Constraining topological and temporal uncertainty in evolutionary history

Abstract

Resolving phylogenetic relationships and divergence times among clades remains a
central challenge in evolutionary paleobiology. Fossil records, despite being the most direct
evidence of ancient life, are often limited by their sparse preservation, heterogeneous
quality, and taphonomic biases, complicating efforts to reconstruct evolutionary histories.
Integrating the fragmented fossil evidence with molecular datasets is critical for addressing
persistent uncertainties in both topological reconstructions and temporal estimates of clade
origins.
This thesis addresses two interrelated questions in paleobiological research. First, to
what extent can morphological data alone resolve phylogenetic relationships consistent with
molecular hypotheses? Focusing on Panarthropoda, I conducted empirical evaluations using
published morphological datasets to test their ability to recover well-supported molecular
phylogenies. I used a diversity of phylogenetic inference methods to test the three classic
hypotheses within panarthropods: the Lobopodia hypothesis, the Tactopoda hypothesis, and
the Protarthropoda hypothesis. The results show that phylogenetic analyses revealed that
while there is some support for Lobopodia, morphology-based analyses of living
panarthropods are challenging, rendering phylogenetic analyses of their living relatives
problematic. The results suggest that the most progressive and effective solution is to stop
discriminating between molecular and morphological data and instead marshall all data
relevant to the phylogenetic question, thus providing an integrated understanding of the
evolutionary relationships and history.
The second question I explored is how we can reconcile conflicting divergence time
estimates derived from fossil evidence, molecular clocks, and biomarkers to construct robust
temporal frameworks. To address this, I used an integrative approach combining the
model-based fossil interpretation model, the Bayesian Brownian Bridge (BBB) model, with
molecular clock dating methods, to generate the estimation of the origin time of the clade of
1
interest. Applied to two critical clades, angiosperms and eukaryotes, this framework
quantitatively incorporates comprehensive fossil data to be informative calibrations prior
densities to the molecular clock. This approach utilises the best understanding of the
massive fossil data, providing more constraint estimation of these clades. By using this
approach, I estimated a late Jurassic angiosperm origin, and a late Paleoproterozoic origin
of eukaryotes. The results suggested that the novel approach to integrating the wealth of the
fossil record into molecular clock analyses we introduce is well-operated. The results of both
projects provide the revised timelines of the two clades.
Overall, by systematically evaluating morphological-based datasets and related
hypotheses, and by incorporating novel Bayesian methods to provide a model-based
interpretation of fossil records, my PhD project explores the topological and temporal
uncertainty in major evolutionary clades, offering protocols for examining and reducing the
uncertainties in evolutionary studies. This project underscores the necessity of fossil records
in paleobiology studies, and the importance of better and correctly integrating and utilising
fossil information in the modern era.

Date of Award	9 Jul 2025
Original language	English
Awarding Institution	University of Bristol
Supervisor	Philip C J Donoghue (Supervisor) & Davide Pisani (Supervisor)