A feather in a dinosaur dropping puts modern plumage architecture in a lineage that died out

The dating claim is the one that travels furthest. A vane with a water-repellent structure comparable to a modern pelican's ^[9], and shafts with lightweight spongy cores inside a harder outer layer, the combination that makes modern feathers both light and stiff ^[10], is now on record at 66 million years ^[1]. That is ten million years earlier than the previous evidence for feathers like those on living birds ^[2], which places the older benchmark at roughly 56 million years ago, after the impact rather than before it ^[18]. And the bird wearing it, on the team's reading, was a hesperornithiform: related to Neornithes, but not one of them, and therefore not among the survivors ^[8][4][21].

The identification runs through two leg-bone fragments, 18 and 12 millimetres long, that resemble material previously collected at Hell Creek and attributed to the group ^[7]. That is a resemblance argument on fragments, which is what a partially digested carcass leaves you.

Two fluffy body feathers in the same dropping had loosely arranged branches, echoing primitive feathers described from 100-million-year-old Myanmar amber in a lineage that also died out ^[11]. Taken with the wing feathers, Jingmai O'Connor of the Field Museum reads this bird as occupying a middle ground: flight and tail feathers good enough for diving, body feathers that would have insulated poorly, leaving the animal exposed to the post-impact winter ^[12]. Her framing is explicitly conditional, that plumage good enough for a Mesozoic greenhouse was not good enough for the catastrophe ^[13], and she calls plumage differences a contributing factor rather than the cause ^[14].

Insulation here is inferred from branch arrangement, not measured, and it is not clear whether that insulation deficit was thermally consequential; the comparison that would test it, body feathers from a neornithine of the same age and place, is not in this specimen. The amber material is what anchors the word "primitive" at all, since this plumage passed through a predator's gut before it was buried ^[1], and morphology read off digested feathers needs an external reference point.

One more inference sits inside the same object. The dropping probably came from a Nanotyrannus or a young Tyrannosaurus rex ^[15], and it also held two diamond-shaped scales, 3 to 4 millimetres long, assigned to a gar and judged too small to have been the big predator's meal, so the bird likely ate the fish ^[16]. Frank Muzio of the University of Connecticut notes that a single coprolite is teaching palaeontologists about at least three species ^[17]. The trophic chain is a side note; the load-bearing result is chronological. Modern feather architecture existed before the asteroid, in a branch the asteroid removed, and any account of why only neornithes came through has to start from there.