Dinosaur feathers found in amber

April 21, 2015 • 10:00 am

UPDATE: I’m a real dummy; I failed to check the dates of any of these items and a eagle-eyed reader noted that they’re all from 2011! I should have seen that from the dates on the Science paper, if not the links. Oh well, it’s still interesting stuff.


A paper in Science by Ryan McKellar et al. (reference and link below; no access to full paper without $$!) reports the amazing discovery of feathers preserved in amber from the late Cretaceous (around 65-90 million years ago). The amber (fossilized plant resin) was discovered in deposits at Grassy Lake, Alberta, Canada. There’s also a nice piece in The Atlantic which summarizes the importance of the findings and a series of photos at io9.com with explanations of what each photo shows. The striking finding is that there are all stages of feather evolution seen in the amber, from simple filaments to very complex feathers.

Now we’re not sure what creatures these feathers belong to, so I’m jumping the gun a bit with the title. Our uncertainty is because there are no fossils of anything in the layer where these feathers were found. But the authors justify the conclusion that these are dinosaur feathers because they’re in layers near those containing near dino remains:

Although neither avian nor dinosaurian skeletal material has been found in direct association with amber at the Grassy Lake locality, fossils of both groups are present in adjacent stratigraphic units. Hadrosaur footprints are found in close association with the amber, and younger (late Campanian and Maastrichtian) strata of western Canada contain diverse nonavian dinosaur and avian remains. There is currently no way to refer the feathers in amber with certainty to either birds or the rare small theropods from the area. However, the discovery of end- members of the evolutionary-developmental spec- trum in this time interval, and the overlap with structures found only in nonavian dinosaur com- pression fossils, strongly suggests that the proto-feathers described here are from dinosaurs and not birds.

Since we already have all kinds of fossil dinosaurs showing imprints of feathers, and those feathers seem to range from simple filaments to more complex feathers similar to those of modern birds, what does this finding add to what we already know? The answer is that these are 3-dimensional feathers that aren’t compressed, and give us a much better look at what early feathers were like. And that view supports a previously-suggested scenario of where bird feathers came from.

That scenario, as sketched in the paper, is diagrammed below. It begins with single filaments (I) that branch out to form a tuft of filaments (II), and some of these either coalesce to form a central shaft, or “rachis” (IIIb), or develop secondary branches (IIIa), with the next step the development of “tertiary” branches (IIIa + b).

Then those tertiary bits can either develop hooks (“distal barbules”: “d.b.”  in the middle figure) or unhooked “proximal barbules” (“p.b.”).  At this point, stage IV below, we pretty much have a modern feather with a central shaft and side filaments that hook together to make the feather into a unit (essential for flying).  In stage V, other specializations develop. Remember, this figure represents a hypothesis about how feathers evolved before the amber feathers were found.

Screen Shot 2015-04-21 at 8.32.16 AM

What the authors found was that basically every stage of feather development could be seen in the eleven specimens of amber analyzed in the paper. This, then, supports the scenario given above. I’ll show some photos from io9, but let me first add four things.

First this scenario for feather development suggests that feathers evolved, as I and many others long suspected, for thermoregulation. The filaments are, according to the authors, present in densities that would help thermoregulation and “protection” (I’m not sure what they’d protect), militating against any use in gliding (filaments don’t help you glide) and perhaps against ornamentation as their sole function (though they could also have served to ornament the bird).

Second, some of the feathers are so well developed that, according to the authors, they would have enabled the dinosaurs bearing them to fly. I’m not an expert on this, so I’ll take their word for it.

Third, some of the feathers are pigmented, and in ways similar to those of modern birds. Further, some of the filaments are coiled at their base, a feature that modern birds like grebes still have, and use to trap water to ferry to their young. Apparently the coiling of a straight filament enables the groove in the middle to trap water through capillary action.

Finally, all of these stages of feather evolution were found in roughly the same time period (same deposit), implying that there were all sorts of dinos coexisting with different degrees of feather evolution. Certainly not all of them gave rise to modern birds; most surely went extinct along with the rest of the dinosaurs. In fact, it’s likely that none of these specimens are from a species that was ancestral to modern birds.

Now to the feathers; all indented captions are from io9:

An isolated barb from a vaned feather, trapped within a tangled mass of spider’s web in Late Cretaceous Canadian amber. Pigment distribution within this feather fragment suggests that the barb may have been gray or black. Image via Science/AAAS

This is one of the later stages of feather evolution.


Below is an earlier stage when there were just filaments. Notice that there are many all together, which supports the notion that they could have been involved in thermoregulation.

Numerous individual filaments in Late Cretaceous Canadian amber. These filaments are morphologically similar to the protofeathers that have been found as compression fossils associated with some dinosaur skeletons. Pigment distributions within these filaments range from translucent (unpigmented) to near-black (heavily pigmented). Image via Science/AAAS


Here are the water-retaining feathers with filaments coiled at their base:

Cross-section through a feather with basally-coiled barbules, accompanied by a microphysid plant bug. The helical coiling observed within these barbules is most obvious in isolated barbules within the image, and is directly comparable to coils found in modern bird feathers specialized for water uptake. The high number of coils in the amber-entombed feather is suggestive of diving behavior, but similar structures are also used by some modern birds to transport water to the nest. Image via Science/AAAS


Pigmented feathers!

Series of six feather barbs in Late Cretaceous Canadian amber. Localized pigmentation creates a beaded appearance within each barbule: This has implications for the structural interpretation of fossil feathers exhibiting this general morphology. Pigment distribution within the specimen suggests that the feather would have originally been medium- or dark-brown in color. Image via Science/AAAS


More colored feathers.

A feather barb within Late Cretaceous Canadian amber that shows some indication of original coloration. The oblong brown masses within the dark-field photomicrograph are concentrated regions of pigmentation within the barbules. In this specimen, the overall feather color appears to have been medium- or dark-brown. Image via Science/AAAS


And more complex feathers.

Overview of 16 clumped feather barbs in Canadian Late Cretaceous amber. Image via Science/AAAS



McKellar, R. C., B. D. E. Chatterton, A. P. Wolfe, and P. J. Currie. 2011 A diverse assemblage of late Cretaceous dinosaur and bird feathers from Canadian amber. Science 333:1619-1622

56 thoughts on “Dinosaur feathers found in amber

  1. The pictures are stunning, but the idea that we can find (dinosaur) feathers preserved for millions of years is amazing. You could have knocked me over with a … .

  2. The third photograph also shows a fly insect embedded in the amber sample. Hopefully McKellar and his colleagues will coordinate with the entomology community as well.

    It will be interesting to read the seemingly inevitable responses from the intelligent design and creationist camps.

    1. If I were a creationist, I’d go with “Amber is a hoax planted by Satan to mislead us.” Far fewer gymnastics involved in the explanation, and no less rational than whatever ark-dependent theory they might propose.

      1. They will hit on this idea eventually. First they will come up with some simple idea and try it. It will fool a few but then be exposed as a crock. By accident though one of the apologists will have slightly modified that fiorst idea, and it will enjoy greater flummoxing success, and spread. Different variants of the excuse will proliferate and be culled until finally this one emerges …
        “See, evolution is impossible!” they will cry.

    2. The insect is a ‘true bug’ or from the order Hemiptera. These have piercing mouthparts. I cannot id this to family, but it looks like one that would feed on plant sap.

    1. It would be fun, but even if amber preserved DNA it does not preserve it sufficiently well to get a whole genome. Maybe they can patch in amphibian and crocodilian DNA to fill in the gaps.
      Sound familiar?

  3. My daughter recently gave a presentation on T. rex to her second-grade class. The teacher had a tape measure to demonstrate dimensions from the kids’ presentations. It’s a Spanish-immersion program and my Spanish isn’t good enough to get everything he was saying, but when he stretched out the tape to show the size of the dinosaur’s “boca” and had five or six kids come up to show how many T. rex could take in one chomp, the meaning was clear enough.

    We have a hard time grasping that kind of scale, and an even harder time appreciating how many millions of generations of dinosaurs transpired during their time on earth. It’s really mind-boggling that we have any fossil record, much less such marvelously preserved transitional elements like these feathers. It must have been a fulfilling experience for the researchers to work on these specimens and thus paper!

    1. You illustrate something that I wish I knew how to get into the curriculum. I knew from my experience with my father (a chemist) that one could enjoy the thrill of finding things out. But I never experienced it in school until high school at the earliest. I regard this as a missing piece of our education, and of science education in specific.

      (Mind you, I also got the reaction when learning how to computer program, but mixed also with the feeling that I had discovered the most wonderful art medium ever.)

      1. It may be that younger children are naturally discoverers so you wouldn’t remember being abnormally thrilled by finding out things at that age. You sort of take it for granted. It’s only when you have an older perspective with lots of grown-up, dull tasks to do that discovery suddenly becomes fascinating.

  4. Those are fascinating specimens, and the post is a fascinating time warp (or maybe a test to see how closely your readers really pay attention to the sciency stuff?).

    I can’t tell for sure, because my university provides access to Science, but I think the paper is now readable without $$. Another interesting tidbit is that the paper has only been cited 10 times, and 3 of those were an editorial in the original issue, a complaint about their interpretations and the authors’ (pretty satisfactory) response to the complaint.

      1. Not my field either (I’m not even a biologist), but isn’t Feduccia the guy arguing that birds aren’t even dinosaurs, but instead another type of archosaur that branched off earlier? I know I should address his arguments and not his credibility, but he’s certainly well outside the mainstream.

        1. It is certainly true that Feduccia is “outside the mainstream”. But that does not invalidate his call to look at these fossils from a perspective that does not pre-suppose them to be dinosaur feathers. For example, what is missing from any of the amber specimens? (the dinosaur) Does that make Feduccia right about everything else, or even right in his interpretation of these fossils? Of course not.

          1. Well, of course they’re dinosaur feathers. The main question is which dinosaurs. Just about everyone except Feduccia knows that birds are dinosaurs. What we should be cautious about is assigning these feathers to non-avian or non-flying theropods. We don’t know that much about the body feathers of early birds. Confuciusornis, for example, seems to have quite a lot of “dinofuzz”, which is probably a decayed form of whatever its body feathers looked like. Maybe it had stage III feathers. Who knows?

    1. I had to renew my (free) registration on Science’s website, but … yep, I’m getting it downloaded now, along with the SOM.
      I did a double take on this because I knew that I’d heard a podcast recently with Phil Currie talking about this discovery, but people are normally fairly tight-lipped about un-published stuff.
      Which was the podcast? Probably one of Dave Marshall’s “Palaeocast

  5. Exciting stuff, Spielberg be damned! Some day hopefully we’ll find some amber with an actual chunk of feathered dino embedded in it like that great baltic amber lizard from the Eocene.


    BBC also has an article arguing for a sixth ancient mass extinction, called the Capitanian, 262 MYA. There is disagreement of course, but it is an interesting discussion based around the 87% decline in brachiopod species. Do we have any paleontologists/geologists in the house who care to weigh in?


      1. I guess I was too obtuse (above) to get credit as “eagle-eyed”. WordPress (or whatever generates the “Related” links) actually gets first credit, because it linked a shorter version that you did post in 2011.

      2. I had never seen these fascinating amber samples with the detailed chronology of traits. Thanks for resurrecting it.

  6. This is really really really cool. I vaguely remember hearing about it, but I am sure I have never seen it and I wonder why I did not hunt it down. Thank you for this post.

  7. I feel like I’m standing next to a feathered reptile with these images on my screen; like it’s a window and I could put my hand outside and touch the plumage.

    I like this feeling.

    1. I know, I thought kind of the same thing when I read about the hadrosaur foot prints. Some hadrosuar was just walking to go get lunch one day and then BAM, there are a bunch of scientists looking at it’s footprints in 2011.
      That’s pretty cool.

  8. Pro forma announcement: of course they’re dinosaur feathers. Dinosaurs are the only animals known to have feathers. The question is whether they’re non-avian dinosaurs, and that gets a bit sticky, because the definition of “avian” is a bit loose, and there’s a lot of ambiguity around the area of the tree in which flight, modern-type feathers, and birds (no two of which are the same thing) arose.

    Also, readers interested in feather evolution might want to check out the seminal paper in feather evolution, Prum & Brush 2002. I think it’s pretty accessible to a non-scientist, but feel free to disabuse me if it isn’t.

    Prum R.O., Brush A.H. The evolutionary origin and diversification of feathers. Quarterly Review of Biology 2002; 77:261-295.


  9. present in densities that would help thermoregulation and “protection” (I’m not sure what they’d protect)

    Protect the skin from biting insects? Of course insects such a fleas would then evolve a way to use the hair to their advantage.

  10. Thanks for this post…I read that it has been posted before, but I’m relatively new here so this is the first I’ve read of this amazing find. Truly unbelievable! Amber has given humans so many interesting treasures. I wonder what other mysteries are yet to be found in amber.

  11. It seems that there have been a lot of things found in amber. I’m curious how widespread, bountiful and large these amber deposits can get.

  12. There are a few common theories on the evolutionary origin (the “why”) of feathers: thermoregulation, flight, display. I’ll add another: sense. I think flight and display can be ruled out at this point, especially if we accept the Prum & Brush model of feather evolution (the one your figure shows), in which they start as tubular structures.

    Let me make a little argument for a sensory function. First, hairlike structures in animals (and plants too) most commonly have a sensory or defensive purpose. Second, it allows for co-evolution of endothermy and thermoregulatory feathers, while an initial thermoregulatory function requires endothermy first. Third, a sensory function is easier to evolve, since even sparsely distributed, small structures can work as sensory organs, but not as insulation. Fourth, the earliest evidence of hair is from little holes in some therapsid skulls suggested to support vibrissae. In fact, let me suggest that avian feathers and mammalian hair may have begun as vibrissae, limited to a few spots on the head. Many birds that catch insects on the wing have feathers, and mostly unbranched ones at that, that act as vibrissae around the mouth. As do, for example, cats.

    None of these arguments are very good, and we are unlikely ever to have good evidence in favor of any of these theories. But I like my speculation anyway.

    1. I think that you make a pretty good case, but then I was going to make a similar post.

      Another example is human body hair, which is far to sparse for thermoregulation, but does serve to signal us when we are being crawled upon by even quite small arthropods. Not feathers, I admit, but then the earliest stage of your scenario would feature animals with those very undeveloped Stage 1 pinfeathery things anyway.

      For that matter, lizards frequently have hairlike sensillae associated with their scales, and these have been demonstrated act as mechanoreceptors, giving the lizard information on what is touching it.

  13. Feathers are amazing. When the genetic engineers get their act together I want feathers instead of head hair.

    1. Try reading Signs of Life (science fiction) by M John Harrison. The protagonist’s girlfriend does exactly that. It doesn’t end well…

  14. The paper may have come out in 2011 but I didn’t hear of this until now so I’m glad you posted it.

    Maybe one day I will be preserved in amber (now I’m thinking about Fringe).

  15. Reblogged this on My Selfish Gene and commented:
    Dinosaur feathers in amber? This caught my eye the other day and I intended to write about it but the folks here at Why Evolution Is True have done such a great job I defer to their expertise. Enjoy!

  16. It’s particularly impressive to find pigmentation in fossil amber, because if there’s one thing that’s notoriously hard to pin down in dino reconstructions, it’s colour. So many modern animals have their own distinctive colours and patterns that it seemed a shame we’ll never see an iconically coloured dino specimen without some guesswork. And then Planet Dinosaur gave us Sinornithosaurus, a dinosaur whose colour scheme can be extrapolated from its fossilized feather pigments. 🙂

    On dino feathers in particular, I really like Mark Witton’s take on the variability of dinosaur integument (outer covering). It was really an eye-opener for me, as I realized before reading that I’d fallen into a restrictive, one-dimensional mindset of feather evolution in dinosaurs:


    So to find amber fossils from the same time period showing off such variety of feather forms is brilliant! Thank you for posting this, anachronistic as it is. 😉

Leave a Reply