A fragment of fossil thigh bone was the only piece carrying this specimen's collection number, and losing it hid where the animal came from for more than 20 years — until the fragment resurfaced in a Porto Alegre university collection and the three bones proved to be a Middle Triassic reptile new to science

Three bones sat in the fossil collection at the Pontifícia Universidade Católica do Rio Grande do Sul, in Porto Alegre, with no verifiable record of where they had come from. A left shoulder blade, a right hip and a left thigh bone, found in association with one another and all the right size to have belonged to a single animal. The piece that would have answered the question, the upper end of the femur with the specimen’s collection number on it, was missing.

It was found again during a visit to that collection by two of the authors of a new study. With the number back in hand the provenance could be verified, and the three bones have now been described in Scientific Reports as Silescelida acristata, a reptile new to science from the Middle Triassic of southern Brazil. It is the first animal of its evolutionary grade reported from the Triassic anywhere in the country.

The provenance of MCP 4186-PV

The paper, by Maurício S. Garcia, Gabriela M. Cerqueira, Francesco Battista, Marco B. de Andrade and Rodrigo T. Müller, states that a portion of the specimen bearing its collection code had been accidentally lost, and that this hindered verification of its provenance and associated data for over 20 years. Garcia, writing about the find for The Conversation, dates the rediscovery to a technical visit in 2022. The paper itself says only that the visit was recent.

The recovered code pinned the specimen, catalogued as MCP 4186-PV, to the Posto site in the municipality of Dona Francisca, Rio Grande do Sul, at 29°37′38.9″S 53°22′06.5″W. That places it in the Dinodontosaurus Assemblage Zone of the Pinheiros-Chiniquá Sequence, part of the Santa Maria Supersequence in the Paraná Basin, which the authors treat as Ladinian in age. The Ladinian is the last stage of the Middle Triassic, running from about 241 million years ago to about 237 million years ago on the current International Chronostratigraphic Chart.

The chain of evidence therefore runs through a catalogue number. Recovering the code is what the paper says restored the provenance and the associated data; the acknowledgements also thank a curator at another museum for sharing fieldwork details.

The animal is named for the gap the loss created. The genus combines the Latin siles, silence, with the Greek skelēs, hind leg, a reference to the missing proximal femur that carried the code. The species epithet acristata means without a crest, and points to a single anatomical feature.

The femur has no fourth trochanter

On the back of a typical archosauriform femur, the muscle that pulls the leg backwards attaches at a raised structure known as the fourth trochanter. On this femur there is no mound and no crest, only faint striations in roughly the right place, barely raised above the shaft at all.

That is the one feature the authors single out as unique to the specimen, labelling it a local autapomorphy in the diagnosis. They note that most eucrocopodans carry a low crest, flange or mound-like structure in that position, and that a complete absence of a distinct attachment site is otherwise seen in the early dinosaur relative Gondwanax, in lagerpetids and in pterosaurs.

The remainder of the diagnosis is a combination of traits: a scapular blade that does not narrow before it flares, an asymmetrical flare at its top, an elliptical tuber above the shoulder socket for the triceps, a hip blade as deep as the socket beneath it, a notch on the lower rim of that socket, a medially expanded femoral head with no lengthwise groove on its upper surface, and distal femoral condyles that hang equally low.

The femur is the most complete element, at 172.9 millimetres long, with a head 43.3 millimetres across its long axis and a distal end measuring 36.1 millimetres. The preserved part of the scapular blade runs 48.4 millimetres and the preserved part of the dorsal iliac blade 45.2 millimetres. Three-dimensional surface models of all three bones are available on Figshare alongside the paper.

Those are the largest dimensions the paper reports. It offers no estimate of body length or body mass anywhere. The small-alligator comparison that has followed the story into the press comes from Garcia’s own popular account of the find rather than from the description. And the description has a good deal less to work with than a headline suggests: there is no skull, no jaw, no teeth and no vertebrae, the genus rests on three postcranial bones from one individual, and its single unique feature is an absence, here on a surface the paper describes as preserved, but an absence still, and absences carry less weight than presences.

The placement moved between analyses

Silescelida sits outside Archosauria, the group that contains crocodiles, dinosaurs and birds. It is an early-diverging member of Eucrocopoda, the wider set of archosauriforms that approach the ancestral archosaur body plan and that contains the archosaurs themselves. The authors describe the early-diverging eucrocopodans, meaning the forms that fall outside Proterochampsia plus Archosauria, as semi-erect, slender, quadrupedal predators.

Where exactly it sits is what the study cannot settle. The team scored the specimen into a matrix of 287 operational units and 919 anatomical characters and ran four analyses, varying how the contested Chinese and Polish material was coded, the Chinese specimens lumped or split and the Polish material as holotype only or full hypodigm. Every analysis put Silescelida at the base of Eucrocopoda. Beyond that the position moved: two of the four recovered it in a clade with the Chinese material and the German Marcianosuchus angustifrons, while some individual trees placed it inside Euparkeriidae, the family named for Euparkeria capensis of South Africa.

The formal listing in the paper records that uncertainty. The specimen is assigned to cf. Euparkeriidae, not to Euparkeriidae.

The instability is not confined to this specimen. Euparkeria remains the only definite member of the family, and whether Euparkeriidae contains anything else has been argued both ways for years. The authors suggest the family may be a more restricted group than has been assumed, or that the euparkeriid-looking anatomy is simply ancestral for eucrocopodans and was retained by several stem-archosaur lineages. They call for a comprehensive reassessment of the putative members.

None of that revises the ancestry of dinosaurs, which some coverage has claimed for it. The paper’s own claims are narrower: a first record of this grade in the Triassic of Brazil, a first putative euparkeriid for the Triassic of South America, and evidence that South America matters more to early archosauriform evolution than the literature has reflected.

Southern Pangaea’s records are all younger

The specimen also sharpens a pattern in the dates. Every early-diverging eucrocopodan known from northern Pangaea comes from the Early Triassic or the early Middle Triassic. In southern Pangaea, Euparkeria is the oldest, from Early Triassic sediments, and everything else including Silescelida comes from the late Middle Triassic or the early Late Triassic.

Two explanations fit. The southern animals may record a later dispersal into southwestern Pangaea, or they may be late survivors of a dispersal that happened as early as the Early Triassic and has no body fossils yet to show for it. Three bones do not distinguish those, and the paper does not try to.

Silescelida is among the youngest records of its grade anywhere. Cuyosuchus of Argentina is currently the youngest and the only one unambiguously Late Triassic, while Asperoris of Tanzania may be about the same age as the Brazilian specimen. Neither Asperoris nor Wangisuchus tzeyii can be compared with this material at all, because both are unequivocally known only from cranial remains.

The Pinheiros-Chiniquá Sequence has already produced proterochampsians, pseudosuchians including Parvosuchus aurelioi and Prestosuchus chiniquensis, and early dinosaur relatives including Gondwanax paraisensis. All of those sit inside the Proterochampsia plus Archosauria group. Silescelida is the first from the unit to fall outside it.

A specimen with no verifiable place of origin is not evidence of much. That is the unremarkable part of this story and the part that took longest: for more than two decades the three bones were an interesting shape in a museum, and what stood between them and a name was a number written on a piece of femur. Collections hold a great many specimens in that condition, and the comparison list in this paper names several of them, taxa now treated as nomina dubia or judged too fragmentary to reassess without new material.