Australasian Science: Australia's authority on science since 1938

Flight of the Kiwi

What came first: the kiwi or its enormous egg? Credit: Rod Morris Photography

What came first: the kiwi or its enormous egg? Credit: Rod Morris Photography (

By Michael Lee & Trevor Worthy

How did large flightless birds such as emus, ostriches and kiwis disperse around the globe? Surprisingly, it seems they flew everywhere.

The presence of large flightless birds across all of the southern continents has been one of the enduring mysteries of evolution. The emu and cassowary (Australia), kiwi and extinct moa (New Zealand), rhea (South America) and ostrich (Africa) form a close-knit group of birds called “ratites”.

Scientists have long wondered how these landlubbing avians managed to disperse across landmasses separated by large swathes of water. A plausible answer proposed a few decades ago has now been overturned by Australasian scientists, with accumulating evidence supporting a more outlandish and exciting scenario. This new research also resolves the long-standing question of what came first: the kiwi or its gigantic egg?

Deep time and continental drift have been integral to a rather prosaic explanation for the current wide distribution of ratites. It was proposed that ratites were an extremely old group that roamed across the ancient southern supercontinent Gondwana. When Gondwana started breaking up, every major continental fragment that split away contained some ratites, which gradually evolved into distinctive new endemic forms while being “rafted” in slow motion via continental drift.

This idea, pioneered by leading ornithologist and biogeographer Joel Cracraft, was intuitively satisfying and initially had some support from evolutionary trees. The ostrich was the first ratite lineage to split away from the others, mirroring the fact that Africa was the first continent to break away from Gondwana. Similarly, the emu and cassowary were closest kin, consistent with them both evolving from a ratite lineage that was later isolated when Australia split from Gondwana.

However, dispersal by continental rafting required lots of time. Gondwana began fragmenting 130 million years ago, and its shards moved at the speed of fingernail growth. So, for this idea to fly, the ratites would have to be more than 130 million years old. This means they would have lived alongside dinosaurs for half of their history – very old birds indeed.

Recently some cracks began to appear in this explanation. Close genetic similarities between all ratites seemed to discount the idea of ancient divergences that coincided with the rifting of Gondwana.

More tellingly, the location of some of the branchings in the ratite evolutionary tree were inconsistent with this idea. The tiny kiwi and enormous moa – both predicted to have evolved from an ancestral ratite trapped on New Zealand as it unzipped off Gondwana – were not close kin. Rather, starting around 1990, DNA analyses by Alan Cooper (now at The University of Adelaide), Matthew Phillips (now at Queensland University of Technology) and colleagues suggested a different story.

First, the genetic evidence grouped the kiwi with Australia’s emus and cassowaries, not moas. Then, analysis of DNA from moa bones revealed – surprisingly – that moas were actually most closely related to tinamous, a group of partridge-like flighted birds from South America.

This pairing of flightless moa and flighted tinamou suggested that loss of flight happened multiple times within ratites, while the huge oceanic distances separating these cousins strongly suggested that their ancestors dispersed by flying. Perhaps a tinamou-like ancestor flew across the Pacific from South America to New Zealand, before completely losing its wings and growing enormous in size. And if moas did this trick, why not other ratites?

The latest twist in the story comes from studies of the “forgotten ratites” – the elephant birds of Madagascar. The largest of these behemoths stood 3 metres tall, weighed 300 kg, and laid eggs more than 100 times the size of chicken eggs. They became extinct only a few hundred years ago – recently enough to be preserved in folk memory and in written records by explorers such as Marco Polo.

The “continental rafting” hypothesis predicted that elephant birds should be a long-isolated lineage of ratites, of similar age to ostriches (since Madagascar separated from Gondwana shortly after Africa split off). But with only bones to work with, scientists have been unable to conclusively resolve the evolutionary affinities of elephant birds.

Despite numerous attempts, only tantalising small snippets of DNA have ever retrieved. Unlike moas – often preserved in cold dry caves in New Zealand that act as “DNA-refrigerators” – elephant bird bones in Madagascar are exposed to warm, wet conditions that rapidly degrade DNA. Their rarity in collections further stymied efforts.

However, with technologies for DNA sequencing advancing at breakneck speed, we worked with a team led by Kieren Mitchell and Alan Cooper that managed to extract the equivalent of blood from a stone – the complete genetic code of the cell mitochondria from subfossil elephant bird bones. The genes provided a surprising revelation: elephant birds were the closest cousins of kiwis – closer than even emus and cassowaries. Two of the most bizarre avians to have ever existed were birds of a feather, despite being diametrically different in appearance and inhabiting islands on opposite sides of the globe.

Flighted dispersal was the only plausible explanation for such a disjunct distribution of closest cousins. A smallish flighted Madagascan bird could have dispersed to colonise New Zealand (or vice versa). Or perhaps both dispersed from that other forgotten realm, Antarctica, a land of forests and abundant fauna at the time.

Regardless of where they flew from, both populations then lost their wings and increased their body size in parallel, with the Madagascan lineage becoming gargantuan.

This idea was further supported by archaic fossil kiwis recently discovered in New Zealand. These are much smaller than modern kiwis, suggesting evolution from a smaller, flighted ancestor.

The new genetic data also suggests that the fossil history of all living ratites needs to be totally reevaluated. The ancestor of emus and cassowaries, for instance, must have been a small flighted bird 40–50 million years ago. This is completely different from the current search image of most palaeontologists! Would it be recognised if found?

The previous picture of ratite evolution – an ancient and uniformly flightless group that dispersed passively by slow continental rafting – is now being replaced by a view that is more dynamic, both evolutionarily and biogeographically. There is now strong evidence that the ratite group evolved more recently, and were initially smallish birds – perhaps similar to partridges – that flew rapidly to far-flung corners of the globe. Founders of the various ratite lineages landing on different continents then lost their wings and increased in size, perhaps five times in parallel.

Modern biologists have had difficulty explaining how a close-knit group of giant flightless birds managed to disperse across widely separated continents, but the answer is that they “cheated”. They flew everywhere, and then lost their wings after arrival, erasing evidence of how they got there.

Many Pacific islands contain smaller-scale analogues of these phenomena. Flightless rails can be found on these islands, having flown there and then rapidly lost their wings. The barred-wing rail group (Gallirallus) has produced large flightless species on New Zealand, Lord Howe, New Caledonia and many other islands of Oceania, as well as tiny, fluffy, near-wingless forms on Chatham Island.

There was never any doubt that these now-flightless rails flew to their destinations. Many of these islands (e.g. coral atolls) have never been connected to other landmasses, and the existence of very similar but flighted rails indicated that flightlessness has been a very recent and repeated invention. On small land masses with few or no predators and shorter travel distances, flight is an inefficient indulgence that is metabolically expensive and requires delicate, injury-prone wings.

The convergent evolution of large, flightless ratites on separate continents highlights the ability of evolution via natural selection to shape similar forms in different parts of the world. Many traits have evolved repeatedly in ratites: wing reduction and increased body size in all forms; long legs for fast running in ostriches, rheas and emus; and pillar-like limbs to support massive bodies in moas and elephant birds. The genetic data – including ancient DNA from moa and elephant birds – have been instrumental in revealing just how often these convergent similarities have evolved.

The end-Cretaceous extinction of non-avian dinosaurs might have paved the way for the evolution of these big, flightless birds by clearing global ecosystems of big predators as well as potential competitors (e.g. large grazers and browsers). Why did such characteristics evolve repeatedly in ratites but rarely in other bird groups? The early, flighted ratites might have been a bit larger than most of their avian contemporaries, giving them an evolutionary head start when the dinosaurs suddenly disappeared.

Finally, these new genetic and fossil discoveries about ratites also resolve the long-standing enigma about how the kiwi came to have its famously large egg. The earlier view of ratite evolution implied that ratites had large bodies and large eggs throughout most of their evolutionary history. Kiwis were thought to be an aberrant ratite lineage that evolved dwarfism relatively recently but retained the large eggs of moas, emus and their kin. Thus, the large egg came first, representing a non-adaptive evolutionary holdover.

The emerging view of ratites inverts this story. Kiwis, like other living ratites, evolved from flighted ancestors, and have always been small – indeed the new fossils indicate they were previously even smaller. These flighted kiwi ancestors would not have been burdened by large eggs.

The enormous kiwi egg more likely represents a recent adaptation to the New Zealand environment. In stable, predictable habitats saturated with competitors, reproductive success is maximised by investing in fewer, stronger offspring rather than many feebler ones. The baby kiwi hatches from its egg, extracts energy from an enlarged yolk sac for a couple of days, then heads off into the world, ready to feed and fend for itself. Dad, who incubated the egg for nearly 3 months, is long gone.

Thus, the small kiwi probably came first, and its gigantic egg was a later adaptation.

Michael Lee is a senior research scientist at the South Australian Museum and the University of Adelaide. Trevor Worthy is an ARC research fellow at Flinders University.