Hidden Species Unmasked: DNA and Morphology Split a Puzzling East Asian Plant Complex Into Three

For decades, botanists studying a group of East Asian shrubs and small trees in the buckthorn family have been arguing over a deceptively simple question: how many species are they actually looking at? A new study published in the journal PhytoKeys has now delivered a striking answer, using a combination of population-level morphological statistics, DNA barcoding, and geographic mapping to untangle the Rhamnella franguloides complex, a notoriously slippery group within the family Rhamnaceae. The research not only splits what was previously treated as two species into three, but also corrects a long-standing case of mistaken identity involving a specimen that belongs to an entirely different plant family.

The story begins with a fundamental problem in biodiversity science: species delimitation. Species are the basic units of taxonomy, ecology, and conservation, yet drawing their boundaries is anything but straightforward. Over the past several decades, the proliferation of species concepts has exposed the weaknesses of relying on any single criterion. Morphological diagnosability, reproductive isolation, phylogenetic relationships, and ecological divergence each emphasize different attributes, and any one concept may fail when applied to cryptic species, recently diverged taxa, or lineages distributed side by side. A growing body of evidence indicates that single-evidence approaches can introduce systematic bias or outright errors, which is why an integrative framework combining at least two independent lines of evidence is increasingly recognized as essential for robust species recognition.

Morphology remains the most practical and widely applied criterion, but the new study shows just how unreliable it can be when applied carelessly. Morphological characters often vary across developmental stages, hybrids frequently exhibit traits that differ from those of their parents, and environmental heterogeneity can generate extensive phenotypic variation, potentially leading researchers to mistake variants for independently evolving lineages. Analyses based on only a few specimens may confound intraspecific variation with interspecific differences. To avoid these pitfalls, the researchers behind the new work compiled an extraordinary dataset: 412 herbarium specimens for a first clustering analysis of vegetative traits, and 372 fruiting specimens for a second analysis that added reproductive characters, covering the entire geographic range of the complex across China, South Korea, Japan, and the Ryukyu Islands.

The genus Rhamnella comprises 12 accepted species distributed mainly in East and Southeast Asia, and morphological distinctions among its members are often subtle and easily overlooked. The type species, Rhamnella franguloides, has accumulated numerous synonyms over more than a century and a half, and its type specimen, collected in Japan, bears glabrous petioles and longer leaf apices, whereas the type of a Chinese synonym, Rhamnella obovalis, bears densely pubescent petioles and shorter leaf apices. Meanwhile, the type of Berchemia congesta from China closely matches the typical morphology of R. obovalis, and the type of Rhamnella japonica from Japan matches R. franguloides itself. These observations raised the question of whether some of these supposedly merged names should actually be reinstated as distinct species.

The study also resolved a bizarre taxonomic detour. Microrhamnus taquetii, a name described from Korea and long treated by some authorities as a synonym of R. franguloides, turned out on direct re-examination of its type specimen to be something else entirely. The specimen displays opposite leaves and branches, branches bearing at least two rows of corky wings, subquadrangular branchlets, and rounded winter buds with irregularly and sharply denticulate bud-scale margins. These are hallmark characters not of Rhamnella, nor even of the family Rhamnaceae, but of Euonymus alatus, the burning bush, a species of Celastraceae. The researchers searched online databases for specimens matching this morphology and confirmed the identification, vindicating earlier treatments by Kim and Ma and an annotation written on the type sheet itself back in 1993. A plant name had been hiding in the wrong family for years, and only direct examination of the century-old type material exposed the error.

For the complex proper, the statistical results were dramatic. A first principal component analysis of seven quantitative and two qualitative vegetative traits recovered two groups: one comprising all Chinese specimens previously identified as R. franguloides, including the types of R. obovalis and Berchemia congesta, plus a single Korean specimen, and another containing Rhamnella caudata together with all specimens from Japan, the Ryukyu Islands, and the remaining Korean material. A second PCA based on 372 fruiting specimens, incorporating 13 quantitative and two qualitative traits such as fruit size, fruit pedicel length, and lenticel density on old branchlets, split that second group further, yielding three major clusters in total. Cluster A corresponds to the Chinese plants, Cluster B comprises only R. caudata from Guangdong, and Cluster C includes the Japanese, Ryukyu, and Korean material.

The three clusters are distinguishable by concrete, measurable characters. Cluster A, the Chinese plants, has pubescent leaf petioles and annual branchlets, short petioles of 1.0 to 3.8 millimeters, and short leaf apices of 2.2 to 9.0 millimeters. Cluster B, R. caudata, has glabrous petioles, dense lenticels of 87 to 144 per square centimeter on old branchlets about two millimeters in diameter, and large dry fruits of 9.0 to 10.9 by 3.6 to 4.4 millimeters. Cluster C, the newly circumscribed R. franguloides, has glabrous petioles, sparse lenticels of 54 to 87 per square centimeter, and smaller fruits of 5.5 to 8.7 by 2.9 to 4.0 millimeters. Within Cluster C, the two varieties, var. franguloides and var. inaequilatera, are separated almost entirely by petiole length, 3.8 to 7.6 millimeters versus 7.8 to 15.8 millimeters, a far narrower set of distinguishing characters than previously described, since traits like leaf-base symmetry and fruit striation proved variable or non-diagnostic.

Geography added a crucial corroborating line of evidence. The three species and two varieties are completely or largely allopatric, meaning they occupy separate ranges: Cluster B is restricted to Ruyuan in Guangdong, southern China, while Cluster C is distributed in Korea, the Ryukyu Islands, and Japan, and Cluster A occupies central, eastern, and southeastern China, rarely extending to South Korea. Allopatric speciation is considered a common model through which reproductive isolation develops, and geographic discontinuity is itself an important isolating mechanism that maintains morphological and genetic boundaries. The newly defined morphological boundaries and corresponding taxonomic treatments differ substantially from those proposed previously, underscoring how population-level statistical comparison can reshape taxonomic decisions.

Molecular evidence then sealed the case. The researchers generated new sequences of the nuclear internal transcribed spacer, or ITS, region, the standard plant DNA barcode, from 34 individuals, and combined them with sequences retrieved from GenBank for a total of 33 sequences of the complex. Sequence variation strongly supported the genetic distinction among the three species. Cluster A was distinguished by fixed diagnostic nucleotide sites at five positions in the alignment, Cluster B by five different fixed sites, and Cluster C possessed nine private sites compared with either of the other clusters, while Cluster B had ten private sites. A maximum likelihood phylogenetic analysis showed that the well-supported R. congesta clade was sister to another well-supported clade comprising R. franguloides and R. caudata, meaning the presence or absence of pubescence, petiole length, and leaf apex length are congruent with the inferred evolutionary relationships.

The taxonomic consequences are substantial. Because Berchemia congesta was published in 1875, predating R. obovalis in 1914, the epithet congesta has priority, and the researchers formally proposed the new combination Rhamnella congesta, with R. obovalis as its synonym. Rhamnella caudata remains a distinct species with redefined boundaries, and R. franguloides is newly circumscribed to comprise only the Korean, Ryukyu, and Japanese plants, with its two varieties retained on the strength of their petiole-length difference and pronounced geographic separation. The study highlights the importance of integrating multiple independent lines of evidence in species delimitation, and the authors note that further work with additional molecular markers and expanded sampling of Cluster C could help clarify whether the Ryukyu variety might eventually warrant resurrection at species rank. For now, three species stand where botanists long saw two, a quiet revolution in herbarium drawers that reveals how much hidden diversity still waits to be unmasked in even the most familiar corners of the plant kingdom.

Subject of Research: Species delimitation and taxonomic revision of the Rhamnella franguloides complex ( Rhamnaceae )

Article Title: Species delimitation and taxonomic revision of the Rhamnella franguloides complex ( Rhamnaceae )

Article References: An, Q., Lu, Z., & Zhou, X. (2026). Species delimitation and taxonomic revision of the Rhamnella franguloides complex (Rhamnaceae). PhytoKeys, 279, 179-200. https://doi.org/10.3897/phytokeys.279.204054

Image Credits: AI Generated

DOI: 10.3897/phytokeys.279.204054

Keywords: Species, delimitation, taxonomic, revision, Rhamnella, franguloides, complex, Rhamnaceae, scientific research