The discovery matters far beyond a simple country tick. Odorrana nagao belongs to the O. lipuensis species group, a phylogenetically distinct lineage of frogs whose members are intimately tied to limestone landscapes — karst streams, cave mouths and subterranean aquatic habitats. All six species currently recognised in the group have been described since 2015, a rate of discovery that suggests scientists are only now beginning to catalogue the diversity sheltered within the fragmented, rugged and often dangerously inaccessible karst blocks of southern China and northern Vietnam. With the new record, the number of Odorrana species known from China rises from 47 to 48, and Yunnan — the province with the highest documented amphibian richness in the country — now harbours 17 species of the genus.

The genus Odorrana, the Asian cascade frogs, sprawls across an enormous range: from north-eastern India and southern mainland China, through mainland Southeast Asia, to the Malay Peninsula and the Greater Sunda Islands, and eastwards to Hainan, Taiwan and the Ryukyu Archipelago. Most species live along forested streams, but several have evolved a striking affinity for limestone. That affinity is no accident of geography. Karst systems are environmentally heterogeneous and fragmented, scattered across the landscape like islands, which restricts dispersal and promotes local differentiation of populations. Add the sheer difficulty of surveying many limestone areas, and it becomes clear why the amphibian diversity of these terrains remains incompletely documented.

The trail to the new record began in May 2026, during fieldwork in Laoshan Provincial Nature Reserve in Malipo County, Yunnan, at an elevation of 1,030 metres above sea level. Three adult female Odorrana, referable to the O. lipuensis species group on external appearance, were collected after complete darkness on limestone rocks near the entrance of a cave — a microhabitat closely matching that reported at the species’ type locality in Vietnam. The specimens were euthanised with tricaine methanesulphonate, fixed in buffered formalin and preserved in 75% ethanol in the herpetological collection of Sun Yat-sen University in Guangzhou, with thigh-muscle tissue retained in 95% ethanol for molecular work.

In the laboratory, the team amplified partial fragments of two mitochondrial markers — the 16S ribosomal RNA gene and the cytochrome c oxidase subunit I gene, COI — following established primers and protocols. The final concatenated alignment spanned 1,630 base pairs, comprising 1,036 base pairs of 16S and 594 of COI, and included 23 ingroup samples of Odorrana plus two samples of the related genus Nidirana as outgroups. Phylogenetic relationships were inferred under both Maximum Likelihood and Bayesian Inference, with substitution models selected separately for 16S and the three COI codon positions using the Bayesian information criterion. Nodal support was assessed with 1,000 SH-like approximate likelihood-ratio test replicates and 1,000 ultrafast bootstrap replicates, alongside Bayesian posterior probabilities from two independent MrBayes runs of 100 million generations each.

The results were unambiguous. The six sampled species of the O. lipuensis species group formed a well-supported lineage distinct from other Odorrana, with combined support values of 1.00 for Bayesian posterior probability, 99.4 for SH-aLRT and 100 for ultrafast bootstrap. Within that lineage, the three Yunnan specimens clustered with two specimens from the type series of O. nagao from Tuyen Quang Province in a strongly supported clade (1.00/98.3/99). Genetic distances sealed the case: the Kimura two-parameter distance between the Chinese frogs and the Vietnamese type series was just 0.38% for 16S and 1.73–1.90% for COI, whereas the distance to O. feii, the genetically closest sampled congener, was 2.11% for 16S and 6.25% for COI. The three Chinese specimens showed almost no divergence among themselves — 0 to 0.19% for 16S, with identical COI sequences.

Morphology agreed with the molecules. The three females are of moderate size, with snout–vent lengths of 56.6, 57.0 and 59.5 millimetres. Their heads are longer than wide, their tympana distinct, round and roughly two-thirds of the eye diameter, and their snouts obtusely rounded in dorsal view, projecting beyond the lower jaw. Their fingers bear the group’s signature features — the discs of Fingers III and IV prominently enlarged and transversely oval, with the disc of Finger IV the widest, and the tip of Finger I lacking a lateroventral groove. Their toe webbing follows the diagnostic formula I0–1II0–1III0–1⅓IV1⅓–¼V, and the tibio-tarsal articulation reaches the tip of the snout when the hind-limb is adpressed along the body. In life, the dorsum is moss-green to olive-green with irregular dark-brown blotches, the flanks brownish-green with darker mottling, the upper lip mottled brown and cream, the iris golden brown and the ventral surfaces creamy-white. No diagnostically relevant differences separate them from the type series.

The Chinese specimens do, however, push the species’ known physical limits. They are larger than the females of the Vietnamese type series, which measured 47.1–54.7 millimetres in snout–vent length, extending the known female range of the species to 47.1–59.5 millimetres. The record also raises the documented upper elevational limit by 257 metres, from 773 metres at the type locality to 1,030 metres in Yunnan. Mature ovarian oocytes were present in all three females collected in May, hinting at reproductive activity during this period, though the species’ breeding season, reproductive microhabitat and larval ecology remain unknown. Minor variation was noted in the ratio of internarial to interorbital distance, which is slightly greater than one in the Chinese specimens but was reported as 0.94 in the male holotype — a difference the authors attribute to intraspecific variation.

Biogeographically, the record is entirely plausible. Malipo County and the type locality in Cham Chu Nature Reserve lie within the same transboundary limestone belt, unseparated by any obvious major ecological or geographical barrier. The country record is new, but the presence of the species in southern China is not unexpected and probably reflects incomplete survey coverage in the border region rather than a genuine distributional gap. Extensive limestone formations and potentially suitable karst forests occur between and around the two known localities, and additional populations may well occur in Tuyen Quang and Cao Bang Provinces of northern Vietnam, in Wenshan Prefecture of south-eastern Yunnan, and in other karst landscapes along the international border. The finding adds to a growing body of evidence that the herpetofaunas on either side of the China–Vietnam border are closely connected, with recent studies repeatedly revealing comparable transboundary distributions and previously unrecognised diversity.

Conservation questions remain wide open. Odorrana nagao is currently known from only two confirmed localities, both within protected areas — an encouraging circumstance, but one that does not by itself demonstrate stable populations or the absence of threats. Nothing is yet known about population size, trends, reproductive success or connectivity between the two populations, and the species’ extent of occurrence and area of occupancy cannot be estimated reliably. Potential threats to limestone ecosystems, including quarrying, forest degradation, tourism development and hydrological alteration, have not been evaluated across its possible range. The team provisionally recommends assessing the species as Data Deficient under the IUCN Red List Categories and Criteria, and calls for targeted surveys across the limestone landscapes between and surrounding the two confirmed localities. Notably, while the present manuscript was under peer review, a second team independently published the first record of O. nagao from China, based on a single female collected in Tianbao Town, Malipo County, in 2021 — a study the authors acknowledge as preceding their account, with the two works complementing one another and jointly strengthening the evidence for the species’ occurrence in south-eastern Yunnan. Coordinated cross-border surveys integrating voucher specimens, morphology, molecular data and acoustic observations, the authors argue, will be essential for refining the distribution of O. nagao — and may yet reveal overlooked populations or undescribed diversity within this enigmatic limestone lineage.

Subject of Research: Taxonomy and Biodiversity

Article Title: Range extension of Odorrana nagao ( Anura , Ranidae ) to south-western China, with comments on the limestone-associated Odorrana lipuensis species group

Article References: Song, H.-M., Zhou, D.-Y., Qi, S., Lu, M.-X., Poyarkov, N., Nguyen, T., & Wang, Y.-Y. (2026). Range extension of Odorrana nagao (Anura, Ranidae) to south-western China, with comments on the limestone-associated Odorrana lipuensis species group. Biodiversity Data Journal, 14, Article e213369. https://doi.org/10.3897/bdj.14.e213369

Image Credits: AI Generated

DOI: 10.3897/BDJ.14.e213369

Keywords: Range, extension, Odorrana, nagao, Anura, Ranidae, south-western, China, comments, limestone-associated, lipuensis, species

The vegetation in question belongs to the class Cytisetea scopario-striati, communities dominated by woody members of the pea family such as Cytisus, Genista, Adenocarpus and Spartium. These shrubs share a distinctive strategy: reduced leaves and flexible, green photosynthetic branches that allow rapid growth on nutrient-poor, acidic soils. Thanks to their symbiosis with nitrogen-fixing bacteria, brooms enrich the soil and accelerate ecosystem recovery, typically marking successional stages on the way to oak, chestnut or beech forests. Their optimal distribution spans the thermo- to supramediterranean bioclimatic belts under subhumid to humid conditions, with the Iberian Peninsula hosting the greatest concentration of these communities.

Northwards, winter cold progressively strips these shrublands of their characteristic species, confining them to a narrow Atlantic coastal strip. South-eastwards, their post-glacial expansion followed a Mediterranean corridor through southern France and the Tyrrhenian slopes of the Apennines, with Sicily representing the south-western limit of their establishment. In Sicily, they occur mainly in the north-eastern mountains on acidic soils derived from metamorphic siliceous rocks, quartz sandstones or volcanic scoriae, becoming increasingly sporadic towards the west of the island.

The research team, led by Gianguzzi and Guarino, sampled 207 vegetation plots across the Tyrrhenian ridge of northern Sicily, recording 398 vascular plant species using the classical Braun-Blanquet phytosociological method. Between May and June 2019, they sampled 50 new plots, adding them to 78 unpublished surveys conducted between 1996 and 2018, plus 79 relevés from the literature covering related communities in Calabria. Plot sizes ranged from 50 to 150 square metres, with canopy heights between 0.7 and 2.8 metres and ground cover between 80 and 100 percent.

The classification relied on a modified TWINSPAN algorithm, a two-way indicator species analysis that divides plots based on floristic similarity. Species fidelity was evaluated using the standardized phi coefficient, with diagnostic species defined as those showing a phi coefficient of at least 0.25 and a probability of occurrence under random expectation below 0.05. Non-metric multidimensional scaling (NMDS) based on Bray-Curtis dissimilarity was then used to visualize the main compositional gradients, while ecological indicator values from the EIVE 1.0 database helped infer site conditions difficult to measure directly in the field.

The analysis revealed a fundamental split. The first group comprises shrublands developing under harsh ecological conditions between 1,000 and 1,500 metres on exposed, wind-swept ridges with shallow, skeletal soils. These are stable, long-lasting vegetation types maintained by topographic exposure and soil constraints, hosting endemic or relict species such as Adenocarpus commutatus, Viola aethnensis subsp. messanensis, Genista brutia and Silene sicula. Rarefaction curves confirmed that this group carries significantly more endemic species than the second group. These communities belong to the alliance Violo messanensis-Adenocarpion commutati, whose name was formally corrected in the study.

The second group tells a very different story. These are secondary shrublands, mostly the result of land abandonment after centuries of cutting, burning and grazing that prevented forest recovery. They occur at lower elevations, between 300 and 1,000 metres, on deeper and more evolved soils, and are dominated by widespread species such as Cytisus villosus, Spartium junceum, Teline monspessulana and Pteridium aquilinum. In the absence of further disturbance, they can progressively facilitate oak forest regeneration by improving soil conditions and microclimate. The study proposes a new alliance for these communities, named Achilleo ligusticae-Telinion monspessulanae, interpreted as a more thermophilous and mesophilous vicariant of the montane alliance.

Within these two groups, the team identified 15 terminal clusters, eight of which represent newly defined plant communities. Among the new associations are Carlino globosae-Cytisetum scoparii, Origano viriduli-Cytisetum villosi, Carlino globosae-Telinetum monspessulanae and Achilleo ligusticae-Spartietum juncei. One cluster, dominated by Cytisus infestus and the calciphilous Phlomis fruticosa, occurs exclusively in the central Peloritani Mountains on calcschist outcrops of the Mandanici unit, where local carbonate enrichment partially suppresses the typical acidophilous flora yet allows the class’s species to persist.

The classification of these secondary shrublands poses intrinsic difficulties, as they lack narrowly defined characteristic taxa and show broad floristic overlaps with gradual transitions between neighbouring units. The authors acknowledge that their proposed associations should be regarded as interpretative constructs aimed at organizing continuous ecological variation into a coherent framework, rather than as sharply delimited entities. This pragmatic approach, based on physiognomic dominance and ecological coherence, aligns with recent perspectives in vegetation science that emphasize ecological realism over rigid typological subdivision.

Beyond the science, the study carries a cultural dimension. Until the mid-twentieth century, broom shrubs were widely exploited by rural communities for poles in vineyards, firewood for bread ovens and lime kilns, and bracken was cut for livestock bedding. Rich vernacular nomenclature testified to their importance: in the Nebrodi village of Frazzanò, Cytisus infestus was known as Làstra and Spartium junceum as Jnèstra. The abandonment of these traditional practices after World War II altered disturbance regimes, favouring shrub encroachment and paradoxically increasing fire frequency. The broom shrublands of northern Sicily, the authors conclude, should not be regarded merely as degraded vegetation but as key components of Mediterranean cultural landscapes shaped by the long-term interaction between ecological processes and human management.

Subject of Research: Broom-dominated shrublands of Sicily: revisiting the syntaxonomy of the Cytisetea scopario-striati at the south-eastern margin of their range

Article Title: Broom-dominated shrublands of Sicily: revisiting the syntaxonomy of the Cytisetea scopario-striati at the south-eastern margin of their range

Article References: Broom-dominated shrublands of Sicily: revisiting the syntaxonomy of the Cytisetea scopario-striati at the south-eastern margin of their range. (n.d.). https://doi.org/10.3897/VCS.183815

Image Credits: AI Generated

DOI: 10.3897/VCS.183815

Keywords: Broom-dominated, shrublands, Sicily, revisiting, syntaxonomy, Cytisetea, scopario-striati, south-eastern, margin, range, scientific research