How does a frog that calls loudly from bromeliads at night, across a swathe of western Ecuador, reach 2026 without a scientific name of its own?

The answer, published in ZooKeys by a team at the Pontificia Universidad Católica del Ecuador, is that it had a name. The name belonged to a different frog. Researchers have now described the animal as Pristimantis milpe, and in the same paper they have removed the name it had been wearing since 1980 from the list of valid species altogether.

The Milpe Rain Frog

The species is named for the Milpe Bird Sanctuary in Pichincha Province, a roughly 100 hectare private reserve where the holotype, an adult male 24.77 millimetres long, was collected in February 2020. The paper gives 21 paratypes, drawn from five Ecuadorian provinces, though its enumerated list runs to 22 catalogue numbers. The authors give the English common name as the Milpe Rain Frog.

Females are markedly larger than males, averaging 33.87 millimetres against 25.80. They also differ in colour, which the authors call unusual: female groins and hidden thigh surfaces carry cream, light blue or orange with contrasting black marks, while male groins are plain apart from one specimen with sparse dark markings. One male changed between dark brown and greenish brown within twenty-four hours, a plasticity the authors call significant.

The call is a single loud click averaging 0.075 seconds, with a dominant frequency of 2,581 hertz measured across five individuals, delivered at roughly three to fifteen calls a minute. Males call from bromeliads and leaves at heights from 30 centimetres to more than 10 metres, and at one site the average calling height was 6.6 metres. A separate 2025 study cited by the paper reports calling from dusk to sunrise, with a peak between 20:30 and 22:00. Recordings are deposited in the audio archive of the QCAZ museum and published through BIOWEB Anfibios del Ecuador. The lead author records that a calling male disturbed on a bromeliad sinks into the water held in the plant.

The figure that has travelled furthest with this frog is 67,000 square kilometres. Its basis is an extent of occurrence of 67,072 square kilometres, calculated as a minimum convex polygon drawn around all known localities. A convex hull of that kind encloses ground the species has never been recorded on, so the number measures the spread of the record rather than the area the animal occupies. On that extent, together with the species’ local abundance and its presence in disturbed habitats, the authors recommend a Least Concern listing.

Colombia carries a further hedge, and the unevenness is the paper’s own. Its Distribution section states plainly that the species occurs in the Pacific basin of Ecuador and southern Colombia, and press summaries followed it. Three sentences later the same section reports that the Colombian records apparently all rest on a single specimen, AMNH 86359, whose identification needs to be confirmed. The authors expect the species does occur there, because localities in northern Ecuador are just 18 kilometres from the border, and expectation is not a record.

Genetically, the paper names Pristimantis degener as the sister species and places the new species in a strongly supported group with both P. degener and P. mindo. Average uncorrected distance in the 16S gene is 9.2 per cent to the first and 10.5 per cent to the second. The phylogenetic matrix, 375 individuals and 4,584 base pairs, is deposited in Zenodo, where the record is public but, as of this writing, the file itself is embargoed until 31 December 2027.

Two numbers inside that analysis come with conditions, one from the method and one from the authors. The oldest divergence within the species is dated to roughly 1.6 million years ago from a time-tree built on a single calibration point, the 47.1 million year split between Pristimantis and its outgroup. And the highest genetic distance recorded between two individuals of the new species, 6.6 per cent, is one the authors suggest may be an artefact of sequence quality and a short 589 base pair overlap.

The specimen that a name is attached to

Every species name is anchored to a physical specimen. That specimen, the name-bearing type, is the only objective thing the name refers to. Everything else, including every field guide entry and every collection label, is an argument that a given population belongs with that specimen.

Santiago Ron and colleagues went back to the anchor. The name in question was Hylodes subsigillatus, described by George Boulenger in 1902 from a single adult female collected at Salidero, in Ecuador’s Esmeraldas Province. That holotype, catalogued at the Natural History Museum in London as BMNH 1947.2.17.1, is 34.0 millimetres from snout to vent and carries eggs in its oviducts.

For nearly eighty years after Boulenger described it, the name went essentially unused. Then in 1980 the herpetologist John Lynch redescribed the species, working mainly from specimens collected in Esmeraldas, Pichincha and Santo Domingo de los Tsáchilas provinces. Lynch listed the London holotype among the material he examined. Every subsequent treatment followed his characterisation, with one late exception: while this revision was under way, Coloma and Duellman suggested in 2025 that the name might belong to a different species from the one it was being applied to, though without molecular evidence and without spotting the link to latidiscus.

The new study finds that most of Lynch’s populations are not the holotype’s species. They differ from it in the shape of the head in profile, in the ventral colour pattern, in the relative length of the third and fifth toes, and in the presence of folds behind the eye, which the holotype has and Lynch’s frogs do not.

Resolving that exposed a second problem. Boulenger had described another frog from the same province four years earlier, Hylodes latidiscus, from two syntypes collected at Cachabé by the collector W. Rosenberg. The authors report that those two syntypes do not even belong to the same species as each other.

One, BMNH 1947.2.15.66, is an adult female 55.02 millimetres long with clear postocular ridges and a visible tympanic ring. The other, BMNH 1947.2.15.67, is 50.7 millimetres long, has no postocular ridges, more tuberculate skin, large tubercles on the eyelids and a tympanum only 20 per cent of the eye diameter. The authors assign the second to Pristimantis cisnerosi, a species named in 2020.

To stop the older name floating between two animals, the team designated BMNH 1947.2.15.66 as the lectotype of latidiscus. They then compared that lectotype with the subsigillatus holotype and concluded the two belong to one species. The type localities are 13 kilometres apart in a straight line, and both animals share the distinctive postocular folds, the wide finger discs and the head shape.

Because latidiscus was published four years earlier, priority applies and Pristimantis subsigillatus becomes a junior synonym. The two lineages are not close: the paper places latidiscus in the subgenus Hypodictyon, the sister group of all other Pristimantis, and puts its split from the rest of the genus, P. milpe included, at about 24 million years ago, citing a 2023 phylogeny by Portik and colleagues. The name is gone. The frogs that had been carrying it needed a new one.

Where the coverage and the paper each slip

A confusion lasting more than a century is the framing that has settled around this study. Applied to this frog it is wrong. The name subsigillatus was published in 1902 and then went unused in the literature until 1980, so the misapplication that hid this animal is 46 years old.

There is a 124-year error in the case, and it belongs to the two names rather than to the frog. Latidiscus, named in 1898, and subsigillatus, named in 1902, were carried as separate species until this paper collapsed them into one. That span is arithmetic on the paper’s own dates rather than a figure the paper states. What is genuinely long for the animal itself is its wait to be described at all, which is the paper’s framing and the reason its title begins “Long overdue”.

Two smaller drifts are worth recording. Pensoft’s announcement calls the animal the Milpe Robber Frog; the paper’s stated English name is the Milpe Rain Frog. And the division of labour in the press account needs stating precisely: the London specimens showed that the name these populations carried belonged to another lineage, while the sequences, calls and measurements from living Ecuadorian frogs supplied the description.

The paper carries one loose end of its own. Its title places the genus in the family Strabomantidae while its taxon classification block gives Craugastoridae, and the two are never reconciled.

Other names the authors want re-examined

Ron and colleagues argue that this is a recurring failure in twentieth century Pristimantis taxonomy rather than an isolated slip, and they name three further cases. P. ockendeni was redescribed in 1974 from Ecuadorian populations more than 1,200 kilometres from the type locality, and those populations later turned out to be three distinct species. P. riveti was redescribed in 1979 from populations 400 kilometres south of the type locality, a characterisation the literature then carried for forty years. And P. cajamarcensis was reviewed in 1969 from Ecuadorian populations 250 kilometres north, which recent phylogenetic work indicates should no longer be considered an Ecuadorian species at all.

Lynch described almost 200 species, and the authors are explicit that most of his Pristimantis remain valid and that his work is the pillar of current systematics in the genus. The paper’s recommendation is narrower and testable: go back to the old binomials and compare the name-bearing types with the populations now assigned to them. The genus holds 635 species on the count the paper cites, and the older names rest on specimens that can, where they survive, still be compared with the frogs now carrying them. That comparison is what the authors are asking taxonomists to do next.