A Blog by Ed Yong

Amber trapped dinosaur feathers at different stages in their evolution

This is the first stage. They’re simple, single filaments, found in dense forests and unrecognisable as feathers. They’re different from the hairs of mammals because they don’t have any scales along their length. Nor are tubes from fungi or plants for they lack the thick cell walls that these groups have. They look a lot like the downy “dino-fuzz” that covered Sinosauropterx, the first dinosaur to be found with evidence of feathers. Sinosauropteryx’s fuzz was largely a reddish-brown colour, while those found in the amber ranged from dark to almost transparent.

In the second stage, the filament has turned into a cluster of ‘barbs’. In a modern feather, the barbs branch off from a central stalk or ‘rachis’ but at this stage of feather evolution, the rachis doesn’t exist yet. Instead, the barbs all emerge from a shared base. They’re closest to the fuzz of Sinornithosaurus, another feathered meat-eating dinosaur – possibly venomous, and one of the earliest members of the group that included Velociraptor and Deinonychus.

In stage three, the barbs partially unite to form a central rachis, and even smaller filaments called barbules start to branch off from the barbs. This type of feather is also found in McKellar’s amber and some of them have barbules wit tight coils at their bases. Grebes – modern diving birds – have similar coils in their barbules. When they dive, the coils expand and allow the feathers to absorb water, making the grebe less buoyant and allowing it to sink. It looks like this adaptation was already around in the Cretaceous period.

In the final stages of feather evolution, the barbules develop small hooks, which allowed neighbouring barbs to interlock like strips of Velcro. These hooks turned the feathers into flat vanes, which in turn allowed them to generate lift. These tiny structures helped birds to conquer the air, they’re the ones that cover the bodies of most modern birds, and they too are found in the Cretaceous amber.

Reference: McKellar, Chattertton, Wolfe & Currie. 2011. A Diverse Assemblage of Late Cretaceous Dinosaur and Bird Feathers from Canadian Amber. Science http://dx.doi.org/10.1126/science.1203344

Images courtesy of Science/AAAS

More on feathered dinosaurs:


11 thoughts on “Amber trapped dinosaur feathers at different stages in their evolution

  1. Wow. With all the feather fluff that must have been floating in the air, I shouldn’t be surprised that we have these fossils. I wonder if similar fossils have been ignored because someone didn’t, or couldn’t, realize they were looking at dinosaur feathers inside.

  2. The authors make some interesting claims in this paper. Among them are that their “Stage II” amber-preserved filaments are “generally comparable” in size to S. millennii protofeathers.. oddly, S. millennii protofeathers are at least ~1.2 mm while their amber filaments are ~0.2 mm.. i guess in that sense a mouse is “generally comparable” in size to a ferret..

  3. I asked Currie that! Sadly:

    “I think it is unlikely in that most of the time, amber preserves high fidelity surface details but not much inside. The biggest problem, however, is that the feathers are still rare enough (and small enough) that it is unlikely that we will even try it without having some assurance of success first.”

  4. When the outlier is an elephant, it is!

    I assume the response is tongue in cheek. they are using size as one of a very few diagnostic characteristics , yet by their standards a bunch of filaments of any of the sizes found in nature would qualify. pretty sloppy.

  5. I haven’t read how you’d distinguish these “proto-feathers” from fluffy ferns. I’m sure the pro’s can figure it out, but I haven’t found any good articles explaining the trail of evidence. Anyone know of any, so that I could read more on the subject?

  6. “The difference between S. millennii protofeathers are at least ~1.2 mm while their amber filaments are ~0.2 mm.. i guess in that sense a mouse is ‘generally comparable’ in size to a ferret.”?

    Not exactly. Not to split hairs–or feathers!–but a largish BALB/c mouse (BALB/c) is about 25 g, and a smallish female ferret is about 750 g, or 30X larger. That difference between S. millennii protofeathers and their amber filaments is merely 6X.

Leave a Reply

Your email address will not be published. Required fields are marked *