis for reconstructing phylogenies from empirical morphological data.
• Simulation studies may not fully capture the complexities of real-world empirical data in phylogenetic analyses.
• Further research is needed to understand the factors contributing to the observed discrepancies, potentially involving data collection or inherent data complexities.