Piceoerpeton: a giant fossil salamander of western North America

Piceoerpeton was a large salamander genus from western North America, represented by different species before and after the end-Cretaceous extinction. Its record is based mainly on atlases and trunk vertebrae, not complete skeletons. The Maastrichtian P. naylori was smaller than the Paleogene P. willwoodense, whose estimated size approaches that of the largest known salamanders. Both species add distinct evidence to the ancient amphibian catalogue.

Quick facts

Two species separated by time and size

Piceoerpeton is a fossil salamander genus known from the latest Cretaceous through the Paleogene of western North America. James Gardner’s 2012 revision recognised two species. The older and smaller P. naylori comes from late Maastrichtian deposits in Montana and Wyoming; some early Paleocene occurrences have been proposed but remain less certain. The larger P. willwoodense is known from the late Paleocene of the Western Interior and early Eocene deposits, including the Canadian Arctic.

These occurrences do not describe one population observed continuously across the boundary. They are separate fossils from different formations and ages. The genus-level range shows that related salamanders occurred on both sides of the Cretaceous–Paleogene transition, but it cannot by itself demonstrate that one species survived locally through the event or that the same lineage persisted without interruption.

The two species are distinguished by vertebral anatomy as well as inferred body size. P. naylori is smaller and has less derived vertebral features; Gardner considered it potentially ancestral to P. willwoodense. “Potentially ancestral” is a hypothesis based on relative age and anatomical pattern, not a direct family tree preserved in the rock.

Atlantes and the limits of an incomplete skeleton

The atlas is the first vertebra and articulates with the back of the skull. In fossil salamanders, it can preserve characters useful for identifying taxa: the shape of the odontoid process, the depth of the anterior articular surfaces and the proportions of the neural arch. Gardner described a suite of features that separates Piceoerpeton from other scapherpetontids.

Most of the known record consists of isolated atlantes and trunk vertebrae. Such bones can establish that a salamander was present and can support comparisons among species, but they do not provide an associated skull, limbs and tail. Even fossils from the same formation may represent different animals and species. Size comparisons therefore rely on the anatomy of the preserved elements and on the reliability of their referral.

P. willwoodense has been estimated at about 100 centimetres in snout-to-vent length, making it a very large salamander. A figure near 1.8 metres sometimes repeated in summaries is a broader extrapolation for total length, including the tail, and should not be mistaken for a directly measured complete specimen. The smaller P. naylori should not inherit the large species’ estimate.

Classification and relationships

Piceoerpeton has traditionally been placed in Scapherpetontidae, a family of fossil salamanders often regarded as paedomorphic. Paedomorphosis describes the retention of juvenile characteristics in sexually mature animals, but applying that interpretation to fragmentary fossils requires specific anatomical evidence. It should not be treated as a complete life-history observation for every species in the group.

Phylogenetic work has questioned whether the traditional scapherpetontid genera form a single natural group. One analysis placed P. naylori near the living siren Siren lacertina, while P. willwoodense remained in an unresolved set of relationships involving proteids, amphiumids and the other Piceoerpeton species. These results show sensitivity to character sampling; they do not mean the fossil species was a modern siren or amphiuma.

The fossils are especially useful because the atlas provides a repeatable basis for comparison. Yet the same anatomical concentration creates uncertainty: most analyses have few body regions to work with. Adding securely associated skull and limb material could change the picture.

Habitat, growth and behaviour

Hell Creek and Lance deposits preserve river, floodplain and wetland environments near the end of the Cretaceous. Willwood Formation fossils document early Eocene floodplains and channels; the Arctic records come from a younger, high-latitude setting. The geology makes freshwater and damp habitats plausible for these salamanders, but an isolated vertebra does not pinpoint the microhabitat where its animal lived.

Growth lines in some Paleocene bones have been discussed as evidence of periodic changes in growth. Such marks can record changing bone deposition, but converting them into exact ages or proving annual hibernation requires more than counting lines. A fossil does not preserve the temperature, season or activity of the living animal directly.

No stomach contents identify the diet, and no complete skeleton establishes gait or swimming performance. A large aquatic or semi-aquatic salamander could have eaten invertebrates or small vertebrates, as living analogues do, but those prey remain comparative expectations. The evidence is strongest for the genus’ vertebral anatomy, separate species and broad geographic history; details of colour and daily life remain unknown.