The first half of the project was focused on two major objectives: data collection and development and testing of surface morphometric approaches and analysis of modularity. To accomplish the first task, we purchased three portable scanning systems and travelled to over 50 international collections to scan >2500 specimens representing over 1800 species, of which approximately 500 are extinct. During the first three years of this project, we spent extensive time trialing and comparing different approaches to dense morphometric data collection and analysis, ultimately selecting surface semilandmark analysis. Our work included comparisons with standard approaches for morphometric data capture and analysis of modularity, including development of the first likelihood based method for evaluating modularity, and use of morphometric data for estimating divergences, which we published in a series of papers. We also conducted theoretical work, simulating the effect of integration on macroevolutionary patterns, resulting in the development of the Fly-in-the-Tube model of the evolution of integrated phenotypes. While our work relied on already collected material, there are large gaps in the fossil record, so we also conducted fieldwork in poorly studied regions to discover new species and specimens relevant to this work. Our fieldwork during this project focused on the Cretaceous and Paleogene of India and Argentina, as well as the Neogene of Kenya, and resulted in the naming of new vertebrates including dinosaurs.
With the establishment of our data collection and analytical pipeline, we shifted focus to analysis of empirical datasets in the second half of the project, again with each team member focusing on a specific group. To date, this work has benn published over 30 papers and further disseminated through numerous international presentations for scientific and public audiences, through open sharing of our 3D data via Phenome10k.org and via articles in mainstream press and popular science journals, including Scientific American, the Guardian, and BBC. Our published works focus on both the species level and macroevolutionary scale, with papers focusing on every major clade of tetrapods. Our results have produced transformative new understanding of the processes underlying cranial evolution across tetrapods, including identifying a novel model of attenuated evolution, describing a likely common pattern of deep time clade evoution in response to major shifts in ecosystems and climate. Both ecology and life history are primary influences on evolutionary tempo, with nectivorous birds, social mammals, and paedomorphic salamanders show particularly high rates of evolution. Different cranial regions, e.g. feeding structures vs the braincase, also show divergent evolutionary trajectories, epitomising mosaic evolution, which also reflects developmental associations of traits. One of the most surprising results is that there is a clear dichotomy in where variation in the skull is concentrated - in the rostrum of birds and mammals, but in the suspensorium of amphibians and non-avian reptiles. Why this divergent pattern exists is an open question for a future project, but it likely reflects fundamental developmental processes associated with neural crest migration. We have also demonstrated that, in some groups, high integration of traits limits the evolution of structures and that cranial integration is highly conserved within vertebrate classes, but shifts between them, with the biggest differences observed between amphibians and amniotes. The final phase of our project involved unifying all of the datasets into a single cross-tetrapod analysis. This work demonstrates that birds are extremely unusual compared to all other tetrapods, but that this distinctiveness is not associated with a higher rate of evolution, demonstrating the remarkable complexity of evolutionary processes.
