Railing about rails: new paper on flightless birds is grossly misreported and distorted by the popular media

May 13, 2019 • 9:15 am

Note: If you can’t access the paper, judicious inquiry might yield you a copy.


Many readers called my attention to new paper in the Zoological Journal of the Linnaean Society (click on the title screenshot below, and see the pdf here), reporting on a case of convergent evolution— two independent cases of evolution producing organisms that looked the same. (In this case, however, “the same” refers only to the size and shape of two bones; it’s not a broad similarity like convergent evolution in the placental mole and Australia’s “marsupial mole”).

The evolution occurred in rails in the genus Dryolimnas, which evolved flightlessness (and similar bone morphology) in two independent colonizations of Aldabra, a large coral atoll (155.4 km2 or 60.0 sq mi) between Africa and Madagascar. As you may know, birds on remote oceanic islands often evolve flightlessness, most likely because the islands have no predators they need to avoid and flight is expensive.

My beef here is that the popular press grossly misreported what this finding means, claiming that the same species evolved twice independently. And that’s just not true.

Here are the players. First, the island, with its location and then a photo. The red dot is Aldabra, which is also famous for having giant Galapagos-like tortoises that descended from small tortoises that migrated to the island.

And here’s a photo of Aldabra:

Here’s the paper (click on screenshot; reference at bottom); it’s short and not to hard to read:

The situation is this. Aldabra is home to the only flightless rail in the Indian Ocean: the white-throated rail subspecies Dryolimnas cuvieri aldabranus. There’s also a flying subspecies on Madagascar and the islands of Mayotte, D. cuvieri cuvieri. (These are subspecies of the widely distributed D. cuvieri, which inhabits many islands in the region.)  Note that the flightless and “volant” (flying) rails are classified as members of the same species, which puzzles me a bit as they are quite different. 

Here’s the flying rail, D. c. cuvieri, photo taken on Madagascar (from Wikipedia):

Here’s the flightless subspecies on Aldabra. The Seychelles News Agency says that biologists are trying to get it classified as a full species (after all, it has reduced wings and can’t fly).

The Aldabra Rail (Janske van de Crommenacker)

As the paper reported, they’ve found ancient fossil rails at two locations on Aldabra, or rather fossil bones: two humeri (wing bones) and a tarsometatarsus (leg bone). The shape of these bones indicates that the rails were flightless. Here are the two locations (stars) where the fossils were found (compare to island photo above; the “Bassin Cabri” is in the foreground:

The kicker in the story is that geological evidence shows that Aldabra was inundated completely about 120,000 years ago, after the first flightless rails had evolved. That inundation would have wiped out the species as they can’t escape rising waters. Thus the fossil wing and leg bones are from an earlier species that went extinct.

Then, some time in the last 100,000 years, the flying white-throated rail re-colonized the island, and once again gave rise to a flightless subspecies. Comparison of the bones of the living flightless bird with the extinct flightless one shows that they are very similar. Here are the photos from the paper.

This first photo shows, from left to right, the humerus of the living flightless rail on Aldabra, that of the extinct flightless rail (note the similarity of the two bones from flightless groups), and then of a fossil flying white-throated rail from Madagascar, and then a living flying rail. I won’t confuse you with the subspecies names; it suffices to show the size difference in wing bones between flying (right) and flightless rails (left), and, more important for our purposes, the similarity of the two flightless-rail humeri on the left.

The tarsometatarsi (leg bones) of rails. Left to right: fragment of bone from extinct flightless rail, modern flightless rail, fossil of flying rail (from Madagascar), and modern flying rail. The authors note that the two flightless rails (remember, we’re not sure that the fossil rail on Aldabra was flightless, but it’s a good bet given its reduced wing bone) show similar features of the tarsometatarsi that differentiate them from the two flying rails on the right and also suggest flightlessness of the two on the left (see paper for details).

So what we have is the independent evolution of flightless rails on Aldabra—twice. The ancestor was probably in the lineage of the white-throated flying rails found more widely. And in both cases, independent evolution resulted in bones that were similar. Note that all we have here are bones from the extinct species: we have no other information about the extinct species’ appearance, behavior, and so on. All we can conclude, then, is that we have two cases of independent evolution that led to probable flightlessness (in the extinct species) accompanied by similar changes in two bones accompanying the loss of flight. The authors do not conclude much beyond this, which is good.

But the popular press is different: they are claiming or implying that the same species of flightless rail appeared via evolution after it had gone extinct. Here are some headlines (click on them to see the misreporting). It’s a “miracle” says the Daily Mail, “an extinct species came back from the dead”, says CNN and many other venues. I could show a lot more examples, but these should suffice:



US News and World Report:

ITV News:

From Mashable:

The Daily Mail:


Species status in biology is generally judged by reproductive compatibility: if two taxa can coexist without producing fertile offspring, they are different species. Chimps and humans, who live in the same area but do not mate and cannot produce offspring if they do (yes, some scientist inseminated female chimps with human sperm, and nothing happened) are different biological species. Neanderthals and modern H. sapiens were probably the same species, as they did mate where they coexisted and produced fertile offspring (we carry some Neanderthal genes from those ancient matings). We simply know nothing about the reproductive compatibility of the modern flightless rail and its extinct flightless relative on Aldabra. The experiment can’t be done. It’s possible that these two rails had very different physiologies or reproductive systems, and, if we could somehow test their ability to hybridize and produce fertile offspring, we might find that they couldn’t.

So I question whether the modern white-throated rail that can fly and its flightless relative on Aldabra are the same species (remember, they’re classified as subspecies). You simply can’t test their reproductive compatibility as they live in different places, and mating them in zoos is not a good way to test that (many species that can produce fertile offspring in captivity will not do so in the wild). So the journalists have failed badly here: mistaking a case of convergent evolution with “the same species being resurrected.”

Remember, it’s only two bones at issue, and that’s all. What we see once more is science journalists overblowing and distorting a nice paper to make it seem even more exciting. That is not, I emphasize, the authors’ fault: they are very careful in drawing conclusions, and say absolutely nothing about the reappearance of an extinct species.


Hume, J. P. and D. Martill. 2019. Repeated evolution of flightlessness in Dryolimnas rails (Aves: Rallidae) after extinction and recolonization on Aldabra. Zoological Journal of the Linnean Society. zlz018, https://doi.org/10.1093/zoolinnean/zlz018

68 thoughts on “Railing about rails: new paper on flightless birds is grossly misreported and distorted by the popular media

  1. Are the bones all hollow in the same way?

    Perhaps the flightless bones could be of higher density.

    I don’t have access to the paper but I guess they cut the bones open or X-rayed them…. maybe I’ll inquire judiciously…. interesting work – never heard of rails or Aldabra.

  2. Thanks for posting this commentary and analysis. I have an assignment in my Honors Biology class where the students find an item in the general media and then have to backtrack to the primary research paper. They are supposed to assess the accuracy of the media account and/or the degree to which the reporting has oversimplified the conclusions. Most of the time they find that things are at least oversimplified. Unfortunately, this case goes further to misrepresent evolutionary processes. The real question for me (and I realize it is almost impossible to answer) is if this were another case of convergent phenotypes achieved via different underlying genetic processes.

    1. As I indicate below at #5, the causative events for the extinction could be recurrent too. These very low-lying islands (actually, I have to check that … Wikipedia says

      Highest elevation 16 m (52 ft)
      Highest point unnamed dune

      which gels with what I know of the region) are susceptible to over-washing by a number of types of events. The 2004-12-26 tsunami had about 1.5m height on the Tanzania coast, and would have been about the same here. Nearer earthquakes (volcanic Comoros, Mayotte, Martinique and Reunion are nearby) can generate tsunami too, as well as megatsunami from sector collapse. All of these are potentially recurring events.
      Recurring extinctions, followed by recurring re-invasion from a “mainland” population, are almost a laboratory setup.

  3. When different popular-press articles get the same thing wrong they have usually been misled by a badly-written press release.

    This might be the culprit:

    “New research has shown that the last surviving flightless species of bird, a type of rail, in the Indian Ocean had previously gone extinct but rose from the dead thanks to a rare process called ‘iterative evolution’.”

    Blame the University of Portsmouth!

    1. That just goes to show that science reporters rely on puffery and publicity of universities rather than reading the original paper. That’s reprehensible and a dereliction of journalistic duty. No other type of reporting would get away with being this shoddy.

      Carl Zimmer is a welcome exception!

      1. Well, you could argue that this dereliction of journalistic duty is what journalists do all the time.

        They basically recycle press releases and press conferences because nowadays they don’t have the time, people or expertise to do anything more.

      2. Well that wouldn’t surprise me. Reading the actual papers is hard work and also might be beyond the technical competence of a lot of science reporters. Couple that with the pressure to meet deadlines, and they are guaranteed to cut corners.

    2. Further down the article says:

      “This means that one species from Madagascar gave rise to two different species of flightless rail on Aldabra in the space of a few thousand years.”

      They know perfectly well what actually happened; they also know what will sell well.

    3. I thought it was the job of journalists to get their facts straight. I still think they do better with science than with political reporting.

    4. The headline and that first paragraph are highly misleading. The rest of the article, which presumably reflects more accurately the researchers, not so much.

      This is what happens when ‘technical writers’ are let loose on anything scientific/technical. Not only do they sensationalise it, they also distort it by over-simplification, and leave out significant technical details because ‘the public won’t understand’ or ‘the public will be confused by it’. (It’s never the technical writer who can’t understand or gets confused, of course).

      Given that press release, I can’t blame the general press too much for getting it wrong.


  4. You should entitle the article ‘In which I rail about rails!

    Thiat was some very dumb sub-editing by the news outlets! It is pretty obvious that they do not understand the difference between ‘the same’ & ‘similar’…

    But why is it that rails so often end up as island species? thinking of Lord Howe Island swamp & wood hens (galinule = extinct)…
    though I seem to recall other rail island species – why? Why rails?

    Oh – I see Wikipedia mentions that… they are rubbish fliers so presumably once they land on an island they can’t be arsed to fly away… I suppose we might feel like that if we washed up on some tripical paradise!

    “It is paradoxical, since rails appear loath to fly, that the evolution of flightless rails would necessitate high dispersal to isolated islands.[13] Nonetheless, three species of small-massed rails, Gallirallus philippensis, Porphyrio porphyrio, and Porzana tabuensis, exhibit a persistently high ability to disperse long distances among tropic Pacific islands,[13] though only the latter two gave rise to flightless endemic species throughout the Pacific Basin.[14]”

  5. We simply know nothing about the reproductive compatibility of the modern flightless rail and its extinct flightless relative on Aldabra. The experiment can’t be done. It’s possible that these two rails had very different physiologies or reproductive systems,

    The incompatibility can be seemingly trivial – for example, nesting at different times of the year. There’s not a huge amount of climate differentiation through the year there, but the rainfall level does vary a fair amount, which is a big deal in the high 30s degC.
    There’s good diving along the Tz coast between Mtwara and Kilwa Masoko.

    Aldabra is home to the only flightless rail in the Indian Ocean: the white-throated rail subspecies Dryolimnas cuvieri aldabranus. There’s also a flying subspecies on Madagascar and the islands of Mayotte, D. cuvieri cuvieri.

    The species name suggests that it was named by the “Grand Vieux Homme” of French taxonomy, Cuvier (or named for him by one of his students) in the early 1800s. Mayotte (to the SSW), Martinique (on the mid-ocean ridge E of Madagascar) and Reunion (a bit further along the ridge, over a mantle hotspot) are départements of France to this day, so Cuvier getting his paws on specimens in the 1800s is no surprise.

    Mayotte island near the N tip of Madagascar is currently undergoing a bit of an earthquake swarm, concentrated under it’s east flank. It’s not at all clear what is going on (an underwater array of seismographs has been deployed, for retrieval in a few months), but it could be a new flank volcano starting up. Or it could be a “sector collapse” starting up, which might well result in the current flightless rails regretting their parent’s choice to stop flying. That’s the sort of collapse which gives Western Europe communal kittens over the collapse of one of several Canary Islands (which may wipe out Washington and New York too, if it triggers). And not to leave the Pacific out of it, there are seriously unstable sectors on the east coast of Hawaii, which has California in it’s gunsights.

    Come to think of it, the “Boxing Day” tsunami of 2004 certainly swept over Tanzanian coastal islands at about 1.5m height, so that’s another threat to flightless birds on atolls worth considering.

  6. The real phenomenon is just as interesting as the fake press-release-science-“journalism” phenomenon. But there are much better instances outside birds. I’m thinking of the Caribbean anoles studied by Jonathan Losos and the three-spined sticklebacks studied by many folks.

    It would be good to insert a plug for Losos’s book Improbable Destinies, which is all about the causes and limits of convergence. I forget if you’ve ever mentioned it here.

  7. Nice, clear debunking of overexcited press reports. Thank you
    BTW, I like Erik’s biology assignment.

  8. I am astonished at the number of news outlets that picked this up – is that just because I do not scour all the news outlets?

  9. Why do I have a mental image of a miscreant Australian emu jockey being told that he’s going to be run out of town on a rail, looking at the bird and going “huh, cobber?”

  10. I just installed Unpaywall in Chrome. Should this “just work”? Because it isn’t. Next up : shutdown/restart/relogin to Chrome.

    1. I get a gold colored padlock icon that does not lead to the full text.

      I used the test article with Unpaywall – it has a green padlock icon that leads to the fulltext.

      this suggests the Zoological Journal article is unfortunately not covered by Unpaywall.

  11. Very interesting story even without the bad journalism. Might mention – Doris Day died today at the age of 97. She was a great friend of animals.

  12. Well, when even professors of science journalism can be people like the famous vindictive fraud Connie St. Louis, I guess this is what we should expect.

    Science journalism has been a problem for a very, very long time. Most journalists who write about science have no education in the subjects they’re writing about, and the issue has only become worse with the rise of clickbait and identity politics.

    Potholer54 has many great videos going back years about the problems with science journalism (among other things). He worked in the field for decades and really knows his stuff. I highly recommend his channel. He also loves debunking climate change denialists, anti-evolution nitwits, and other promulgators of poppycock.

  13. “Evolved into existence again.”

    It seems to me that the “science writers” are presenting this case as an example of orthogenesis, the mention of which is like fingernails on a chalkboard to evolutionary biologists.

  14. I saw these bad headlines over the weekend and immediately thought “No way.” I didn’t realize until reading this post that the only similarity between the earlier flightless rail and the new one is based solely on the similarity of two bones. So irresponsible!

    I wonder if authors of scientific papers like this have to explicitly head off such misinterpretations? I know that they do this already but it is usually too subtle for these mainstream publishers to pick up. I’m thinking that they need a “non-conclusions” section right after the abstract. Perhaps even make it a bulleted list for easy consumption. Would it help? Perhaps not.

    1. I wonder if authors of scientific papers like this have to explicitly head off such misinterpretations?

      In the paper? Irrelevant, since the journalists won’t read it.

      In the press release? No, the whole point of a press release is to do the opposite, to over-hype the research to try to get it reported!

      1. Yes, I see that the PR was to blame in this case. I suppose we’re stuck with this kind of thing as the people writing the PR are motivated to mislead.

        We could encourage readers of the MSM pieces to read the original paper but that’s not going to work. Even understanding the abstract requires a level of knowledge that most won’t have.

        So sad.

        1. I wouldn’t *expect* journalists to read the original paper, if there’s enough ‘meat’ in the press release to make an article out of.

          Hence in this instance I blame the writer of the press release. Who probably has no scientific training either, *should* have asked the original authors to vet the press release, but probably didn’t.

          I wouldn’t go so far as to call the PR (is that ‘Press Release’ or ‘Public Relations’?) people failed journalists, but – oh, I think I just did.


    2. I saw these bad headlines over the weekend and immediately thought “No way.”

      My brain autocorrected it to “species that looks extremely similar”. Didn’t even realize they were implying that the exact same species evolved there twice since “Nobody could possibly believe that!”


  15. To be honest, I think all four bones in the 2 sets quite resemble each other. It would have been helpful if it were pointed out more precisely in the illustrations where they differ and where they resemble, for the interested layman -as most of us are, I guess.

  16. I’d put this down to the training of journalists. They must be told that to reach your reading audience, you have to tell a good story. Even if that means stretching the truth. Stories must have a beginning, middle, and a big dramatic end. With a little encouragement from editors, they become elastic as hell.

    1. Hey, we should be relieved that they all managed to post a picture of the right species of bird!

      As any tech geek will tell you, the correlation of a newspaper photo with the actual subject of an article is often spectacularly wrong.


    2. For a good story you can change the focus and expand here and there. But I think it is more the problem that a lot of journalists have not read the research and are just copying from each other and changing everything so it sounds more and more dramatic. They have sometimes even changed my quotes.

      1. That’s totally illegitimate in my view. I’m sure someone like Carl Zimmer doesn’t need to distort reality to produce fascinating books and articles.

  17. [sigh]

    I am admittedly the sort of person who hangs around sites like these, but I have *1* CEGEP level course in biology (so 1 roughly university level) and I know the distortion for what it is. So one does not need much background to know this!!

  18. Thanks, Jerry. I had noted one of those stories the other day, but had not had time to read it. Now I don’t have to.

  19. Excellent post. Too bad the press appears to have distorted this. I was also perplexed when I read the news outlets: how could a species evolve after it’s gone extinct!

    1. No, since DNA easily falls apart when the circumstances are not perfect and these bones are very old. It is already a real struggle to get any DNA from the Dodo that died out a lot more recently.

  20. Another reason to take note of such papers (beyond how popular press coverage tends to hype and misunderstand) is certain types of papers get riffed on in the antievolution literature, and the Hume paper is no exception. The Canadian ID groupie Denyse O’Leary has already bounced off Uncommon Descent’s spin on this, themselves running secondarily off the ScienceDaily coverage.
    O’Leary tweeted on May 10 “Bird ‘goes extinct’ twice Note that loss of the ability to fly is treated in this story as a form of evolution, as if the loss resulted in greater complexity rather than less. As if it wasn’t fatal when the island was inundated.” (The extensive science lit on bird evolution, including the paleogenomic reconstructions of bird skulls, limbs, feather systems, has escaped her attention, of course, possibly because she keeps her eye fixed on the Uncommon Descent postings.)

    1. If one adds up all the Genesis numbers as to how long the rain lasted (40 days), how long they spent bobbing around (150 days), and how long the drying up took (150 plus some more 40 days and 40 days episodes, the Bible is fairly vague on these details), you get a bit over a year for the whole process. Literalist creationists stick by that, then snag on trying to shoehorn in some of the data (which is never going to fit in the cartoon borders, so the task of observers is to spot all the stuff they’re leaving out, which is almost all of it).

    2. A long time.
      A third major marine incursion is represented by the Aldabra Limestone, and all available geological and palaeontological evidence supports a complete inundation event during this time. Deposited between 118,000 and 136,000 YBP ± 9,000, it covers most of the outer rim of the atoll and rises to around 8 m above PSL, forming some low cliffs [13]; the sea level had risen to ~10 m above PSL


  21. this is why i hate yellow journalism – distorting the facts for the sake of making a sensation!

  22. The Continental Telegraph got it right.

  23. So reproductive compatibility which differentiates species depends more on whether they do and not whether they can? So are polar and brown bears different species? But that can change in the future if their ranges overlap more?

  24. Note that the flightless and “volant” (flying) rails are classified as members of the same species, which puzzles me a bit as they are quite different.

    Given Mayr’s biological species concept, I’m not sure why you find it surprising. I assume the reproduction test hasn’t been done, so shouldn’t one remain agnostic as to whether they are separate species?


  25. Why is this so different from three-spined sticklebacks repeatedly evolving into freshwater forms (with big changes in morphology and behavior) from a marine ancestor?

  26. If someone wanted to go hunting for evidence for epigenetic changes, seems like sub-species like flightless birds, blind cave salamanders and the like would be prime places to look.

  27. If someone wanted to find evidence for epigenetic changes, sub-species like flightless birds on islands, blind salamanders in caves, and the like would seem to offer fertile grounds.

      1. I know you did not mean it negatively towards us. I meant to say that in agreement. I am glad you noticed the mismatch between the media craze and the actual paper. Since my research focusses on extinctions, the last thing I want is to let people think extinctions can be reversed.

    1. Excellent, concise description. Julian Hume briefly describes the smaller bones of the flightless rail.

      I am wondering if the bone density is the same? Aren’t bird bones “hollow” to allow flight, and could the flightless rail have solid bone?


      1. Good suggestion. I don’t know if the bone density is the same. Some things evolve faster than others. For instance, egg size takes a lot longer to change then the whole body size of a bird. That sometimes leads to small birds with huge eggs, like the Kiwi. Perhaps bone density changes take longer too. But the legs were a bit “sturdier” in the flightless rails.

        1. “… that sometimes leads to small birds with huge eggs, like the Kiwi.“

          Very interesting!

Leave a Reply