Keeping seen from inside another colony — a slave-maker of 1922 that holds no territory of its own, whose range is the range of the species it raids, and whose one conservation verdict was filed under a genus name it no longer carries
The last page ended inside someone else’s nest: one colony of the newborn Temnothorax marae, at the Vendicari reserve, was found hosting the social parasite Temnothorax muellerianus (Finzi, 1922). Marae waits for a census of its narrow ground. This name has no ground to census. Ask instead what a species holds when every colony it has is a colony it took.
First, it is a raider, not a lodger. Muellerianus is an active slave-maker: its workers raid nests of other Temnothorax, and in the raids of its old genus the workers sting their opponents to death and carry the brood home, where the captured young emerge and do the work of the colony that stole them. It is not workerless, and it does not merely move in — it restocks itself by force, again and again, because a workforce taken as brood ages and is not replaced from inside. Among the eight or so species once grouped in its old genus, it is one of only a few studied in any detail; several of its relatives are known from type material or scattered finds alone.
Second, it began under another genus. Finzi described it in 1922 as Chalepoxenus muellerianus, and for most of a century that was its address — a whole small genus of Mediterranean slave-makers, spread across southern Europe, North Africa, and western and central Asia, all of them living off Temnothorax hosts. The genus has since been folded into Temnothorax itself, so the parasite now sits inside its hosts’ own genus, under the name Temnothorax muellerianus. The raider and the raided file on the same shelf.
Third, its map is borrowed. The species is reported from Spain to Turkey — a range no narrow endemic could match — but that breadth is not its own. It goes exactly where its hosts go, and only where they are dense enough to raid. In the host tally assembled for its old genus, one host dominates: Temnothorax unifasciatus accounts for roughly three quarters of the records, with recedens and nigriceps far behind, and a long tail of occasional hosts — flavicornis, exilis, tuberum, affinis and others named in ones and twos. A distribution like that is a shadow: wide because the thing casting it is wide, and liable to thin wherever the hosts thin, whatever the map seems to promise.
Fourth, the newest host on its list is the newest species on the last page. The 2022 paper that described marae recorded one of its colonies, at Vendicari, hosting muellerianus — so a host list built mostly on a widespread, common species now also carries a Siculo-Maltese endemic known from eleven localities, on coastal ground its describers judged likely much reduced. The parasite does not choose between the secure host and the fragile one; it takes what is there. But every colony it founds on a rare host is a colony whose whole future is that host’s future, twice over.
Fifth, it was assessed — once, long ago, under the old name. Here the contrast with marae turns. Marae has never been assessed at all. Muellerianus carries a Red List verdict of Vulnerable, given in 1996 by the Social Insects Specialist Group — filed under Chalepoxenus muellerianus, the genus name that no longer stands, and under the older version of the Red List criteria. So the assessment exists and does not quite attach: look for the species under the name it now bears and the verdict sits one synonym away, thirty years old, pronounced before the host list included a species that had not yet been described.
And so its keeping is doubly borrowed. Land, it borrows from its hosts. Labour, it borrows from their brood. Even its one act of official concern is borrowed from a former name. What it holds in its own right is only the behaviour — the raid, repeated. A free-living species can in principle be counted where it stands. This one can only be counted where something else stands first, which may be why a slave-maker known from Spain to Turkey can still, plausibly, be vulnerable: the map is large, and none of it is its.
A walk, schematic only — five stages, tap each one:
An active slave-maker: raids host nests, kills defenders, carries brood home to emerge as the workforce.
Four 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, and now a name that keeps no colony it did not take and no verdict under the name it wears. The thread running through them is that a record is only as good as the name and the ground it is filed against — and this species files everything against other species, in their nests, under their range, once even under a genus that has ceased to exist. If its hosts hold, it holds. That conditional is the whole entry.