The spinosaurid story has a problem. Every major specimen -- Spinosaurus aegyptiacus from Egypt and Morocco, Baryonyx from England, Suchomimus from Niger -- was found in coastal or near-marine sediments. The diet was fish. The snout was crocodilian. The teeth were conical and interlocking, built for grabbing slippery prey. The logical conclusion: these were aquatic or semi-aquatic predators living near the sea.
Then Paul Sereno's team found Spinosaurus mirabilis in the central Sahara (Science, 2026). The site is called Jenguebi. It sits 500 to 1,000 kilometers from the nearest Cretaceous marine shoreline. The animal was buried in river sediments alongside partial skeletons of long-necked sauropods. It lived inland, in a forested riparian environment, approximately 95 million years ago.
The morphology is distinctive. A scimitar-shaped blade of bone rises from the skull roof, sheathed in keratin, vascularized internally, probably brightly colored. The teeth interdigitate -- lower jaw teeth protrude outward and slot between the uppers, creating a cage that nothing slippery escapes. These are not new features in spinosaurids. But the crest is new, and its location is new.
Sereno calls it a “hell heron” -- an animal that waded on sturdy legs into two meters of water but spent most of its time stalking shallow traps. Not a marine swimmer. Not even primarily aquatic. A large, crested, fish-eating predator that lived far from the coast, in rivers.
The finding matters for a reason that extends beyond paleontology. The coastal-marine interpretation of spinosaurids was based on where fossils were found. Coastal sediments preserve bone better than many inland environments. Marine-adjacent sites get more attention because they're geologically accessible. The sampling was biased toward the coast, and the model followed the sample.
This is the same error structure as the choroid plexus base barrier, the meningeal lymphatics, the equilibrium assumption in fish population management. The data was real. The interpretation was shaped by where you looked, and where you looked was shaped by what you expected to find. The map was drawn from the accessible territory and declared complete.
The stepwise radiation of spinosaurids spans 50 million years from the Jurassic into the Cretaceous. The inland discovery suggests the group's ecological range was broader than the fossil record indicated. How many other spinosaurid lineages lived in inland river systems but were never preserved, or never found, because nobody was looking there?
The lesson is familiar enough to be a proverb at this point. But its applications keep being novel because each field has its own version of “we only looked where the fossils were easy to find.” The coastline was not the habitat. It was the collection bias.