A Flock of Genomes Reveals the Toothy Ancestry of Birds

A tidal wave of genomic information seems to wash up on shore each year.   In 2014, an especially large wave – in those days long ago! – came ashore in the form of 45 entire genome sequences of birds.   The relevance of these genomes and their hundreds of billions of letters of code representing all the major groups of birds were discussed in a series of scientific papers in the journal Science.  Prior to 2014 only three bird genomes had been completed which were the chicken, turkey and finch.

To give you some perspective on the sheer magnitude of this data set, in 2000 the first genome of a mammal was reported.  That genome was the human genome and it took 10 years and over a billion dollars to complete.  The chicken genome took years to finish and cost millions of dollars.  In 2014 nearly every day a genome of an organism was published and the pace of sequencing has only accelerated since them.  Many of these genomes are much larger and more complex than the human genome.  DNA sequencers have become so fast and sophisticated that they can now read the entire genetic code of an organism in a few days and in some cases a single day. The challenge is no longer obtaining DNA sequences but rather processing the terabytes of data that a single DNA sequencing machine can produce in a few hours.

When I produced my first DNA sequences as a graduate student I was thrilled to be able to generate a portion of a single gene for a few organisms in a few weeks.  Even in 2000 when I set up my independent lab I never dreamed it would be possible to sequence whole genomes in a matter of hours.  Today, billions of nucleotides–the ATCGs of the code–can be obtained for a few thousand dollars.

What can be learned from a flock of 45 bird genomes? 

The published papers in Science (see references) report on the genome structure, relationships among birds.  They also delve into comparisons of specific genes that are involved in unique bird characters such as bird songs, feathers, and teeth.

Photograph of Sulcavis geeorum skull, a fossil bird from the Early Cretaceous (120 million-years-ago) of Liaoning Province, China with scale bar in millimeters. Credit: Stephanie Abramowicz Read more at: http://phys.org/news/2013-01-fossil-bird-teeth-specialized-tough.html#jCp
Photograph of Sulcavis geeorum skull, a fossil bird from the Early Cretaceous (120 million-years-ago) of Liaoning Province, China with scale bar in millimeters. Credit: Stephanie Abramowicz
Read more at: http://phys.org/news/2013-01-fossil-bird-teeth-specialized-tough.html#jCp

Teeth?  But, birds lack true teeth don’t they?  It is true, no modern bird has true teeth, but nonetheless there is an entire paper devoted to looking at the genetics of teeth in modern  birds.  Birds do not produce enamel or dentin, the hard materials that make up teeth in mammals and some reptiles.   However, we can readily observe that there are very ancient bird fossils from the time of the dinosaurs that clearly had true teeth.   From that we can infer that although no birds have teeth today their ancestors did.

Until recently it might be possible to object to this inference of ancestral connections by claiming that although fossil birds had teeth that there is no connection between those fossil birds and toothless birds today.  Some creationists might counter that God could have created toothed birds as separate kinds of birds but then all of them perished in a global flood and only non-toothed birds survived to become the birds that we have today.

Bird genomes reveal the toothy history of all birds

The case of a toothed ancestry of all living birds got a big boost from the bird genome sequencing project.  Our ability to peer into the genomes of any organism on earth is rapidly becoming a reality.  In those genomes we find the history of species written in their DNA.  This history is not too dissimilar to reading the fossil record to reconstruct that past.   Genes that were used in the past by a species are frequently found discarded in the genome where they decay over time.  But that genomic decay process takes a long time and so when we read the genome we not only read all the genes that are actively doing things today to make a bird a bird we are also reading all of its code that it has used in the past but are no longer using.

This process of discovery of discarded genes would be similar to looking at the hard drive of a 10-year-old computer.  After a few years you start to use new programs and others go into disuse. Even when your “erase” a program you really only erase the information about where the files are and allow that space to be written over again with new code. Many files from those old programs are still there but you have no way of using them because they are just broken pieces.

So what do these new genomes reveal about teeth and birds? 

They reveal what we already knew from a few genomes but now we know from all major groups of birds:  All birds contain the genes for making true teeth in their genomes!

From many studies of mammalian teeth we know there are at least six genes involved in constructing the dentin and enamel of teeth.  All six of those genes are found in modern birds.

But if they have all these genes for making teeth why don’t they have teeth?  It is very simple, although they all have the genes, those genes all have serious errors in their code or in some cases large pieces of the genes are missing.   These are programs for making enamel and dentin that are not used by the birds because they are too messed up to work.

With 48 entire genomes of birds scientists now have the data to show that the main gene responsible for enamel has the same function-losing mutation in every bird. This is very strong evidence that the ancestral genome, and by extension the ancestral bird, that gave rise to all modern birds experienced this mutation and then passed that same problematic gene on to all its descendants.  Without the ability to produce enamel the other genes for tooth formation were not needed. Without any natural selection to maintain the integrity of those gene sequences, they have accumulated hundreds of mutations rendering all of these genes non-functional in all birds.

So the genetic evidence strongly supports what was already inferred from the fossil record – that ancestral birds were able to produce teeth but then lost that ability before the diversified into the species that we see today.   Birds aren’t the only animals that have tooth-making genes but don’t use them. Turtles, anteaters, and baleen whales also do not have teeth but, like birds, they also have broken versions of teeth-making genes.

Addendum:  Implications for young earth creationist’ view of the origin of birds

Let’s explore briefly some of the implications of these genomes on the view of young earth creationists (YECs) view of origins.  YECs don’t accept common descent of all birds from a common ancestor much less from an ancient theropod dinosaur.  They accept that birds have speciated within biblical “kinds” or “baramins.”  These are groups of species, generally at the taxonomic level of a family that have a common ancestor but these ancestors were specially created separately from one another very recently.  For example, all penguins may be a kind and thus God may have created one penguin species originally that was preserved on Noah’s ark and then diversified (ie. evolved) into all of the species of penguins that exist today and are in the fossil record.  For more see my article on bird diversity on the ark:  10 Invoking Super-Speed Evolution: How to Squeeze 10,000 Bird Species on Noah’s Ark.

If chicken, eagles, penguins, ostriches, finches, etc.. are each members of separately created bird kinds then each was created with their special attributes as we see them today.  None of these bird groups have teeth but they all possess broken genes for making teeth.

Why then do all these bird groups have genomes with genes for making teeth in them?   I don’t think that any YEC would a priori predict that these birds would have genes for making teeth.   Such a prediction is made from the assumption of common ancestry.  Sometimes creationists appeal to common design for similarities between organisms they believe don’t share a common ancestor.  Most often this claim is made in reference to why humans and chimpanzees share so many features and have DNA sequences that are so similar.

Bird beaks are very diverse. They are adapted for many different tasks.
Bird beaks are very diverse. They are adapted for many different tasks yet none retain or use true teeth despite the genetic “roots” to do so.

Does common design explain these toothy gene in birds?  No! Common design might have some appeal when their common design makes sense but why would God need to make birds with broken genes for making teeth.  Aren’t birds “perfectly” designed for this world with beaks?   Common design of non-functional parts makes no sense.  Our “design intuition,” which intelligent design advocates such as Douglas Axe espouse, doesn’t tell us that modern birds with no teeth should all have broken genes for making teeth.  This is not common sense for the creationists. However, it is very much common sense from an evolutionary biology perspective.

There is one other possibility.  A YEC could appeal to the fossil record and say since there were some birds that had teeth maybe all birds were created with teeth.  Maybe birds lost their teeth because of Adam’s sin. Maybe beaks are the result of a degraded and less-perfect world.  I think this is a bit silly. Beaks are amazing and clearly are adapted to serve birds amazingly well.   How could a YEC possibly want to say that God didn’t create beaks but only gave them the ability to make beaks later because he knew they would lose their ability to make teeth.

Birds aren’t the only example of these problems for creationists.   See these blog articles by two other biologists,  Birds, turtles and other toothless vertebrates have remnants of tooth genes,  and Pseudogenes testify to the evolutionary ancestry of animals each of which also discuss the loss of teeth at a broader scale showing that other groups of mammals that don’t have teeth also have the genes in their genomes.


Featured image by: Julie McMahon / University of Illinois


Merideth et al.  Evidence for a single loss of mineralized teeth in the common avian ancestor

Science 12 December 2014: 346 (6215), 1254390 [DOI:10.1126/science.1254390]  Abstract,

Full Text (PDF), Supplementary Materials

Abstract:  Edentulism, the absence of teeth, has evolved convergently among vertebrates, including birds, turtles, and several lineages of mammals. Instead of teeth, modern birds (Neornithes) use a horny beak (rhamphotheca) and a muscular gizzard to acquire and process food. We performed comparative genomic analyses representing lineages of nearly all extant bird orders and recovered shared, inactivating mutations within genes expressed in both the enamel and dentin of teeth of other vertebrate species, indicating that the common ancestor of modern birds lacked mineralized teeth. We estimate that tooth loss, or at least the loss of enamel caps that provide the outer layer of mineralized teeth, occurred about 116 million years ago.

This article is an updated and revised version of an article that was first published in 2014.

26 thoughts on “A Flock of Genomes Reveals the Toothy Ancestry of Birds

  1. Great post, is it just a coicidence that you published this a couple of weeks after my blog post that touched on this subject, or did I perhaps inspire you to write about this?
    I’ll have to update my original post with a link to this post and Meredith et al. 2014. Unfortunately when I was writing my blog post I was kind of rushing it so didn’t have time to do a more complete search of the literature that would have no doubt uncovered it, so I had to rely on Meredith et al. 2009 and 2013 instead.


    1. I honestly can’t remember for sure. Your 2014 post was from Nov 8 while the original version of this post was from December of 2014. However, I remember reading the Science article and writing this post. I think I wasn’t aware of your blog at that point and so we independently wrote about nearly the same thing making the same point. I think I would have referred to your post if I had know about it earlier. BTW, I just told my genomics class about your “third eye” paper as an example of searching genomes for ancient genes.


      1. I think you might have confused me with Christopher Emerling! But if you originally wrote this post at the end of 2014 I obviously couldn’t have influenced it as I wrote mine in January 2018. I’m glad you linked to Christopher’s blog through, I wasn’t aware of it before and I’ve been binge-reading it over the last couple of days.


        1. Ah, yes, I did get you confused. I did read your post just recently. My post on the birds was written in 2014 and I was looking at some of my older posts for something to update an publish since I have been so busy with other things and can’t write new stuff. Your post was great. Maybe I was subconsciously influenced to revise this one because it reminded of what a great story psuedogenes are. I’ve just added a link to your article to the blog post. BTW, I’ve been anxiously awaiting your reviews of the next two chapters of Jeanson’s book. JD


          1. Thanks! I know, it really bugs me that I haven’t been able to finish off Jeanson’s book yet. I was able to publish the blog posts on the first 2/3rds of the book in unusually quick succession because I had a fairly light workload, but since November I’ve much busier working on my thesis and teaching. I don’t have the mental energy to trawl through Jeanson’s claims about an entirely different field of biology at the moment!
            It helps that there wasn’t really much of substance in those first chapters of Jeanson’s book too. When tackling the final ones I want to make sure I do them justice by being as thorough as possible.

            Liked by 1 person

    1. I think that is what they would have said 20 years ago not knowing that there would fossil birds with teeth found and now the genes for making teeth in modern birds. That should be their prediction but since they know their prediction isn’t confirmed they will modify their prediction to be something like, all birds have a common design in the creators mind and thus used comment genetic elements. That really doesn’t help understand seemingly wasted genetic pieces though they could just pull out the catch-all “we just don’t know what those genes are doing today and someday we will find out that what appears to be tooth genes are doing some other function” This their generic response to most pseudogenes.

      Liked by 1 person

      1. I’ve heard the “It’s not a broken gene for X; it’s a functioning gene for something we’ll discover later” claim before. But how do we identify them as broken versions of existing genes to begin with? Is it that their locations and sequences are so similar to the functioning versions that calling them pseudogenes is the most reasonable explanation? Are there genes that might be pseudogenes but are so damaged/degraded that we can’t make a solid inference?


        1. Hi Mel, essentially yes: with pseudogenes like these (unitary pseudogenes) they are in the same location as the functional genes and have relatively high similarity to the functional genes in other species. In this case, they used a crocodilian for reference (what we believe to be the closest living relatives to birds) and found quite a bit of similarity despite the fact that they were pretty trashed. Furthermore, when you estimate their evolutionary history using molecular phylogenetic methods (basically using sequence similarity and assuming models of mutation) they come out exactly where you’d expect them to based on their inferred evolutionary history. For instance, I don’t think Meredith et al. did this, but what we’d expect that the bird tooth pseudogenes would come out closest to crocodilians, just like functional genes do.

          Liked by 1 person

          1. If we look at the phylogeny of ENAM gene sequences, we clearly observe two clades: Bird/Crocodile and mammals/marsupials. If we would not look any further and only include the common sequences, Darwinains would be right. However, if we look at gene structure, we observe functional blocks that form different and unexpected clades. In crocodile, lizard, and frog ENAM exon 3 is absent. Exon 8b found in the lizard and crocodile is also present in the marsupial ENAM but is absent in the other mammalian species. This shows again that evolution is frontloaded and modulated and the proces Darwinians pretend it is. Evolution is like creating with lego blocks. Also show me whether selection played a role. I am pretty sure it did not. Evolution is not a Darwinian process. Natural selection is not important. It is all frontloaded and drifting.


              1. Hi Peer:

                First, thank you for attaching the article!

                “If we look at the phylogeny of ENAM gene sequences, we clearly observe two clades: Bird/Crocodile and mammals/marsupials.”

                Not sure what you’re getting at here, so perhaps you can elaborate? It’s not just birds + crocodiles but birds + crocodiles + lizards. Then at the base of the tree there are frogs. If you looked at fishes, you would almost certainly get those branching even earlier in the tree.

                “However, if we look at gene structure, we observe functional blocks that form different and unexpected clades. In crocodile, lizard, and frog ENAM exon 3 is absent. Exon 8b found in the lizard and crocodile is also present in the marsupial ENAM but is absent in the other mammalian species. This shows again that evolution is frontloaded and modulated and the proces Darwinians pretend it is.”

                Also uncertain what you mean by this. Yes, we believe that exons can be gained and lost in genes, but in my experience this is comparatively rare. Some of the tooth genes seem to do things like this a bit more frequently (AMBN and ODAPH are others), and it might depend on how those exons are related to tooth formation. ENAM is an enamel matrix protein that provides a sort of scaffold for crystal seeding and growth, but it is then quickly digested by enzymes, so it would be interesting to learn if those exons affect this process much, if at all.


                1. Christopher, I also was puzzled by “peer” citing this paper. To me, it seems to argue for the existence of exon 8b in the common ancestor of sauropsids and marsupials, but its loss in the placental mammal clade, completely consistent with the conventional scientific account, and indeed an interesting story of DNA evolution.

                  Not my area. Have I understood correctly?


                    1. It sounds like peer is trying to say that this somehow proves that all of the exons had to be extant before whatever divergence they accept, and that none of them could have arisen by means of a beneficial mutation. In one line (birds, lizards, crocs), exon 3 disappears. In the line that contains placentals, exon 8b disappears. But in none of the cases do we see a new exon type develop out of nothing. This probably assumes some prior ancestor that had all of them to work with, but I’m not sure because it seems to me that peer’s argument is incomplete; if I’m understanding this correctly, nowhere do they show that exon 3 can’t be novel to the line leading to mammals.


            1. Peer, I’ve been too busy to follow the comments section carefully. I apologize for not being able to spend the time especially given your effort to explain your ideas. But, I am struck by one thing looking now through some of your comments. You keep saying that NS is not important. I’m not sure why you are so against God using such a mechanism when you want to simply put another mechanism in its place. What is to be gained really from “frontloading and drifting” that natural selection could also not play a part or a major role? Really, your proposal read to me as deism and leave God less of a participant in creating in the present than does the OEC or EC advocate. What I mean by that frontloading and drifting feels like God simply providing all the instructions at the beginning and then either species are generated as a result of playing out a preprogrammed set of instructions or the basic framework for how to adapt was frontloaded and now organisms are randomly changing over time (genetic drift).

              Let’s put this to an example. Would you say that the original canine (or whatever your inclusive kind group is) was created with the specific instructions for making the exact features of my dog Lassie? Or is my dog Lassie simply one of billions of possible products of the original kind but my exact dog was not planned specifically? Put another way; Did God create the original kinds foreknowing (predestining) every single individual that would live and their characteristics or did he create the kinds with the ability to change but not knowing exactly what changes would occur though they were given their limits (kind parameters you could say)? It’s a predestination vs open theism question. Another nuance on the predestination angle would be: Did God create the original canine and then 1) step back and watch his plan unfold with all the information for how it would unfold all contained in the original animal or 2) guide/ be involved in some way with, every stop of the process along the way to create each individual organism that has lived. Your mechanism sounds more akin to an open theism approach. A creator supernatural frontloaded variation and mechanism for adaptation but then once created individuals drifted and, without natural selection apparently, filled various niches but the exact forms that filled those niches are not necessarily exactly programmed but chanced into their present circumstances.

              It’s a question of God’s sovereignty over creation and I’m not sure how drift fits into a sovereign plan. Now, it could be you only use drift in the sense that we describe drift as random because we can’t see the control behind it but God doesn’t play with dice. If that is the case then why no us natural selection as a tool? I think it stems from the fact you are committed to God not being able to create anything new after the 6 days of creation. You want God to be a passive participant in the unfolding of creation rather than an active participant. If he is active and his tools are natural selection and mutation then it would be difficult to deny that new characteristics can’t be formed over time.
              At the end of the day, you seem to be trying to find a “natural” mechanism to fill a void that doesn’t really exist. You need species to form more quickly to possibly fit into a short chronology of the Earth and to maintain a particular reading of scripture. Natural selection doesn’t fit that bill and you would be right, it doesn’t have the power to effect that kind of change that rapidly and it requires mutations as a source of variation (as an aside, I don’t understand why you can’t incorporate NS into your VIGES and have the VIGES be the source of variation that NS acts upon). So its frontloading because you can’t have “random” mutations be the source of variation though paradoxically you are willing to use random genetic drift to create divergences of species.

              Lastly, I get the strong impression you haven’t collected population data over many generations and calculated selection coefficients nor have you used real-world wild population data to measure rates of genetic drift of alleles. Genetic drift does not how the power you are ascribing to it under most real world conditions and natural selection though loci have lower selection coefficients than most people realize is very powerful over time.

              I’m going to copy this response to one or two other posts comments sections because I think it might elicit some feedback from others.


  2. I’m curious if bird ancestry can be mapped by tracing the specific ways these genes are corrupted. Is that something that is studied?


  3. I think you’ve anticipated almost all of the YEC responses, and they could almost hold together according to a just-so story:

    Toothy birds were another kind that simply died out. They were similar enough to birds that we would expect to see similar genomes according to similar design. These “broken tooth genes” you speak of could be explained one of two main ways: 1) parts of them could have been necessary in order to keep the genome coherent in some way we don’t know yet, or they could function in some other way besides tooth production. 2) Since we know that genomes can only lose information over time due to mutation [at least, according to YEC lore], it’s likely that all the individual baramins of birds had teeth before the Flood but didn’t need them afterward; so bird teeth disappeared in every bird kind as a result of mutations that didn’t reduce fitness.

    Liked by 1 person

    1. Not that you’re advocating for YECs but nonetheless…

      If they countered with point #2, it raises the question of why some (all?) of the tooth genes in birds belonging to different baramins have identical mutations. For instance, basically all birds that have had the ENAM (enamelin) gene examined have an eight base pair and 13 base pair deletion in the exact same location, both of which are incredibly rare. This would seem to run counter to the YEC claim that convergent evolution should be expected to be exceedingly rare.

      Liked by 1 person

  4. Could the creationist respond by saying the broken genes are in-built design for future diversification? That is, rather than having it be from the past, could they claim its part of God’s design for the bird “kinds” to have teeth in the future?

    Liked by 1 person

    1. Ooohh. I like it. I should have thought of that since I’ve written about the YEC pre-loaded genetics for the future before. Yeah, the birds we have today are still adapting and in the future each “kind” will need teeth. Todd Wood proposed that looking for evidence of hidden genetic traits would be evidence for creation. It seems he need not have to look very far:-)

      Liked by 2 people

      1. As said, people who believe the universe was created for&with a purpose are creationists. Instead of being condescending towards fellow creationsists, you better team up all together to fight the Darwinian nonscience. There are DNA sequences in the non-transcribed intergenic regions, my dear friends, which fold into functional protein structures (alpha helices and beta sheeths). That is not what we would expect when we are dealing with random sequences, isn-t it. What it means is that the genome was indeed frontloaded (preloaded) and novel genes are fairly easy to produce (through acquisition of regulatory elemenst, which are all over the genome in the form of transposable promoters and enhancers). As said before, our current understanding of the genome is islands of functionality in a sea of regulatory elements. It was designed to vary, adapt and bring forth new species. And all without the need for novel information. It explains why we observe the patterns everywhere and independent of common descent. The genome is prepared to change and adapt. It is a masterpiece of design.


  5. Was losing the ability to make dentine neutral, or advantageous? Do we know? What does that tell us about the lifestyle of species at the root of the “toothless bird” clade? And does that include any bird groups other than the groups that survived the K/Pg event?

    Liked by 1 person

Comments are closed.

Up ↑

%d bloggers like this: