Return of the lost birds

Since humans arrived in New Zealand, we’ve lost nearly half of our native terrestrial bird species. Some of those extinct icons are well known, while others are recalled only by myth and bones. We will probably never know the full polyphony of that primordial dawn chorus, but old bones and new science are giving us a richer picture of life in the land of birds, back when they still ruled the roost. For the first time, we’re able to answer questions about what they ate, where they came from, how they were related to each other, and how they got so much bigger, heavier, and weirder than their ancestors.

Written by       Photographed by Neil Silverwood Illustrated by Henrietta Harris

In a shallow cave on Mt Nicholas Station, perched above Lake Wakatipu, four scientists are looking for treasure. It’s early spring, and a kilometre away, shearers are working their way through the farm’s 30,000 merinos. Three days ago, a heavy snowstorm painted the mountains white, and the cave’s open mouth frames a spectacular sweep of the Southern Alps, from Mt Earnslaw, shining silver-blue at one end of the long lake, to Queenstown at the other.

Inside, Janet Wilmshurst’s focus is ten centimetres from her face. Crouching in the dirt in a Swanndri and gumboots, her spiky grey hair almost brushing the cave roof, she peers into a sieve with her headtorch, hoping to find the partially digested remains of a laughing owl’s lunch.

She shakes the sieve over and over, sifting through the tiny particles with practised fingers. Then—“Ooh!” She picks something minuscule out of the rubble, and passes it to Jamie Wood.

“Wing bone,” he says. “Clearly been chewed by an owl.”

These are the moments that make the days of sifting worth it.

“After endless hours of this, someone says, ‘Ooh,’ and everyone looks up,”
says Wilmshurst.

“Even if you just sat on a sharp rock,” adds Wood.

This morning, there have been plenty of ‘Ooh’ moments. Along this small crack in the rock, hidden among the stones and the soil kicked up by rabbits, are dozens of tiny bones.

Wilmshurst and Wood, both from Manaaki Whenua–Landcare Research (Wilmshurst is also an associate professor at the University of Auckland), describe themselves as palaeoecologists. Their job is to reconstruct the lives of New Zealand’s birds and animals, many of them now extinct.

The tools that this team from Landcare Research use in the field are straight out of Indiana Jones: sieves, a dustpan and brush, a trowel and paintbrush. “It’s cutting edge analysis meets old school excavation,” says Janet Wilmshurst (right). “You just never know what’s there until you dig a hole.”

Dry caves in Central Otago such as this one provide the perfect preserving environment for the traces of long-dead birds. One extinct species in particular, the laughing owl, has made the task even easier—it has helpfully preassembled a collection of bones.

The last laughing owl was found dead on a South Canterbury road in 1914, but before that, says Wilmshurst, generations of them perched in this cave.

The laughing owl, or whēkau, was more than twice the size of a morepork, with a cackling cry that was likened to “doleful shrieks” and “two men cooeeing to each other over a distance” by the amateur naturalist William Walter Smith.

The owl hunted a wide array of small animals—birds, lizards, frogs, tuatara, insects, and, after the arrival of humans, rats. It brought its prey back to its roost and ate it whole, regurgitating the hard bits—bones, beaks and claws—onto the cave floor. What’s left is an invaluable record of a lost ecosystem.

“If it wasn’t for the owls, we’d know very little about any of these little birds and frogs,” says Wood.

He and Wilmshurst first visited this area in 2009, looking for moa coprolites—preserved poo (see sidebar). Searching the literature, the researchers had discovered an account by Taylor White, Esq., who in 1875 described to the Otago Institute a cave “two miles east of the Von River… in a small conical hill about a quarter-of-a-mile from the Lake” where he had found moa feathers, bones, eggs and droppings.

Wood and Wilmshurst followed his directions. Beneath the muscular green flanks of Mt Nicholas, they came to a low ridge of schist, and, just as White had said, at the top of a “tolerably steep rise”, they soon found a cave matching his description. To their great disappointment, it was riddled with rabbit burrows, the layers mixed up and any coprolites likely destroyed.

Just along the ridgeline, though, was another cave, and another—a series of small crevices and rock shelters. Inside, the researchers unearthed dozens of coprolites and a collection of tiny bones. They spent only a couple of days there, and barely scratched the surface.

Now, almost a decade later, Wilmshurst and Wood have returned to the site, along with two young researchers—Landcare’s Jessica Rivera-Perez, who is studying kiore coprolites, and Luke Easton from the University of Otago, who’s hoping some extinct frog bones will turn up.

As they pick bones and droppings out of the sieves, they place them in little screw-top plastic bottles, ready for analysis at Landcare’s ancient-DNA laboratory. These findings don’t look like much—bone fragments the size of staples and matchsticks, droppings not much bigger than a grain of rice—but they have stories to tell.


Over the past decade, advances in DNA extraction and radiocarbon dating, alongside other new techniques, have transformed the information scientists can extract from artefacts like these, leading to a host of new discoveries about our native species: where they came from, how they lived, and why they died.

Researchers can now reconstruct aspects of New Zealand’s deep past that we once knew nothing about. They’re also shedding light on the process of evolution itself—how, why, and how quickly one species changes into another.

According to American biologist Gareth Nelson, the weirdness of New Zealand’s plants and animals is a key question of biogeography—the study of which species end up where in the world. “Explain New Zealand,” he wrote in Systematic Zoology in 1975, “and the rest of the world falls into place around it.”

[Chapter Break]

Exploring caves and sifting tiny bones out of the dust is rather romantic (and certainly gets you out of the office) but it’s no longer the only way to make dramatic new discoveries. In fact, most recent breakthroughs regarding New Zealand birds have come from applying the new techniques on bones that were collected decades or even centuries ago.

In the basement of Te Papa’s back-of-house building on Tory Street in Wellington, hundreds of thousands of biological treasures huddle on shelves and in filing cabinets, an assembly of feathers and fur, teeth and claws.

Whale bones in wooden boxes line one wall of the huge warehouse-like room. Spiky antlers protrude from another, where the mounted heads of deer, moose and bison jostle for space. Underneath the trophies, the fluorescent lights glint off a fibreglass Māui dolphin, improbably placed next to a taxidermied tiger.

“What’s the tiger doing here?” I ask Alan Tennyson, Te Papa’s vertebrate curator, who is giving me a tour. The exotic fauna is a bit of a relic, he explains—in part, a legacy of New Zealand’s most iconic extinct bird. Some of the strangest international oddities were traded in the 19th century for sought-after moa bones. At that time, he says, museums were trying to showcase the diversity of the world’s interesting things, and moa bones were good currency.

These days, museums are more focused on the local. Tennyson takes me over to another part of the building, where precious avian artefacts are stored in cardboard boxes. Here, Te Papa holds samples from all of New Zealand’s living and recently extinct bird species. It’s like a morbid version of Noah’s ark—except the museum wants more than two of every kind, says Tennyson. At minimum, it aims for at least 30 males and 30 females of each species, a record for posterity of the genetic diversity both within and between species.

“We’ve got one of the best recent bird-bone collections in the world,” says Tennyson, waving a long arm. “These are all albatrosses. I don’t think anyone in the world could compete with that many albatross specimens.”

A whole row of shelves is reserved for moa bones, still the single-most-researched thing in this entire room: “There’s always something new to be done on moa.”

Advances in ancient DNA have resulted in a new era of discovery for New Zealand’s extinct birds. “People have tended to focus on moa—and although they’re interesting, they’ve been studied to death,” says Jamie Wood, who runs the ancient DNA laboratory at Landcare Research. The techniques honed on moa are now being used to explore other species, with unexpected results.

In the early 1990s, moa and kiwi bones were among the first in the world to receive the full treatment with new DNA tools, and the results were full of surprises. Moa and kiwi had been assumed to be close cousins, as they’re both members of the ratite family of birds.

But an early DNA study conducted in 1992 suggested that the ancestors of the kiwi and the moa had flown to New Zealand separately and, over millennia, had each adapted to become flightless. Kiwi turned out to be relatives of the giant, extinct Madagascan elephant bird, while moa were most similar to the South American tinamou, which is small, quail-like and can fly.

In the early 2000s, other studies revealed that what were thought to be separate species of giant moa were in fact the males and much-larger females of the same species. Females were half as tall again as males, and almost three times as heavy—an example of sexual dimorphism so extreme that no one had picked it up by just looking at the bones.

Moa and kiwi are the sexy subjects that everyone in ornithology wants to study, says Tennyson, but in the past few years, researchers have turned their attention to lesser-known extinct birds. Bones that had sat largely undisturbed for decades began to reveal new secrets—and new species.

“DNA sequencing is revolutionising taxonomy,” says Tennyson. “Most people thought the world’s birds were already well described and known, with maybe one or two new discoveries a year in the Amazon jungle or outback Indonesia. But actually, genetics is discovering far more diversity than was known about, and revealing relationships between species we’d never considered.”

He bends down to a lower shelf and opens a sturdy cardboard box. Under a layer of cotton wool nestles the disassembled skeleton of a large bird, each bone inscribed with an identifying number. There’s a row of vertebrae, long thin wing bones, a pair of heavy femurs—and one huge talon.


Tennyson picks up the talon, and it stretches three-quarters of the way across his palm. It looks like it should belong to the stuffed tiger, but these are the bones of Haast’s eagle, the apex predator of prehistoric New Zealand. Weighing up to 17 kilograms, with wings three metres across, it was the world’s largest eagle—and the original moa hunter.

When Julius Haast first described the bird in 1871, he called it Harpagornis moorei, from the Greek word harpax, meaning ‘grappling hook’—and that seems a good description for this formidable claw.

The skull is smaller than I’d pictured, but Tennyson says this is common with bird skeletons. “They look quite scrawny when you take their feathers off them.”

A nearby box contains an Eyles’s harrier specimen. Researching this story was the first time I learned that New Zealand had not one giant raptor, but two. Circus teauteensis looked a lot like kāhu, the swamp harrier common around New Zealand now—but it was four times as heavy, with a two-metre wingspan. At 3.5 kilograms, it was significantly lighter than Haast’s eagle, but it was still the largest-known harrier in the world.

Like Haast’s eagle, Eyles’s harrier is believed to have died out shortly after Polynesians arrived, its extinction the result of habitat change, the demise of its prey, and possibly predation by rats and humans.

To reconstruct the history of these aerial terrors, all scientists had to go on until recently was the shape of their bones—their morphology—and a few other tantalising sources of circumstantial evidence: moa pelvises with huge holes torn through them, early Māori rock art depicting a giant bird of prey, a few tools carved from eagle bones. There were also myths and oral histories, including several stories of a bird so big that it carried children away in its talons.

The evolutionary origin story of these two birds is even more dramatic. Looking at the bones, researchers had assumed Haast’s eagle was most closely related to Australia’s large wedge-tailed eagle. Then, in 2005, Michael Bunce and colleagues from the University of Oxford compared the mitochondrial DNA of two Haast’s eagle specimens with that of 16 living raptors. The results surprised everyone.

Excavating moa bones at Glenmark in north Canterbury in 1871, Frederick Fuller, the Canterbury Museum taxidermist, unearthed the first example of what Julius Haast later described as “a raptorial bird of enormous dimensions”—a bird that would soon be known as Haast’s eagle. These are the same bones, still in Canterbury Museum’s collection.

It turned out Haast’s eagle did have a close relative across the ditch, but it wasn’t the wedge-tailed eagle. The DNA evidence was clear: Haast’s eagle belonged with a group of tiny eagles in the genus Hieraaetus, including the Australian little eagle and the Eurasian booted eagle. Both weigh less than a kilogram—a tiny fraction of the size of their immense cousin.

Their genomes were so similar that Bunce decided they should share a genus, and Haast’s eagle has been reclassified as Hieraaetus moorei—making the evocative name of Harpagornis as extinct as the bird it
once described.

The results indicated Haast’s eagle had probably evolved from a very small ancestor in less than a million years. It was a 15-fold increase in size, and a record-breaking example of island gigantism—the phenomenon where species isolated on islands evolve to become larger than their mainland relatives.

So how did one of the world’s smallest eagles become the biggest ever known, in an evolutionary blink of an eye?


Michael Knapp, a senior lecturer at the University of Otago, is trying to find out. He first read Bunce’s paper as a PhD student, having just arrived in New Zealand from Germany. He was fascinated: “Whether you know anything about birds or not, that’s a spectacular change. I thought, ‘How can it be? How can you have two species just two million years apart and so different, when evolution is supposed to work slowly?’”

Knapp now has an idea why it happened. Around two million years ago, the tiny ancestor of Haast’s eagle arrived in New Zealand, perhaps blown across the Tasman in a westerly storm.

“They came into this environment where there were huge chunks of meat running around, and nothing eating them: moa. If you’re a small bird, you have a lot of competition—all the owls, the falcons, everything looking for reasonably small prey. But nothing was going after moa.”

The bigger an individual eagle was, the more success it was likely to have—and that exerted what’s called ‘strong evolutionary pressure’ in favour of large size. With every generation, the biggest individuals were favoured, and the larger they got, the more food there was available for them.

The next question is how, and Knapp’s been awarded a five-year Rutherford Fellowship to try to answer it. He wants to pinpoint what genetic changes underpin the dramatic size increase by comparing the Haast’s eagle genome with that of its Australian and Eurasian relatives. Did the genome itself change dramatically as the eagle evolved—or do a handful of genes make the difference between a tiny eagle and a giant one?

Two developments over the past decade have made this research possible. Next-generation DNA sequencing gets cheaper and faster every year, and isolating the sequences themselves is also getting easier.

Ten years ago, scientists could only analyse fragments of DNA longer than about 100 base pairs. But DNA degrades over time, and most fragments of ancient DNA are very short—only 50 base pairs, perhaps. This means the vast majority of the DNA extracted from ancient bones tends to be contaminated by bacteria living inside and on them. But recent advances have made it possible to target sections of 30 base pairs, greatly increasing the amount of useable DNA that Knapp can get out of a single bone.

Not that it’s an easy process.

“With modern DNA, you take a blood sample, you chuck it into some buffers, mix them a couple of times, rinse them, and done. The complete process is maybe an hour or two, and then it’s basically ready to send off to a sequencing service,” says Knapp. “With ancient DNA, I have to spend two days in the lab, pipetting like mad, to get the DNA out of the bone. After two days, you’ve got your Haast’s eagle DNA—or more accurately, your bacterial DNA with traces of Haast’s eagle. The next step is to make it ready for sequencing, and that’s another two days of pipetting like mad.”

Timing is everything—the longest break that Knapp can take in each of those days is half an hour.

“Even if you go flat out, it takes at least a week to get from the bone to something that is sequenceable.”

The next challenge is finding a fossil that’s well enough preserved to contain a complete genome.

Early naturalists placed the laughing owl in its own genus, but Jamie Wood’s DNA studies revealed it actually belongs in a wider family of owls, Ninox, which includes the morepork and Australian barking owl. Like the Haast’s eagle, the laughing owl rapidly adapted to fill an empty ecological niche. “Even though they were closely related to other owls, they looked and behaved quite differently,” says Wood.

Knapp has already sequenced the genomes of the little eagle and the booted eagle—both of which are still living—but Haast’s eagle has proved harder than expected.

“At the moment, I’m just hopping from one Haast’s eagle fossil to the next, trying to find one with good enough DNA,”
he says.

So far, he’s tested around 25 of the 70 known fossils, to no avail. But he’s planning to visit Alan Tennyson at Te Papa to drill a few small holes in some of the bones I saw—and he’s optimistic that one of these will have useable DNA.

Once he has sequenced the trifecta of Hieraaetus eagles, he’ll be able to compare the three genomes, looking for places where the two small birds’ genes are the same, but the Haast’s eagle ones are different. It’s a starting point for pinning down the genes responsible for the super-sizing of the New Zealand bird.

Knapp is also working on Eyles’s harrier—but in this case, he got lucky. Canterbury Museum natural history senior curator Paul Scofield sent him a harrier specimen that a museum and Landcare team had found in a South Canterbury cave in 2011.

“When I saw how well preserved the DNA was, I thought he was playing a trick on me,” says Knapp. “It was like he had given me a bone from a swamp harrier off the road.”

A 2015 paper had already laid out the evolutionary relationships of living harriers using mitochondrial DNA, so Knapp plugged his Eyles’s harrier data into that.

“To my great surprise it fitted with the Australian spotted harrier—again, a very small harrier. It was the story of the eagle, repeating itself.”

Not only that, but timing of the divergence between species was almost identical. Like Haast’s eagle and the little eagle, Eyles’s harrier and the spotted harrier diverged around two million years ago. And, it turned out, so did pūkeko and takahē, black stilt and pied stilt, the Australian raven and the extinct New Zealand raven.

It almost seemed like something was wrong with his methods, says Knapp—it was a bit too neat. But then he started comparing notes with Scofield and University of Otago colleague Nic Rawlence, and they came up with a new theory of New Zealand bird evolution.

“We’re affectionately calling it the Australianisation of New Zealand,” says Rawlence. “We’re trying to test whether the onset of the ice ages was this watershed moment in bird evolution in New Zealand.”

Geologists divide Earth’s history into eras, periods, epochs and ages. During the Miocene Epoch, which took place between about 23 million years ago and 5.3 million years ago, New Zealand was much warmer.

“If you were in New Zealand in the Miocene, it would have felt like the Sunshine Coast or even north Queensland,” says Scofield.

There were casuarinas, crocodiles and turtles, and nearly the whole country was covered in lush forest. But by the Pleistocene Epoch, which began about 2.6 million years ago, the ice ages had set in, and New Zealand was much colder and drier. Many tropical species died out, and the dense bush retreated, replaced by steppe or open woodland—more like the forests of southern Australia. Some birds blown across the Tasman found an environment they recognised and could thrive in—ecological niches were empty and large predators such as crocodiles were absent.


Over the following millennia, as ice ages came and went, and sea levels rose and fell by as much as 100 metres, the migrants adapted to their new environment. One of those new arrivals was the black swan.

Researchers had always assumed that swan bones turning up in early Māori middens and sub-fossil deposits on the Chatham Islands and the mainland were the same species as the black swans we now see on lakes, estuaries and ponds all around New Zealand. But when Rawlence and colleagues from the University of Otago, assisted by Tennyson and Scofield, compared the DNA of ancient and modern swans, they discovered they came from completely different lineages, separated by—yet again—one to two million years.

The researchers went back to the bones and started measuring them, and the differences became clear. What had been thought to be a wide variation in size within one species was in fact two species. They had found a completely new swan.

When you discover a species, you get to name it. Along with the swan’s scientific name—Cygnus sumnerensis, after the Christchurch suburb where the first remains were discovered—the researchers also suggested a common name: poūwa.

In 1892, Moriori leader Hirawanu Tapu told Henry Forbes, director of the Canterbury Museum, of a legendary giant black bird called the poūwa that had once roamed the Chathams. Its bones could still be found protruding from the Te Whanga lagoon, he said.

“It’s a dead ringer for a swan,” says Rawlence.

The poūwa was 20 to 30 per cent larger than the Australasian black swan—an All Black forward compared with a soccer player, says Rawlence—providing yet another example of island gigantism. Like so many other New Zealand birds, the poūwa was on the path towards flightlessness when it became extinct. Its wings were shorter than its Australian relatives, and its legs were much longer and stronger.

“If you’re spending more time on the ground, and you’re a lot heavier, you need increased power for take-off,” says Rawlence.

The discovery of the poūwa raises all sorts of interesting conservation questions, he says. Now that we know the Australasian swan is not a replica of the original New Zealand swan (it arrived in the 1860s, either under its own steam or introduced by Europeans), should it be protected, or is it more properly a pest? Can one species fulfil the same role in an ecosystem as a close relative?

[Chapter Break]

Ancient DNA has been revolutionary, but it isn’t the only tool that’s brought New Zealand’s extinct fauna into sharper focus over the past decade.

Medical technologies are giving researchers fresh insights into the biology of these animals, says Paul Scofield. With colleague Ken Ashwell, he’s been using CT scans to learn more about Haast’s eagle.

“CT scans give us a look into its spinal cord, its circulation system, into its brain and nerves,” he says. “You can virtually reconstruct the nerves that go to the eyes and the olfactory parts of the brain.”

Scofield and Ashwell found the eagle had a relatively small brain and eyes for its size, another piece of evidence supporting the idea that it evolved rapidly from a tiny ancestor. Its brain didn’t show signs of a keen sense of smell, and the high blood supply to its legs indicated it was a strike killer, with a lot of force in its talons—proof it wasn’t a vulture-like scavenger, as some had suggested, but an aerial hunter, truly capable of carrying off a child.

Other new techniques include bombarding buried bones with neutrons in a nuclear reactor to form a 3D image—avoiding the need to painstakingly chip away rock or metal—and using infra-red and ultra-violet light to reveal the imprint of mineralised scales on the leg bones of a moa.

Janet Wilmshurst, Jamie Wood and Jessica Rivera-Perez use fine-grained sieves to hunt for bones and coprolites in a cave on Mount Nicholas Station. Field expeditions are rare these days, says Wood: “We only tend to collect on occasions when there’s a very good reason to do it—keeping bones in caves where they’ve been sitting for thousands of years is actually the best preservation site for that material.”
Canterbury Museum has the world’s largest collection of moa bones. These are all from the same site, Pyramid Valley in north Canterbury—a natural moa cemetery. Discovered in 1938, the ancient swamp yielded a treasure-trove of 183 complete moa skeletons, trapped for centuries in the mire. The moa’s continued mystique ensured that many new research techniques, including ancient DNA, were pioneered on moa, says natural history curator Paul Scofield (pictured with the skull of a giant moa).

For all the new discoveries, Scofield says, we still know so little about New Zealand’s deep past. Compared with other countries, our fossil record is incredibly sparse. Partly, Scofield says, it’s because “we just don’t dig enough holes”.

Many fossil finds overseas have come from quarries, mines, and infrastructure excavations, which aren’t common in New Zealand. Many of the places most likely to contain fossils, Scofield says, are incredibly remote—“it’s a really difficult job to even get there, let alone dig a hole”.

Another reason is that the landscape is so geologically active.

“In Australia, there are fossil reefs that are 200 million years old but which still look like coral reefs; virtually nothing’s happened for that long. You can’t say that about any part of New Zealand.

“There are huge parts of the country, especially in the volcanic North Island, where we have no idea what used to be there, because there are no fossils or even subfossils preserved—everything is buried under at least ten metres of volcanic tuff.

“And though the kauri swamps of Northland preserve kauri wood beautifully, they’re unbelievably acidic, and no bone has ever been found there.”

We do have an excellent record of the past 20,000 years, from cave deposits and middens. Other than that, we have only glimpses of prehistoric New Zealand, tiny slivers in
the timeline.

The St Bathans area in Central Otago reveals a detailed snapshot of New Zealand around 17 million years ago—more than 70 new species, including birds, bats, fish, lizards, molluscs, a turtle, and a mammal have recently been discovered.

And we also know a little about life 60 million years ago, thanks to a handful of giant penguin species discovered in the Waipara Greensand in Canterbury.

But it’s like trying to put together a 500-piece landscape puzzle with only six of the pieces, says Scofield.

“Imagine an old-fashioned jigsaw of the forest, the beach and the sea. We’ve got three pieces from the sea, two pieces from the beach, and one tiny piece from the forest.

“The tiny windows that we have into our past are just unbelievable pieces of chance.”

Even from the era we know most about, the past 20,000 years, the entirety of our knowledge about some species comes from a handful of tiny bones.

When Wilmshurst and Wood first visited the Mt Nicholas caves in 2009 looking for moa coprolites, they scooped up a box of sediment. Going through it later under the microscope, Wood realised it was full of little bones. He picked them out and gave them to the Canterbury Museum.

Several years later, the bones were finally examined. “When the list came back it was quite eye-opening—tons and tons of little birds, a huge diversity of things.”


One of those little birds was a long-billed wren, an incredibly valuable find. The species was previously known only from the remains of six individuals, and this was the seventh. The long-billed wren is a member of the Acanthisittidae, the New Zealand wren family—a group that Wood has been studying through ancient DNA.

The story of the New Zealand wrens was turned upside down—and rendered even more tragic—by the new technology. “The DNA brought up relationships that we never expected,” says Wood.

There were originally six species in the family—today there are two, the rifleman and the rock wren being the only ones to survive the onslaught of mammalian predators. Despite being a small family of tiny birds, New Zealand wrens can reveal a lot about evolution, says Wood.

“About half the bird species on Earth are passerines, or perching birds—there are about 5000-odd species, and they fall into three main groups: the oscines, or song birds, the suboscines, and then there’s the New Zealand wrens.

“They’re a really ancient, distinct group of one of the most abundant bird types in the world. Knowing when the New Zealand wrens split off from the other 5000 species can tell you something about every other group.”

Wood’s DNA studies revealed that Lyall’s wren—also called Stephens Island wren—was not a close relative of the rock wren, but was in fact a distinct, ancient species that had diverged from all the other wrens 30 million years ago. It was a tiny, flightless creature that ran around on the ground like a mouse—but it survived dramatic climate changes, the near-drowning of the entire New Zealand continent during the Oligocene, and the uplift of the Southern Alps. It did not survive the lighthouse keepers’ cats.

According to an apocryphal tale, the entire species was discovered, and then demolished, by a single cat named Tibbles. The real story is a little more complicated, though no less tragic. In the early 1890s, a lighthouse was constructed on Stephens Island/Takapourewa in Cook Strait, and three lighthouse keepers moved onto the island with their families. In the summer of 1894, a pregnant cat escaped, and during the following winter, began bringing small birds home to the settlement.


One of the keepers, David Lyall, was interested in natural history. He saved the tiny carcasses, skinned them, and sent one off to the ornithologist Walter Buller in Wellington. Buller was delighted: “There is probably nothing so refreshing to the soul of a naturalist as the discovery of a new species,” he wrote in 1895 in the Transactions and Proceedings of the New Zealand Institute. “You will readily understand, therefore, how pleased I was at receiving… the skin of a bird from Stephen Island [sic] which was entirely distinct from anything hitherto known.”

Any pleasure in the bird’s existence was bound to have been short-lived—because so was the wren. Just 18 months later, the species was considered extinct.

“This is probably a record performance in the way of extermination,” scolded an 1895 editorial in the Christchurch Press.

The odd bird may have clung on for another year or two, but by 1898, the island was reportedly swarming with cats, and the wren was definitively eliminated.

“When it went extinct, we lost 30 million years of evolution,” says Wood. “No one had expected that before we did the DNA.”

For that project, Wood succeeded in extracting the DNA of three of the four extinct New Zealand wren species. But one, the enigmatic long-billed wren, is still holding onto its secrets.

With so few existing specimens, museums were obviously reluctant to let scientists crush up the bones to extract the DNA. Instead, Wood used a new, non-destructive technique pioneered by Te Papa’s Lara Shepherd—soaking the DNA out of the bones in a special solution. So far, though, it hasn’t worked.

“Looking at them under the microscope, they’re in quite good condition, it’s just that they’re so tiny, there’s such a small amount of material to start with,” he says. “It might just be a matter of waiting another 10 years until we can retrieve amounts of DNA smaller than we currently can—or else finding more bones.”

That’s one of the reasons Wood has returned to the caves at Mt Nicholas. Buried beneath a laughing owl’s perch, perhaps, are more long-billed wren remains. “If we were able to get DNA for it, and fit it into the family tree of wrens, it would tell us a lot more about how they evolved.”


When I arrive at the cave, around lunchtime, Wood—bright eyes shining beneath his googly-eyed frog beanie—shows me the collection of bones the team has already found. He tips them into a dish—a skull, bits of beak, a tiny wishbone.

“This thing here, I’m quite certain now that I see it in the light, is the sternum of a New Zealand quail; it went extinct in the 1870s,” he says. There’s also a minuscule tibia—a bone so small it could only belong to a wren. Could it be a long-billed wren? Wood says he’ll need to compare it to museum specimens to be sure.

Over the next two days, the researchers find more treasures—moa leg bones, vertebrae and tracheal rings, and many more small bones belonging to other extinct birds—parakeet, Finsch’s duck, and wrens, although they’re not sure which.

Before I leave the cave, Wilmshurst points out the spot near the entrance where she and Wood uncovered dozens of moa coprolites in 2009. They spent two days scraping away the surface soil and carefully extracting the layer of ancient poo underneath.

“Three different species parked up here. You can imagine them sort of reversing in out of the weather,” she says, shuffling bum-first into the cave wall and fluffing up imaginary feathers.

In evolutionary time, 500 years is an eyeblink, a heartbeat. Not so long ago, Haast’s eagles and Eyles’s harriers swooped through these skies, laughing owls carried tiny flightless wrens back to this cave, a pair of moa sheltered here from a storm. Under our feet, there are probably hundreds more coprolites, recording what the birds ate, how they lived.Isn’t Wilmshurst tempted to get in there and have a look?

“It’s nice to leave something for another generation of scientists,” she says. “In 30 years’ time they’ll probably have super wands or something—they’ll come in with new technologies that can do all kinds of things we can’t even imagine.”

More by

More by Neil Silverwood

More by Henrietta Harris