Data associated with: Global ecomorphological restructuring of dominant marine reptiles prior to the K/Pg mass extinction

Mosasaurid squamates were the dominant amniote predators in marine ecosystems during most of the Late Cretaceous. Here, we use a suite of biomechanically rooted, functionally descriptive ratios in a framework adapted from population ecology to investigate how the morphofunctional disparity of mosasaurids evolved prior to the Cretaceous-Paleogene (K/Pg) mass extinction. Our results suggest that taxonomic turnover in mosasaurid community composition from Campanian to Maastrichtian is reflected by a notable global increase in morphofunctional disparity, especially driven the North American record. Ecomorphospace occupation becomes polarised during the Late Maastrichtian, with morphofunctional disparity plateauing in the Southern Hemisphere and decreasing in the Northern Hemisphere. We show that these changes are not strongly associated with mosasaurid size, but rather with the functional capacities of their skulls. Our novel approach indicates that mosasaurid morphofunctional disparity was in decline in multiple provincial communities before the K-Pg mass extinction, highlighting region-specific patterns of disparity evolution and the importance of assessing vertebrate extinctions both globally and locally. Ecomorphological differentiation in mosasaurid communities, coupled with declines in other formerly abundant marine reptile groups, indicates widespread restructuring of higher trophic levels in marine food webs was well underway when the K/Pg mass extinction took place.

Methods
Raw measurement data collected from:

3D surface scans / photogrammetry models
2D images from published articles
2D images taken by authors

Raw measurement data collected using linear measurement tools in ImageJ, Geomagic Wrap, and MeshLab softwares

Raw measurement data is arranged in MS Excel spreadsheet format (.csv)

Converted trait data (averaged per species) is arranged in MS Excel spreadsheet format (.csv)

R-code for processing trait data built in RStudio