Second place

Crematogaster ariadnae, described new in 2026 — the Cretan endemic that stands second in another species’ file again and again, ruled out each time on colour and birthplace, and first in the revision’s own key

R1009 closed the table of type workers and ended on the runner-up those workers kept pointing at. This page reads the runner-up as the subject. Everything below is summarised in my own words, with no quoted text beyond bare collecting and bibliographic data.

First, how it appears in someone else’s file. Table 2 of the revision runs six old type workers as wild cards against four clusters built from 201 workers in 68 nests. The ariadnae column is the only one that never wins a row. It is second three times: 0.379 behind the Pazardzhik christowitchii syntype’s 0.620 for schmidti, 0.088 behind the Sliven syntype’s 0.898, and 0.031 behind the Dalmatian stand-in’s 0.945. It is present, at some small value, in every row — 0.011 behind the Smyrna syntype, 0.002 behind medispina, 0.001 behind the excluded Patras worker. And where it matters most, in the two Bulgarian rows, the revision’s results passage excludes it by hand: the syntypes’ second values were for ariadnae, and conspecificity with it is set aside on coloration and zoogeography. A worker pinned in inland Bulgaria in 1891 cannot be the animal that lives only on Crete. In schmidti’s file, ariadnae is the answer the measurements keep suggesting and the map keeps forbidding.

Second, what it is on its own. Crematogaster ariadnae was described as new in the same 2026 revision (ZooBank registration 09733294-7739-4659-8FDA-C396B2E775E8). Its holotype is a worker collected on 29 March 2024 by H. C. Wagner, 0.8 km west of Maronia on Crete (35.1404° N, 26.0728° E), in broadleaf wood along a swale, held at the NHMW in Vienna under the codename WAG3285; the type paragraph gives that spot as 180 m above sea level, while the account’s own distribution list gives the same locality as 222 m — the paper carries both numbers and this page keeps both. Its paratypes are two workers from the holotype nest and twelve more from four other nests, spread across four collections. Behind the name stand 66 measured workers from 21 nests — a sample as large as schmidti’s own, and larger than either of the other two species in the revision. It was found at six localities, all on Crete, from the lake at Agiás in the west to Maronia in the east, and never alongside any other member of the complex: in this study no locality held two of the four species.

Third, second in the table, first in the key. The identification key does not treat ariadnae as a residue to be reached by elimination. Its very first couplet separates it from all three others at once, on the shortest pubescence and hairs of the first gastral tergite in the complex: its mean pubescence index (PLG/CS 0.054, range 0.046–0.064) and mean longest-hair index (GHL/CS 0.088, range 0.072–0.113) are the lowest of the four species, and the couplet’s discriminant function misplaces 3.0% of its workers and none of its 21 nest means. Its colour is given as often homogeneously dark brown to blackish, the head only slightly lighter than the gaster. The perpetual second of the type table is the first name the key can hand you.

Fourth, the sister it keeps almost being. The revision names ariadnae the sister species of schmidti, and of all six species pairs theirs differs in the fewest measured traits — five, against eleven between schmidti and ionia. In the unsupervised PCA their nest samples largely overlap, and in the pair-only plot two ariadnae samples fall inside schmidti’s polygon. That closeness is exactly why Bulgarian workers lean toward it: it is the nearest shape to schmidti in the file. Yet the supervised analyses hold the line the PCA blurs — the four-species LDA classifies every nest mean correctly, and the pair’s cross-validation error is 3.0%. Near enough to attract another species’ strays, distinct enough that none of its own nests is ever misread. Its separation is given the same deep-water explanation as its two fellow island-and-mainland splits: ranges divided by the Aegean for millions of years, speciation by geographic isolation.

Fifth, the name. The epithet is Ariadne’s. In the etymology, in paraphrase, Ariadne on Crete gave Theseus the thread that led him out of the labyrinth, and the name alludes to this ant’s pronounced pheromone trails — laid by the colony and followed not only by its own nestmates but by five other ants: Camponotus gestroi, Ca. kiesenwetteri, Ca. lateralis, Ca. rebeccae and Colobopsis imitans. The species that spends the type table as the following answer is, in the field, the one whose line the others follow.

So what does second place hold? A species entire, held the ordinary way — a fresh holotype, a large sample, a first couplet, an island — whose strangest property is visible only from inside another species’ paperwork: it is the standing alternative for every weak schmidti reading, close enough to collect 0.379 of a Bulgarian syntype and never once allowed to keep it. Even its endemism is held with the revision’s own hedges kept on: currently endemic to Crete, expected to be endemic, with the account itself noting it is uncertain whether it also occurs on neighbouring islands, or whether another member of the complex also lives on Crete. Nothing in its own account is borrowed, proxied or pending. The secondness was never in the ant. It was in the table it happened to be read from.

The walk — five moments in one species’ double life, schematic only. Click each stage.

summarised under ionia, unnamed
second in the table, ruled out by the map
a holotype of its own, 2024
first in the key
the thread the others follow

Pick a stage.

The excluded worker was certain and lost; the lowest was doubtful and carried; this one was never anyone’s type at all, and needed none — she arrived with her own.