THE USE OF GEOMETRIC MORPHOMETRICS AND PHYLOGENETIC COMPARATIVE METHODS TO ASSESS THE EVOLUTION OF THE ABELISAURID MAXILLAE

SECULI PEREYRA, EMANUEL; MÉNDEZ, ARIEL; PASCHETTA, CAROLINA

General Roca

Jornada; REUNIÓN DE COMUNICACIONES DE LA ASOCIACIÓN PALEONTOLÓGICA ARGENTINA; 2023

Asociación Paleontológica Argentina

The morphological variation in the skull and the vertebrae of the clade Theropoda hasbeen studied through the implementation of geometric morphometrics (GM). Until now,only the abelisaurid Majungasaurus crenatissimus was studied using GM, evaluatingontogenetic morphological changes. Furthermore, the fossil record of cranial elements inAbelisauridae provides an excellent opportunity to study the morphological diversity ofthe skull, how the skull morphology changed through the time and which evolutionarypressures have driven the skull evolution in this clade. In this work, we employed GM toquantify and describe the shape of abelisaurid maxillae, and Phylogenetic ComparativeMethods to evaluate the phylogenetic signal and the fit of different evolutionary models.The first four PCs (Principal Components) explain the 83,25% of the variance in the dataset, in which the PC1 explains the 37.35% of the variance and it was related to long(PC1+) and short (PC1-) maxillary shape. Almost all abelisaurids evaluated are in thenegative values of the PC1 showing a short and tall maxilla shape in whichSkorpiovenator, Ekrixinatosaurus and Carnotaurus are in the extreme of this PC beingthe extreme phenotypes within Abelisauridae. The PC2 explains the 22.08% of thevariation in the data and it is related to the anterior part of the maxilla, in which thenegative values of the morphospace are derivative abelisaurids that have straight anteriormaxilla shape. Dilophosaurus, noasaurids and Ceratosaurus occupy this part of themorphospace too, being Noasaurus the extreme phenotype. Despite of the phylogeneticordination methods like Phylomorphospace showed that species more related have similarmaxilla shape, the phylogenetic signal test was not significant. This has two possibleexplanations: 1) the maxilla shape did not follow Brownian motion evolution model 2)the phylogeny used is poorly resolved. When we hypothesized abelisaurids as aphenotypic optimum and fixed different evolution models, a simple Ornstein-Uhlenbeck(phenotypes evolving under stabilizing selection with one phenotype optimum) fit betterto the data set, following a corrected AKAIKE criterion information, suggesting thatabelisaurids evolved to only one phenotypic optimum (short and tall maxilla shape).When we used Bayesian methods to estimate how many phenotype optimums can bedetermined in the abelisaurid phylogeny, we found at least eight regimens showing thatFurilesauria abelisaurids and noasaurids could be a regimen with a probability greaterthan 0.4.