Keeping as a border crossed by counterfeit — a large blue caterpillar whose surface chemistry reads closely enough as ant brood that the guards themselves carry it into the brood chambers, where it spends most of its life eating the brood it impersonated
R952 closed with a society whose line had stopped being drawn from inside: narrow the recognition cue far enough and almost every stranger passes. This page takes the turn it pointed to, to a nest where the line is still drawn, still patrolled — and is passed anyway, from outside, by something wearing a copy of the signal. The large blue, Phengaris arion (long known as Maculinea arion), does not break into a Myrmica nest. It is carried in, by the nest’s own workers, through the front of the recognition system itself.
First, it begins as an ordinary caterpillar, and almost none of it is built there. In the well-studied British populations the female lays on wild thyme. The young larva feeds on the plant for roughly two to three weeks and stays tiny — on the usual life-table accounting, only a small opening fraction of its final mass is gained on the plant, and the overwhelming share, around ninety-eight percent in that model, is gained underground at the ants’ expense. Then the final-instar larva leaves the plant, drops or falls to the ground, and waits. It does not search for a nest. It waits to be found by a foraging Myrmica worker, and everything after that depends on which worker finds it.
Second, the password is chemical, and it is tuned to one host, not to ants in general. Ant recognition rides largely on cuticular hydrocarbons, the waxy profile carried on the surface. Work on these butterflies and their hosts — Akino and colleagues on chemical mimicry and host specificity, Elmes and colleagues on the hydrocarbon profiles of different Myrmica species — found that the profiles differ between ant species consistently enough to explain why each butterfly has its own host, and that a caterpillar’s surface resembles its host’s brood closely enough for workers to treat it as a larva of theirs and carry it below. The copy is a resemblance that passes a test, not an identity: later genus work found that after adoption a caterpillar’s signature changes further, with host hydrocarbons acquired by contact alongside mimetic ones it synthesises itself — the forgery is touched up from inside, once it is already in.
Third, the wrong password-holder is a death sentence, and the numbers are blunt. A caterpillar can be picked up by whatever Myrmica worker meets it first; adoption is not survival. In the comparative work summarised across the literature, P. arion caterpillars survived at an average of about fifteen percent in nests of its principal host, Myrmica sabuleti, and about two percent in nests of the close relative M. scabrinodis — a near neighbour, chemically and otherwise, and still the wrong address. Even in the right nest mortality is high. The mimicry does not make the caterpillar invisible or safe; it buys it an entry and a probation, in a nest fully capable of killing and eating it if the resemblance stops convincing.
Fourth, inside, this species is a predator, and that word separates it from its famous relatives. Once installed among the brood, the arion caterpillar feeds directly on the ants’ own larvae and brood, tended and guarded by the colony it is eating, for some ten months underground, through the winter. It then pupates inside the nest and the adult crawls out and away the following summer. That is the predatory strategy. Its relatives the cuckoo species — P. alcon and P. rebeli are the studied cases — mostly do not hunt brood at all: workers feed them mouth to mouth, by trophallaxis, as if they were favoured brood. Same genus, same doorway, two different ways of being kept once the password has worked, and this page does not lend either species the other’s method.
Fifth, there is a second password, in sound — and its strongest evidence also belongs chiefly to the cuckoo side. Myrmica queens make stridulatory sounds distinct from workers’. Barbero and colleagues (2009) showed that the caterpillars and pupae of the cuckoo species M. rebeli mimic the queen’s acoustics closely enough that played-back or live signals drew royal treatment from workers — attendance and protection of the kind normally reserved for a queen, which helps explain reports of cuckoo caterpillars being favoured even over real brood. Later recording work comparing a predatory species, M. teleius, with a cuckoo one found larval sounds in both were more queen-like than worker-like, with the workers responding most strongly at different stages in each. So sound is a real second channel in the genus, unevenly demonstrated species by species. It is not, on the evidence used here, a licence to give the predatory arion the cuckoo species’ documented privileges.
Last, the forgery’s precision is why the butterfly vanished from a country that still had thyme and ants. The large blue was lost from Britain in 1979. The mechanism reconstructed by Thomas and colleagues was not a shortage of food plant or of Myrmica in general: small changes in grazing let the turf grow slightly taller and cooler at ground level, the warmth-loving M. sabuleti gave way on former sites to unsuitable congeners, and caterpillars kept being carried into nests whose password theirs did not quite match. Restore the grazing, restore the host ant’s narrow niche, and reintroduction from Swedish stock from 1983 onward worked. A counterfeit this specific is only as good as the one lock it was cut for.
A walk from flower to brood chamber. Five stages: a larva feeding openly on thyme, still almost nothing of its final size; the drop to the ground and the wait for whichever worker arrives first; the chemical resemblance read as “brood” and the guards carrying the stranger in themselves; ten months underground feeding on the brood it impersonated, its survival still hostage to having landed in the one right species; and pupation in the nest, the adult leaving as a creature the password was never written for. Schematic, compressed from the published accounts.
Walk it in order.
Schematic only: the stages compress plant feeding, ground adoption assays, hydrocarbon profiling, in-nest survival work and a national reintroduction history, so no single caterpillar is promised this exact account. Labelled cautions: do not call P. arion a cuckoo species — it is predatory in the nest, feeding on brood itself; the trophallaxis feeding, rescue priority and famine privileges reported in the genus belong to the cuckoo species’ studies and are not transferred to it here. Do not say any Myrmica nest will do — survival in the principal host and in a close congener differ by an order of magnitude in the figures used. Do not call the chemical copy identical, perfect or permanent — it is a sufficient resemblance, sharpened after adoption in genus work by acquisition and synthesis, and it still fails often, including in the right nest. Do not say the caterpillar destroys whole nests as a rule — the claim here is brood predation over about ten months, at high cost to itself. And do not merge the plant, the chemistry, the sound and the conservation history into one study: they are separate programmes, decades apart, that agree on the shape of the life.