Abstract

Bayesian phylogenetic methods integrating simultaneously morphological and stratigraphic information have been applied increasingly among paleontologists. Most of these studies have used Bayesian methods as an alternative to the widely-used parsimony analysis, to infer macroevolutionary patterns and relationships among species-level or higher taxa. Among recently introduced Bayesian methodologies, the Fossilized Birth-Death (FBD) model allows incorporation of hypotheses on ancestor-descendant relationships in phylogenetic analyses including fossil taxa. Here, the FBD model is used to infer the relationships among an ingroup formed exclusively by fossil individuals, i.e., dipnoan tooth plates from four localities in the Ain el Guettar Formation of Tunisia. Previous analyses of this sample compared the results of phylogenetic analysis using parsimony with stratigraphic methods, inferred a high diversity (five or more genera) in the Ain el Guettar Formation, and interpreted it as an artifact inflated by depositional factors. In the analysis performed here, the uncertainty on the chronostratigraphic relationships among the specimens was included among the prior settings. The results of the analysis confirm the referral of most of the specimens to the taxa Asiatoceratodus, Equinoxiodus, Lavocatodus and Neoceratodus, but reject those to Ceratodus and Ferganoceratodus. The resulting phylogeny constrained the evolution of the Tunisian sample exclusively in the Early Cretaceous, contrasting with the previous scenario inferred by the stratigraphically-calibrated topology resulting from parsimony analysis. The phylogenetic framework also suggests that (1) the sampled localities are laterally equivalent, (2) but three localities are restricted to the youngest part of the section; both results are in agreement with previous stratigraphic analyses of these localities. The FBD model of specimen-level units provides a novel tool for phylogenetic inference among fossils but also for independent tests of stratigraphic scenarios.

Highlights

  • The use of Bayesian inference methods in phylogenetic analysis of morphological features (Lewis, 2001; Nylander et al, 2004, see Lee & Palci, 2015) is a relatively novel approach in paleontology (Pyron, 2011; Lee et al, 2014a; Wright & Hillis, 2014; O’Reilly et al, 2016)

  • The Maximum Clade Credibility Tree (MCCT) of the combined tree samples supports the monophyly of the non-Tunisian species included in the analysis (Fig. 1): each least inclusive node containing the representatives of these species does not include any member of the other species

  • Focusing on the MCCT topology, the analysis found Ferganoceratodus jurassicus as the basalmost lineage of the ingroup, as sister-taxon of the node containing the specimen of Ceratodus africanus and a clade including all other specimens

Read more

Summary

Introduction

The use of Bayesian inference methods in phylogenetic analysis of morphological features (Lewis, 2001; Nylander et al, 2004, see Lee & Palci, 2015) is a relatively novel approach in paleontology (Pyron, 2011; Lee et al, 2014a; Wright & Hillis, 2014; O’Reilly et al, 2016). Stadler (2010) and Heath, Huelsenbeck & Stadler (2014) introduced a method for fossil calibration in phylogenetic analysis that integrates extinct and extant species with a single macroevolutionary model, named the ‘‘Fossilized Birth-Death (FBD) process’’ (Heath, Huelsenbeck & Stadler, 2014). Another significant area of application for Bayesian phylogenetic analyses is the reconstruction of evolutionary patterns among a set of taxa where both sister-taxon (cladogenetic) and ancestor-descendant (anagenetic) relationships are involved. As outlined by Gavryushkina et al (2014) and Gavryushkina et al (2016), failing to account for sampled ancestors may lead to significant bias in parameter estimation, in particular in nodal age inference, in the quantification of cladogenetic events and in the estimation of the fossil diversity

Objectives
Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call