Deep Time Narrative
Triassic origins through Mesozoic, Chicxulub K-Pg impact boundary, forest collapse, and Paleogene radiation.
Five Events That Made a Bird
A timeline is a promise about resolution, so here is ours up front: this one starts where the direct evidence starts, and each entry says which part of it is fossil and which part is inference. Dates are pulled from the primary geochronological and stratigraphic literature rather than copied off a chart.
BRACKETED 1 · Protofeathers, before anything could fly
Filamentous integument in coelurosaurs is well established and directly preserved. The live question is how much deeper it goes. Kulindadromeus is an ornithischian — the other great branch of Dinosauria — and it preserves filaments; so, in various forms, do Tianyulong and Psittacosaurus. If those structures are all the same thing, simple protofeathers sit at or near the base of Dinosauria.
That “if” is doing real work. A single deep origin is a bracketed inference, and it remains debated against the alternative that filamentous coverings arose more than once by convergence.
CONTESTED Pushing further still: pycnofibers in anurognathid pterosaurs have been reported with feather-like branching and preserved melanosomes, which would place feather-like structures outside Dinosauria altogether. The structures themselves are directly preserved; whether they are homologous with feathers is disputed, and we are not resolving it here.
DIRECT 2 · The paravian radiation
This is the window in which the near-bird body plans appear, and appear in variety. Anchiornis huxleyi (~160 Ma, Tiaojishan Formation) is a four-winged paravian with long pennaceous feathers on both arms and legs, known from over 200 specimens and the first Mesozoic dinosaur to have a near-complete life coloration reconstructed from melanosomes.
Alongside it, the scansoriopterygids were running an entirely different experiment: Ambopteryx longibrachium (~163 Ma) preserves a membranous, bat-like wing supported by a styliform element, corroborating Yi qi. At the young end of the window sits Archaeopteryx, from the Solnhofen lithographic limestone.
Read the interval this way: not a march toward flight, but several lineages independently trying out ways to move through air, at least one of which went nowhere.
DIRECT 3 · The Jehol Biota
Most of what this volume can say about feathers with a straight face comes out of two formations in northeastern China. The Jehol deposits preserve integument on animal after animal: Sinosauropteryx with its countershading and banded tail, Microraptor with four feathered wings, Caudipteryx with symmetrical non-aerodynamic vaned feathers, the confuciusornithids, and Yutyrannus huali (~125 Ma, Yixian Fm), a basal tyrannosauroid with long filaments preserved on multiple specimens.
Yutyrannus is the size record-holder for direct feather evidence — the holotype is estimated at around 9 m and roughly 1,414 kg, though that mass figure traces to secondary reporting and should be checked against the primary description before it's quoted as hard.
DIRECT 4 · Chicxulub
The single best-dated moment on this page. Renne and colleagues place the impact at 66.038 ± 0.011 Ma and show that impact and extinction were — in Paul Renne's phrase — “synchronous to within a gnat's eyebrow.” Global forest collapse across the boundary is documented palynologically as a fern spike.
Non-avian dinosaurs end here, and so do the arboreal stem birds. The sorting mechanism, and how much of it is observed versus reconstructed, gets its own room.
INFERRED 5 · The Neoaves radiation
Modern bird orders diversify explosively in the wake of the extinction. Two large phylogenomic studies anchor this: Jarvis and colleagues across 48 genomes, and Prum and colleagues across 198 taxa. That an abrupt post-K-Pg radiation happened is robust — it survives different datasets, different taxon samples, and different analytical approaches.
CONTESTED The fine branching order at the base of Neoaves is not. The deepest splits may represent a hard or near polytomy — lineages separating so fast that the genome may simply not preserve a resolvable order of events. Published topologies conflict, most visibly in the Columbea/Passerea arrangement versus Columbaves. Treat any confidently drawn tree at the base of Neoaves with suspicion, including the pretty ones.
What would change this: a genuine consensus on the basal topology. Until then, this entry stays a fan, not a ladder.
The one thing to carry out of this room. The sequence is well constrained — feathers long before flight, several flight experiments rather than one, a hard boundary at 66.038 Ma, an explosion afterward. The order of branching inside that explosion is a different kind of claim, and it is genuinely open. A timeline can be confident about when things happened and still honestly refuse to say what split from what.
Notes on the Dating
Where these numbers come from
Ages on this page are taken from the primary geochronology and the formation-level literature for each taxon — Renne et al. (2013) pins the K-Pg — rather than from reproducing a published chronostratigraphic chart. The descriptive paper behind the international chart is available if you want the underlying stage boundaries, but the chart graphic itself is copyrighted and not freely reusable, so we build the axis from the dates rather than the picture.
CONTESTED The handoff to Volume I
Where this timeline ends, the parrot volume begins — but the join is loose, and saying so is the honest move. Eocene stem parrots are known from Messel and the Green River Formation, yet no crown parrot is known from the Paleogene at all, and the placement of the halcyornithids is explicitly described in recent literature as elusive. Psittacopes, long treated as parrot-like, has migrated toward the passerine stem.
Molecular dating supports a Gondwanan/Australasian origin for crown parrots, Late Cretaceous under the favored calibration at roughly 80 Ma — but the estimate is calibration-dependent across a 50–82 Ma range, so treat it as a range, never a date.