Decelerated dinosaur skull evolution with the origin of birds

Abstract

The evolutionary radiation of birds has produced incredible morphological variation, including a huge range of skull form and function. Investigating how this variation arose with respect to non-avian dinosaurs is key to understanding how birds achieved their remarkable success after the Cretaceous-Paleogene extinction event. Using a high-dimensional geometric morphometric approach, we quantified the shape of the skull in unprecedented detail across 354 extant and 37 extinct avian and non-avian dinosaurs. Comparative analyses reveal fundamental differences in how skull shape evolved in birds and non-avian dinosaurs. We find that the overall skull shape evolved faster in non-avian dinosaurs than in birds across all regions of the cranium. In birds, the anterior rostrum is the most rapidly evolving skull region, whereas more posterior regions-such as the parietal, squamosal, and quadrate-exhibited high rates in non-avian dinosaurs. These fast-evolving elements in dinosaurs are strongly associated with feeding biomechanics, forming the jaw joint and supporting the jaw adductor muscles. Rapid pulses of skull evolution coincide with changes to food acquisition strategies and diets, as well as the proliferation of bony skull ornaments. In contrast to the appendicular skeleton, which has been shown to evolve more rapidly in birds, avian cranial morphology is characterised by a striking deceleration in morphological evolution relative to non-avian dinosaurs. These results may be due to the reorganisation of skull structure in birds-including loss of a separate postorbital bone in adults and the emergence of new trade-offs with development and neurosensory demands. Taken together, the remarkable cranial shape diversity in birds was not a product of accelerated evolution from their non-avian relatives, despite their frequent portrayal as an icon of adaptive radiations.

Complex macroevolutionary dynamics underly the evolution of the crocodyliform skull

“…Allometry and phylogenetic history are often considered key factors in shaping and constraining crocodyliform variation and diversification [4,16,31,32]. Allometry has a significant but relatively small effect on overall skull shape (electronic supplementary material, table S1; R 2 = 0.21, p = 0.001), similar to other vertebrate clades [25–27,33]. Phylogenetically informed regressions of skull shape on log-centroid size across the four phylogenetic hypotheses recover a similar significant but weak relationship ( R 2 = 0.07–0.21, p = 0.002–0.02).…”

“…Examining the Procrustes variance and evolutionary rates of the individual landmarks demonstrates that these are diffusely distributed across the crocodyliform skull ( figure 4 ; electronic supplementary material, figure S6), rather than the more concentrated patterns observed in other tetrapod clades [ 25 , 27 , 28 ]. Many skull regions experience high variability, including the pterygoid, ectopterygoid, jugal, quadratojugal and squamosal.…”

“…This pattern is unusual among extant tetrapods, with the other extant archosaur lineage, birds, showing an overwhelming concentration of cranial variation in the anterior rostrum [ 26 , 28 ], while squamates and amphibians show concentrations of high disparity in the elements forming the suspensorium [ 27 , 33 , 40 ]. A more diffuse pattern of cranial disparity and evolutionary rates also characterizes non-avian dinosaurs [ 25 ], which may reflect a greater diversity of trophic niches and food acquisition strategies.…”

“…Interestingly, within-module integration and disparity of the neurocranium (e.g. occipital, postorbital) are low in both non-avian dinosaurs and crocodyliforms [ 25 ], but high in birds [ 26 ], suggesting that this is a derived condition in birds potentially related to changes in development and cranial architecture and function [ 45 ].…”

“…High-density geometric morphometrics seeks to quantify shape data in extremely high detail by distributing landmarks and semilandmarks across the entire surface of the structure of interest, capturing much more nuanced information about phenotypic variation than more simplistic morphometric techniques [23,24]. This approach has been applied to diverse vertebrate clades in recent years and enables the representation of the entire surface of the skull, capturing key structures that are typically overlooked [25][26][27]. With these data, we use phylogenetic comparative methods and evolutionary modelling to reconstruct the evolution of the crocodyliform skull over 200 Myr and test long-standing hypotheses on how convergence and constraint have shaped its evolution.…”