A lime-sized cluster of bones, discovered at the Bromacker excavation site in Germany, is the oldest known fossilized vomit from a land predator — approximately 290 million years old, forty million years before dinosaurs existed. CT scanning and chemical analysis revealed bones from at least three different animals within the cluster: a small reptile, a fast-moving lizard-like organism, and part of a much larger plant-eater. The predator was likely an early mammal relative, possibly Dimetrodon or a similar sail-backed synapsid.
The identification as vomit rather than coprolite (fossilized feces) relies on the bone condition: the fragments are partially digested but not as thoroughly degraded as material that has passed through a complete digestive tract. The bones were rejected before digestion was complete. Published in Scientific Reports, the finding provides the first direct evidence of opportunistic feeding by a Permian land predator — three prey species in one meal implies the predator ate whatever was available rather than specializing.
The structural insight is about what failure preserves. Successful digestion destroys the evidence. A predator that catches, kills, eats, and fully digests its prey leaves almost no direct record of the interaction. The trophic connection is invisible in the fossil record. Vomit — a digestive failure — preserves what success would have erased. The bones survived precisely because the process that should have destroyed them was interrupted.
This is why direct evidence of predator-prey interactions is rare in paleontology. Teeth marks on bones, coprolites with identifiable fragments, and gut contents of exceptionally preserved specimens are the only windows into who ate whom. Each of these windows exists because something went wrong: the predator didn't eat the whole carcass, the feces mineralized instead of decomposing, the animal died before digesting. The functional ecology of ancient ecosystems is reconstructed almost entirely from the residues of malfunction.
The Bromacker vomit specimen is 290 million years old. The trophic information it contains — opportunistic predation, multi-species feeding, partial digestion — could not have been inferred from bones alone. The predator's skeleton tells you what it could eat. The rejected meal tells you what it did eat. The difference between capability and behavior is preserved only in the failures.