Melvius: a large Late Cretaceous amiid fish

A newly described skull roof clarifies an unusual fish, while important questions about its species remain open.

Melvius reconstructed as a large amiid fish near a Cretaceous river mouth
Known skull and vertebral bones inform this reconstruction; the complete body proportions and colour are not established by a full skeleton.

Melvius is a genus of large amiid fishes from Late Cretaceous western North America. Its remains include separated skull bones, teeth and vertebrae from continental deposits. A 2023 description of an articulated skull roof and braincase added substantial anatomical information, but no complete skeleton fixes the fish’s full outline.

Recognised speciesMelvius thomasi and M. chauliodous; their distinction needs further testing
GroupAmiid ray-finned fish; placement within Vidalamiinae is under revision
AgeCampanian and Maastrichtian, Late Cretaceous
Main evidenceSeparated skull bones, teeth and vertebrae; one skull roof and braincase preserve articulation
SizeAn estimate for M. thomasi starts at about 1.6 m standard length; exact total length is unknown
HabitatMostly continental deposits; brackish-water tolerance is possible but unproven
Evidence guide

What can the fossils tell us?

An articulated skull roof and braincase expanded the record

The Kirtland Formation specimen described in 2023 preserved these parts in articulation. It revealed anatomical features not previously known or incompletely documented, but it is not a complete skeleton.

Two species from different formations

The type species Melvius thomasi was described from the Hell Creek Formation of the northern Great Plains; later records also include the Maastrichtian Lance Formation. M. chauliodous is known chiefly from Campanian Fruitland and Kirtland formations in New Mexico. These records differ in age and locality, and isolated remains from other places may be assigned only tentatively to the genus or species.

For that reason, a tooth or vertebra from a new locality is not automatically evidence that a named species lived there. Open identifications such as Melvius sp. or Melvius cf. M. chauliodous preserve uncertainty when the available material lacks diagnostic features.

A new view of the skull

In 2023, Robert Sullivan, Steven Jasinski and Thomas Williamson described a well-preserved, articulated skull roof and braincase of M. chauliodous from the Kirtland Formation. It was the first such skull material documented for the genus. The specimen revealed dermopterotic rays and short bundles of parietal rays that were unknown or incompletely recorded before.

The new anatomy also prompted a reassessment of relationships. Earlier classifications placed Melvius in a particular tribe within Amiidae’s subfamily Vidalamiinae. Their phylogenetic analysis did not support the former division of some tribes as distinct clades. This does not settle the genus’s placement; it shows that more complete skull evidence can change which characters are informative.

The same study treated postinfraorbital ornamentation, previously used to distinguish M. chauliodous from M. thomasi, as a dubious character. The authors retained both species until more diagnostic material of M. thomasi becomes available. Their decision keeps the names in use while making clear that the boundary is not fully tested.

Vertebrae and uncertain body length

Melvius vertebral centra have ventrolateral concavities that help distinguish the genus. Bryant’s original description estimated a standard length of at least about 1.6 metres for M. thomasi. Standard length runs from the snout to the base of the tail; it is not the fish’s full length, and the estimate came from incomplete remains rather than a complete skeleton.

A 2024 conference abstract reported an associated skeleton from Wyoming’s Almond Formation, including a trunk centrum 5.6 centimetres wide. The authors regarded the fish as large and discussed differences between the Campanian and Maastrichtian material. This is preliminary conference evidence, not a peer-reviewed monograph; the bone measurement cannot be converted directly into total length.

Comparative reconstruction of a large amiid fish swimming near a Cretaceous river estuary
The broad body plan is compared with other amiid fishes. Melvius skull and vertebrae indicate a large fish, but no complete skeleton establishes exact proportions or colour.

Continental rivers and a changing coastline

Many fossils come from continental formations that preserve river channels, floodplains and low-lying coastal plains. Fruitland and Kirtland sediments formed in a landscape affected by changing waterways near the Western Interior Seaway. Fossils and locality context have led researchers to consider whether some amiids could tolerate brackish water. Proximity to a seaway is not direct proof that Melvius regularly entered marine water.

Large jaws and teeth are compatible with predation, but known material does not identify a regular prey species or preserve a convincing stomach meal. Comparison with the later amiid genus Cyclurus can help explain shared family features, not establish that both fishes ate or behaved in the same way. The marine teleost Enchodus belongs to another lineage and is a separate part of Cretaceous fish diversity.

Frequently asked questions

When and where did Melvius live?

Its recognised species are Late Cretaceous. M. thomasi is associated with the Hell Creek and Lance formations; M. chauliodous is known chiefly from the Fruitland and Kirtland formations.

What fossils are known?

The record includes separated skull bones, teeth and vertebrae. A 2023 Kirtland Formation specimen preserves an articulated skull roof and braincase, but no complete skeleton is known.

How large was the fish?

A standard-length estimate of at least about 1.6 m was proposed for M. thomasi. Other large bones are known, but exact total body length remains uncertain.

Did Melvius live only in freshwater?

Most fossils come from continental deposits. Brackish tolerance has been proposed for some coastal settings, but regular use of marine water is not directly demonstrated.