TheCommonplace

· plate

When the body changes its mind

A Triassic crocodile relative that stood up on two legs as it grew — the shift was not learned, it was grown into.

Read aloud
Sonselasuchus cedrus at two life stages, with a plot of forelimb-to-hindlimb scaling across the ontogenetic series A two-panel plate. The upper panel shows two silhouettes of the Triassic archosaur Sonselasuchus cedrus drawn to the same linear scale, both facing left on a shared ground line. On the left is stage A, a skeletally immature juvenile: a small quadrupedal animal with a short beaked snout, long neck, slender body, long tapering tail, and four limbs of roughly similar length, all touching the ground. On the right is stage B, a skeletally mature adult, about 2.3 times the linear size of the juvenile, standing bipedally on two long hindlimbs with the trunk held horizontal and the long tail as counterbalance; the forelimbs are held up against the chest and are visibly shorter than the hindlimbs. A shared 50-centimetre scale bar sits between the silhouettes. Below, the second panel plots humerus-to-femur length ratio against femur length across the ontogenetic series; a single rust-coloured line connects a point at femur length seventy-five millimetres with a higher ratio to a point at femur length one hundred seventy-eight millimetres with a lower ratio, showing negative allometric growth of the forelimb relative to the hindlimb. Sonselasuchus cedrus · two life stages, drawn to the same scale stage A · juvenile femur ≈ 75 mm all four limbs bear weight stage B · adult femur ≈ 178 mm hindlimbs bear weight; forelimbs held free 50 cm · same scale for both silhouettes Forelimb-to-hindlimb scaling across growth · schematic, after the paper's allometric analysis femur length (mm) — used here as a growth index humerus / femur length ratio 75 178 higher lower isometry stage A stage B negative allometric trajectory — forelimb grows more slowly than hindlimb
Sonselasuchus cedrus in juvenile and adult form, drawn to the same scale, with the limb-ratio trajectory that separates them.

In a bonebed in the Sonsela Member of the Chinle Formation, under the petrified wood of what is now Arizona, Elliott Armour Smith and Christian Sidor recovered the disarticulated bones of at least thirty-six individuals of a single species. They named it Sonselasuchus cedrus and published it in the Journal of Vertebrate Paleontology in March.1 The paper's contribution is not the animal itself — another small shuvosaurid, another Late Triassic croc-line archosaur built like a proto-ostrich — but what a quarry full of one species at many sizes makes visible. Thirty-three complete femora, the shortest seventy-five millimetres and the longest one hundred seventy-eight, span the animal's life. Across that range, the hindlimb grew faster and more robust than the forelimb; the forelimb's trajectory, measured against the hindlimb's, sat below isometry. The ratio changed as the animal grew.

The abstract's final sentence — the only sentence the press release had to quote accurately — reads: consistent with a transition in locomotory mode from quadrupedal to bipedal during ontogeny. Most of the coverage rendered this as the animal "learning" to walk on two legs. Sci.News and Earth.com and phys.org and the university's own press release reached for learned, as though the four-legged juvenile had one day taken stock and stood up. What the paper describes is not a behaviour the animal acquired. It is a proportion the animal grew into. The body changed its mind about how many legs to use, not by deciding, but by growing one pair longer.

  1. Elliott Armour Smith and Christian A. Sidor, "Osteology and relationships of a new shuvosaurid (Pseudosuchia, Poposauroidea) from the Upper Triassic Chinle Formation of Petrified Forest National Park, Arizona, U.S.A.," Journal of Vertebrate Paleontology e2604859 (published online 8 March 2026). The silhouettes above are composed from the paper's stated measurements, not traced from its figures; the plot is schematic, since the exact allometric coefficient is in the paper's body (paywalled) and not in any public secondary s ↩

back to entries