Keeping sabotaged from inside — captives who tend their captors’ larvae faithfully, then, once those larvae pupate and a chemical signature gives them away, neglect the pupae or tear them apart, so the keeping fails at the last stage, from the workforce itself
R955 closed with a theft timed so recognition could never happen: Polyergus carrying host pupae home, so the workers who emerged had never known another nest and the question of whose brood they were tending could never be asked. This page takes the turn it pointed to, to a system where that question is asked — one stage too late for the captives to leave, but in time for them to strike. The parasite is the small North American slave-maker the older literature calls Protomognathus americanus, now placed in its hosts’ own genus as Temnothorax americanus. The captives are workers of one of its hosts, Temnothorax longispinosus. The behaviour the Foitzik group named, in their own term, slave rebellion is the spine of this page, and the word is kept here as the literature’s word, with no claim made that the workers understand, consent, or resist in any human sense.
First, the parasite is tiny, old, and wholly dependent. A Temnothorax americanus colony lives in a hollow acorn, a nutshell or a twig in the north-eastern North American forest. It is a queen, a handful of her own workers — two to five in the usual account — and thirty to sixty captive host workers. The slave-maker, in the Mainz group’s summary, neither rears its own offspring nor forages for itself: the captives feed and clean its brood, bring in food, feed the slave-makers themselves, and defend the nest. A colony is founded when a slave-maker queen invades a host colony and kills or drives out its adults; workers emerging from the host brood that remains are then, in Achenbach and Foitzik’s framing, tricked into caring for parasite brood instead. Later raids renew the workforce the same way R955’s raiders renewed theirs, by carrying host brood home. So the set-up is familiar: keeping by theft at the brood stage. What happens inside the stolen nest is not.
Second, for as long as the parasite brood are larvae, the captives are faithful nurses. The enslaved workers feed and clean the slave-maker’s larvae exactly as they would their own colony’s, and in the work the Mainz group summarised in 2012, about 95 per cent of that brood survives the larval stage. Foitzik’s stated reading — an interpretation, kept as one here — is that at this stage the captives probably cannot tell that the larvae belong to another species at all. A larva is a soft, undifferentiated charge: fed, cleaned, carried. Nothing in the record on this page says the larvae are recognised and deliberately spared; the faithfulness reads, on the researchers’ own account, as a failure to tell, not a decision.
Third, pupation changes what there is to tell. A pupa already looks like an ant. It is also, in this system, at its most defenceless: it cannot move, cannot fight, and is not protected by a cocoon. And it carries, on its cuticle, chemical cues that the researchers report the captive workers can apparently detect as foreign. At that point the care stops and reverses. Captive workers neglect the parasite pupae, leaving them untended to die; they carry pupae out of the nest; and they attack them directly, several workers at once tearing a single pupa apart. Achenbach and Foitzik’s 2009 study, the first evidence published under the rebellion name, found the enslaved workers killing about two-thirds of the parasite’s female pupae. The timing is the whole point: recognition does not arrive before capture, or during the long larval tending. It arrives exactly once, at the pupal stage, against the brood of the other side — the stage at which, in R955’s system, recognition was supposed to be permanently forestalled because the stolen had emerged already home.
Fourth, the killing is widespread, but its strength is local. Pamminger and colleagues, in the 2012 geographic study, found the same neglect and killing in host populations in three states, with slave-maker pupal survival inside parasite nests of about 27 per cent in West Virginia, 49 per cent in New York and 58 per cent in Ohio — an average of roughly 45 per cent surviving — against about 85 per cent survival for host pupae in their own free-living nests. Those two sets of numbers are not merged here: the 2009 figure is a share of female pupae killed in one study; the 2012 figures are pupal survival across populations, a different measure in a different design. The geographic spread is itself part of the finding. The group reads it through the geographic mosaic of coevolution: local selection differs, so local defences differ. In their account, New York hosts are aggressive at the nest door and often thwart raids outright, while West Virginia hosts gain more from the rebellion route — genetic analyses showing that the free colonies near a parasite nest there are more often close relatives of the captives inside it. A later Mainz study added a turn at the door itself: whether defenders react aggressively depends in part on where the raiders come from, hosts responding to raiders from places where slave-makers are rare and failing to recognise raiders from places where they are common, with the difference reflected in aggression-related gene activity in the defenders’ brains. Defence, on this record, is not one trait. It is a local answer, and it differs by population and by stage.
Last, the killers gain nothing for themselves, and that is what makes the behaviour explicable at all. An enslaved worker will never reproduce; killing the parasite’s pupae does not free her, feed her, or return her to her own colony. The benefit, on the researchers’ account, lands elsewhere: a parasite colony whose pupae are destroyed grows more slowly and stays smaller, and a smaller slave-maker colony mounts fewer and less destructive raids on the free host colonies around it — colonies that may be the captives’ own kin, their sisters’ nests. That indirect route is also the answer to the older objection, recorded in the 2009 paper, that a trait of already-enslaved workers could never be selected for because the enslaved are an evolutionary dead end: selection can act there if the population is structured closely enough that harming the parasite reliably helps relatives nearby. Achenbach and Foitzik drew the parallel themselves: in birds parasitised by cuckoos, egg recognition can be defeated by mimicry and a second line of defence evolves at the chick stage. Here the larva is the stage that gets past, and the pupa is the stage that is finally told apart. Keeping, in this nest, is undone not at its border and not by escape, but by its own workforce, at the last stage before the kept brood would have become keepers in turn.
A walk from faithful nursing to sabotage. Five stages: a workforce stolen as brood into a parasite nest it will work for life; larvae tended so faithfully that almost all of them live; a pupal cuticle that finally carries a detectable foreign signature; neglect, removal and group killing of the helpless pupae; and a benefit that lands not on the killers but on their free kin nearby. Schematic, compressed from the published accounts.
Walk it in order.
Schematic only: the stages compress a founding account, a 2009 laboratory finding, a 2012 three-state geographic study, and later work on defence at the nest door, so no single colony is promised this exact sequence. Labelled cautions: do not present rebellion as conscious resistance, refusal understood as such, or a bid for freedom — the killers do not leave, are not freed, and gain no direct benefit; the literature’s word is kept as its word. Do not say the captives recognise the larvae and spare them — the researchers’ reading is that they probably cannot tell the larvae apart, and the faithfulness is a failure of recognition, not mercy. Do not merge the 2009 and 2012 numbers: about two-thirds of female pupae killed is one study’s measure; 27, 49 and 58 per cent pupal survival in West Virginia, New York and Ohio parasite nests is another’s, against about 85 per cent for host pupae in free nests, not against parasite pupae tended by their own. Do not present the geographic differences as noise — on the group’s reading they are the finding, defences differing locally by stage and by kin structure. Do not revive the dead-end objection as settled in either direction without its condition: the indirect-benefit account depends on nearby free colonies being kin, which held in the studied populations and is not promised everywhere. And do not swap the parasite’s names carelessly: Protomognathus americanus is the name in the rebellion papers; Temnothorax americanus is its current placing, in the same genus as its hosts.