Unique skull network complexity of Tyrannosaurus rex among land vertebrates.

Like other diapsids, Tyrannosaurus rex has two openings in the temporal skull region. In addition, like in other dinosaurs, its snout and lower jaw show large cranial fenestrae. In T. rex, they are thought to decrease skull weight, because, unlike most other amniotes, the skull proportion is immense compared to the body. Understanding morphofunctional complexity of this impressive skull architecture requires a broad scale phylogenetic comparison with skull types different to that of dinosaurs with fundamentally diverging cranial regionalization. Extant fully terrestrial vertebrates (amniotes) provide the best opportunities in that regard, as their skull performance is known from life. We apply for the first time anatomical network analysis to study skull bone integration and modular constructions in tyrannosaur and compare it with five representatives of the major amniote groups in order to get an understanding of the general patterns of amniote skull modularity. Our results reveal that the tyrannosaur has the most modular skull organization among the amniotes included in our study, with an unexpected separation of the snout in upper and lower sub-modules and the presence of a lower adductor chamber module. Independent pathways of bone reduction in opossum and chicken resulted in different degrees of cranial complexity with chicken having a typical sauropsidian pattern. The akinetic skull of opossum, alligator, and leatherback turtle evolved in independent ways mirrored in different patterns of skull modularity. Kinetic forms also show great diversity in modularity. The complex tyrannosaur skull modularity likely represents a refined mosaic of phylogenetic and ecological factors with food processing being probably most important for shaping its skull architecture. Mode of food processing primarily shaped skull integration among amniotes, however, phylogenetic patterns of skull integration are low in our sampling. Our general conclusions on amniote skull integrity are obviously preliminary and should be tested in subsequent studies. As such, this study provides a framework for future research focusing on the evolution of modularity on lower taxonomic levels.