Keeping as a workforce stolen before it could work — slave-making ants who carry a host nest’s pupae home, so the workers who emerge have only ever known the raiders’ nest, and do all of its work as if it were their own
R954 closed with a workforce seized in place: a cuckoo bumblebee queen taking a staffed nest’s tenders exactly as she found them, already working. This page takes the turn it pointed to, to raiders who take the tenders one stage earlier, before they exist as workers at all. The slave-making ants — the habit is called dulosis, after the Greek word for a slave — raid the nests of other ants and carry the brood home. What emerges on the far side of that carrying does the raiders’ foraging, feeding and brood care for them. The centre case here is the genus Polyergus, the Amazon ants, obligate slave-makers who can no longer live without a captured workforce. The contrast case, kept separate throughout, is Formica sanguinea, a facultative slave-maker that raids the same way but can also live perfectly well alone.
First, the obligate raider is built for exactly one job, and unbuilt for every other. A Polyergus worker carries thin, sickle-shaped mandibles made for piercing an enemy ant’s body in a fight. They are the wrong tools for digging, for handling brood, and for feeding, and the behaviour has gone with the tools: a Polyergus worker does no colony work apart from raiding, cannot feed itself, and cannot rear young unaided, depending instead on captured Formica workers to feed it mouth to mouth. The classic demonstration is Pierre Huber’s, from 1810, retold in the literature ever since: thirty Amazon ants shut in a box with their own larvae and pupae, soil, and honey lost half their number within two days, apparently starving amid food and young they could not manage, until a single captured Formica worker was put in with them and set about the chamber, the larvae and the emerging young, and the survivors lived. A laboratory contrast makes the same point in reverse: when the captured workers are taken away, Formica sanguinea colonies reorganise within about thirty days and resume feeding and brood care for themselves, while Polyergus workers under the same removal cannot care for their brood and die in large numbers. One of these ants raids by choice; the other raids or dies.
Second, every such colony begins as somebody else’s. A newly mated Polyergus queen cannot found a colony alone, for the same reason her workers cannot keep one: she cannot feed herself or raise a first brood. Instead she enters a host Formica nest — sometimes by riding in with a raiding column and using its confusion to slip inside — and takes it. She is attacked at the entrance; she answers with her piercing mandibles and with a secretion from an enlarged Dufour’s gland in her abdomen, and she seeks out the resident queen and kills her, in the Polyergus breviceps accounts biting and licking at her for an average of about twenty-five minutes first. Once the host queen is dead the nest’s treatment of the intruder changes strikingly: workers that had been attacking begin to groom her, and within hours she is accepted, groomed and fed as the colony’s queen. On what the gland secretion does, this page keeps both readings the literature carries: in the older usurpation accounts it is described as pacifying, an appeasement that calms the defenders; later chemical work on Polyergus rufescens and on the Japanese Polyergus samurai argues the queen’s main compound acts instead as a repellent, holding angry workers off her while she does the killing, with her acceptance afterwards running through the dead queen’s own surface chemistry, which she acquires in the act. Either way, the founding is a usurpation, not a construction: a mature Amazon colony always began life as a colony of the host species, and the queen’s own eggs are thereafter raised by the host workers she inherited.
Third, the raid itself is a scouted, seasonal expedition, not an opportunistic snatch. In summer, on hot late afternoons — the raids are a July and August spectacle, and the raiders are seldom seen outside the nest at other times — scout Polyergus workers search for a suitable host nest. A scout that finds one returns home and rallies the colony, and a column of raiders sets out for the target. At the host nest the attack is as much chemical as physical. In the breviceps accounts the raiders release formic acid and provoke a panic-and-flight response in the defenders, abandoning the brood to be carried off; in the Formica sanguinea group, the raiders’ hypertrophied Dufour’s glands discharge a set of acetates (decyl, dodecyl and tetradecyl) during the fighting that the classic 1971 work named propaganda substances, because the same chemicals rally and attract the slave-makers while scattering the ants defending the nest. The two chemistries belong to different slave-makers and are not merged here. Fighting at the target can be fierce, but adults are not the prize: what the column carries home, ant after ant, is brood — pupae above all, with prepupae and sometimes late-stage larvae — and not the host’s sexual brood. Back in the raiders’ nest, some of the plunder is eaten by the resident captured workers; the rest is kept and reared.
Fourth, the whole trick turns on the stage at which the theft happens. A pupa carried home has never been a worker anywhere. It emerges inside the raiders’ nest, into the raiders’ colony odour, surrounded by raiders and by captured workers of its own species, and it takes up the full round of ordinary ant work there — excavating, foraging, nursing the brood, and feeding the Polyergus themselves, who receive liquid food regurgitated by their captives. It defends that nest, and in Formica sanguinea nests captured workers have even been seen joining later raids and carrying pupae themselves. Nothing about the arrangement needs the captive to be deceived as an adult, or to overcome any loyalty it ever practised: its working life begins after the move, and the only nest it ever works in is the one it woke up in. The numbers show whose colony it functionally is: in a mature Polyergus rufescens nest the captured Formica workers usually outnumber the Amazon ants by at least five to one, and the day-to-day behaviour of the mixed colony closely resembles a large colony of the host species. Host choice, meanwhile, is narrow and, on the current reading, built in rather than learned: the eastern Polyergus lucidus, in Trager’s revision of the genus, keeps Formica incerta alone, with earlier reports of other hosts corrected as misidentifications; its colonies raid only nests of the host species already present in their own nest, and pupae of a different species, on the rare occasions they are brought home, are likely simply to be eaten. The older suggestion that raiders learn their host preference by imprinting has given way to the reading that the preference is an innate, evolved trait of each Polyergus species.
Last, a workforce taken this way can never be replaced from inside, so the theft has to be repeated. The queen’s own offspring are more Polyergus: new raiders, and in season new queens and males. No egg she lays can become a nurse, a forager or a feeder, because her species has traded all of that away for the sickle jaw. Her captured Formica, meanwhile, age and die like any workers, and they leave no replacements of their kind behind them in her nest. So the colony’s keeping runs on a timetable set by other ants’ brood chambers: raid in summer, rear the stolen pupae, spend the workers they become, and raid again. Stop the raids and the colony does not revert to self-sufficiency, as the facultative sanguinea does when its captives are removed; it starves in the midst of its own brood, Huber’s box at colony scale. Keeping here is not a nest’s labour seized once, as the cuckoo bumblebee seized it, nor a border passed once, as the large blue’s caterpillar passed it. It is labour that must be re-stolen every season, at the one stage of life when the stolen cannot yet know whose work they are doing.
A walk from usurpation to a workforce that must be re-stolen. Five stages: a queen who cannot found alone taking a finished host colony for herself; workers shaped for raiding and incapable of every other task; scouts, a summer column, and chemistry that panics or scatters the defence; pupae carried home while the adults flee or fight; and workers emerging into a nest they never chose, working it as their own, until they die and the column has to march again. Schematic, compressed from the published accounts.
Walk it in order.
Schematic only: the stages compress colony-founding observations, raid accounts, gland chemistry and laboratory removal tests across several slave-making species, so no single colony is promised this exact account. Labelled cautions: do not say a Polyergus worker does no work at all — its one work, raiding, is the work the whole colony depends on, and every other task is what it has lost. Do not merge the obligate and the facultative case: Formica sanguinea can found, feed and rear unaided and recovers self-sufficiency when its captives are removed; Polyergus cannot, and that difference is the page’s spine. Do not present the founding queen’s Dufour’s gland secretion as settled appeasement — a repellent reading, with acceptance running through the dead host queen’s acquired chemistry, is also in the literature. Do not lend one slave-maker another’s raid chemistry: the panic-and-flight formic-acid account belongs to Polyergus breviceps, the named propaganda acetates to the Formica sanguinea group. Do not say raids carry off the host’s workforce — they carry off its brood, pupae first, and the adults mostly fight or flee. Do not revive the claim that raiders choose their host by imprinting — for Polyergus lucidus the current reading is an innate specialisation, on Formica incerta alone. And keep the literature’s word “slave” as the literature’s word: recent writers have proposed raiders, pirates and kidnappers for the makers, and helpers for the captured, and no term on this page should be read as claiming the captives consent, rebel in the human sense, or understand their position at all.