Prehistoric motion capture: arched footprints record the motions of fossil hominin feet


Hominin footprints can immediately capture anyone’s interest and imagination – fossil bones might seem abstract or unrelatable to a casual observer, but with footprints we experience palpable connections to those we have generated ourselves when walking along the beach or when stepping through mud. For paleoanthropologists, fossil footprints are recognized as remarkable opportunities for insights to the evolution of human anatomy and locomotion.

In 1979, Mary Leakey and Richard Hay announced in Nature the landmark discovery of 3.66 Ma footprints at Laetoli, Tanzania1. With these, our field gained its first glimpse of the intact feet of living Pliocene hominins. One prominent and oft-cited observation made by Leakey and Hay from these footprints was that “the longitudinal arch is well-developed and resembles that of modern man” (p. 320)1. Thirty years later, in 2009, 1.5 Ma hominin footprints from Ileret were announced in Science. That paper also focused heavily on inferred longitudinal arch anatomy, concluding that whichever hominin created those tracks, presumably Homo erectus, had an even more human-like foot than the hominins at Laetoli2

The fact that interpretations of these footprint discoveries focused heavily on their arched shapes makes a lot of sense – longitudinal arches distinguish our feet from those of nonhuman primates, and have long been thought to be critical for human-like walking and/or endurance running3,4. And intuition plays a key role, too. Longitudinally-arched hominin footprints (such as those from Laetoli and Ileret) look a lot like the arched shape we would see if we looked at the soles of our own feet. But for decades we have lacked critical data on whether and how footprints preserve information (anatomical or otherwise) about their creators – because feet and substrates are opaque, we have not been able to observe the process of footprint formation and untangle the complex records of foot anatomy and foot motion held within the resulting print.

My colleagues Steve Gatesy and Peter Falkingham pioneered techniques that rely on biplanar X-ray to “lift up the hood” and see inside substrates as footprints are formed (two roughly perpendicular X-ray fields allow 3-D visualization in the area of overlap5). In concert with biplanar X-ray experiments with living birds, they have used 3-D animation and particle simulation to study the mechanics of footprint formation, and gain insights to dinosaur footprints6

For the past 7 years, we have worked together to develop similar methods for the study of human footprint formation, with an eye towards interpreting fossil hominin footprints. In 2018 we published our method for using biplanar X-ray and 3-D animation to observe the process of human footprint formation7, and in 2021 we presented the integration of particle simulation, which provides a window to directly study how substrates deform in response to human feet moving through them8. This workflow, which we call “track ontogeny”, allows us to study how a footprint develops from start to finish, and allows us to connect specific aspects of human footprint morphology to specific anatomical or biomechanical causes (Fig. 1).

Fig. 1. Left: 3-D animation of a moving foot making a footprint in mud, derived from data acquired in biplanar X-ray experiment. Right: visualizations of particle simulations used to study substrate deformation during footprint formation.

In our new paper in Nature Ecology & Evolution, we applied this approach to study how longitudinally-arched human footprints form. We ran biplanar X-ray experiments in which we acquired 3-D scans people’s feet and placed tiny lead beads on their outsides, such that we could track the motions of the beads in X-ray recordings (Fig. 2). People walked across rigid carbon fiber but also through a variety of custom-made artificial muds that allowed X-rays to easily pass through them (the water and mineral content of normal muds do not allow X-rays to easily pass through). We used specialized software to derive 3-D animations of each person’s 3-D foot models moving and deforming in direct accordance with the motions of the lead beads we could see in biplanar X-ray videos.

Fig. 2. A marked foot making a footprint in biplanar X-ray view.

Using new virtual tools that we developed for measuring the arches of feet and footprints, we observed that footprint morphology not a good match for the anatomy of the sole of the foot, especially in deeper muds (akin to the substrates in which the Laetoli and Ileret footprints are preserved). Regardless of their arch anatomies, people produced highly arched footprints in deep muds. This negative result was, however, accompanied by an exciting positive one – when we looked “inside” substrates using particle simulations, we realized that arched footprints instead record very important signals of foot kinematics. Specifically, longitudinally-arched footprints are directly generated by the heel-sole-toe rollover pattern that is fundamental to human bipedal walking (Fig. 3).

Fig. 3. Particle simulations demonstrated that the footprint’s longitudinal arch forms as a response to the foot’s rotation through the substrate. There is a poor match between footprint and foot morphology - a footprint’s arch can appear 2-3 times larger than that of the foot that made it.

Once we understood how the arches of footprints develop, we were able to apply that understanding to comparative analyses of fossil hominin footprints. We demonstrated that the Laetoli footprints - while arched - preserve direct evidence of foot kinematics different from those of modern humans. We think this most likely has to do with the way they used their forefoot to push off against the ground. Meanwhile, footprints from Ileret appear to record kinematics that were indistinguishable from those of our experimental participants. This evidence supports the hypothesis that the earliest kinematic records of fully human-like bipedalism come from Ileret, presumably from early Pleistocene members of the genus Homo.

Not only do our results offer new perspectives on prominent hominin fossils, but the experimental methods used to acquire them offer great potential for accessing kinematic data from the rapidly growing sample of fossil footprints known the human fossil record. We look forward to continuing to apply this approach ourselves, and to sharing it with others, as we seek to uncover new information about how and when humans came to walk the way we do today.

Read the paper here.


  1. Leakey, M. D. & Hay, R. L. Pliocene footprints in the Laetoli beds at Laetoli, northern Tanzania. Nature 278, 317–323 (1979).
  2. Bennett, M. R. et al. Early hominin foot morphology based on 1.5-million-year-old footprints from Ileret, Kenya. Science 323, 1197–1201 (2009).
  3. Morton, D. J. Evolution of the longitudinal arch of the human foot. J. Bone Jt Surg. 6, 56–90 (1924).
  4. Bramble, D. M. & Lieberman, D. E. Endurance running and the evolution of Homo. Nature 432, 345–352 (2004).
  5. Brainerd, E. L. et al. X-ray reconstruction of moving morphology (XROMM): precision, accuracy and applications in comparative biomechanics research. J. Exp. Zool. Part Ecol. Genet. Physiol. 313A, 262–279 (2010).
  6. Falkingham, P. L. & Gatesy, S. M. The birth of a dinosaur footprint: Subsurface 3D motion reconstruction and discrete element simulation reveal track ontogeny. Proc. Natl. Acad. Sci. 111, 18279–18284 (2014).
  7. Hatala, K. G., Perry, D. A. & Gatesy, S. M. A biplanar X-ray approach for studying the 3D dynamics of human track formation. J. Hum. Evol. 121, 104–118 (2018).
  8. Hatala, K. G., Gatesy, S. M. & Falkingham, P. L. Integration of biplanar X-ray, three-dimensional animation and particle simulation reveals details of human ‘track ontogeny’. Interface Focus 11, 20200075 (2021).

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in