Keeping at the middle of the slide — a slave-maker whose own workers have thinned to almost none, whose raids were seen under glass and long doubted in the field, and whose stolen workforce may be betrayed only by its genes
R961 closed by pointing along a slide: from active slave-makers with many workers, through degenerate ones with few, to parasites with none at all. This page stops at the documented middle of that slide and stays there — Myrmoxenus kraussei (Emery, 1915), long also carried under the genus name Epimyrma. Where bikara on the last page was a name held by inference, kraussei can be counted: colonies opened, workers tallied, raids watched in the laboratory, and host workers genotyped. What those counts show is a species that has not finished losing its workforce — and may not have finished using it, either.
First, its genus holds the whole slide in one frame. In Myrmoxenus, a young mated queen does not found a colony alone; she enters a Temnothorax nest and eliminates the resident queen by slowly throttling her, a process the accounts put at days to weeks, during which the intruder apparently takes on the colony’s odour and is integrated. At one end of the genus, active slave-makers such as M. ravouxi keep numerous workers of their own — up to about two hundred in the genus — which pillage brood from neighbouring nests in organised raids, so the borrowed workforce is steadily renewed and the queen can produce offspring over several years. At the other end, workerless forms such as M. corsicus and M. adlerzi produce no workers at all and only one or a few broods of sexuals before the inherited host workers die. Kraussei is placed between them, and the placement has a name the literature uses for it in so many words: a degenerate slave-maker, intermediate in the loss of the worker caste.
Second, how few workers is few can be said in numbers, from one intensively studied place. In olive groves near Tignale, above the western shore of Lago di Garda in Italy, where the ants nest between rocks in stone walls, an earlier population study found colonies holding a median of only one and a half kraussei workers, ranging from none to ten. A later collection there, in May 2011, took ten colonies, each with a single kraussei queen, none to three workers of its own (median one), and three to thirty-three workers of its sole host, Temnothorax recedens (median fourteen and a half), plus brood. Set those medians side by side and the shape of the colony is plain: one queen, roughly one worker who shares her species, and a workforce of fifteen or so who do not. The two medians come from different samples and are kept unmerged on this page; together they say the same small thing twice.
Third, the raid survives in the laboratory even at that thinness. The accounts report that parties of four to five kraussei workers, assembled under laboratory conditions, could raid a recedens colony successfully and showed the full behavioural repertoire of active slave-makers. That is a real observation, and its rank matters: a raid elicited, not a raid found happening. On the strength of the tiny worker numbers, the same literature assumed field raids would be exceptional if they occurred at all. For decades the honest summary was therefore a strange one — a slave-maker that demonstrably still knew how to steal, and was believed, in nature, essentially not to.
Fourth, the genes suggest the belief may have been too tidy. Free recedens colonies at Tignale are genetically plain: their workers are full sisters, offspring of one singly mated queen, with relatedness close to the value expected for such sisters. In four of the ten kraussei colonies examined, the host workers inside were nothing like so plain — many more distinct genotypes and alleles, an estimated five to seven full-sister lineages living together, and relatedness falling far below the free-colony value. In the other six colonies no clear sign of mixed lineages appeared, and three of those six held no kraussei workers at all when collected. The authors weigh three readings: that kraussei queens preferentially take over already-mixed host colonies, which the uniformity of free colonies makes unlikely; that parasitised colonies adopt stray foreign workers more readily, for which there is at present no evidence; and, as the most likely reading, that even a handful of kraussei workers does raid in the field, bringing home brood from more than one nest, whose daughters then stand in the same colony as unrelated fellow workers. Small raiding parties are not in themselves implausible — other slave-makers in these pages have raided with parties as small.
Last, notice what kind of keeping the middle is. An active slave-maker keeps by renewal: steal again, and the workforce never runs out. A workerless parasite keeps by inheritance: take the workforce once, spend it, and end when it ends. Kraussei sits exactly where those two economies blur — too few workers to raid as a matter of obvious routine, still workers enough that, under glass, the whole old repertoire comes back intact, and with a genetic trace in four nests out of ten that looks like renewal happening quietly anyway. Nothing about it is settled by that trace, and this page does not settle it. The middle of a slide is not a compromise between the ends; it is the place where you can watch one economy turning into the other, one or two workers at a time.
A walk through a keeping halfway lost. Five stages: a host nest in a stone wall; a queen who throttles the resident queen; a colony with almost no workers of its own; a raid party of four or five under glass; and mixed bloodlines among the host workers, hinting the party may also work outdoors. Schematic, compressed from the published accounts, with the inferred stage labelled as inferred.
Walk it in order.
Schematic only: the stages compress a genus survey, a population study and a genetic study, so no single colony is promised this exact story. Labelled cautions: do not present field raiding by kraussei as observed — raids were observed in the laboratory, and field raiding is the most likely reading of a genetic pattern, not a watched event. Do not merge the earlier Tignale median of one and a half workers (range none to ten) with the 2011 sample median of one (none to three) — different samples, kept apart on this page. Do not read the six colonies without mixed lineages as proof no raiding happened there — the finding is no clear evidence in those six, three of which held no slave-maker workers at collection. Do not give kraussei a second host past Temnothorax recedens — the genetic study calls it the species’ only host. Do not present the throttling founding, described for the genus generally, as a kraussei-specific timed observation. And do not present the mixed lineages as explained — two rival readings are weighed and set aside as unlikely or unevidenced, not disproven by direct observation.