The story begins in January and February of 1924, when von Huene conducted a field survey on the Patagonian coast north of the Pico de Salamanca, roughly 80 kilometres north of Comodoro Rivadavia in Chubut Province. The specimens he recovered came from the uppermost brackish and freshwater layers of the Salamanca Formation, known informally as the ‘Schwarze Leitschicht’ or ‘Banco Negro’, which overlie older marine deposits. Alongside the crocodile bones, the same beds yielded fragmentary turtle remains, tooth plates of the lungfish Ceratodus, and shells of the bivalve Corbicula and the gastropod Viviparus. Although von Huene initially assigned these layers a Late Cretaceous age, later studies revised the horizon to the upper Danian, making it lowermost Palaeocene in age, approximately 62 to 66 million years after the dinosaurs vanished.

The material, now catalogued in the Palaeontological Collection of the University of Tübingen under the GPIT prefix, consists of 23 isolated fragments: a fused frontal, a fused parietal, a partial left dentary and right angular from the lower jaw, seven vertebrae from the neck, back and hip regions, three femora, four metatarsals, two osteoderms and a few unidentifiable scraps that may belong to jugal bones. The first description was published in 1933 by Oskar Kuhn, who assigned the vertebrae to the genera Bottosaurus and Holops, the skull fragments and osteoderms to Leidyosuchus, and the remaining pieces to indeterminate crocodylians. At the time, no caimanine fossils had been described from the Palaeocene of South America, so Kuhn could only compare the material with North American taxa, a limitation that shaped his conclusions in ways the new study now overturns.

One of the most striking observations from the re-evaluation is that the bones almost certainly do not belong to a single animal. The size differences among the vertebrae are dramatic: one cervical vertebra is four times the size of another from the same region, and the four dorsal vertebrae likely represent at least two, and probably four, different individuals. The bones also differ in colour, ranging from whitish and reddish to bluish, and show varying degrees of abrasion and weathering, suggesting distinct taphonomic histories. Whether the fragments came from slightly different horizons or were scattered laterally within the upper part of the formation cannot be determined, particularly because the exact position of von Huene’s collection site within the survey area is no longer known.

The anatomical details preserved in the fragments are nonetheless remarkably informative. The frontal bones, fused along a straight midline suture, form a low, flat, trapezoid skull-table element with a relatively smooth dorsal surface bearing small round to oval pits. Between the orbits, the ventral surface preserves the canal for the olfactory tract and a triangular impression of the cerebral hemisphere. The parietals, also fused, show a well-developed M-shaped suture for the supraoccipital and deep rectangular to oval pits on their dorsal surface, with an open supratemporal fenestra that is not overhung by the skull roof. The dentary fragment preserves four alveoli and a short mandibular symphysis extending only to the first half of the fourth tooth socket, with the splenial excluded from the symphysis. The angular contributes a relatively large external mandibular fenestra and a deep groove that once housed the musculus abductor mandibulae posterior, a key jaw-closing muscle.

Comparisons with the two other crocodyliform groups known from the Palaeocene of South America, the long-snouted marine dyrosaurids and the terrestrial, ziphodont sebecids, rule out both as candidates. Dyrosaurids differ in having frontals with a more rectangular outline at the orbital margin, a much more posteriorly oriented frontoparietal suture, narrower parietals between the supratemporal fenestrae, no supraoccipital exposure on the skull table, a much longer dentary symphysis and a smaller external mandibular fenestra. Sebecids, meanwhile, have a broader frontal between the orbits, a dentary constricted behind the fourth tooth, and an external mandibular fenestra that is very small or entirely closed. The procoelous vertebrae of the Patagonian material, with their concave anterior and convex posterior articulation surfaces, further confirm its eusuchian identity, since amphicoelous presacral vertebrae reported from the same locality indicate that at least one non-eusuchian crocodyliform also lived there.

With dyrosaurids and sebecids excluded, the comparison turns to the five caimanine species now known from the Palaeocene and Eocene of South America: Eocaiman cavernensis from the Eocene of Patagonia, Eocaiman itaboraiensis from the Palaeocene of Brazil, Eocaiman palaeocenicus and Necrosuchus ionensis from the Salamanca Formation, and Protocaiman peligrensis, which, like the von Huene material, comes directly from Punta Peligro. The lower jaw elements show subtle but consistent differences from all of these species. The dentary symphysis of the Tübingen specimen ends at the first half of the fourth alveolus, whereas in E. cavernensis and E. itaboraiensis it extends to the posterior margin of the fifth, and in E. palaeocenicus to the middle of the sixth. The anterior dentary margin is broadly rounded rather than pointed as in E. cavernensis, E. itaboraiensis and N. ionensis, and the angular-surangular suture runs straight rather than curving ventrally as in E. palaeocenicus or sitting high on the fenestra margin as in E. cavernensis.

The skull bones tell a different and tantalising story. The frontal and parietal of the von Huene material are nearly identical to the corresponding elements of Protocaiman peligrensis, the only other caimanine from the Palaeocene or Eocene of South America in which skull-table bones are preserved. The proportions of the interorbital bridge, the flush orbital margins, the narrow supraoccipital exposure and the open supratemporal fenestrae all match closely. The differences are minor: the ornamentation on the frontal is less pronounced, which could simply reflect a younger individual, and the supratemporal bridge of the parietal appears slightly narrower. The problem is that P. peligrensis has no described lower jaw, while every other species lacks skull-table material, so the two halves of the skeleton can never be compared directly. This gap in the fossil record forces a conservative conclusion: the material is assigned to Caimaninae indeterminate, a reminder of how much remains unknown about the earliest caimans.

The re-description also carries a broader message about the value of historical collections. Many of the fossil sites excavated decades or centuries ago are no longer accessible, and the significance of some finds was simply not recognised when they were first described, either because comparative material was missing, the specimens were misidentified, or the appropriate analytical methods did not yet exist. Recent years have seen important discoveries emerge from old drawers, and the von Huene material is a case in point: it represents one of the stratigraphically oldest records of the caiman lineage, yet it was largely omitted from recent phylogenetic studies of early caimanines. Notably, the eminent Argentine palaeontologist Zulma Gasparini, examining the collection in Tübingen in 1978, left a handwritten note concluding that the material was undiagnostic beyond Eusuchia, a judgement the new study refines rather than contradicts by demonstrating a clear caimanine affinity.

The open and closed neurocentral sutures visible on different vertebrae, the seams where the neural arches fused to the centra during growth, reveal that several individuals of different ontogenetic stages, from subadult to adult, lived in the Punta Peligro area during the earliest Palaeocene. The largest vertebra indicates a medium-sized crocodylian comparable to E. palaeocenicus and P. peligrensis, falling well within the size range of the known early caimanines. Although the material cannot yet be pinned to a genus or species, its re-evaluation underscores that the key to understanding how caimans rose from small, fragmentary Palaeocene survivors to the giant Miocene forms such as Purussaurus and Mourasuchus lies in exactly these kinds of incomplete, century-old bones. Future discoveries, and perhaps further re-examinations of museum collections, will be needed to test whether the von Huene fragments ultimately belong to Protocaiman peligrensis or to an additional, still-unnamed member of Patagonia’s ancient crocodile fauna.

Subject of Research: Re-evaluation of historic early Palaeocene caimanine crocodylian fossils from the Salamanca Formation of Patagonia, Argentina

Article Title: Re-evaluation of historic putative caimanine material from the lower Palaeocene Salamanca Formation in Patagonia (Argentina)

Article References: Massonne, T., & Augustin, F. J. (2025). Re-evaluation of historic putative caimanine material from the lower Palaeocene Salamanca Formation in Patagonia (Argentina). Fossil Record, 28(1), 147-163. https://doi.org/10.3897/fr.28.150567

Image Credits: AI Generated

DOI: 10.3897/fr.28.150567

Keywords: Caimaninae, Crocodylia, Palaeocene, Salamanca Formation, Patagonia, Friedrich von Huene, Protocaiman peligrensis, Eocaiman, Necrosuchus, paleontology, Fossil Record, Argentina