Taphonomic controls on a multi-element skeletal fossil record

Animals with multi-element skeletons, including vertebrates, echinoderms, and arthropods, are some of the most biodiverse and ecologically important animal groups. Understanding the relative impact of the myriad geological and biological factors that impact the quality of multi-element skeletal fossils is thus crucial for disentangling perceived changes in biodiversity through time and shedding light on gaps in the fossil record. We have characterised the specimen-level taphonomic history of virtually the entire Palaeozoic fossil record of echinoids, the class of echinoderms which includes the living sea urchins. We find that the majority of this fossil record consists of disarticulated skeletal elements, and as preservational quality increases, so does the proportion of specimens that can be identified with higher taxonomic precision. We then assessed the relative impacts of multiple geological and biological factors on our specimen quality, identifying that fine-grained host matrix, as well as siliciclastic lithology, are the biggest factors in determining the quality of fossil echinoid preservation, while aspects intrinsic to specific taxonomic groups also play a role. Differential sampling of the fossil record seems to play little role in influencing the distribution of taphonomic grades, and fluctuations in the North American record of Siliciclastic rocks are positively correlated with fluctuations in taphonomic grades. Our results highlight that the factors controlling the animal macrofossil record are varied, and that the interplay between these variables, taphonomic grade, and taxonomic precision impacts our ability to use the fossil record to understand macroevolution.

Methods
This dataset includes the specimen dataset used for analysis, as well as custom R scripts for generating results and figures. 

The specimen dataset was collected from 33 museums by the lead author (Jeffrey R. Thompson) and bolstered by additional specimens found in the available literature. The associated lithology and grain size were recorded for each specimen, alongside data on taxonomy (including most-specific taxonomic rank possible), stratigraphic provenance and locality of collection, and an estimate was made of the associated Taphonomic Grade (see manuscript for full details). All data were collected by one individual (J.R. Thompson), who also vetted all the taxonomic designations of all specimens. Specimens without precise geographic coordinates were georeferenced by locating text-based locality descriptions on Google Maps (available from http://google.com/maps); we additionally included the likely precision of this estimate in the degrees-minutes-seconds format. Palaeo-coordinates for these locations were obtained using the ‘PALAEOMAP’ rotation model through the GPlates API, using the ‘palaeorotate' function of the ‘palaeoverse’ package (Jones et al., 2023). Data were additionally assigned to stage, series, and period level time bins using the ‘majority’ method of the ‘bin_time’ function from the ‘palaeoverse’ package (Jones et al., 2023).