One frog becomes three: hidden horned frog species rediscovered in Brazil’s Atlantic Forest

Deep in the leaf litter of Brazil’s Atlantic Forest lives a master of disguise: a horned frog whose bumpy skin and fleshy eyelid appendages let it vanish among dead leaves and twigs. For two centuries, scientists lumped populations of this frog, Proceratophrys boiei, into a single species stretching across six Brazilian states. Now a sweeping taxonomic revision published in the journal Herpetozoa has shattered that picture. By combining genetics, body measurements, frog calls, and even the fine details of skull bones, researchers have shown that what was long called P. boiei is actually three distinct species. Two of them, P. fryi and P. intermedia, had been buried as synonyms for decades and are now formally brought back to life.

The horned frogs of the genus Proceratophrys belong to the family Odontophrynidae and currently comprise 42 recognized species, making them one of the most species-rich frog groups in the Neotropics. These frogs are famous for the palpebral appendages, small fleshy flaps on their upper eyelids that resemble dead leaves, which help camouflage them in the forest floor. Species in the P. boiei complex bear these appendages but lack a rostral appendage on the snout, a combination that separates them from relatives like P. appendiculata. Yet despite their distinctive appearance, the frogs are morphologically conservative, meaning that different species can look remarkably alike. This has made the genus a persistent headache for taxonomists relying on appearance alone.

The trouble with P. boiei began long ago. The species was originally described as Ceratophrys boiei by Wied-Neuwied in 1824, but the original description lacked information on the collector and the type locality. Over the years, the type locality was restricted first to Bahia, then to Rio de Janeiro, after researchers compared populations with the lectotype specimen held in the Netherlands. Two other names, Ceratophrys fryi from Serra da Mantiqueira in Minas Gerais and Ceratophrys intermedia from Santa Katarina, were described in 1873 and 1908 respectively, but both were later synonymized with P. boiei, one 50 years ago and the other 15 years ago, because no clear morphological differences could be found between them.

The turning point came in 2012, when researchers using molecular markers and cytogenetics identified three deeply divergent lineages within what was then considered P. boiei. Two northern lineages emerged: North 1, found in Espírito Santo and Minas Gerais, and North 2, spanning northern São Paulo, southern Minas Gerais, and Rio de Janeiro. A third southern lineage ranged from southern São Paulo to southern Santa Catarina. Crucially, the cytogenetic work revealed a striking difference: the southern lineage has its nucleolar organizer region, a chromosomal feature, on the short arm of subtelocentric pair 4, whereas both northern lineages carry it on the short arm of submetacentric pair 8. Later studies confirmed this pattern and showed that tadpole chondrocranial morphology also varies geographically, with specimens from Minas Gerais exhibiting unique subdivisions of the frontoparietal fontanelle and ventral fusion of suprarostral cartilages, while those from Paraná present a reduced hypobranchial plate and a more developed ceratobranchial IV.

To settle the question, the new study conducted a full taxonomic review of populations across the entire geographic range of P. boiei. The researchers evaluated external morphological and morphometric characters in 132 individuals, 80 males and 52 females, from 48 localities across southeastern and southern Brazil. They analyzed 708 advertisement calls from 66 males collected at 27 localities, and examined osteological traits in nine males from seven localities using a clearing and staining protocol that renders bones visible in otherwise translucent specimens. Genetic distances were estimated from 94 individuals across 40 localities, using partial sequences of two mitochondrial genes: 93 sequences of the 16S ribosomal RNA gene and 59 sequences of cytochrome b.

The results revealed a paradox that will resonate with anyone who has tried to identify cryptic species. External morphology, osteology, and advertisement call parameters varied among the three lineages, but most of the variation overlapped, even among individuals within the same lineage, indicating high polymorphism. Individuals from all three lineages could exhibit either short or long palpebral appendages, canthal crests ranging from faint to pronounced, and ocular-dorsal ridges that were either continuous or discontinuous. The advertisement calls of all three lineages consist of a multipulsed note emitted sporadically, and a multivariate analysis of acoustic variables found no significant differences among them. In other words, no single adult character could reliably separate the species.

Yet some differences did emerge when the data were examined statistically. Morphometric variation was significantly influenced by both lineage and sexual dimorphism, with sexual dimorphism explaining a greater proportion of the variation than lineage. Pairwise comparisons showed that the southern lineage differed significantly from both northern lineages in males and females, whereas the two northern lineages did not differ from each other. Among the lineages, specimens from the North 1 lineage had the largest mean snout-vent length, 51.5 millimeters in males and 60.3 millimeters in females, while the southern lineage had the smallest, 46.0 millimeters in males and 51.6 millimeters in females. The two most important variables contributing to lineage separation were interorbital distance and eye-nostril distance for females, and interorbital distance and palpebral appendage length for males.

The genetic data provided the clearest signal. Mean uncorrected mitochondrial p-distances among the three lineages were relatively similar for 16S, ranging from 2.7 to 2.9 percent, whereas for cytochrome b they ranged from 5.6 to 7.3 percent. These divergences, combined with the previously identified differences in cytogenetics, tadpole external morphology, and chondrocranial characters, convinced the researchers that three valid species are currently grouped under the name P. boiei. The type locality historically associated with the species corresponds to the populations of the North 2 lineage, which should therefore be recognized as P. boiei. The North 1 lineage, with a more northern distribution in Espírito Santo and Minas Gerais, is associated with P. fryi. The southern lineage, occurring in Santa Catarina, Paraná, and southern São Paulo, is linked to P. intermedia.

The revalidation carries significant implications for conservation. The recognition of more narrowly distributed taxa requires a reassessment of conservation strategies, as populations previously attributed to a widespread species may in fact be at greater risk due to habitat fragmentation and the presence of specific geographic barriers. The researchers suggest that rivers such as the Tietê, which separates P. intermedia from P. boiei, and the Muriaé, which separates P. boiei from P. fryi, may have acted as biogeographic barriers limiting gene flow and promoting lineage divergence, a pattern documented for other vertebrates in eastern Brazil. Rather than representing simple geographic discontinuities, these rivers may have shaped the evolutionary history of the entire species complex.

Beyond the immediate taxonomic outcome, the study reinforces the importance of integrative taxonomy in groups with morphologically conservative species. Similar difficulties have been documented in other Proceratophrys species complexes, where species delimitation required the combined evaluation of multiple sources of evidence rather than a single external character. Advertisement calls are currently known for 33 species of the genus, genetic data are available for 28 species, and tadpole morphology has been described for 21 species, yet none of these sources alone is sufficient to resolve all taxonomic problems. The case of P. boiei, P. fryi, and P. intermedia shows that robust species limits often emerge only when morphology is interpreted together with molecular, cytogenetic, and acoustic datasets, a lesson that will guide future explorations of Atlantic Forest amphibian diversity.

Subject of Research: Biology

Article Title: Taxonomic revision of the horned frog Proceratophrys boiei (Wied-Neuwied, 1824) ( Amphibia , Anura , Odontophrynidae ) with revalidation of P. fryi (Günther, 1873) and P. intermedia (Barbour, 1908)

Article References: Castro, L., Guimarães, C. S., Carrillo, J. F. C., Santana, D. J., & Mângia, S. (2026). Taxonomic revision of the horned frog Proceratophrys boiei (Wied-Neuwied, 1824) (Amphibia, Anura, Odontophrynidae) with revalidation of P. fryi (Günther, 1873) and P. intermedia (Barbour, 1908). Herpetozoa, 39, 177-202. https://doi.org/10.3897/herpetozoa.39.e186246

Image Credits: AI Generated

DOI: 10.3897/herpetozoa.39.e186246

Keywords: taxonomy, amphibians, Atlantic Forest, cryptic species, integrative taxonomy, herpetology, Brazil, conservation, Taxonomic, revision, horned, frog