Do present rates of biodiversity loss imply that we are entering a sixth 'mass extinction' comparable to the 'big five' of the geological past? If so, can we predict which species are likely to be hardest hit, and is it possible to identify those groups most likely to re-radiate and diversify in the wake of environmental catastrophe? If so, can we use these findings more broadly for global conservation planning? One way to answer these questions is to follow the evolution of a large group that has repeatedly been hit by mass extinctions, but in which multiple lineages have survived in order to re-radiate. Fossils record a complex series of natural experiments that allow us to make generalisations about parallel phenomena occurring in different branches of the evolutionary tree, and at multiple mass extinction events. Ammonites - perhaps the most iconic and instantly recognisable of all fossil groups - have an exceptional fossil record that is ideally suited for this purpose. Originating in the Devonian they transit the end Devonian, end Permian and end Triassic mass extinction events prior to their demise at the end Cretaceous. Ammonites are uniquely suited to a study of extinction selectivity for several other reasons. Firstly, their shells grow by accreting chambers into a (typically) spiral form, such that all post-embryonic stages are fossilized together. This is important because it allows us to account for developmental changes in morphology (and associated shifts in mode of life). Secondly, the external shells of most ammonites can be modeled very simply in a theoretical morphospace, as well as in more complex empirical morphospaces. The manner in which these spaces are depopulated (e.g., the extinction of extreme morphologies first versus random extinction) at successive mass extinction events will reveal the nature of extinction selectivity. Additionally, by modeling developmental trajectories in allometric and other developmental spaces, we can further test whether particular growth patterns increase the risk of extinction, or promote the radiation of lineages after environmental crises. A third advantage of using ammonites is that (unlike virtually any other swimming animals) their bodies do not deform for locomotion. This makes them particularly suitable for hydrodynamic studies. Do any such biomechanical properties correlate with extinction risk? Ammonites were able to swim by jet propulsion; repeatedly squirting water from a siphon within the mantle. We will model how well, fast and efficiently they were able to swim and manoeuvre using two complementary methods; computational fluid dynamics and physical modeling in water tanks. For the former approach, we will write new software much better-suited to simulating the complex flow around moving and rotating bodies than standard computational fluid dynamics (CFD) packages. For the latter we will use structured light or CT scanning coupled with 3D printing in a variety of media to yield realistically weighted and balanced models. Computer design will further enable us to virtually and physically test any potential ammonite morphology from within the theoretical morphospace. Were many physically possible geometries not realised because they had undesirable hydrodynamic properties, and did ammonites repeatedly converge on the same small sample of efficient designs? Or does the dense packing of ammonite genera in certain regions of morphospace reflect wide variation in hydrodynamic parameters in these same regions (and an associated finer subdivision of niche space)? A central objective of this project is to make all of our data, computer code, software, representative video and results available to the widest possible community of academics, educational users and the public. All will be released under Creative Commons (CC0) and OSI approved licenses, and we will promote their re-use and repurposing in other fields of the natural sciences.