Diet diversity might explain the differencial survival of Notosuchia (Crocodyliformes) at the Cretaceous-Palaeogene crisis

<p class="western" align="justify"><span lang="en-GB"> The 24 species of crocodylians, including crocodiles, alligators, caimans, and the gharial, represent the extant diversity of Crocodyliformes. These species poorly reflect the past diversity of this group. Indeed, extinct crocodyliforms include hundreds of species adapted to semi-aquatic but also to terrestrial and marine environments. Characterization of the factors driving the macroevolutionary history of Crocodyliformes has been difficult partially because of this ecological disparity. Indeed, studies working at a global scale (</span><span lang="en-GB"><em>i.e.</em></span><span lang="en-GB"> at the scale of the Crocodyliformes or more inclusive) have yielded ambiguous or even contradictory results. This highlighted the need for smaller-scaled studies, both from a phylogenetic and stratigraphic point of view.</span></p> <p class="western" align="justify"><span lang="en-GB"> The work presented here is part of this new framework. It focuses on an extinct group of crocodyliforms, the Notosuchia, at a specific time interval, the Cretaceous-Palaeogene (K-Pg) crisis. This group was chosen because of three characteristics that made it relevant for deciphering the factors explaining extinction/diversification events: (1) it is homogeneous regarding the living environments: almost all notosuchians were fully terrestrial; (2) it exhibits a high degree of ecological diversity with diets ranging from strict herbivory to specialized carnivory, including omnivory; and (3) it crosses a major crisis event, the K-Pg crisis. We tested the impact of body size, using skull length as a proxy, and local palaeotemperature, on the survival/extinction, coded as a binary response variable, at the K-Pg crisis using the phylogenetic logistic regression (PLR). We also investigated the evolution of body size throughout notosuchian evolutionary history and its relationship with diet. The analyses were performed on several sets of dated supertrees of Notosuchia: two different node dating methods were used on two different topologies. Furthermore, each species was dated by randomly picking an age included in its stratigraphic interval and 100 supertrees were produced for each sets (400 in total). This allowed to account for the effect of the stratigraphic and phylogenetic uncertainties. Finally, to assess the effect of the Adamantina Formation, of uncertain age and from which a high number of specimens come from, we performed the PLR analyses on all 4 sets of supertrees, including or not all the species belonging to this formation.</span></p> <p class="western" align="justify"><span lang="en-GB"> We found that local palaeotemperature does not explain the survival/extinction while body size do when all the species from the Adamantina Fm are considered to have faced the crisis: the larger notosuchians were, the higher their probability of survival was. Furthermore, Notosuchia showed a trend towards larger body sizes during the Cretaceous. This trend is driven by the apparition of specialized carnivorous species having significantly larger body sizes than omnivorous ones. Thus, diet rather than just body size might explain the survival/extinction of notosuchians at the K-Pg crisis. Because the relationship between survival and body size is only significant when all the species belonging to the Adamantina Fm are considered to have faced this crisis, we corroborates previous studies which found this formation to have a Laggerst&#228;tten effect.</span></p>