The curved crest of Spinosaurus mirabilis is preserved as bone; its social use has to be inferred. Researchers favor visual display, but its exact signal, color and behavior remain uncertain. Follow the evidence in that order—preserved structure, reconstructed covering, proposed function—before treating a sword-shaped outline as a weapon.
The useful question is what would distinguish those possibilities. A conspicuous outline could affect another animal without touching it. A weapon would also need to withstand and deliver forces in a particular way. Calling a crest sword-shaped describes its appearance, leaving both jobs to be tested.
A discovery assembled over several years
The University of Chicago account describes expeditions in 2019 and 2022. A local Tuareg guide led researchers to a remote fossil field in Niger. An initially puzzling curved bone became recognizable after further finds and digital assembly of a skull. The team eventually had three crests.
The new species was named in February 2026. That date records the scientific publication, rather than the first moment anyone saw its bones. Keeping those events separate helps make sense of discoveries whose headlines arrive years after fieldwork.
The setting matters too. Fossils were buried in river sediments alongside long-necked dinosaurs, far inland from the ancient coast. An animal found in today’s Sahara need not have lived in a desert. Here, the rocks preserve a different landscape with rivers and large fish.
What belongs to the fossil, and what belongs to the reconstruction?
The Science paper distinguishes preserved pieces from reconstructed anatomy in its figures. The crest is principally formed by joined nasal bones, with neighboring prefrontal bones contributing at its base. It rises along the middle of the skull.
Surface grooves and vascular canals support the authors’ inference of a keratin covering. That covering could have extended the crest beyond the bony core. Its complete outline and color are not preserved. The comparisons with living crested birds help propose a covering; they do not supply a photograph of the extinct animal.
These are different kinds of evidence. A bone can be scanned and measured. A missing covering requires an inference from the bone and comparison with living anatomy. The inference can be well motivated without becoming a direct observation.
The thumbnail is an illustration of that reconstructed animal and its crest, including a bone-like inset. It should not be read as a field photograph or a record of preserved skin color. A realistic surface texture can make uncertainty harder to see, so the distinction is worth keeping beside the picture.
Why a display explanation fits
The Natural History Museum’s discussion describes possible roles in attracting mates or deterring rivals. Both belong to visual signaling: the structure changes what another animal sees.
Consider a hypothetical encounter at a river edge. One animal can notice the other’s outline before either gets within striking distance. A conspicuous headpiece could therefore matter in a contest even if the animals never collide. Competition does not necessarily mean combat.
That example explains the proposed function rather than reconstructing an actual encounter. We do not know which observer mattered most, whether an animal adopted a particular posture, or whether the crest signaled age, condition, identity, or something else. “Display” leaves those questions open.
It also does not prove the structure never made contact with another animal. Establishing a primary visual role and ruling out every occasional physical use are different tasks. The cautious conclusion is that the shape alone cannot demonstrate a striking technique.
Size and sex are especially easy to overread
The paper estimates the subadult holotype’s reconstructed skeletal length at about eight meters. All known individuals are immature, so adult size remains uncertain. Its much-discussed larger crest estimate is conditional: scaling an adult to a large S. aegyptiacus produces roughly 40 centimeters of bony crest length, potentially more than 50 centimeters with a covering.
That is not a measurement of a complete adult mirabilis crest. A scale estimate answers what a structure might measure under specified growth assumptions. It cannot establish that those assumptions describe an animal we have not found.
The same problem affects comparisons between individuals. Three crests are enough to show variation, but different growth stages complicate explanations based on sex. A robust crest on a larger individual does not establish that males had bigger crests. More similarly mature specimens would help separate age-related changes from other differences.
When reading a reconstruction, ask whether a stated dimension comes from preserved material, restored missing pieces, or scaling. Each can contribute useful information. They carry different uncertainties.
A river predator still leaves a swimming debate
Chicago’s account emphasizes interlocking upper and lower teeth, a plausible arrangement for retaining slippery fish. It also reports an inland river habitat. Those findings support a fish-eating animal connected to water, while the researchers favor a wading predator.
The museum includes an outside paleontologist’s caution that the new fossils leave many questions about spinosaurs unanswered. Fragmentary skeletons limit the precision of full-body and behavioral reconstructions. An inland river occurrence does not, by itself, tell us how often an animal entered deeper water.
Habitat, food capture, and locomotion need separate explanations. A jaw can help retain prey without revealing the entire hunt. A river deposit establishes an environment without recording every activity within it. Treating those as one answer would give the fossils more detail than they preserve.
The next useful evidence would narrow a particular uncertainty: better adult skulls for growth and variation, more complete skeletons for proportions, or injuries that help test contact behavior. Until then, mirabilis gives us a striking structure and a plausible display interpretation. The crest is real; its social life still has to be reconstructed.
