Richard Robinson

The future of the world is written in penguin blood

Erect-crested penguins are one of the most mysterious birds on the planet. We have little idea how many there are, what they eat, where they forage, or how their environment may be changing as the Southern Ocean warms. No one has even visited the Bounty Islands where they breed in three years. Scientist Thomas Mattern chartered a yacht and mounted a mission to answer some of these urgent questions before it’s too late.

Written by       Photographed by Richard Robinson

The Bounty Islands are a huddle of 25 rocky outcrops in the Southern Ocean, more than 600 kilometres from the mainland. They were first sighted by Captain Bligh in 1788, who named them for his ship, mutinied just seven months later. He described them as covered in “patches of snow”, which is in fact guano—the output of more than 30,000 erect-crested penguins and 68,000 Salvin’s albatross that nest here.
The Bounty Islands are one of the hottest real estate markets on the planet. Every inch is occupied—the erect-crested penguins distributed evenly between the sparring Salvin’s albatrosses, both species benefiting from the social buffer provided by the other. But while the albatross can fly onto the 40-metre high islands, the penguins must land using a breaking wave then climb, laboriously, to their nest sites. These islands were once home to more than 50,000 fur seals, but they were slaughtered—largely to make top hats—in two years from 1807. Numbers have since recovered to around 20,000.

The Bounty Islands rise from the horizon like dorsal fins. After sailing for two days, we’re 670 kilometres east-south-east of Dunedin, closing in on a scattering of 25 granite rocks so remote that no one has seen them since 2019.

Steve Kafka, master of the 82-foot expedition ketch Evohe, alters course to come up in the lee of the islands. He describes the anchorage as “terrible”.

“I see sheer cliffs,” says scientist Thomas Mattern. “It just looks like something you shouldn’t land on. It looks so hostile.”

Mattern, a tall, ebullient German, has watched with growing dismay for 10 years as the population of yellow-eyed penguins around Southland and Rakiura/Stewart Island withers, falling victim again and again to gill nets. He co-founded and leads the Tawaki Project, focused on the football-sized Fiordland crested penguin, or tawaki, which appears to be thriving.

Why are some penguin populations stable while others are in free fall? Like a canary in a coal mine, penguins are a reflection of the ocean around them, and the ocean is changing in ways that are difficult to predict.

Some of New Zealand’s penguin species live so far from the mainland they are rarely studied. Despite being classified as endangered, we know almost nothing about the erect-crested penguins that breed on the Bounty and Antipodes islands. They are the least-studied penguin in the world. Today, Mattern hopes to correct that, revealing not only the changing circumstances of the penguins, but the changing patterns of life in the Southern Ocean, too.

[Chapter Break]

“Once you’re on the island everything is just birds,” says Mattern. “Look at Spider Island,” he says, waving at a 60-metre-high Tetris block of rock, waves pounding white at its foot. “There are penguins breeding on top of that island. Albatross can fly there, but the penguins have to climb 30, 40 metres of sheer cliffs, almost vertical, to the plateau on top. It’s insane.

“It just goes to show how valuable every inch of real estate is out here. There aren’t many places on Earth that are comparable.”

We gingerly approach Proclamation Island, named for the day New Zealand eventually (and reluctantly) annexed these rocks on July 9, 1870. Captain George Palmer raised a flag aboard HMS Rosario rather than attempting to land, and reported back on the new territory with muted enthusiasm, describing the islands as “rocky islets, perfectly barren, destitute of any covering and exposed on every side to the fury of the sea”. Landing was impractical, he suggested, “even on the calmest summer day”.

There are just two practical landing sites at the Bounty Islands, and this is one of them—scaling the cliff at Proclamation Island. Packs for the gear, sticks to fend off marauding seals, a rope to keep you attached to the rock, and not a small amount of pluck.

On this calm summer day, we survey a ledge in Bucket Cove, trying to figure out if Palmer was right after all. A three-metre swell lifts the rigid inflatable boat up and down the rock wall, sucking at the kelp. The skipper, Murray Watson, strokes his whiskered chin. “I think we can make this work,” he says, leaving room in his tone for failure.

A loose plan develops, requiring the passenger to leap at the peak of a wave and scramble on to the ledge as the boat disappears beneath. From there, you locate a gumboot-size furrow in the wall, and slither, as if up a drainpipe, to the next ledge. Then again, up an easier pitch, to a sort of gutter in the rock that runs across the cliff face, full of seal urine. It’s a useful clue, because here is where you usually meet your first fur seal—the small ones gaze blankly, the large ones snort and charge. Sometimes there’s a couple of seals, other times almost a dozen. You never know your luck.

You’re now headed more or less straight up, about six metres next to a granite escarpment. You grip a rope (rigged earlier by Mattern) and haul yourself hand-over-hand up the incline to a ledge about the width of a couch, and perch there, gathering your wits. The Southern Ocean thrashes 25 metres below. Above, a host of Salvin’s albatross orbit in hypnotic circles.

Now, the traverse of fear, five metres along a polished granite footpath—a wall on one side, a cliff on the other. On good days, like today, the surface feels sticky and safe under gumboots, so you may saunter along the ledge, avoid spooking the albatross nesting immediately around the corner, and step over the oozing seal carcass to reach the relative safety of a rocky slope.

On bad days, wet days, it’s like ice skating, and you rely entirely on a length of rope strung up along the wall.

There are a thousand birds on the slope between you and the dome of the summit, every square metre occupied, if not by a squawking erect-crested penguin, then by a Salvin’s albatross, fulmar prion or fur seal.

Salvin’s are dapper birds. Head silver from the wing, a lemon-coloured beak, charcoal shoulders, a Karen Walker colour scheme.

They sit atop turrets of reconstituted poo and feathers, sometimes other things—one bird has the mummified foot of another bird sticking out the side of its nest, another has requisitioned the pelt of a seal pup, staring back with holes for eyes.

At the summit of Proclamation Island, Otago University scientist Thomas Mattern launches a drone through ‘the albatross layer’ that swarms the island, with Dave Houston acting as an air traffic controller. Mattern (and his daughter Hannah who flew a second drone) took 8,145 photos on 44 white-knuckled flights to map and model the entire island group; data he will use to count every single bird that nests here.

Between each albatross poo-tower are the erect-crested penguins, nesting directly on the ground. They’re about the size and shape of a rugby ball, jet-black on the back, alpine-white on the front, glossy black buttons for eyes—all business, but for a lipstick beak and a straw-coloured tiara above each eye, giving the colony a party atmosphere.

Some perch on their feet bolt upright, beaks tucked back revealing folds of white chins, crests a-quiver in the southerly. Others are propped on their fronts, as if they’ve blown over while awaiting a conversion kick. All view the visitor with passing contempt.

Some are as clean as fresh snow. Others have poo-stained fronts, muddied with the muck that passes for soil on the Bounty Islands—the rotting remains of dead birds, moulted feathers, bird guano, regurgitated food, seal dung, seaweed and freshwater algae. The islands have an overwhelming, stomach-turning stench of ammonia—rich with nutrients and yet starved of the structuring, balancing medium of soil that allows plants to survive. There isn’t any greenery here at all, save a handful of specimens of Cook’s scurvy grass clinging to just two of the 25 rocks.

To reach the summit of Proclamation Island is a bit like the board game Operation, where a player negotiates a metal maze with metal tongs rigged to a loud buzzer. If you climb slowly, maintaining an equal distance between the birds, you get little more than a nod of the head, a clack of a beak, a pair of enormous wings unfurling. But where the gap is too narrow, the bird too easily offended or your route too ambitious, a penguin will let out a scream that could shatter a windshield—wings abruptly cocked back, head thrown in the air and a soaring screech of descending register, like a Formula-1 car passing at full noise. Neighbours immediately join in, an orchestra of outrage that you can’t suppress or escape.

I have never felt so much like an alien, never been so aware of my presence and my influence. After two days on the island, I concluded that this is not a place for people. There is no room for us here. We bring you this story, therefore, in the hope you will never feel the need to visit, and that the Bounty Islands remain unseen by humans, unheard, uninfluenced and undiminished.

[Chapter Break]

At the top, Mattern and his team already have drones in the air, zigzagging back and forth over the island as if mowing a lawn. Every few metres the drone takes a 20-megapixel photograph. These will be jigsawed into an aerial image gigapixels in resolution, including every penguin, every albatross, every rock and crevice of the 50-hectare archipelago.

“Something like this was virtually impossible a few years ago because we didn’t have the technology,” says Mattern over the cry of albatrosses. “Up until 2015 or so, we were just landing on these islands, counting a certain block of the colony, then extrapolating that to the entire archipelago. But our drone data already shows that you can’t do that because the densities differ between these islands. This is a totally new chapter for population science.”

Mattern’s data reveals that the population of erect-crested penguins at the Bounty Islands is relatively stable over time, with about 30,000 birds, most of them breeding pairs. It’s good news.

“It could even mean that the endangered classification is wrong,” he says.

The erect-crested penguin has a lemon-coloured quiff and a call that could shatter a windshield. It is also the least-studied penguin on the planet. The islands are almost devoid of vegetation and soil, the albatrosses making nests out of poo, feathers, sometimes bits of plastic or rope. It’s a hard existence, and those that fail also end up as nest material.

In 1978, a team counted both albatrosses and penguins on Proclamation Island, then extrapolated, concluding there were 115,000 breeding pairs of penguins—an “incredible amount”, says Mattern.

It was that figure that formed the baseline for the threat classification, and subsequent (more realistic) counts therefore appeared to show a dramatic decline. Today’s finding highlights how little we know about the species.

“If they are really endangered, we should know way more about this species so we can develop strategies to get them out of this rut,” says Mattern. “But because it is so difficult to get to these islands, we have basically been sitting on our hands.”

Are these penguins endangered? We still don’t know. What are their greatest threats? Not entirely sure. What is the impact of climate change? We can’t measure it. How deep do they dive? What do they eat? Are they limited by food or habitat? Where do they forage? The scientists shrug.

“We know that they exist, and that’s about it,” says University of Miami PhD candidate Jeff White. “But all that changes this week. This is the last penguin in the world to be studied properly, so all the data we get back will be new to science.”

The research could not come soon enough. The penguins’ foraging environment is changing more rapidly than at any time in their history—including the two years from 1807 which saw the total extermination of fur seals on the Bounty Islands.

Scientists have recorded dramatic declines of other species elsewhere in the subantarctic over the past half-century: At Campbell Island, rockhopper penguins have declined more than 90 per cent, grey-headed albatross at least 82 cent, elephant seals 87 per cent. On South Georgia, the closely related Macaroni penguin decreased from a population of 5.4 million to 2.7 million in the two decades to the mid-1990s. We know that the productivity of the seas declined in that time, and we also know that sea surface temperature increased significantly, but scientists are unable to make a direct connection.

So, if penguin populations are crashing across the Southern Ocean, why are they stable at the Bounty Islands? The anomaly “sends the wheels spinning in my head”, says Mattern.

“From a scientific perspective, an anomaly like that tells you so much more about how species work, how they interact with their environment, and what drives what.”

This science is a matter of life and death, not just for the individual penguins, not really even for the species, but for an entire ecosystem on the brink of epoch-defining change. Erect-crested penguins are a living barometer for everything around them—the albatross, the seals, the troop of life clinging to this rock, and even the ocean itself.

[Chapter Break]

Everything seems pitched against the penguin.

While the albatross soars through the air where molecules are so sparingly associated that they rarely need to beat their wings, the penguin occupies salt water which is 800 times more dense, an environment that envelopes and suffocates, that impedes progress for all but the most streamlined and powerful of bodies.

As a result, penguins are round and muscled, stout and powerful, the geometric solution to the problem of hydrodynamic drag.

They also seem brave. To journey alone, sometimes hundreds of kilometres out to sea, to dive more than a hundred metres into the dark, and return to share your catch with your offspring is a staggering and admirable journey.

It also appears to be getting more demanding.

A raft of erect-crested penguins make the bull-rush to shore on Proclamation Island. They time their approach with a breaking wave, using the energy to hoist themselves on to the rock platform as if riding a liquid escalator. These birds are optimised for the ocean—efficient predators, lithe and fast, a bird-shaped torpedo. On land, however, they waddle like a broken duck. Nevertheless, with claws like crampons and a beak like an ice axe, they can clamber up rock stacks some 40 metres high to roost.

The Southern Ocean isn’t just a monolithic body of immobile water but a complex, swirling and interconnected environment with currents and upwellings transporting food and nutrients to the warmer, poorer waters at the surface, like a series of escalators.

Erect-crested penguins rely on these ocean escalators to deliver the high-protein, high-calorie food that is critical to their survival through the long period of incubation ashore, when they will essentially be starving.

As sea surface temperatures increase as a result of climate change, the food escalators slow down, and there are less nutrients to be found at the surface—less fish, less swarms of krill, less food to bring home to hungry chicks.

The penguins at the Bounty Islands, for some reason, and for the moment, seem to be coping with these changes. But 200 kilometres to the south, on the Antipodes Islands, colonies of both erect-crested and rockhopper penguins are in trouble—the rockhoppers, especially, are declining rapidly.

Is there a difference in the productivity of seas around the two island groups? Or differences in how the oceanic currents are responding to climate change? The answer to these questions may be written in the penguins’ blood.

[Chapter Break]

A northerly swell rises to Proclamation Island, blue and unstoppable, and thunders into its flank, issuing an avalanche of spray.

Erect-crested penguins raft up in a tide line a hundred metres out to sea, bobbing as they orientate themselves to the island, then begin porpoising towards the landing platform—a 45-degree ramp of granite. They might take half a dozen such runs until their arrival on the shore coincides with a breaking wave of enough volume to hoist them on to the rock platform.

The raft, now a flock, preen themselves thoroughly, before embarking on a slow parade up the rocky slope to where White sits motionless on a guano-covered rock.

He takes careful note of their weight, the length of their beaks and feet (indicators of sex) until his eyes settle on his preferred subject. He eases forward, then pounces, with a turn of speed unusual for a PhD candidate, to grip the bird around the abdomen, pinning its thumping wings. A hood goes on and the bird settles as White weighs, samples, needles.

He is leading the ‘blood team’, taking blood and feather samples from 30 erect-crested penguins here on the Bounty Islands, and comparing them with samples of both erect-crested and rockhopper penguins he will be taking over the next four weeks at Antipodes Island.

“Blood tells us the story of what a bird was eating, and when,” he says. “It gives us a picture of their whole world.”

Robin Long, ‘ground truths’ Mattern’s drone data by physically counting every single nest on Proclamation Island—marking each with a daub of spray paint to avoid counting twice. The drone data is indexed against the physical count to check its accuracy.

Analysis of ‘stable isotopes’ in whole blood provides a time window of 20–30 days, but break a blood sample into its constituent parts with a centrifuge and you get two sets of data—red blood cells represent the feeding activity of the bird two weeks ago, and plasma two days ago.

Stable isotopes of carbon-13 can tell a scientist where a bird was feeding, and nitrogen-15 reveals which trophic table they’re feeding at—whether they’re eating fish or invertebrates or subsisting on jelly-like salps. Combine samples from penguins, their prey and the immediate environment and you have a picture of the entire food web.

In this way, White can answer the what and when of the penguin’s diet. Add tracking data from GPS tags and he can tell you where, too. (There has never been any tracking of erect-crested penguins, something that Mattern hopes to address next year.)

The data offers a near-perfect picture of the present, but because erect-crested penguins haven’t really been scientifically studied before, we have no idea of how their environment or behaviour is changing, if at all. That mystery might lie in their feathers.

[Chapter Break]

Penguin feathers are weird. Rather than growing outwards over time like most birds’ feathers, penguins experience a ‘catastrophic moult’, losing all their feathers and growing new ones at the same time. They can’t feed during this period, and rely on accumulated body reserves. As a result, the new feathers are an exact reflection of the diet of the penguin in the critical period two months before the moult.

Feathers on specimens stored in museums may be able to tell us what the birds were eating a century or two ago—data scientists desperately need to understand changes in the ocean’s productivity and the escalator system of food distribution.

In 2009, Mattern used the same technique to show that the diet of Snares crested penguins hadn’t changed in 135 years. This indicated stability not only in the marine ecosystem of The Snares (islands south of Rakiura/Stewart Island), but in the prey species the penguins rely on.


Not so the rockhopper population—the smaller cousins of the erect-crested and Snares crested penguins. Sixteen years ago, scientists from the UK discovered that the seas in the areas that rockhopper penguins were feeding had decreased significantly in productivity. In response, the rockhoppers had been forced to switch from a diet dominated by fish to lower value food sources such as krill.

It’s all red flags for Mattern. “Humans are changing the ocean in a massive way,” he says. “Heating up the ocean screws up primary production and all the subsequent food webs.”

As well as changing the depth of food sources, the warming sea surface may also be shunting things sideways, changing where the warm northern currents meet the colder, richer upwellings. The air is warming, too, which may affect breeding success on land, and more violent storms may alter the viability of exposed nesting sites. The questions are piling up, even while Mattern is scrambling to understand what ‘normal’ looks like.

His science has turned into a race, with global consequences. The set of circumstances in New Zealand, with populations of penguins both around the mainland and far out to sea, provide the experiment-and-control mechanism science needs to provide clear evidence.

“From a global perspective, New Zealand can explain a lot to the world about how ocean ecosystems work. It’s just a pity that we don’t do more of this sort of science.”

[Chapter Break]

When Mattern finished school in 1993, he and a friend bought round-the-world tickets. He ended up in New Zealand, “almost broke”, and spent the last of his money on an annual hut pass and a supply of two-minute noodles. He was sitting in the marram grass at Long Harry Bay on Rakiura/Stewart Island, hoping to spot kiwi, when he witnessed a miracle. A group of yellow-eyed penguins tottered down the sand, slipped into the waters of Foveaux Strait and disappeared from view.

“And I went like, ‘What? They don’t come up?’” Mattern makes an exploding head gesture. “Why on earth would a bird give up flying and learn to hold its breath for prolonged periods of time? What’s the deal there?”

It was the first time he had seen a penguin in the wild, and he was transfixed.

“That thought just wouldn’t let me go. And when I pursued that, I realised these penguins can tell us a lot about the oceans. It was an epiphany for me.”

Meanwhile, scientist Jeff White (right) takes blood from a penguin which will reveal what the bird was feeding on, and when. Klemens Pütz, whose organisation Antarctic Research Trust funds the five-year research project, keeps the penguin blindfolded and calm.

The Southern Ocean completely encircles the planet—connecting the Atlantic, Pacific and Indian ocean basins—and is uniquely important in driving global ocean circulation and climate. Despite covering just 15 per cent of the world’s total area of ocean, it absorbs more heat due to climate change than any other ocean on the planet. Understanding how it works has become critical to climate science, and climate adaptation.

“You wouldn’t try to repair a car without understanding how an engine works, right? We can only see what we do to our natural world if we understand how that natural word functions,” says Mattern.

Advances in micro-computing, GPS technology and the ‘internet of things’ that allows your refrigerator to add milk to your shopping list has also allowed for smaller, lighter tracking devices. We’re living in a new era of surveillance science, so-called ‘biologging’.

When a penguin scientist waits four years to sail 670 kilometres for their first glimpse of the world’s least-studied penguin, they don’t leave any tools at home. Mattern has GPS loggers, temperature and depth recorders, and a network of automatic timelapse cameras to be deployed within the colonies to record the passage of life on these remote rocks in greater detail than in most urban centres.

He even has a penguin camera, about the size of a matchbox, which can record high-definition Netflix-ready video for three hours while attached to the back of a penguin at a depth of 160 metres. He designed it himself.

Anchorage Bay, 1998
Anchorage Bay, 2022. The ‘colony’, pictured here in 2022, now numbers just half a dozen birds. Dave Houston’s 1998 photo at the same site (above) reveals the scale of the decline in the past 14 years.

Over the next four years, Mattern will peer into their private lives, sap their blood, analyse its chemistry and sequence its DNA. The team will inject subcutaneous transponders (like the microchip in your cat) into every adult and fledgling they handle and may also construct some automatic transponder gates that will read the birds as they pass by, like a roasted chook going through a supermarket checkout. Boop!

The work programme for the Bounty Islands alone is staggeringly ambitious. It’s barely possible to land on these islands, and yet Mattern is determined to aerial-survey every millimetre of the 50 hectares, count every penguin, albatross, shag and prion; capture, tag, test and juice 30 adult penguins; recover GPS tags from half a dozen Salvin’s albatrosses (which need to be found first); physically count every penguin nest on Proclamation Island to ground-truth the drone data, and install the timelapse cameras.

After four days of six, the science team has done less than half of the work, largely because a southerly swell and northerly gale put the expedition to sea for two days—sailing in circles waiting for the breeze to abate. Mattern looks hollow.

By midday on day five, however, the sea has dropped sufficiently for a small team to get ashore, and they lean into their work—counters out counting, crewmembers mounting trail cameras, the ‘blood team’ rustling up donors. Mattern is determined to drone-survey distant Centre Group, 1.5 kilometres away, despite the stiff breeze.

By the end of the afternoon, they are all spreadeagled on the summit, bathed in sweat.

Mattern is giddy with joy. In two frantic days, he has made good on four and a half years of planning and permit-writing. The final flight of his mechanical bird, a cannonball run over kilometres of Southern Ocean to take 700 photographs, has been a triumph.

“I died a thousand deaths,” he says. But the plastic pigeon has returned, with gigabytes of data stored on a millimetre-thick wafer of silicon the size of a fingernail. It seems hard to believe that it might hold the answer to some of the biggest questions about this species and the vast Southern Ocean stretching out before us now, silver and quixotic.

[Chapter Break]

A day’s sail south we anchor beneath a wall of rock forming the north-eastern tip of Antipodes Island. Unlike the rocky Bountys, the ‘Antips’ are clothed in silver tussock, with verdant tufts of megaherbs and swathes of fern, mosses, lichen—a 2000-hectare bonsai garden.

A palisade of basalt columns describe the landing at Anchorage Bay. Elephant seals the length of a station wagon are parked up, snorting.

In four muddy patches, each the size of a volleyball court, hundreds of erect-crested penguins are arrayed about a metre apart. In one corner, tucked up under a sandstone cliff, there are four rockhopper penguins, all that remains of a colony of 300 once found in this bay.

Dave Houston was last here in 1998, and sucks some air through his teeth when I ask him how the population of the colony compares today. “About half,” he says, wincing at the sound of his own calculation. He thumbs through images on his phone until he finds pictures he shot 24 years ago, carefully scanned from film and archived for posterity. We swivel around until we can match up caves, rocks, features of the bay. This is spitballing, not science, but it confirms his first impressions—the historical photos show about double the number of individuals.

“It’s not what I expected here. The decline is real,” says Mattern. “We knew that rockhopper penguins were declining, but we hoped it wouldn’t be true for erect-crested penguins, too.”

An eastern rockhopper penguin eyeballs the camera at Anchorage Bay on Antipodes Island. Mattern hopes that the differences between rockhopper and erect-crested penguin populations, and between the Antipodes and Bounty Island locations, will reveal the underlying reasons for the decrease. “It’s important to take a look at these remote places,” he says, “so we have an idea of what’s going on, what we are doing to our oceans.”

Mattern points out the poo around the rockhopper nests. “It’s green,” he says. “From bile. It’s because they have empty stomachs now. They’re starving, and they’re waiting desperately for their mate to return so they can go out and feed.”

This isn’t unusual—it’s the same stress that this species endures year on year. The difference is that for many years their mate simply hasn’t returned, and the ‘colony’ of rockhopper penguins here now numbers just four nests.

Is the thermocline now too deep for the rockhoppers to reach food? Are the erect-crested penguins declining because they’re starving, too? Have the upwellings of nutrients and food remained within reach of Bounty Island birds but shifted out of range of those that nest at the Antipodes?

Over the next month, Mattern and his team use drones to survey every penguin colony on the island in a desperate bid for answers. White weighs, samples, needles. The results come back by email from the island, a tragedy unfolding in weekly episodes—penguins guarding dead chicks, their mates lost at sea. The surveys show many colonies have shrunk in size, some have been abandoned, not a penguin to be seen. Though it all needs more analysis, it’s a grim picture.

Mattern’s fear is that, in a changing ocean, the penguins on the Antipodes are simply the first to fall. The ecosystem appears to be failing the penguins here, which makes it crucial to understand why the population at the Bounties appears to be stable.

“We live in one of the few countries on this planet that have the privilege to be guardians of places like this,” he told me, as we stood on the granite turret of Proclamation Island, albatrosses whirling over our heads. “This is nature at its purest. We should do everything in our power to keep it that way.”


New Zealand Geographic is grateful to Live Ocean Foundation and Whakatupu Aotearoa that contributed to the production of this feature. The five-year science programme is funded by the Antarctic Research Trust, with support from Vontobel Foundation.

A colony clings to the volcanic slope of Bollons Island, on the north coast of the Antipodes. There are 81 such colonies at the Antipodes, and many have decreased in extent. Some 47 colonies have disappeared altogether. Scientists hope that it is consistent with a population that is limited by resources, and when resources rebound, so will the birds. So far, however, no rebound.