PhD Defense - Paul Aubier

La soutenance aura lieu sur le campus Pierre et Marie Curie de Sorbonne Université (4 place Jussieu 75005 Paris), amphithéâtre Durand (Bâtiment Esclangon)
Jury:
Abstract:
Metasuchia appeared during the Jurassic. At the Cretaceous, its diversity included semi-aquatic (e.g. Eusuchia) as well as marine (Tethysuchia) and terrestrial (Notosuchia) organisms. During its evolutionary history, it crossed two major crisis event, the OAE2 and the K/Pg boundary. The response of the Metasuchia to each was different as the first one produced a global decrease in diversity, contrary to the second one. Large scale studies, both temporally and phylogenetically speaking, yielded contrasted results regarding the main drivers of Metasuchian diversification, pleading for more refined studies. TShe second chapter of this manuscript was focused on the differential survival of the Notosuchia at the K/Pg crisis. This group radiated during the Upper Cretaceous but was strongly affected by this crisis and the Sebecidae is the sole clade to have survived it. Investigating the reasons explaining this survival, we showed that the extinction/survival that occurred in this clade during this event was not randomly distributed with respect to the phylogeny, therefore suggesting that there were specific factors, with a strong phylogenetic signal, that may explain it. Using PLR analysis, we showed that large notosuchians were significantly less prone to become extinct than smaller ones. Using PGLS, we were able to show that body size was associated diet in this group, with larger species being carnivorous and smaller species being omnivorous. This led us to hypothesize that the survival of the Sebecidae is explained by their hypercarnivorous diet. According to our analyses, palaeotemperature did not played a significant role in the differential survival of the Notosuchia at the K/Pg crisis. This is explained by the small time scale of our study. Indeed, this variable was probably a major driver of notosuchian diversity as the radiations of this group seems to correlate with palaeoclimatic variations. The work presented in this chapter corroborates the value of smaller-scaled evolutionary studies. Furthermore, it highlights the importance and usefulness of the PCMs in such sutdies. However, such methods require a strong phylogenetic framework. Multiple controversies exist regarding the phylogeny of the Metasuchia. Regarding the Crocodylia, the conflict between the Longirostres and Brevirostres hypotheses not only concerns the topology, but also the dating of the nodes which could strongly affect PCM analyses. Thus, treating this phylogenetic controversy the same way we treated that of the Notosuchia would probably yield strongly contradicting results depending on the phylogenetic hypothesis, therefore impairing their discussion. The third chapter of this manuscript investigates this Longirostres/Brevirostres conflict. We used the 3ta to develop a new method of phylogenetic support assessment. Because of character optimization in parsimony, the link between the characters as hypothesized in the matrix and as optimized on the optimal tree is lost. Because this is not the case in 3ta, phylogenetic support measured directly from the matrix can provide explanation regarding the yielded topology of the optimal tree. We present the Contradictory Index, a metric that computes the support of a relationship hypothesis relative to a second, contradicting one. We analysed the most recently published morphological matrix. We showed that despite the Longirostres hypothesis being the most supported, it is highly contradicted in the matrix. We were able to quantitatively target the Tomistominae as the main providers of support to the Longirostres hypothesis. Furthermore, we showed that characters considered by previous study to be correlated to longirostry were not those that support the most the Longirostres hypothesis. Finally, our results suggests that resolving conflict between characters describing the skull will be more helpful in the resolution of the Longirostres/Brevirostres conflict than the description of new postcranial ones.