A Blog by

The Surprising Closest Relative of the Huge Elephant Birds

The largest birds that ever lived—the now-extinct elephant birds—looked a lot like super-sized ostriches. They were fast-running and flightless, just like ostriches are. And their island home of Madagascar was just a short distance from mainland Africa, where ostriches live.

If you had to put money on the identity of the elephant birds’ closest living relative, the ostrich would be a safe bet. It would also be a spectacularly wrong one.

Elephant birds have been extinct for centuries, but many of their skeletons reside in museums around the world. By extracting DNA from these specimens and comparing it to DNA from living birds, Kieren Mitchell from the Australian Centre for Ancient DNA has discovered that the elephant birds’ closest living relative is… the kiwi.

“That blew us away completely,” says Alan Cooper, who led the study. “You have to try pretty hard to get a more disparate pair.”

He’s not exaggerating. Elephant birds probably plucked fruit from trees, while kiwis rummage through leaf litter for grubs and worms. Elephant birds lived in Madagascar, around 7,000 miles away from the kiwi’s home in New Zealand. The biggest elephant birds great up to 3 metres tall and weighed up to 275 kilograms; kiwis would bump against your shins and smaller species could fit inside an elephant bird’s gargantuan egg.

“Geographically, it didn’t make any sense. Morphologically, it didn’t make sense. Ecologically, it didn’t make any sense,” says Cooper. “We tested it pretty exhaustively because we were so surprised, but there’s no doubt in the genetic data.”

All the scientists I contacted agreed that the result is surprising. But it actually fits with a new narrative about the origin of ostriches, kiwis and their kin, which has been gaining support over the last few decades.

Aepyornis maximus, the elephant bird.
Aepyornis maximus, the elephant bird.

The elephant bird and kiwi belong to a group of birds called the ratites. These include the ostrich from Africa, the rhea from South America, the emu and cassowary from Australia, and the extinct moas of New Zealand.

Kiwis aside, these species are all big and flightless. Many scientists (quite reasonably) assumed that evolved from a common ancestor that was itself already big and flightless. This ancestral ratite probably lived at a time when all the southern continents were fused into a single land mass called Gondwana, and diverged into separate forms when the super-continent broke apart.

This ‘rafting’ story seems intuitive but it has crumbled in the face of genetic evidence.

As scientists compared ratite DNA, they found that geographical neighbours aren’t necessarily evolutionary neighbours. The moas and kiwis, for example, both hail from New Zealand. But when Cooper sequenced moa DNA in 1992, he found that kiwis are closer to the Australian emus and cassowaries than to their island neighbours. These birds arose after Australia and New Zealand had split so if they all evolved from an already flightless ancestor, the kiwis must have somehow rafted over a huge stretch of Pacific.

The ratites aren’t all flightless either. Genetic studies revealed that a group of flying, South American birds called tinamous are part of the ratite group. Stranger still, the closest relatives of the small, partridge-like tinamous are the huge, towering moas—a fact that Allan Baker from the Royal Ontario Museum confirmed earlier this month.

Cooper’s discovery mirrors the moa-tinamou relationship. Two groups of giant birds (moas and elephant birds) are more closely related to small, chicken-sized ones (tinamous and kiwis) from the other side of the world, than to similarly large neighbours (ostrich and rhea). Time and again, physique and geography have proved to be poor guides to ratite evolution.

There’s only one plausible explanation: the ratites evolved from small, flying birds that flapped their way between continents and independently lost the ability to fly on at least six separate occasions.

A) The break-up of Gondwana into separate continents. B) The ratite family tree, as you'd predict from the rafting hypothesis. C) The actual ratite family tree. Credit: Mitchell et al, 2014.
A) The break-up of Gondwana into separate continents. B) The ratite family tree, as you’d predict from the rafting hypothesis. C) The actual ratite family tree. Credit: Mitchell et al, 2014.

The rafting hypothesis is dead, and the kiwi-elephant bird is the “final nail in the coffin”, says Michael Bunce from Curtin University, who studies ancient DNA. “A number of text-books need to be re-written.”

Indeed, in his 2004 book The Ancestor’s Tale, Richard Dawkins writes, “I take delight in the power of natural selection, and it would have given me satisfaction to report that the ratites evolved their flightlessness separately in different parts of the world… Alas, this is not so.” Cheer up, Richard. It is so.

The ratites are an incredible example of convergent evolution—the process where living things turn up to life’s party wearing the same clothes. “They all started as these small, flighted, partridge-like things and most of them became these large, giant forms that were so close that everyone assumed they must have started off like that,” says Cooper.

Cooper thinks that the rise of the ratites took place shortly after the extinction event that wiped out most of the dinosaurs. Their absence created an ecological vacuum—there were lots of plants around and no big animals to eat them. The ratites filled those niches. Time and again, they evolved into big plant-eaters, losing the ability to fly in the process.

After around 10 million years, the mammals started doing the same thing, and their success stopped other birds from following in the ratites’ footsteps. “The window of opportunity had gone,” says Cooper. “No bird after that point could try and become big and flightless again, or they’d get eaten. The ratites survived by running like hell.”

This idea also explains why the kiwis and tinamous stayed small. Cooper thinks that they diversified in places that already had large flightless ratites—the moas and rhea. “These guys turned up after someone else had taken up the big and flightless niche, forcing them to do something alternative,” he says. “The kiwis became small, nocturnal insectivore. The tinamous kept flying.”

The discovery raises other puzzles about kiwi evolution. Some scientists believed that kiwis lay disproportionately huge eggs because they evolved from a much larger ancestor. “This new paper proposes that kiwis have always been small, suggesting that egg size independently became large in kiwis even when body size did not,” says Rebecca Kimball from the University of Florida. “That may stimulate some new and interesting research.”

If there’s a crack in this story, it’s that Cooper’s result is based on mitochondrial genomes—small, secondary sets of DNA within our cells. In the past, scientists have had to revise conclusions based on mitochondrial DNA after getting their hands on the main nuclear genome.  “The mitochondria only give us part of the picture,” says Bunce. “The next challenge is to peek into the nuclear genome.”

“We’re working on that,” says Cooper. It’s not easy. Elephant birds tend to die in hot, humid, swampy conditions, which are awful at preserving DNA. Cooper’s team struggled for years to get enough material to sequence, before finally recovering enough from specimens stored in a New Zealand museum. It’ll be even harder to sequence the nuclear genome of these titans. Cooper adds, “The nuclear DNA from the other ratites, including the moa, confirms what we see from the mitochondria, so we’re not expecting too many surprises.”

Reference: Mitchell, Llamas, Soubrier, Rawlence, Worthy, Wood, Lee & Cooper. 2014. Ancient DNA reveals elephant birds and kiwi are sister taxa and clarifies ratite bird evolution. Science http://dx.doi.org/10.1126/science.1251981


21 thoughts on “The Surprising Closest Relative of the Huge Elephant Birds

  1. Thanks Ed Yong for another well written and very accurate report.

    @ Lee J.
    Your question about the egg size is a good one! Our study suggests moas arrived first in NZ and occupied the niche for large flightless herbivores, forcing kiwis to remain much smaller and adopt another role (much like rheas, giant and flightless, and tinamous, small and volant, in South America). The most favored hypothesis is the ancestor of the kiwis was a giant bird and while body size shrunk on the course of evolution, the egg size remained more or less the same. But was the kiwis’ ancestor really a giant? Well, Trevor Worthy and colleagues recently published a paper in Paleornithological Research where they show the kiwis’ ancestor was most likely a small volant bird. So what alternative hypothesis could explain the disproportionate size of the kiwi egg? Problem is when you are a small-ish bird, predation becomes a serious issue especially for the young. While tinamous retained the ability to fly (although badly, they usually prefer to run) maybe kiwis opted for another solution: allowing the chicks to grow as much as possible in the egg, to the point where they could run as soon as they hatch and be less vulnerable to predators. Of course one can argue a big egg laid on the ground is also pretty vulnerable… And tinamous lay their eggs on the ground as well, their chicks can run as soon as they hatch, but although the relative egg size is on the large side, it does not reach an extreme size such as in kiwis… I guess more research is needed!

  2. The graphic of the Tinamou is slightly misleading. Looking at pictures of it on Wikipedia, it looks like it shares morphological features with the other Ratites. So it seems logical all modern Ratites shared a common ancestor that had the morphology to begin with but the current species specialized into unique niches and became larger or smaller.

    I find the flying from Australia to NZ theory a little hard to stomach, unless these two landmasses were much closer during some time. The Kiwi and Moa may have had their own ancestors in New Zealand that we havn’t discovered yet. These ancestors could’ve entered New Zealand over land.

  3. If the Kiwi feeds at night, what keeps the egg warm? Nights tend to be cooler than days: are the eggs so large as to avoid heat loss? Or do the parents take it in turn to incubate as other species do?

  4. I enjoyed reading this, but I don’t think it does anyone any favors to attribute conscious agency to evolving species, even metaphorically. No animal ever “tried” to fill a nice, or was discouraged from doing so. Such discourse is lazy on the part of scientists, confusing to the public, and encouraging to those looking for intelligent design.

  5. Verg much agree with Zigth. There is need for a much more precise wording that is much closer tuned to current biological thinking. It will require a departure of popular phrasing and images, but the sooner that is started the better.

  6. The fact that they used mitochondrial DNA is not a crack in the story. It’s a gaping chasm.

    Anyone remember how polar bears became brown bears due to mtDNA testing, and then weren’t when the tests were done properly with nuclear DNA?

    It’s just too easy for a single hybridization event to carry through immense ages on mtDNA for it to be reliable for these purposes. That doesn’t mean these results are wrong, but it does mean it’s entirely reasonable to seriously doubt them.

  7. It’s funny; it never occurred to me that kiwis had to be related to other flightless birds. I always assumed New Zealand had flightless birds because a) flying is expensive and b) life before people and rats showed up was pretty easy in NZ, and so the birds that arrived there simply gave it up as an unnecessary luxury.

    [Andrew, that’s pretty much right. NZ has lots of other flightless birds that lost flight for those reasons, like the kakapo. The kiwis (and the moas) are exceptions in that they’re part of a whole order of (mostly) flightless birds) – Ed]

  8. The fact that mitochondrial DNA is used is not a serious problem in this case. If you’re talking about closely related organisms that can hybridize in the relatively recent past, then mitochondrial DNA can give misleading signals (as in Polar vs. Grizzly/Brown Bears). There are two reasons to think that mitochondrial DNA might give an accurate clue to the phylogeny of ratites. First, the birds diverged by crossing significant water barriers, which would inhibit hybiridization by the early members of a given lineage, who might otherwise have hybridized. Second, once organisms are separated for a long time, they can’t hybridize. There is good reason to think that the divergence times for ratites are so long ago that hybridization hasn’t been happening recentlly enough to confuse the phylogeny.

    Of course, this a hypothesis that should be tested if enough nuclear DNA can ever be extracted from elephant bird remains, which existed in hot, often humid conditions that allow DNA to degrade relatively rapidly.

    [Good point, well made. Thanks, Barbara – Ed]

  9. It is interesting to see how a small flying bird could evolve into several large flightless birds.
    Has any one looked further back in the evolutionary tree to see if there was a large ancestor in the ratite’s family history? It would be great if the birds started out with a large ancestor which evolved into a small flier, and regained size in new locations as the opportunity presented itself. This could also help explain why the kiwi has such a large egg, as it was leftover genetic code from the large ancestor of the flying ratite.

    [Hi, Tony. All birds evolved from small feathered dinosaurs. Archaeopteryx, which was either an early bird, or a dinosaur very close to the root of the bird family tree, was about the size of a raven. So the group as a whole started small. And since the ratites are one of the earliest bird groups, it’s likely that they started small too. – Ed]

  10. While ratites may be an early split in the Neornithes, that still puts that event between 60 and 90 million years ago. Archaeopteryx lived 150 million years or so ago. That gives us 60-90 million years of evolution between the two, or just about as much as we’ve had since. Using Archaeopteryx and other small theropods as a model for the LCA of neornithes seems rather problematic to me. As an example, by the mid-Cretaceous or so, some of the Hesperornithes had reached 6 feet in length — hardly a small bird. (They weren’t direct ancestors, just an illustration that the timing would certainly allow a larger ancestor.)

  11. @Patrick You may have just solved the disproportionate egg question – A large egg might hold more heat through the cool night than a small egg, facilitating nocturnal feeding

  12. This is just hypothetical (probably unethical as well), but can any different species of Ratites mate and produce viable offspring? Also , if there is enough DNA to understand the common relationships, is there also enough DNA to bring back the Moa? They did bring back the Kiwi and the Kakapo from near extinction.

  13. @ Wade; that’s true, Wade; also the Black Robin. But there’s quite a difference between “NEAR extinction” and FULL extinction.

Leave a Reply

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