Where the birds are

May 24, 2014 • 6:06 am

by Greg Mayer

[JAC note: read this! It’s good biology!]

One of the key sources of evidence for evolution is biogeography– the geographic distribution of plants and animals. The great 20th-century biogeographer P.J. Darlington famously said that biogeography showed Darwin evolution, and Jerry has long said that the facts of biogeography are among the most  compelling evidences of descent with modification. Carl Zimmer in the New York Times has called attention to a paper published yesterday in Science by Kieren Mitchell of the University of Adelaide and colleagues which addresses an old problem in biogeography with new data. Another paper, in press in Molecular Biology and Evolution, by Allan Baker of the Royal Ontario Museum and colleagues, addresses the same problem– and specific example– in a similar way.

Artist’s reconstruction of an extinct elephant bird (New York Times).

The problem is that of disjunction— related forms being found in widely separated places. The problem posed by disjunction is this: how did the related forms come to be in the disparate locations.? For land animals, a number of answers have been suggested over the years, all of which seem to be true in some cases:

The intervening space was once inhabited, but the creatures have since become extinct there. This is a well-known phenomenon: for example, species expand their range during climatically favorable times, and then contract during climatic deterioration, leaving relict disjuncts behind in locations that remain suitable. The wood frog (Rana sylvatica), a North American species found in colder areas and high latitudes, has disjunct populations in the Rockies, left behind as the general range shifted north as the climate warmed in interglacial periods (which for the wood frogs was a climatic deterioration).

Disjunction by extinction of intervening populations, while applicable to terrestrial animals separated by land, cannot account for disjunctions across water. This led to a once very popular explanation:

A land bridge connected the disjunct areas, allowing organisms to spread between them, but the bridge has now foundered and sunk beneath the sea. Speculative biogeographers once criss-crossed the oceans with hypothetical land bridges, but most of these land bridges are now seen as geologically improbable and biogeographically unnecessary. Nonetheless, land bridges definitely have existed and subsequently disappeared, and organisms crossed them when they were above water. Mammoths, for example, crossed between Asia and America during times of lower sea level when the Bering Strait became the Bering Land Bridge.

If there never was a land connection, then how did the organisms get from one place to the other?

The organisms dispersed across the water– flying, floating, rafting, swimming, etc.– by what Darwin called “occasional means of transport”. For land masses never connected, this is the only way a disjunction could have arisen, and numerous instances of over water dispersal have been observed, most frequently by flying, but even the rarer forms (e.g., lizards moving from island to island in the West Indies, an Aldabra tortoise coming ashore in Tanzania) are occasionally directly observed.

And finally, there’s a possibility that became popular after the establishment of plate tectonics in geology:

The disjunction was caused by the separation of an initially continuous landmass into two or more parts, so that the geographic distribution of organisms dispersed throughout the landmass became separated as well. In other words, the organisms didn’t move—their continents did. Once geology demonstrated the reality of continental drift, it became clear that a disjunction could occur not by the animals moving around (although they would have to have initially moved around throughout the contiguous primeval landmass), but by pieces of the earth moving around. So, to use a fossil example, the mammal-like reptile Cynognathus of the Triassic was found in both Africa and South America. This stout, meter-long animal was not a good candidate for trans-oceanic travel; but when it was realized that at that time the two continents were joined, the problem disappeared: continental drift accounted for the apparent disjunction.

The problem studied by both Mitchell and Baker is that of the distribution and relationships of the ratites—the group of large flightless birds that includes ostriches, rheas, cassowaries, kiwis, emus, and the very recently extinct moas and elephant birds. All of these birds are primarily found on the southern continents that were once joined together into the super-continent of Gondwana. Here is their present distribution:

Ratite distribution (Bronx Zoo).

Ratites have long been known to be related to one another and to the tinamous, a group of poorly-flying birds from South America. How the various flightless forms got their present distributions has long been a puzzle. Overwater dispersal seemed out (they’re flightless, after all), so perhaps they were once spread across the connecting northern continents, but have since gone extinct there. With the emergence of plate tectonics, a favored hypothesis was that the ancestral ratite was widespread on Gondwana, and, as the super-continent split up during the Cretaceous, the flightless birds on their respective chunks of land each went their own way and evolved into the modern forms. This view was championed by Joel Cracraft, now of the American Museum.

This view, however has come under criticism, with various authors suggesting that the birds actually had dispersed over water (see, for example, this precis of a talk by Mike Dickison from 2008). Both of the new papers, by Mitchell and Baker, support the over-water dispersal view. Baker, using ancient DNA techniques to study the DNA of the extinct New Zealand moas, confirms earlier work by him and his collaborators showing that the closest relatives of moas are the flighted tinamous, thus showing that flightlessness has evolved more than once within the ratites. Mitchell, also using ancient DNA techniques, shows that the Madagascan elephant birds’ closest relatives are the kiwis of New Zealand, and furthermore estimates the time of divergence at 50 million years ago—long after Gondwana had broken up to the point that wide seas separated Madagascar from New Zealand, thus implying that the common ancestor of the two was flighted.

Both research groups conclude that most or all of the distributional pattern of the ratites is due to over water dispersal by flying ancestors, occurring long after Gondwana had broken up into its modern parts, and that flightlessness evolved convergently within the ratites. According to the Times, Cracraft is not quite convinced yet, but thinks further work will soon reveal a fuller story of ratite evolutionary history.


Baker, A.J., Haddrath, O., McPherson, J.D., and A. Cloutier.2014. Genomic support for a moa-tinamou clade and adaptive morphological convergence in flightless ratites. Molecular Biology and Evolution in press.

Cracraft, J. 1973. Phylogeny and evolution of the ratite birds. Ibis 116:494-521.

Mitchell, K.J., Llamas, B. Soubrier, J, Rawlence, N.J., Worthy, T.H., Wood, J., Lee, M.S.Y., and A. Cooper. 2014. Ancient DNA reveals elephant birds and kiwi are sister taxa and clarifies ratite bird evolution. Science 344:898-900.

40 thoughts on “Where the birds are

  1. Was the title of this article really an allusion to the Connie Francis hit (in the movie of the same name) ? 😉

  2. Well done, Greg! Succinct and to the point. JAC is correct – this is good. It does point up one aspect perhaps not readily apparent, that of what fossils will be found on Antarctica when Global Climate Change melts all the ice. That is THE only upside I can think of for that process.

    1. Thanks, Jim. You’re absolutely right about the importance of Antarctic fossils (which, as you also note, are sadly becoming more accessible). There have already been important finds of marsupials and reptiles there, and there will certainly be more to come.


      1. is it possible to measure the sea level, relative to today, in the period when Antarctic fossils were living creatures?

        1. Yes, there are eustatic sea level curves for throughout the Phanerozoic (ca. 550 million years back). But there has also been much vertical (as well as horizontal) shifting of the crust since then, so it is necessary to consider these shifts, as well as sea level change, in coming up with a a paleomap.


  3. Thanks for the post. I didn’t have enough room in my piece to go into much detail about why Cracraft has reservations. While some of the most interesting clades (elephant bird + kiwi ) have strong support, some of the deeper nodes are not so strong, and so it’s possible that some relationships will move around a bit. But there’s lots of genome sequencing going on now, which promises to tighten those deeper relationships up nicely. Then we can look anew at how the phylogeny and geology line up (or not).

  4. Well fancy that, biology never fails to amaze. I look forward to the next instalment of this research.

  5. The Gondwana breakup hypothesis was so simple and elegant. I guess it had to be wrong. Evolution is usually pretty messy with all sorts complexities entering the picture that could have been left out of the story, if it had been planned properly.

    “tinamous, a group of poorly-flying birds from South America”

    They actually make it well into N Am. (Mexico) and I believe occur not terribly far S of the Texas border. They are very poor flyers, at least the species that occur in wooded habitats are (chickens may be better at it). I’ve never seen one fly at all. Mostly they seem to skulk in spiny thickets. I hear the S Am. grassland species are better at getting off the ground though.

    1. I guess the tinamous are very different from their common ancestor with moas. I can’t imagine something like a tinamou crossing an ocean by flight. I have seen them fly, but their flight looks like a pheasant’s, bursting into the air a bit and then immediately gliding down.

      1. I agree that the ancestor can’t have been much like a tinamou either. Maybe the immediate common ancestor was flightless and the weak “pheasant flight” of tinamous is derived? It’s conceivable the flying ancestor that dispersed across the southern hemisphere was back another step or 47. Pure speculation, of course.

        I wonder if the ancestry of the ratites is parallel to what’s happened with the rails: another group with apparently poor flying ability but which has dispersed widely anyway — and which then frequently generates flightless species after colonizing some isolated oceanic island.

        1. Ignore most of what I just said. The tinamou-moa common ancestor must have been flying, of course, since those two birds are on opposite sides of the Pacific.

          1. At the relevant time South America and Australia were nearly or quite connected to Antarctica, so the major barrier was the Tasman Sea rather than the actual Pacific basin. Unfortunately the Australian fossil record is crappy, but we can predict that members of the tinamou/moa and elephant-bird/kiwi lineages will eventually be found there (i.e. here, from my perspective).

    2. I wouldn’t say that the Gondwana break up is irrelevant to the new hypothesis. Surely the fact that Gondwana components were once in close proximity is still important.

      1. Sorry. I was trying to be brief and as a result wasn’t clear. I should have said the “Gondwana break up only” hypothesis. Now we have a messier hypothesis involving rafting, dispersal and weird long distance evolutionary relationships (Argentina-New Zealand?)

  6. The big problem is that they used mitochondrial DNA, which has been demonstrated to get results like these absolutely wrong.

  7. What I find extraordinary about the rattites is how, even after 50m years along with geographical separation , it is still obvious that they ( or at least a fair number of them ) are related to each other.

    1. Since it is now suggested that different ratite species were each evolving to to be ‘Big Bird’, perhaps it is a case of convergent evolution. That is, their similar appearance is how this clade tends to evolve large flightless forms.

  8. The Monkey’s Voyage has now moved up a few notches on my list of books to read.
    It seems to me a nice follow up on this would be to find 40-50 myo fossils of the transitions in the relevant locations. I know they find fossil finches in the Galapagos so I would think finding fossil birds on these much larger islands would be a possibility.
    As a long time ID-watcher I couldn’t help but to remember Casey Luskins long piece on biogeography where he discussed ratites at length as a problem for evolution. I have a hunch as to how he’ll respond to this.

  9. I took a course in zoogeography in 1963. We used Darlington’s 1957 text. It was set in a world with immobile continents. A large part treated distribution of freshwater fishes. Darlington’s hypotheses of dispersal and replacement was shown to be untenable by a study of Central American fishes by George S. Myers in 1966. The book is still useful if you want to know ranges of various animal groups.

    1. Same here Jim, that book was used in a course I took, probably around 1966. Darlington was a fine field biologist and especially knew his stuff with regard to distribution of insects.. I corresponded with him a few times as a student, particularly concerning his ideas of post-glacial “suture zones” in North America, and was struck by his generous responses. Darlington had the misfortune at the end of his career to publish several comprehensive works assuming stationary continents well after the oceanographers had pretty well put the nails in the coffin. I own a copy of Biogeography of the Southern End of the World [published 1968, I believe], which was essentially special pleading for organisms with “Gondwanan” distribution arising from waves of settlement, migration, and extinction and replacement in the original homeland…

      About the same time I read Willi Hennig [father of cladistics, at least in zoology] on the complexity of relationships in groups like chironomid midges — it seemed clear that vicariance wasn’t the whole answer, that temperate South America and Australis/Tasmania especially remained close enough for multiple cross-ocean dispersal events.

  10. Ha! I guessed convergence at the beginning of the article but the idea of widespread ratites throughout Gondwana seems very appealing. Imagine a super continent bustling with ratites!!

    1. I wouldn’t want a lot of cassowaries around my place.
      When I first saw one in a nature reserve I thought what a lovely looking bird, and then I read their description.

      From WP –
      “The inner or second of the three toes is fitted with a long, straight, murderous nail which can sever an arm or eviscerate an abdomen with ease. There are many records of natives being killed by this bird.”

      It does however note there is only one documented human death.

  11. Mitochondrial DNA phylogenies have sometimes proved spectacularly wrong — when dealing with closely related species that hybridized occasionally long after they diverged. With the major ratite groups, one would expect mDNA to work well because any hybridization would have occurred so long ago, presumably before the over-water dispersal events.

    Of course, this is a hypothesis worth testing — if sufficient nuclear DNA can be extracted from the elephant bird remains, which were preserved in situations that lead to the degradation of DNA.

  12. Now comes the issue of why would birds repeatedly evolve large flightless forms in the first place? What are the evolutionary pressures?
    I have some ideas, but I would like to learn what other people think.

    1. The Ars article on the elephant bird/kiwi connection noted that the giant forms evolved before or when mammals were present. That would nicely predict how the kiwi couldn’t become a grazer, because that niche was already taken, so it became a more diminutive insect eater. The dodo may then be the latest “no mammals” giant form.

      So I guess the pressure was availability of grazing and absence of other predators than birds. The latter would presumably help force giant forms, like they did in other dinosaurs.

      1. Oops. When mammals were _not_ present.

        And if it wasn’t obvious, I simply recapitulate the pressures that have been postulated to drive other giant forms of dinosaurs during earlier times. No need to invent new pressures if the old was still around at times.

        1. Your point about no large mammals is apt, but also this ratite divergence occurred soon after the dinosaurs went extinct, so there was a paucity of large predators overall.

          Flightlessness is an easy development to explain, flight takes a lot of energy and limits body mass, so if flight is not absolutely necessary to survival, it is quickly discarded. Large body size helps reduce heat loss, as well as making you a stronger fighter.

  13. I found the interesting Aldabra tortoise here. Notably in this context, “this occurrence is the inverse of the mainland-to-island rafting mechanism central to the island biogeography of terrestrial vertebrates.”

    That individual is a ringer for ocean transport – look at those barnacles and their position! Good think the poor animal has a long neck.

    I guess my favorite land bridge/missing land bridge/rafting is the arctic fox of Island, discussed on WEIT before. “The arctic fox (Vulpes lagopus) ventures way out on to the sea ice, and can float out on ice floes, having been recorded as turning up occasionally in eastern Canada, far to the south of its native range. It is also the only native land mammal of Iceland, an island which has never had a continental land connection, and which it must have reached on ice floes.”

  14. Great post – tetrapod zoology is discussing ratites. When I used to go to Madagascar (a long time ago) I knew a local guy in the north who used to wind up the few tourists by telling them elephant birds still roamed the forests of the Masoala Peninsula. Also, as I have nothing ‘scientific’ to add to the discussion, I’ll just recommend ‘Aepyornis Island’ by HG Wells.

Leave a Reply