Hidden in Plain Sight: Botanists Unveil a New Clematis Species Endemic to Korea

For more than a century and a half, a slender climbing vine with nodding purple flowers has been hiding in the forests of the Korean Peninsula, masquerading as a plant it closely resembles. Now, a team of Korean botanists has pulled back the veil: what field botanists had long labeled Clematis fusca var. violacea actually includes two distinct entities, and one of them is a species entirely new to science. The discovery, published in the open-access journal PhytoKeys, introduces Clematis pedunculata, a androdioecious woodland vine known only from South Korea, and it underscores how much unrecognized biodiversity can persist in well-studied temperate floras.

The story begins with a name coined in 1859 by the Russian botanist Carl Johann Maximowicz, who described C. fusca var. violacea based on plants collected along the Sungari River in the Amur region. His protologue emphasized violet sepals with whitish hairs largely confined to their margins, and for decades Korean floristic treatments applied the name broadly to any Korean clematis sharing that combination of nodding purple flowers and whitish sepal hairs. Thirteen taxa of Clematis, a genus of roughly 300 species distributed mainly in temperate regions, are recognized in South Korea, and the C. fusca complex has generally been treated as comprising three varieties: fusca, flabellata, and violacea. But preliminary field observations and herbarium work hinted that something was off. Among Korean collections identified as var. violacea, two morphologically coherent groups kept turning up, differing consistently in characters that, at first glance, few collectors would think to record.

To settle the question, the researchers mounted an unusually thorough investigation. They examined living plants in cultivation alongside no fewer than 452 herbarium specimens belonging to the C. fusca complex, including material housed at the Korea National Arboretum, and scrutinized digital images of type specimens and other relevant material at the Komarov Botanical Institute of the Russian Academy of Sciences and the Royal Botanic Garden Edinburgh. Measurements were taken with digital calipers, and vegetative and reproductive structures were dissected and examined under a stereomicroscope, with particular attention to leaflet number and size, petiole morphology, and the size, shape, and hair covering of bracts, sepals, stamens, and carpels. Crucially, the team also compared Korean material against Maximowicz’s original description, ensuring that at least one of the two groups could be anchored to the historical name.

The morphological evidence quickly proved decisive. One group matched var. violacea: hermaphroditic plants bearing only bisexual flowers, with distinctly grooved and densely villous stem internodes, subsessile or barely pedunculate flowers, and stamens clothed in brown hairs toward their tips. The other group was a different animal altogether. It is androdioecious, meaning individual plants bear either staminate flowers or bisexual flowers, a sexual system that is rare and taxonomically significant in the genus. Its stem internodes are smooth to only shallowly grooved and sparsely villous to nearly hairless. Most strikingly, its flowers sit on conspicuous stalks, or peduncles, ranging from 1.5 to 12.1 centimeters long, whereas in var. violacea the peduncles are essentially absent or under seven millimeters. And its stamens carry short whitish hairs rather than brown ones. The species epithet pedunculata celebrates that most eye-catching feature, and the proposed Korean vernacular name, Jang-dae-jong-deong-gul, likewise nods to the conspicuously elongated flower stalks.

Morphology alone might have sufficed for some taxonomists, but the team went further, adding molecular phylogenetics to the case. Fresh leaves from eleven individuals representing the complex were collected from sites across South Korea, from Danyang in Chungcheongbuk-do to Mount Baekhwa in Gyeongsangbuk-do and Mount Chilbo in Gyeonggi-do. Genomic DNA was extracted from silica-dried leaves, and three regions were amplified and sequenced: the nuclear ribosomal internal transcribed spacer, or ITS, the plastid matK gene, and the atpB–rbcL intergenic spacer. These were supplemented with sequences of additional Clematis species downloaded from GenBank, with Pulsatilla koreana serving as the outgroup based on prior phylogenetic studies of the tribe Anemoneae.

The sequence data were aligned and analyzed under a maximum-likelihood framework using IQ-TREE, with the best-fitting substitution models selected according to the Bayesian information criterion and branch support assessed through 1,000 ultrafast bootstrap replicates. The results were congruent across the two genomes. In the ITS phylogeny, the three accessions of the putative new species formed a moderately supported clade, with 83 percent ultrafast bootstrap support, separate from the sampled Korean members of the C. fusca complex, although the deeper position of that clade within the genus remained weakly resolved. In the combined plastid phylogeny of matK and atpB–rbcL, the same three accessions again cohered as a distinct, moderately supported clade, with 82 percent support, separate from all sampled Korean members of the complex. Nuclear and plastid data thus tell the same story: this is a cohesive lineage, not a color variant.

The concordance between the morphological discontinuity and the molecular separation is what gives the new species its scientific footing. Neither line of evidence alone would have been airtight; moderate support values leave room for uncertainty about the closest relatives of C. pedunculata, and previous molecular studies have repeatedly shown that deeper relationships within Clematis are sensitive to taxon and marker sampling. But when independent nuclear and plastid markers both recover the same accessions as a distinct lineage, and when those same plants differ from their relatives in sexual system, stem anatomy, inflorescence architecture, and hair type, the case for species status becomes compelling. The authors are careful, however, not to overclaim: the analyses do not identify the closest relative of the new species or establish its definitive placement within the infrageneric classification of Clematis, and broader sampling across northeastern Asia will be needed to resolve those questions.

The formal description reads like a portrait of a plant that managed to stay invisible for 165 years. C. pedunculata is a perennial herbaceous vine, scandent and reaching about four meters in length, with yellowish green to purplish green stems and opposite leaves that may be simple or divided into three to seven leaflets, borne on tendril-like petioles that help the plant climb. Its nodding, actinomorphic flowers carry four ovate to oblong-ovate sepals, purplish white to purple, almost hairless on their inner surfaces and densely whitish villous along the margins, which is precisely the feature that led generations of botanists to file it under var. violacea. Flowering runs from June to July and fruiting from July to September. The type specimen was collected in June 2024 at Pocheon-si in Gyeonggi-do and is deposited at the Korea National Arboretum herbarium.

Even its geography is revealing. Verified records of C. pedunculata span eight South Korean provinces, from Gyeonggi-do and Gangwon-do in the north to Jeollanam-do and Gyeongsangnam-do in the south, with the species growing in inland forests, along forest margins, and occasionally on roadsides. By contrast, true C. fusca var. violacea, whose range extends through northeastern China and the Russian Far East, has been recorded in South Korea only from Gyeonggi-do. The recognition of the new species therefore does not expel var. violacea from the Korean flora; rather, it refines its circumscription, carving out of the broad old concept a genuinely endemic lineage. The authors provisionally regard C. pedunculata as endemic to South Korea, though they caution that herbarium specimens identified as var. violacea from adjacent regions should be re-examined to determine whether the new species also occurs elsewhere in northeastern Asia.

Beyond the novelty itself, the discovery carries a broader lesson for field botany and conservation. Sepal color and marginal hairiness, the characters that defined the old variety, turn out to be simply insufficient for identifying Korean plants once assigned to C. fusca var. violacea. Reliable identification now demands attention to floral sexuality, peduncle development, stem groove morphology, and the indumentum of stems and stamens, details easily overlooked on a pressed sheet or in a hurried survey. That a distinct species could persist unrecognized across eight provinces in one of the world’s most intensively botanized temperate regions suggests that species complexes with superficially uniform flowers may still conceal surprises, and that revisiting old names with modern tools, from stereomicroscopy to phylogenomics, remains one of the most productive enterprises in plant taxonomy. For now, the long-stalked clematis of the Korean woods finally has a name of its own.

Subject of Research: Taxonomic description of the new Korean plant species Clematis pedunculata using morphological and molecular phylogenetic evidence

Article Title: Clematis pedunculata ( Ranunculaceae ), a new species from Korea previously confused with C. fusca var. violacea

Article References: Park, B. K., Lee, D.-H., Kim, S.-C., & Son, D. C. (2026). Clematis pedunculata (Ranunculaceae), a new species from Korea previously confused with C. fusca var. violacea. PhytoKeys, 280, 29-50. https://doi.org/10.3897/phytokeys.280.213893

Image Credits: AI Generated

DOI: 10.3897/phytokeys.280.213893

Keywords: Clematis pedunculata, Ranunculaceae, plant taxonomy, new species, South Korea, phylogenetics, ITS, chloroplast DNA, androdioecy, biodiversity, Clematis fusca complex, endemism