On Squirrels

Start with the fact that American cities had no squirrels. By the middle of the nineteenth century they had been hunted out of urban areas, and the animal that now defines a park was put there on purpose. The first documented introduction was Philadelphia’s Franklin Square in 1847, three squirrels with food and nest boxes; Boston Common followed in 1855 and New Haven Green by the early 1860s. The large releases came with the Olmsted-era parks of the 1870s, and Central Park had about 1,500 squirrels by the mid-1880s^[4]. The historian Etienne Benson, who traced this, found the reasoning in the record: the squirrel was “living proof of the rewards of extending charity and community beyond the bounds of humanity,” and feeding it was adopted as a way of encouraging humane behavior in children. People had to stop shooting squirrels and start feeding them^[4][5].

Britain imported the ornament. The first verifiable release was a pair let go in Henbury Park, Cheshire, by Thomas Brocklehurst in 1876. In 1890 ten squirrels from New Jersey went to Woburn Park, and from Woburn the 11th Duke of Bedford gifted them around the country for thirty years, including a release in Regent’s Park that likely seeded London; 100 more went into Richmond Park in 1902, and the last documented release was Needwood Forest in 1929^[6][7]. A 2016 genetic study of 381 squirrels from fourteen sites confirmed that the British population descends from a handful of deliberate human moves rather than from the animal spreading on its own: Aberdeen’s grays are closest kin to the New Forest’s, four hundred miles away^[7]. Just over seventy percent of the successful introductions began with fewer than ten animals^[6].

Ireland got a wedding present. In the summer of 1911 a wicker basket was opened on the lawn of Castle Forbes, County Longford, and about a dozen grays from the Woburn stock jumped out^[8][6]. Cecil Rhodes put the first batch on his estate at Groote Schuur, Cape Town, soon after 1900, and a 1957 bounty of sixpence a tail paid out two guineas in total^[9]. Italy’s came to Piedmont in 1948; when a trial eradication began at Racconigi in 1997 and the early results showed it was feasible, animal-rights groups sued, the biologists were acquitted three years later, and by then the window had closed^[10]. Only Australia reversed it: Melbourne’s colony, introduced around 1900, was gone by 1973^[11].

The numbers first. Britain had around 2.5 million red squirrels a century ago and has about 287,000 now, three quarters of them in Scotland; England is down to roughly 15,000. Grays, from zero, number at least 2.7 million^[2][12]. The easy explanation is size, and size matters: a British gray averages around 550 g against a red’s 280 to 300, and grays pilfer red caches so effectively that reds in shared woods go into spring lighter and breed less^[2][13]. But competition alone is slow. What made it fast is a virus.

Squirrelpox is carried by grays without symptoms. A 2000 survey found antibodies in 61 percent of 223 apparently healthy grays and in 2.86 percent of reds, four animals out of 140, which is the signature of a reservoir host^[14]. In reds the disease is almost invariably lethal, with death typically within two weeks of symptoms; an outbreak at Formby in 2007 killed an estimated 85 percent of the colony^[15]. Stephen Rushton’s group, modelling a decade of Cumbrian data, found red declines “17–25 times higher in regions where SQPV is present in grey squirrel populations,” and that stopping the decline by culling would need a kill rate above 60 percent^[16]. The gray does not have to outfight the red. It has to stand near it.

Grays strip bark. In 2021 the Royal Forestry Society commissioned a model of what that costs in England and Wales, assuming damage on 15 percent of broadleaf and 5 percent of conifer area: £37 million a year, £1.1 billion over forty years, more than 90 percent of it on broadleaves, with sycamore taking 40 percent of the damage, beech 26, birch 9 and oak 6^[18]. The figure is a model, not a count, and the report says so. It is also the number that made the Treasury care, which is why the contraceptive got funded.

The UK Squirrel Accord is a partnership of 48 forestry and conservation bodies, government agencies and companies. Since 2017 it has paid the Animal and Plant Health Agency about £1 million, under Giovanna Massei, to turn an injectable immunocontraceptive vaccine into something a wild squirrel will eat^[3][19]. The vaccine works on the immune system to stop the body producing sex hormones, making both sexes infertile; the active ingredient is sealed in plant-spore capsules, the same delivery used for vitamin D in human trials, so it survives the gut^[3]. The bait is hazelnut butter. The feeder is the clever part: a hopper that only grays can open, one design with a weighing platform that unlocks the door only for an animal inside a set weight range. Reds are too light^[3].

Everything published so far uses a dye instead of the drug, to measure who eats. In six trials across five English woods, 271 squirrels were trapped and tagged and 151 of them, 56 percent, turned up dye-positive, spring highest and summer lowest; no other animal got into the hoppers^[20]. A parallel study in three woods found 49 percent, with sex, age and breeding status making no difference to who ate^[21]. Earlier modelling had used a 75 percent efficacy threshold, and the preliminary finding is that at least 60 percent of animals given the oral dose could become infertile^[22]. In 2025 the Accord announced the formula had produced infertility in rats by free feeding and “a reaction” in grays, with field-scale trials, registration and rollout still ahead; its chair called it the world’s first oral free-feeding immunocontraceptive^[3]. A Clemson trial of the injectable version in 2009 shrank testes and raised antibody levels but also left abscesses at the injection site, which is the argument for a spread you eat^[23].

There is one place the gray has lost on its own. The pine marten, a native predator trapped nearly to extinction in Britain and Ireland, has been recovering, and Emma Sheehy’s four-year study of the Irish midlands found that “where pine martens had returned to healthy numbers, grey squirrels had all but disappeared,” and reds were recolonising woods that had been gray for thirty years^[26]. The 2018 follow-up in Scotland, 223 feeders across the Borders, Central Scotland and the Highlands, put it formally: gray occupancy “strongly negatively affected by exposure to pine martens,” with an indirect positive effect on reds^[27]. The mechanism is history. Reds co-evolved with the marten over tens of thousands of years and stay off the ground; grays are heavier, less agile, feed on the forest floor and have never met one^[26]. A 2020 all-Ireland survey found grays had all but disappeared from ten midland counties^[28].

Where there is no marten, people have done it by hand once. On Anglesey, grays arrived in the mid-1960s, and by 1998 fewer than 40 reds remained, all of one female bloodline, among some 4,000 grays. A trapping program began that year, reached the coastal boundary in 2013, and the island was declared gray-free in 2015 with around 700 reds^[29][30]. Seropositivity in the island’s remaining grays fell from 75 percent to 4 percent as the cull proceeded^[30]. Anglesey is 710 square kilometres with a strait around it. England is not.

It is worth saying what the gray is good at, since everything above is about what it does to others. It is a scatter-hoarder, burying several thousand caches a season one nut at a time, and it finds them again: hand-raised squirrels tested at two, four and twelve days dug up significantly more of their own caches than of strangers’^[32][33]. Recovery runs from 95 percent in lean years down to 26 percent in heavy mast years, when there is no need to bother^[32]. And it lies. Michael Steele’s group at Wilkes University, in more than 500 trials, watched grays dig a hole, mime burying, and walk off with the nut still in their mouths; the fake burials ran about 13 percent of caching events, rose when another squirrel was within twenty metres, and doubled after the researchers deliberately stole caches. Human observers found the acorn one time in ten when the deception was used^[34]. An animal that keeps a mental map of thousands of points and salts it with decoys is not a pest in any interesting sense. It is a competitor that happens to be winning.

The stories agree on one thing: the squirrel is a go-between, and not a trustworthy one. In the Norse Eddas, Ratatoskr runs the trunk of the world-tree Yggdrasil carrying words between the eagle at the top and the dragon Nidhogg at the roots; the prose version says plainly that he “tells slanderous gossip, provoking the eagle and Nidhogg”^[35]. In the Cherokee story of the ball game between the birds and the animals, recorded by James Mooney in 1902, the birds take a small creature and stretch its skin until it becomes Tewa, the flying squirrel, who catches the ball on almost the first toss and carries it up a tree^[36]. In Wabanaki telling the squirrel was once a giant until Glooscap shrank it; in a Choctaw one a black squirrel eats the sun, which is an eclipse^[37]. The Victorian version is Beatrix Potter’s Squirrel Nutkin, 1903, a red, impertinent to an owl on an island in Derwentwater, who loses his tail for it^[38].

The law is its own lore. Britain’s Grey Squirrels (Prohibition of Importation and Keeping) Order of 1937 made it an offence to import or keep “any animal of the species designated Sciurus Carolinensis,” eight years after the last release; a 2015 deregulation dropped the duty on landowners to report them, on the stated ground that eradication was “neither feasible nor widely supported”^[24]. Since December 2019 a rescued gray cannot legally be released back into the wild, which puts wildlife hospitals in the position of euthanising an animal they have just saved^[39]. The epithet “tree rat” is the everyday version of the same thing, and Jason Gilchrist’s defence of the gray in 2017 notes that it outnumbers the red seventeen to one and that tens of thousands are culled a year^[40]. Benson’s point lands here: a century ago feeding the squirrel was a lesson in kindness, and by the late twentieth century it had become “almost a crime or a sin”^[4]. Same animal.

The gray squirrel is guilty of nothing except being carried. Every colony on the timeline was a gift, a civic project or a rich man’s estate, and the people who made them were acting on the best idea of their time, that a tame wild animal in a park teaches children to be gentle. The idea was right and the ecology was wrong, and the bill came due in a country the animal never asked to visit, paid by a smaller red squirrel that dies of a virus the gray does not notice. The honest accounting runs through the whole piece: a modelled £37 million, a measured 61 percent, a measured 56 percent, a declared island.

What I find myself respecting about the hazelnut program is its temper. For ninety years the only tools were a trap, a gun and a law, and they worked on an island with a strait around it and nowhere else. The alternative is a feeder that weighs the animal, lets it in, and gives it something it wants to eat, and the whole bet is that half the squirrels eating is enough. It may not be: half is not three quarters, and nobody has yet run the drug at scale. But it is the first fix in a century and a half that does not require anyone to hate the squirrel, and in Chicago, where I watch the same animal bury a walnut under the same tree twice and pretend the second time, that seems like the only version of this story I could stand to tell at the end.