Of Mice and Men

When, in 1772, French explorer Marc-Joseph Marion du Fresne sighted the Prince Edward Islands, he believed he had found the mythical Terra Australis (Antarctica). The accidental collision of his two vessels forced him to abort landing attempts.

Instead, sealers were the first to set foot on Marion Island, drawn by colonies of shuffling elephant seals and Cape fur seals that coated the dark, pebbled beaches. The sealers eventually dubbed it Marion after du Fresne to distinguish it from its smaller sibling. (The Dutch sailors who first spotted but mischarted the islands a century earlier were left out in the cold.)

At some point in the early 1800s, mice reached Marion, disembarking from a sealing vessel or carried ashore on shipwreck debris. The tiny European natives may not seem an obvious fit for a sub-Antarctic island. But if albatrosses have thrived by being sub-Antarctic specialists, mice have traveled the world by being generalists.

To escape the cold and almost constant wet, the newcomers followed the islands’ four species of burrowing petrels and headed underground, tunneling into and under cushion plants and other vegetation. They located the entrances away from prevailing winds, creating a habitat 50% warmer at night than aboveground. Gradually, they evolved unique physiological adaptations to the cold: bigger bodies and shorter tails.

Their fate, from the start, was tied to the albatrosses. Invertebrates accounted for most of the mice’s diet, and their favorite grub—the larvae of an endemic flightless moth—thrived in the warmth of albatross nests (despite being hardy enough to survive freezing solid). Seeds, though less palatable, proved a storable supplement in winter.

By the time sealer William Phelps arrived on the island in 1818, the mice had multiplied “until their name was legion.” But in these early years, the island’s wildlife faced a far greater threat from humans than mice.

Phelps and seven companions lived on the island for more than two years. By day, they killed seals for their pelts and oil; at night, they retired to the natural basalt caves. Phelps regarded the island’s multitude of birds with awe and curiosity, but also with a practicality befitting the party’s caterer and chef. He dubbed the coastal plains his “poultry yard.”

By 1850, most of Marion’s seal population had been destroyed. Although hopeful sealers occasionally returned, for 100 years the island’s birds, seals, and mice were mostly left to themselves. That changed in 1947, when South Africa, backed by Britain, annexed the island.

Marion was a sentinel for cold fronts sweeping toward South Africa’s east coast, and meteorological data, officials reasoned, could also inform future air routes between South Africa and Australia.

Unofficially, but more importantly, the annexation was a strategic defense measure. Just a missile’s fire away from the Cape, the islands could make South Africa vulnerable should they fall into Soviet hands. The navy was dispatched to lay claim to Marion and build a meteorological outpost.

Mice quickly overran the island’s new real estate. Sailors dubbed their temporary shelter Mouse Inn after a rodent was found floating in the soup one day. When the first contingent of researchers occupied the completed base, the mice followed.

To deal with their unwanted housemates, researchers brought over a female cat, a castrated tomcat, and three (intact) kittens. These became the founders of a ballooning cat colony. Reports, however, suggest the felines were not especially good at their job. Like Phelps, they tended to visit the “poultry yard” when hungry.

For a while, the cats’ toll on the bird population went mostly unchecked.

A 1965 report concluded the cat population was not large enough to be a threat; in fact, later research suggested the interlopers had driven the common diving petrel to local extinction by that same year. By 1975, the cats’ effects were unmistakable. Ecologist Rudi van Aarde and colleagues from the University of Pretoria found the original five cats had become 2,139, and the population was increasing by 26% annually. The cats were killing an estimated 455,000 birds each year.

“The coastal plains were scattered with wings of petrels killed by the cats,” van Aarde later recalled.

That finding informed a budding plan to tackle the cat problem. Van Aarde and his colleagues now knew what they were up against; at that time, though, no one knew how to eliminate some 3,400 cats on a wild and remote island. Given the daunting topography, they landed on a biological mechanism as the most logical first step.

In 1977, they released 96 cats infected with the highly infectious feline panleukopenia virus around the island. The virus swept through the island’s cats, halving their population in 18 months. Still, the seabirds remained in peril: not a single great-winged petrel chick survived the 1979–1980 breeding season.

The real work of rounding up the cats began from there. Cage and gin traps proved ineffective; researchers tried 20 different kinds of bait, catnip oil, and even jackrabbit distress calls, but nothing could tempt the cats away from the petrels. Poison, solid and gas, was tested and discarded for environmental reasons, as was sterilizer-laced bait. Hunting dogs were sent home with tails between their legs. The felines were devilishly difficult to eradicate.

The team took up arms. Under the leadership of the charismatic mammalogist Marthán Bester, a motley crew of university students, adventure seekers, former soldiers, and seal researchers made Marion their temporary home.

Beginning in 1986, hunters stalked the cats at night with spotlights and shotguns, whittling away at the population over the course of five years. Holdouts were rounded up with traps or killed off with poisoned chicken carcasses. The last cat was captured in July 1991; two years later, the last hunters departed the island. The 19-year effort—launched during the height of apartheid and concluded on the eve of democracy—remains one of the most ambitious projects of its kind.

The Fate of a Remote Island

By the time Marion’s last cat was captured, scientific interest in the ecological impacts of non-native species was growing.

In 1958, Oxford zoologist Charles Elton published a seminal book warning that human travel was facilitating the rapid reshuffling of species globally and that non-native plants and animals could fundamentally damage ecosystems. The consequences of invasive species on isolated oceanic islands, Elton argued, were particularly dramatic. However, it took more than two decades for these and related ideas to crystallize into the field of invasion ecology.

Early researchers in the field recognized that in isolated ecosystems, particularly islands, animals often evolved without entire categories of predators and competitors—such as terrestrial mammals—leaving them ill-prepared when those animals arrived. At the same time, these invading species, now outside their natural range, encountered no co-evolved rivals to keep them in check.

More recent research emphasizes the especially tight interdependencies in island food webs. These interconnections mean invaders can trigger a cascade of destabilizing effects. Seabirds, for example, are among the few agents that transport marine nutrients inland, fertilizing plants and supporting terrestrial communities. Remove them, and those communities will suffer.

Yet the same features that make islands susceptible to disruptive invasions—their small size and distinct boundaries—also make them unusually responsive to lasting interventions. The isolation of islands, though a serious hurdle to eradication projects, means that resident invasive species are easier to target and can’t easily recolonize.

The eradication of cats on Marion Island came to illustrate both sides of that coin.

“In an age where invasion biology is . . . growing in influence and impact, one of the major scientific achievements has been the cat eradication programme,” wrote environmental historian Jane Carruthers in a 2018 review of Pain Forms the Character, a collection of firsthand tales of the cat eradication years.

In 1991, 64% of great-winged petrel chicks survived, and the breeding success of other petrels improved. In 2015, the common diving petrel recolonized the island. The next year, however, researchers revealed that the diminished petrel populations had increased by just 56% between 1979 and 2013—a far cry from the fourfold increase expected. The culprits, scientists concluded, were the mice.

The Rise of the Mice

People often credit the cats for suppressing Marion’s mouse population. However, the island’s sub-Antarctic climate really deserves most of those accolades, says conservationist Anton Wolfaardt. Each year, cold and wet conditions cut the mouse population back by about 90% in winter and suppressed their peak densities in summer.

But recent decades have seen a rapid decline in precipitation, driven by climate change. One study found that annual rainfall dropped by 37%, or around 39 inches, between the 1960s and 1990s. Mice that forage in the rain burn energy just to stay warm; drier evenings mean more capacity for reproduction.

In 2017, researchers determined that as winter nights without rain or snow more than quadrupled between 1960 and 2008, the mice’s breeding season lengthened by about a month. Between the 1979–1980 and 2008–2011 seasons, peak summer mouse populations in the island’s lowlands increased by 430%. That translated to around 1.8 million mice—almost one for every seabird.

As their invertebrate prey dwindled, the mice sought a supplementary food source. The burrowing petrels’ underground nesting behavior essentially brought dinner to their doors. Around the same time, researchers recorded a similar switch in mouse diet on Gough Island, also in the sub-Antarctic.

The mice were slower to turn on wandering albatross chicks, which nest aboveground and are around 400 times a mouse’s size. The first signs of attack came in 2003, when researchers noticed chicks with rump wounds consistent with mouse-inflicted injuries recorded earlier on Gough Island.

By 2015, such attacks were becoming more frequent and fatal, with the island’s other albatross species, the sooty, grey-headed, and light-mantled albatrosses, also targeted. Night footage showed the chicks trying to flap the mice away, before eventually tiring out. The rodents clambered onto their backs, gnawing the backs of their necks and heads where the feathers are thinnest.

In 2023, researchers documented adult wandering albatrosses killed by mice for the first time. While incubating eggs, the parents steadfastly remained on their nests even as house mice swarmed their undersides. At the colony where attacks on adult birds were first reported, the breeding success rate fell to 28%, compared with an island-wide average of 61%.

“Because seabirds generally live so long, adult birds tend to be much more important for a population than chicks,” Wolfaardt says. “And so, if we’ve got mice that are indiscriminately killing both adults and chicks, that will much more rapidly drive those populations to local extinction.”

The attacks on adult birds signaled the severity of the mouse invasion, but as is always the case on Marion Island, much had already been happening beneath the surface.

The flightless moth has almost vanished. And the mice’s preference for large, juicy weevils over smaller specimens has, over many generations, made two weevil species on Marion smaller than their Prince Edward counterparts. All told, invertebrate biomass, which also includes snails, spiders, and earthworms, has collapsed by around 90% since the 1970s. With less insect prey available, the island’s only permanent avian resident, the black-faced sheathbill, is increasingly absent from the shoreline.

Already under pressure from climate change and invasive grasses, indigenous vegetation like the dense cushion Azorella selago—favored by mice for its seeds and suitability for burrows—has also begun to die back, taking tiny critters with it.

The mice are tipping the other end of the scale, too. By eating the insects that break down plant matter, they disrupt the accumulation of rich peat that native plants rely on to thrive. And the daily deliveries of nutrients and energy from the southern seas in the form of guano become less reliable as petrel numbers decline.

The smaller Prince Edward, by comparison, remains relatively pristine, never having contended with invasive mammals. “You go on there, and you just see an absolute abundance of life,” says Wolfaardt. “And you see an impoverished system on Marion.” Even Prince Edward, however, bears effects of climate change and invasive plant species. (The interaction between invasive species and climate change is a focal topic for invasion ecologists today.)

If left unchecked, the knock-on effects of Marion’s growing mouse population are likely to reverberate far beyond its shores. Losing nearly a third of the world’s wandering albatrosses, Wolfaardt says, “doesn’t only impact Marion, it affects that broader sub-Antarctic regional ecosystem.”

A Mouse-Free Marion

The Mouse-Free Marion Project, a partnership between conservation group BirdLife South Africa and the South African government with Wolfaardt at the helm, is determined to finish what the cat eradication teams started. Planners have spent 10 years refining the project, which is expected to cost $32.3 million and aims to rid the island of mice in one fell swoop.

Though a speck in the ocean, Marion’s 115 square miles makes it the largest island conservationists have tried to clear of mice in a single operation. And, unlike the cat eradicators, Wolfaardt’s team does not have the luxury of trial and error. Mice breed so quickly, with females producing around 7 litters of 7 pups each season, that even a single surviving pregnant female could undo the entire effort.

In 2021, an eradication attempt at the smaller Gough Island illustrated how failing to account for any one variable could unravel years of planning.

During the operation, the Gough team realized the island’s nocturnal slugs—another invasive species—had a taste for poison pellets meant to wipe out the mice. Although the team tried to compensate by adding more bait, it was not enough. Six months later, mice were spotted roaming the island. Were a similar oversight to happen on Marion, it could doom the island’s seabirds: given the size of the operation and the costs involved, organizers are skeptical a second chance would get off the ground any time soon.

How can the mouse eradication team expect to reach every last mouse in this expansive, cave-riddled terrain?

Fortunately, eradication knowledge and technology have come a long way since the days of shotguns and poisoned chickens. The Mouse-Free Marion program builds on more than a thousand scientific papers documenting the island’s ecosystem as well as lessons from 700 successful invasive species eradications globally.

Ground operations will be led by Keith Springer, who previously helped eradicate rodents on Macquarie Island and Antipodes Island, remote sub-Antarctic islands south of New Zealand. Trials have shown that mice, unlike cats, readily switch from fresh meat to poison-laced pellets. Based on terrain type, Springer’s team will calibrate bait amounts to mouse densities down to the hectare.

Today’s gold standard rodenticide is brodifacoum, a slow-acting anticoagulant that breaks down naturally over a few months. The poison binds tightly to soil and dissolves poorly in water, which minimizes the likelihood that it will leach into the ocean. Nor does it harm invertebrates, as Gough Island’s insatiable slugs showed. Although seabirds generally aren’t attracted to the pellets, the team is planning to disperse the bait during the winter, when most of the seabirds will be out at sea. Mice also will be fewer and hungrier.

The team has been meticulous in studying the bait’s risks and chances for success.

Like Gough, Marion Island is populated by invasive slugs, although of a different species. In 2022, the team conducted nontoxic bait trials to investigate whether feeding slugs would discourage mice from pursuing the bait; they did not, even when many slugs covered the pellets.

In an April 2025 trial, the team scattered poisoned bait over a nine-hectare grid. To help monitor the mice’s consumption, they treated the pellets with a fluorescent biomarker and laid traps. Every one of the trapped rodents glowed under UV light. The researchers also determined that a new, smaller pellet, designed specifically for mice, could adequately withstand hungry slugs and several days of rain and snow.

The final piece of fine-tuning is planned for 2027. Helicopters will distribute bait across nearly 1,000 hectares of coastal lowlands, rocky slopes, and high-altitude polar desert habitats to test aerial operations and bait uptake.

Though planners have gone to great pains to limit collateral damage, they admit some bird casualties are inevitable, especially among black-faced sheathbills, which stay on the island year-round. And, unsurprisingly, poisoning thousands of animals is not a prospect conservationists relish. Although a 2022 ethics assessment of ways to limit the mice’s impact concluded that eradication was the only viable option, the moral calculus remains uncomfortable.

“We don’t see the mice as the enemy in the way that a lot of people think we do,” Wolfaardt says. “It was humans who facilitated their introduction, accidentally. It’s not the mice’s fault that they’re there.”

Yet, as the 2022 assessment made clear, given the ongoing bird suffering and ecosystem damage, doing nothing is not ethically neutral.

“It would be a dereliction of our responsibility if we said, ‘Well, it’s too difficult or too controversial—let’s just have a hands-off approach and hope that the seabirds somehow manage to survive,’ when we know that they won’t.”

Eradication programs elsewhere show how swiftly island ecosystems can recover once invasive rodents are removed. Marion’s birds will still be faced with climate change, overfishing, accidental capture in fisheries, and pollution. But, Wolfaardt says, by removing the most manageable of the challenges, “you then provide resilience to those populations and ecosystem . . . and allow yourself more time to try and chip away at those other threats.”

If all goes to plan and the rest of the project’s funding is secured, in 2028 or 2029, research vessel S.A. Agulhas II will carry approximately 600 tons of poisoned bait to Marion Island, along with six helicopters, fuel, and enough food to feed a team of technicians, pilots, GPS specialists, and others over a five-month stay.  

During two calm-weather spells—which might take months to materialize—the helicopters will use GPS technology to trace parallel lines above the island, flinging pellets in precise, overlapping circles from motor-driven spinners. A ground-based team will manually bait hard-to-reach mouse hideouts—the nine research station buildings, caves, and lava tunnels. Then, the waiting begins.

After two years of monitoring, they will know definitively whether any mice remain. If the project is successful, flightless moth caterpillars will reclaim albatross nests. Tussock-covered slopes will come alive at dusk with the rattles and coos of petrels. And when an adult wanderer wheels back to Marion after a long foraging trip, it will be to find its chick still waiting below, one of millions of young seabirds readying to fly for the first time.