Amargasaurus: a dicraeosaurid with paired neck spines

Amargasaurus cazaui was a relatively small dicraeosaurid sauropod from Early Cretaceous Patagonia. It is known from a connected partial skeleton from the La Amarga Formation of Neuquén Province, Argentina, broadly dated to about 126–120 million years ago. Its paired rows of extraordinarily elongated vertebral spines make it distinctive, while a nearly continuous section of the neck and back preserves their arrangement in sequence.

The bones establish the spines, not the exact soft-tissue structure that surrounded them. A recent histological interpretation favours a fleshy crest or sail joining neighbouring spines over long, isolated keratinous horns. The precise outline and function remain hypotheses. The known individual is usually reconstructed at about 9–10 metres and 2.4–3 tonnes, but the missing tail and parts of the skull leave those values uncertain.

The holotype preserves much of the neck and back, including the paired bony spines. Their outer covering and function are not fossilised, and the absent snout, jaws, most of the tail and parts of the limbs require reconstruction.

Quick facts

Discovery in Patagonia

The genus name means “lizard from La Amarga”, referring to the stream, locality and geological formation. The species name cazaui honours geologist Luis B. Cazau, who drew attention to the fossil potential of the formation and provided information that helped later fieldwork.

José Bonaparte led the eighth palaeontological expedition to Patagonia in 1984. Guillermo Rougier discovered the skeleton in February of that year in a sandy conglomerate about 2.5 kilometres southeast of the bridge over La Amarga Creek. Several neck and back vertebrae remained articulated. That arrangement was especially valuable: it showed that the unusual spines belonged to successive vertebrae rather than to a scattered collection assembled after burial.

Leonardo Salgado and José Bonaparte named and described Amargasaurus cazaui in 1991. The holotype is catalogued as MACN-PV N 15, also written MACN-N 15 in early literature, in the Museo Argentino de Ciencias Naturales collection. The type specimen remains the principal basis for the genus; additional dicraeosaurid vertebrae from the formation have not been assigned to it with confidence.

Age and geological setting

The skeleton came from the lower part of the La Amarga Formation, specifically the Puesto Antigual Member. The beds are generally assigned to the Barremian–early Aptian interval of the Early Cretaceous, commonly expressed as roughly 126–120 million years ago. That range is a stratigraphic estimate. It is not a direct date measured from the dinosaur's bones, and the fossil horizon does not have one precise numerical age.

Sandstones and conglomerates in the member record river channels and floodplains in a continental landscape. The deposits preserve sauropods, small theropods, crocodylomorphs and early mammals. A formation can accumulate over a substantial interval, so every named animal from it did not necessarily share the same place and moment. The broader geological context belongs to the Cretaceous Period.

What the skeleton preserves

The holotype includes the rear and upper portions of the skull, with a well-preserved braincase, but it lacks the snout, jaws and teeth. Twenty-two presacral vertebrae are preserved: thirteen cervical vertebrae and nine dorsal vertebrae. The skeleton also includes a five-vertebra sacrum, several isolated caudal vertebrae and chevrons, a cervical rib, many fragments of dorsal ribs, parts of the shoulder and pelvic girdles, major long bones from one side, an astragalus and two metatarsals.

The hands, much of the feet, the pubis and ischium, the front of the skull and almost the entire tail are missing. Calling it a “nearly complete skeleton” is meaningful only in comparison with many sauropods. It is unusually informative for a dicraeosaurid, yet large regions of a full-body reconstruction are restored from relatives.

Computed tomography of the braincase has been used to reconstruct the brain cavity and inner ear. Histological sections of the bones preserve at least ten cyclical growth marks. Their distribution has been interpreted as evidence that the individual was not yet at its maximum adult size, but growth marks do not provide an exact age without assumptions about growth cycles and missing outer bone layers.

Classification and relatives

Amargasaurus is a sauropod within Diplodocoidea and Dicraeosauridae. Dicraeosaurids commonly had shorter necks than many other sauropods and distinctive vertebral anatomy, including tall or bifurcated neural spines. Its family placement is well supported, while the exact branching order among dicraeosaurid genera changes in different phylogenetic analyses.

Large modern character matrices often place it among derived dicraeosaurids together with Dicraeosaurus and Brachytrachelopan. Which of those genera is the closest relative depends on the characters and taxa included. It is therefore better to state the stable family-level relationship than to present one nearest relative as certain. Amargasaurus was neither a diplodocid nor a “dwarf Brontosaurus”; the guide to dinosaur classification explains the larger sauropod branches.

Only one species is generally accepted: A. cazaui. Bonaparte had used the name Amargasaurus groeberi in a popular book in 1984, but it was not accompanied by a sufficient diagnosis and is treated as a nomen nudum, an unavailable name. The spelling cazuai, sometimes seen, is a typographical error rather than another species.

Two anterior dorsal vertebrae, MOZ-Pv 6126-1 and MOZ-Pv 6126-2, also show long, split spines and resemble the genus. They are retained as indeterminate Dicraeosauridae because they lack enough diagnostic features for a secure referral to Amargasaurus. They do not turn the genus into a taxon represented by multiple complete skeletons.

Size and body proportions

The holotype is usually reconstructed at about 9–10 metres long. A volumetric estimate places its mass near 2.5 tonnes, while broader estimates are around 2.4–3 tonnes. These values refer to the known individual and depend on the restored body volume. Because most of the tail is absent, a different reconstruction of its length changes the total length substantially. The maximum size of a fully mature animal is unknown.

The preserved trunk and limbs support a four-legged sauropod with a relatively short neck. The skull is incomplete, and the body silhouette beyond the articulated vertebral region depends on comparative anatomy. The missing tail, distal limb elements and soft tissues make a mount a synthesis of evidence, not a life-sized replica of one complete fossil.

The paired spines and their soft covering

The best-known feature is a double row of greatly elongated neural spines on the neck and front portion of the back. The bony projections are directly preserved. They are not automatically equivalent to exposed horns: in a living animal, skin, ligaments, muscle or other soft tissues could have bridged or covered them. The fossils preserve the supports but not the full external outline.

Older illustrations often showed two rows of separate, pointed spikes, sometimes with long keratin sheaths. Bone histology has challenged that version. A study of the spine microstructure reported vascular canals, Sharpey's fibres and patterns of remodelled bone consistent with strong soft-tissue attachments. The authors considered a fleshy sail or crest connecting adjacent spines the best-supported model among those tested.

“Sail” does not mean that its height, thickness, colour or purpose is known. Display or recognition, mechanical support for the neck, and more than one role have been proposed. Defence is possible but not demonstrated. The bone anatomy can test how tissues attached; it cannot preserve a behaviour or identify what another animal did during an encounter.

A reconstruction should therefore distinguish the observable paired bones from the inferred tissue between them. A tall continuous crest is a research-based interpretation, not a fossilised membrane. Separate naked spikes are also not a direct reading of the skeleton. Both the outline and any function remain open to new evidence and alternative biomechanical models.

Feeding, movement and behaviour

Amargasaurus was herbivorous, but its own teeth and jaws are missing. Comparisons with other diplodocoids suggest that it cropped vegetation rather than chewing it thoroughly. Its relatively short neck probably offered a different reach from that of high-browsing sauropods, although an exact preferred feeding level cannot be recovered from the incomplete neck alone.

A biomechanical model using the neutral neck posture and inner-ear orientation estimated that the muzzle could rest around 0.8 metres above the ground and reach roughly 2.7 metres when raised. Those numbers depend on how joints are aligned and soft tissues restored. They describe a modelled range, not a posture maintained all day. The animal could alter head position while feeding, moving or looking around.

There is no direct evidence for herds, courtship displays, parental care, calls, colour pattern or seasonal migration in this genus. One well-preserved skeleton cannot establish sexual dimorphism. Even if the spine-supported crest helped display, the fossils do not show whether it differed between males and females or how individuals used it.

What remains uncertain

The fossil supports a distinctive Early Cretaceous Patagonian dicraeosaurid with paired elongated neural spines and a relatively short neck. Histology makes a soft-tissue crest more plausible than two rows of long keratin-covered horns, but does not settle the crest's exact shape or function. The genus is represented by one accepted species and one principal connected skeleton.

Length, mass, feeding height and parts of the body outline rely on models and comparison. Colour, skin, behaviour and social organisation are unknown. These limits do not make Amargasaurus less unusual; they show precisely where the bones end and reconstruction begins.