Spinosaurids were coastal predators. This was not a hypothesis — it was a premise. Their elongated snouts, conical interlocking teeth, and pressure-sensing pits resembled those of fish-eating crocodilians. Spinosaurus aegyptiacus had a paddle-shaped tail, dense bones, and retracted nostrils — adaptations associated with aquatic pursuit. The fossils came from coastal deposits. The feeding ecology, the morphology, and the taphonomy all pointed the same direction: these were animals of the shoreline.
Spinosaurus mirabilis was found 620 miles inland, in forested river habitat in the central Sahara.
Paul Sereno's team unearthed the first fragments in Niger in 2019 — a massive curved bone so unfamiliar that the team initially failed to identify it. A 2022 expedition recovered additional material: a scimitar-shaped crest sheathed in keratin, heavily vascularized, probably brightly colored, projecting upward like a blade. The teeth interlocked in the same fish-trapping arrangement as other spinosaurids. The feeding mechanism was preserved. The habitat was not.
The finding does not disprove the aquatic feeding ecology. Spinosaurus mirabilis almost certainly ate fish — the dental architecture makes this clear. What it disproves is the coastal requirement. Inland river systems in Cretaceous Africa apparently supported enough aquatic prey for a large spinosaurid, far from any marine shoreline. The niche was riparian, not littoral. “Hell heron” — the researchers' informal name — captures the ecology precisely: a large wading predator of freshwater systems, not an open-water swimmer or a coastal hunter.
The structural point is about how niche assumptions constrain the search. If spinosaurids are coastal predators, then fieldwork targeting spinosaurids focuses on coastal deposits. The logic is efficient: look where the animals should be. But efficient search strategies produce biased samples. Every spinosaurid found in a coastal deposit reinforces the premise that spinosaurids are coastal, which directs the next expedition to another coastal site, which finds another coastal spinosaurid. The confirmation is circular. The premise determines the search space, and the search space confirms the premise.
Finding Spinosaurus mirabilis 620 miles inland required not looking for spinosaurids. Sereno's team was not conducting a spinosaurid search. They were excavating a Cretaceous river deposit for whatever it contained. The discovery was incidental to a premise-free search — or rather, a search operating under different premises about what the site might yield.
This is a sampling problem that recurs across paleontology and ecology. The geographic range of a taxon is inferred from where it has been found, but where it has been found depends on where people have looked, and where people have looked depends on what they expected to find. The resulting distribution is the intersection of the actual range and the search effort, and these two sets can diverge dramatically when the search is theory-driven. The absence of evidence from unsearched areas is treated as evidence of absence, and the distinction collapses.
The first new spinosaurid species in over a century was not hidden by geological time or taphonomic destruction. It was hidden by a premise about where to look.