Keeping seen from the carrier’s side — a free-living ant of 1798 that keeps its own colony on its own ground, lends its brood to a slave-maker in roughly three quarters of that parasite’s assembled host records, and in 2024 took five other names back into itself
The last page was a species that borrows everything. Lift that borrowed map away and the thing underneath it is this: Temnothorax unifasciatus (Latreille, 1798), the slave-maker’s main host. Ask what the carrier holds when it is counted not as someone else’s range, but for its own sake.
First, it is old, wide, and marked by one band. Described in 1798 and carried for a time in Leptothorax before its present genus, it is the small brownish-yellow worker with one clear, continuous dark band across the first segment of the gaster — the one band its name says. Its range runs across southern, western and central Europe and into parts of the east — in the 2024 revision’s words, pan-European, from the Iberian Peninsula to the Caucasus — with records also noted from Morocco and southern Sweden, and a presence in Sicily’s recent checklists where it stands unremarked among the island’s Temnothorax. In Britain it is essentially a Channel Islands species, apart from a single London record treated as a probable import.
Second, its own colony is ordinary in the best sense: its own. It nests in warm, dry, open places — rock crevices, under stones, in dead branches, under bark, in hollow stems. A nest holds a single queen and typically two hundred or more workers. Winged sexuals appear in July and August, and the mating flights fall in the hours around sunrise. Nothing here is borrowed: the ground is chosen, the workforce is raised at home, the queen is its own. This is the baseline the last four pages kept departing from, stated plainly because the sequence had nearly forgotten it exists.
Third, it became a laboratory animal for its own behaviour. Because its colonies are small and keepable, unifasciatus is one of the ants in which social behaviour itself has been measured: dying workers leaving the nest hours or days before death, reducing the chance of infecting kin; workers policing one another’s egg-laying, with the policing’s effectiveness shown to vary with relatedness and colony size. A host that common is counted, yes — but often as an instrument for asking questions about ants in general, which is its own kind of being used.
Fourth, the susceptibility is part of its biology, not an accident of the record. Comparative work on hosts and non-hosts found that species liable to enslavement are shaped more strongly by experience very early in life, when nestmate recognition by odour is forming — the opening a slave-maker’s brood exploits, emerging in the host nest and thereafter treated as kin. So the three-quarters share in muellerianus’s host tally rests on two plain facts laid together: this species is widespread and abundant enough to be found everywhere the raider looks, and its young are built to learn who belongs in a way that can be taught wrong.
Fifth, in 2024 it was tested — and it absorbed. An integrative revision of the unifasciatus complex combined molecular phylogenetics, species-delimitation analysis and multivariate morphometry across its whole range. The trees disagreed with one another; the delimitations inside the complex came back weak and ambiguous; the authors concluded the variation was population structure inside one broad species rather than several species wearing one name. Five names were newly sunk into it as junior synonyms, a lectotype was designated to fix the name to a specimen, and only one neighbour split away cleanly (see the opening below). The revision’s own title calls what came before inflated biodiversity. Where lagrecai was proved and lost ground, and marae was born proved, unifasciatus was proved and gained names — the test ran the other way, and the wide name got wider on paper while the ant, presumably, noticed nothing.
And so, counted for itself, it holds almost everything the parasite lacked. Own ground, own labour, own queen, a name fixed to a lectotype, a verdict of one species delivered in its own right in 2024. What it does not appear to hold, in the sources checked for this page, is a conservation assessment of its own — the parasite that lives off it carries a Vulnerable verdict from 1996, however stale and misfiled, while the carrier that makes that verdict’s map possible has, so far as these sources show, never needed one. Commonness is not a verdict; it is the absence of anyone filing one.
A walk, schematic only — five stages, tap each one:
Described in 1798; one clear dark band across the first gastral segment gives the name. Pan-European, Iberia to the Caucasus in the 2024 revision’s account.
Five keepings, then, in a row: a name handed an island on loan, a name handed nothing and proved, a name born proved and never assessed, a name that keeps nothing of its own, and now the carrier underneath that one — a name that keeps everything in its own right and is mostly read, all the same, as the ground other species’ stories stand on. The thread turns once more: a record is only as good as the name it is filed against, and a very wide name can be a filing cabinet as much as a finding.