Mussaenda is a paleotropical genus of the coffee family, Rubiaceae, comprising roughly 135 species distributed mainly in tropical Africa and Asia. Molecular and biogeographic work suggests the genus originated in Africa and later diversified across tropical Asia, with the Indochina Peninsula emerging as one of its major centers of diversity, harboring more than 50 recognized species. The genus is best known to gardeners and field botanists for its enlarged, petal-like sepals, technically called calycophylls or semaphylls, which flash white or colored among the ordinary green foliage and attract pollinators to otherwise modest flowers. Classification within the group has traditionally rested on morphology: the shape and venation of leaves, the form of stipules, the architecture of inflorescences, the relative lengths of calyx tubes and lobes, and the shape of calyx and corolla lobes.

The newly revised species, Mussaenda wallichii, was described in 1834 by the Scottish botanist George Don from material collected at Tavoy and Rangoon in southern Myanmar, catalogued under Wallich specimen number 6258. From the outset the species carried complications. Don himself noted that the two voucher specimens behind the name did not represent Mussaenda longifolia, as the botanist Nathaniel Wallich had annotated on the labels. When the authors of the new study re-examined those original sheets, they discovered something more troubling: the two specimens, labelled 6258A and 6258B, are morphologically distinct from each other. Specimen 6258A matches the original description, with its terminal, trichotomous inflorescence. Specimen 6258B, a fruiting branch bearing densely congested, subcapitate berries on short pedicels, does not fit at all.

After extensive comparison with numerous herbarium specimens and the relevant literature, the researchers determined that the mismatched specimen 6258B actually belongs to a different species, Mussaenda roxburghii. The two taxa are most reliably separated by inflorescence architecture: M. wallichii bears spreading corymbs, whereas M. roxburghii produces capitate cymes, along with differences in fruit shape and the persistence of calyx lobes. Because an earlier lectotypification of Wallich 6258 had failed to single out one specimen among the original material, and because one of those specimens is now known to belong to another species, the team designated specimen 6258A as a second-step lectotype under Article 9.17 of the International Code of Nomenclature for algae, fungi, and plants. This nomenclatural housekeeping anchors the name M. wallichii to a specimen that genuinely matches its description.

The heart of the revision, however, concerns two younger names. Mussaenda intuspilosa was described in 1963 by Jayaweera on the basis of specimens from Assam in India and from Upper and Lower Myanmar, with a collection made near Clegg Hill on the east bank of the Irrawaddy in Kachin as the holotype. In its protologue, the species was distinguished by larger leaves, larger stipules with distinctive pubescence and glands, larger sepals, a diffuse inflorescence, and calyx segments that fall away from the fruit. Mussaenda lancipetala followed in 2008, described by Deng and Zhang from Mengla County in Xishuangbanna, western Yunnan, China, and differentiated by larger leaf blades, entirely glabrous upper leaf surfaces, more numerous pairs of secondary veins, diffuse inflorescences, and longer corolla tubes and lobes.

When the researchers placed the type specimens and protologues of all three names side by side, and combined this with field observations of living plants across the Indochina Peninsula, the supposed boundaries dissolved. No stable morphological character was found that consistently separates the three taxa. All share oblong-elliptic leaves that are cuneate at the base and acuminate at the apex, terminal trichotomous cymes, and ovoid to elliptic berries. The apparent differences dissolved under scrutiny. M. wallichii had been described as entirely glabrous, yet young branches and leaves of both it and M. intuspilosa carry sparse pubescence just like M. lancipetala, and the calycine leaflets of all three are pubescent along the veins. A re-examination of the M. lancipetala type revealed 8 to 11 pairs of secondary veins, matching the other two taxa rather than the 11 to 13 pairs recorded in the original description, a discrepancy the authors attribute to the counting of minor secondary veins. And although the protologue of M. lancipetala described caudate corolla lobes while that of M. intuspilosa described merely acuminate ones, abundant specimens show that corolla lobes vary continuously from acuminate to caudate, with the length of the terminal tail grading between the two extremes.

Geography tells the same story. The three taxa were described from ranges that partially overlap or run continuously into one another: M. wallichii from southern and southeastern Myanmar with extensions to the Nicobar Islands, Thailand, and Peninsular Malaysia; M. intuspilosa from Myanmar and adjacent Assam; and M. lancipetala from southern and southwestern Yunnan. Together these distributions form an unbroken belt from Assam through Myanmar to Yunnan, with no geographic discontinuity corresponding to any morphological break. The authors conclude that M. intuspilosa and M. lancipetala are heterotypic synonyms of M. wallichii, meaning they are separate names published for what is in reality a single biological species.

The revision also required untangling a nomenclatural irregularity. When M. lancipetala was published, the same gathering, Deng and Ren 410, was cited simultaneously as both holotype and isotype, which does not satisfy the Code’s requirement for a single holotype. Under the rules, the two duplicate sheets deposited at the herbarium IBSC must be treated as syntypes, and the team has now designated one of them, specimen IBSC0733503, as the lectotype. The Chinese name for the species, xia ban yu ye jin hua, is retained, because it was originally applied to M. lancipetala, now subsumed within M. wallichii. The revised species is a shrub with angular young branches, oblong-lanceolate to elliptic-lanceolate leaves 8.5 to 24 centimeters long, white ovate semaphylls up to 10 centimeters long, corolla tubes 2.5 to 3.6 centimeters long, and heterostylous flowers, a breeding system in which long- and short-styled floral forms promote cross-pollination. It occupies a striking breadth of habitats, from tropical lowland evergreen rain forests and dry evergreen forests to coastal forests, limestone hills, lower montane rain forests, and secondary growth, from near sea level to roughly 1,000 meters, flowering from April to October and fruiting from June to December. Given its wide range and numerous recorded populations, the authors assess it as Least Concern under the IUCN Red List criteria.

Why did three species become one? The authors point to a familiar pattern in tropical taxonomy: early treatments were often based on a small number of collections, leading variable characters to be overinterpreted as species-level differences. Pubescence density in Mussaenda varies among individuals and developmental stages, with young organs sometimes hairy and mature organs glabrescent, a pattern documented in other species of the genus as well. Vein counts, corolla lobe tips, and calycophyll hairiness all show continuous variation across populations rather than the discrete gaps that would justify separate species. The lesson extends beyond this one shrub. Previous estimates of species richness in Mussaenda have been shaped by incomplete knowledge of morphological variation and insufficient examination of type materials, and the new synthesis suggests the true diversity of the genus in Southeast Asia may be lower than current checklists imply. The authors argue that continued revisionary work integrating field observations, herbarium research, and molecular evidence will be essential for resolving the remaining uncertainties in this morphologically diverse genus, and for building the stable taxonomic framework on which conservation, ecology, and evolutionary studies ultimately depend.

Subject of Research: Taxonomic revision and synonymization of Mussaenda wallichii (Rubiaceae) in Southeast Asia

Article Title: Taxonomic revision of Mussaenda wallichii ( Rubiaceae ): two new synonyms from Southeast Asia

Article References: Duan, T.-T., Li, Y.-X., & Li, J. (2026). Taxonomic revision of Mussaenda wallichii (Rubiaceae): two new synonyms from Southeast Asia. PhytoKeys, 280, 93-104. https://doi.org/10.3897/phytokeys.280.209108

Image Credits: AI Generated

DOI: 10.3897/phytokeys.280.209108

Keywords: Mussaenda, Rubiaceae, taxonomy, lectotypification, Southeast Asia, herbarium, plant morphology, species delimitation, PhytoKeys, biodiversity, Indochina, heterotypic synonyms

The new species was found at an elevation of roughly 1,614 meters at the edge of alpine forest, growing as a biennial herb that can reach 150 centimeters in height. What immediately caught the surveyors’ attention was the plant’s covering of long-stalked glandular hairs, which clothe the stems, leaves, calyx, corolla, and capsule. At first glance the species superficially resembles V. macrocarpum and V. blattaria, but closer examination revealed a combination of characters not present in any other known Verbascum, prompting its description as new.

Confirming a new species in this genus is no small task. Verbascum comprises more than 360 species distributed mainly across Eurasia, and its taxonomy has been revised repeatedly since Linnaeus first separated the genus into Celsia and Verbascum based on stamen number in his 1753 Species Plantarum. Bentham later divided the group into sections according to anther morphology, and Murbeck subdivided it further using seed characters. Ferguson and Huber-Morath eventually merged Celsia back into Verbascum, a conclusion now strongly supported by molecular phylogenetic studies. Turkey holds the world record for mullein diversity with 254 species, 193 of them endemic, while Kazakhstan hosts nine species and China six, one of which remains incompletely known.

To verify the status of the Gongliu population, the researchers cross-checked their specimens against the Flora of China, Flora Xinjiangensis, Flora of Kazakhstan, Flora of Turkey and the East Aegean Islands, and the Flora of the USSR, as well as the diagnostic key of Huber-Morath. They also compared material held in the Herbarium of Shihezi University and consulted online databases including iPlant, Plants of the World Online, the Plant Photo Bank of China, and the Global Biodiversity Information Facility. Floral and fruit structures were dissected and examined under a stereomicroscope, and photographs and GPS coordinates were recorded in the field.

The morphological evidence proved decisive. Verbascum gongliuensis differs from its closest relative V. macrocarpum by bracts that are shorter than the flowers rather than longer, a smaller corolla, a denser glandular indumentum, and a capsule that conspicuously exceeds the persistent calyx. Compared with V. ponticum, its corolla measures 18 to 23 millimeters across rather than 26 to 33 millimeters, and its indumentum is exclusively glandular. It is separated from V. bornmuellerianum by the two anterior filaments being glabrous near the apex, whereas in that species the violet filament hairs extend right up to the anthers. From V. zerdust, which typically bears four stamens, the new species is readily distinguished by having five stamens, rarely four.

The detailed description records a biennial herb with unbranched or mid-branched stems 30 to 150 centimeters tall, densely glandular throughout. The basal leaves are ovate, elliptic to lanceolate, with bicrenate margins and undulate teeth, measuring 5 to 35 centimeters long and 3 to 6 centimeters wide on petioles of 3 to 6 centimeters. Cauline leaves are sessile, oblong-ovate to triangular-ovate, and dentate. The inflorescence is a simple or few-branched raceme with solitary flowers, each subtended by densely glandular bracts 0.5 to 1.5 centimeters long. The yellow corolla, 18 to 23 millimeters in diameter, carries purple and white appendages at the base of the upper two lobes, and its lobes are unequal, the upper two slightly smaller than the lower three. Stamens number four or five, with purple woolly filaments of unequal length, and the anthers are laterally compressed, curved, narrowly reniform, and decurrent at the base. The ovoid-globose capsule, 6 to 10 millimeters long, exceeds the calyx and bears a short beak about 1.5 millimeters long. Flowering occurs from June to August and fruiting from July to September, with cylindrical seeds 0.5 to 1 millimeters long marked by longitudinal undulate striations.

Geography and ecology add further weight to the species’ recognition. Verbascum gongliuensis occupies alpine steppe at around 1,600 meters in the central Tianshan Mountains, whereas V. blattaria inhabits lower-elevation riverbanks and marshes. Cultivation observations suggest the new species prefers humid, low-temperature environments, while V. blattaria tolerates high temperatures and strong sunlight. Although V. macrocarpum occupies a similar habitat, its nearest recorded population lies in the southern basin of Lake Alakol in Kazakhstan, roughly 800 kilometers from Kuerdening. The authors argue that this geographic isolation has driven divergence in corolla size, bract length, and stamen dimensions between the two taxa.

The type specimen was collected on 30 July 2024 at Kuerdening, at coordinates 43°08’46″N, 82°51’41″E, and deposited as a holotype and isotype at the Shihezi University herbarium under the collection number Wenbin Xu and Yuhao Luo 4546. The species carries the IPNI name identifier urn:lsid:ipni.org:names:77380708-1, and the paper is accompanied by detailed photographs of the plant, its flowers, leaves, capsules, and seeds, together with a distribution map and the holotype specimen image.

Conservation concerns shadow the discovery. The species is currently known from a single location, with an estimated extent of occurrence of less than 20 square kilometers and a population of fewer than 1,000 mature individuals. Its habitat is significantly threatened by grazing and tourism activities, which are projected to cause a continuing decline in the number of mature individuals. Applying the IUCN Red List Categories and Criteria, version 3.1, the authors assessed Verbascum gongliuensis as Vulnerable under criteria D1 and D2, underscoring how narrowly restricted alpine endemism can leave even a newly described species exposed to environmental pressures.

The description also required an updated identification key to the Verbascum species of China. The new species keys out alongside V. blattaria among the yellow-flowered, solitary-flowered, glandular-hairy members of the flora, from which it is separated by its longer-stalked glandular hairs and larger capsule of 6 to 10 millimeters. The full diagnostic comparison with V. macrocarpum and V. zerdust published in the paper documents differences across stems, leaves, bracts, pedicels, calyx, corolla, stamens, and capsules, providing a practical reference for future field surveys in the Tianshan range. As alpine grasslands across Central Asia continue to face grazing and tourism pressures, the discovery highlights both the botanical richness of the central Tianshan Mountains and the urgency of documenting its flora before such narrowly distributed species disappear.

Subject of Research: Verbascum gongliuensis ( Scrophulariaceae ), a new species of Yili, Xinjiang, China

Article Title: Verbascum gongliuensis ( Scrophulariaceae ), a new species of Yili, Xinjiang, China

Article References: Luo, Y., & Xu, W. (2026). Verbascum gongliuensis (Scrophulariaceae), a new species of Yili, Xinjiang, China. PhytoKeys, 280, 73-81. https://doi.org/10.3897/phytokeys.280.201229

Image Credits: AI Generated

DOI: 10.3897/phytokeys.280.201229

Keywords: Verbascum gongliuensis, new species, Xinjiang, Scrophulariaceae, PhytoKeys, taxonomy, IUCN, biodiversity, Verbascum, gongliuensis, species, Yili