VIRTUAL PALAEONTOLOGY: GAIT RECONSTRUCTION OF EXTINCT VERTEBRATES USING HIGH PERFORMANCE COMPUTING

W.I. Sellers, P.L. Manning, T. Lyson, K. Stevens, and L. Margetts

Abstract

Gait reconstruction of extinct animals requires the integration of palaeontological information obtained from fossils with biological knowledge of the anatomy, physiology and biomechanics of extant animals. Computer simulation provides a methodology for combining multimodal information to produce concrete predictions that can be evaluated and used to assess the likelihood of competing locomotor hypotheses. However, with the advent of much faster supercomputers, such simulations can also explore a wider range of possibilities, allowing the generation of gait hypotheses de novo. In this paper we document the use of an 8000 core computer to produce mechanically and physiologically plausible gaits and trackway patterns for a sub-adult dinosaur (Edmontosaurus annectens), evaluating a large range of locomotor possibilities in terms of running speed. The anatomical reconstruction presented is capable of running and hopping bipedal gaits; trot, pace and single foot symmetrical quadrupedal gaits; and asymmetrical galloping gaits. Surprisingly hopping is the fastest gait (17 ms^-1), followed by quadrupedal galloping (16 ms^-1) and bipedal running (14 ms^-1). Such a hopping gait is considered unlikely for this animal, which would imply that either our anatomical and physiological reconstruction is incorrect or there are important constraints such as skeletal loading and safety factor that are currently not included in our simulation. The most likely errors are in joint ranges of motion, combined with the lengths of muscle fibres and tendons since these values are difficult to reconstruct. Thus the process of gait simulation is able to narrow down our predictions of unknown features of the extinct animal using a functional bracket. Trackway geometries derived from the gait models are currently very basic due to the simplicity of the ground/foot contact model used, but demonstrate the future potential of this technology for interpreting and predicting trackway geometry.

W.I. Sellers. Lecturer in Integrative Vertebrate Biology, Faculty of Life Sciences, The University of Manchester, Jackson's Mill, PO Box 88, Sackville Street, Manchester M60 1QD, UK,  
P.L. Manning. Senior Lecturer in Palaeontology, School of Earth, Atmospheric and Environmental Sciences, University of Manchester, Manchester, M13 9PL, UK.
T. Lyson. Graduate Student, Department of Geology and Geophysics, Yale University, New Haven, USA.
K. Stevens. Professor of Computer Science, Department of Computer and Information Science, Deschutes Hall, University of Oregon, Eugene, OR 97403. 
L. Margetts. Visiting Lecturer in Geology, School of Earth, Atmospheric and Environmental Sciences, University of Manchester, Manchester, M13 9PL, UK.