Hidden fungal kingdoms in China’s mountains reveal gaps in nature reserves

Deep in the mountains of southwestern Sichuan, a team of mycologists has spent two field seasons picking their way through subtropical evergreen forests, cold-temperate conifer stands, bamboo groves and alpine meadows, collecting every mushroom, bracket fungus and puffball visible to the naked eye. What they found is forcing an uncomfortable question about how China protects its biodiversity: the country’s existing nature reserves, designed largely around pandas, tigers and trees, may be leaving a huge share of the fungal world outside their boundaries. In a study published in IMA Fungus, researchers report that of 809 macrofungal species documented in the southern section of the Hengduan Mountains, roughly a third occur only in areas that currently enjoy no formal protection at all.

The Hengduan Mountains, where the southeastern Qinghai-Xizang Plateau meets the western Sichuan Basin and the northwestern Yunnan-Guizhou Plateau, rank among the planet’s recognized biodiversity hotspots. Their southern section, a priority zone for conservation in China, spans an elevation gradient from 500 to 4500 meters and packs an extraordinary vertical succession of vegetation into a relatively compact area, from subtropical evergreen broad-leaved forests at the lowest altitudes through mixed conifer-broadleaf and coniferous forests up to alpine shrub and meadow habitats. That climatic and vegetative variety makes the region a natural laboratory for asking how fungal communities are organized, and whether the boundaries humans have drawn around protected land actually capture the full sweep of that organization.

Between July and September 2019 and August and September 2020, the researchers mounted four field expeditions across the Daxueshan-Xiaoxiangling and Liangshan Mountains. Before any sampling began, they divided the region into three elevation groups, high, mid and low, following the geographic transition from west to east, a division that also tracks major climate shifts and the living habitats and economic activities of local communities. In the end, they surveyed 177 sites, each roughly 400 square meters, covering every typical vegetation type in the research region. The high elevation group comprised 70 sites between 1963 and 4010 meters, the mid elevation group 77 sites between 1235 and 2963 meters, and the low elevation group 24 sites between 377 and 1753 meters. At each site, collection effort was deliberately standardized: four experienced investigators spent five hours gathering all visible macrofungi growing on soil, wood, leaf litter, dung and other substrates, so that differences between sites would reflect biology rather than sampling intensity.

The haul was remarkable. A total of 1771 macrofungal specimens were collected, dried on-site in a portable oven at 35 degrees Celsius overnight, frozen at minus 80 degrees for two weeks, and deposited in the fungarium of the Institute of Microbiology, Chinese Academy of Sciences in Beijing. Identifying what they had collected demanded both classical morphology and modern genomics. Under stereomicroscopes and optical microscopes at magnifications up to 1000 times, the team examined macroscopic and microscopic features using cotton blue, Melzer’s reagent and potassium hydroxide. Then, DNA extracted from each dried specimen was amplified for two genetic barcoding regions, the internally transcribed spacer (ITS) and the nuclear large subunit (nLSU), yielding 1460 newly generated ITS sequences and 1477 nLSU sequences. BLAST searches against existing databases, with a similarity threshold above 98 percent and coverage above 90 percent, provided first-pass identifications, and where sequences and morphology proved ambiguous, the researchers built maximum likelihood phylogenetic trees to place difficult specimens in well-supported clades with known species.

The result: 809 species accurately identified, belonging to two phyla, nine classes, 28 orders, 123 families and 336 genera. The dominant orders were Agaricales with 270 species, Polyporales with 161, Russulales with 109 and Hymenochaetales with 103, while the most speciose family was Russulaceae with 50 species and the most species-rich genus Russula with 30. At the other end of the spectrum, 195 genera were represented by a single species each, and 255 genera together accounted for 76 percent of all genera but only 39 percent of species, a signature of extraordinary taxonomic rarity that conservation planners rarely see in such detail for fungi. About half of the 809 species were newly recorded in the study region, and seven are listed as threatened or near threatened on China’s macrofungal Red List, including the vulnerable Engleromyces sinensis and Tricholoma matsutake, the prized matsutake mushroom, along with five near threatened species such as Cordyceps militaris and Ganoderma applanatum.

With species identities in hand, the team turned to community phylogenetics, a set of techniques that treats evolutionary history as data. They selected one specimen per species, aligned the ITS and nLSU sequences, extracted conserved sites with Gblocks, and concatenated them into a single matrix of 2804 characters. Using IQ-TREE, they estimated the best-fit evolutionary model, GTR plus invariant sites plus gamma rate variation, and constructed a phylogenetic tree. Molecular clock dating with the penalized likelihood method in treePL calibrated this topology using three fossil fungi: Archaeomarasmius leggetti, a 90-million-year-old agaric; Quatsinoporites cranhamii, a 125-million-year-old russulalean; and Paleopyrenomycites devonicus, a 400-million-year-old ascomycete-basidiomycete divergence. From the resulting time-calibrated ultrametric tree, they calculated phylogenetic diversity, the net relatedness index and the nearest taxon index for each site, using 999 randomizations under a null model to test whether communities were more clustered or more dispersed than chance would predict.

The ecological verdict was subtle and, in places, surprising. Species richness and phylogenetic diversity showed no significant differences among the high, mid and low elevation groups, and the net relatedness and nearest taxon indices were generally greater than zero, indicating phylogenetic clustering within each group. When the team examined beta diversity, the turnover of phylogenetic composition between groups, the beta net relatedness and beta nearest taxon indices all fell between minus 2 and 2, the statistical signature of stochastic rather than deterministic assembly. The Bray-Curtis-based Raup-Crick index drove the point home: ecological drift contributed 93.11 percent of community assembly in the high elevation group, 85.79 percent in the mid elevation group and 80.57 percent in the low elevation group, with heterogeneous selection, homogeneous selection, dispersal limitation and homogenizing dispersal together accounting for the remainder. In plain terms, macrofungal communities across this mountain landscape are assembled mainly by random processes shaped by ecological habitats, not by the kind of strong environmental filtering that host identity or soil chemistry might impose on other fungal groups.

Yet similarity in richness and assembly mechanism masked profound compositional divergence. The mid elevation group harbored the most unique species, 262 found nowhere else in the study region, followed by the high elevation group with 222, while the low elevation group held 64 unique species, fewer than the 67 species shared among all three groups. Principal coordinate analysis confirmed that species composition differed significantly among the three groups, with the low elevation group the most distinct. The Dufrêne-Legendre indicator value method, run with 999 permutations, identified 11 indicator species, including Ampulloclitocybe clavipes exclusive to the high elevation group and ten species such as Bovista nigrescens, Daedaleopsis confragosa and Peniophorella subpraetermissa characteristic of the low elevation group. Linear discriminant analysis effect size analysis, with a Kruskal-Wallis screen at alpha 0.01 and an LDA score above 4, further revealed that Russulaceae and Russula were disproportionately prominent in the high elevation group, while Rigidoporaceae, Haploporus, Lentinus and Pseudohydnum stood out in the low elevation group.

The conservation implications are the study’s sharpest edge. Of the 177 sampling sites, 103 fell within protected areas and 74 outside them, and although the non-protected sample was smaller, species richness showed no significant difference between the two categories. More strikingly, 261 species, about 32 percent of everything identified across the region, were unique to non-protected areas. Among them were new records for China of endemic species such as Hymenochaete yunnanensis, Coltricia weii and Hyphoderma subsetigerum, found around Xichang City, and the near threatened Laccaria alba, discovered in a heavily grazed, unprotected area of Mianning County. China’s five national parks tell the story of why this happens: two were established for animals, the Giant Panda and Northeast China Tiger and Leopard parks; one for forests, the Hainan Tropical Rainforest; and two for ecosystems, the Three-River-Source and Wuyishan parks. None was conceived with the habitat requirements of macrofungi in mind, and fungi represent just 0.8 percent of the 1.3 million-plus species assessed on the IUCN Red List, a knowledge gap researchers call the Scottian shortfall.

The researchers propose a replicable framework that begins with dividing a region into meaningful units by elevation, climate or habitat, conducting standardized field surveys within each, identifying species through morphology and DNA barcoding with voucher specimens preserved in fungaria, and then analyzing richness, phylogenetic diversity, indicator species and community assembly to flag units most urgently needing protection. They recommend establishing additional nature reserves in newly identified non-protected areas harboring unique macrofungal species, such as Xichang, and building long-term monitoring programs around indicator species like Ampulloclitocybe clavipes and Peniophorella subpraetermissa to track the spatiotemporal dynamics of fungal communities. They also acknowledge the limits of their own work: two years of July-to-September surveys cannot capture species that fruit in other seasons or exist only as invisible mycelia, and combining fruiting-body inventories with environmental DNA sequencing, which neither approach alone can replace, would yield more reliable evaluations. The framework, they hope, will be adapted by scientists to other regions and ecosystem types, turning what has long been an afterthought in conservation policy into a scientifically grounded priority.

Subject of Research: Macrofungal diversity and conservation priority assessment in the southern Hengduan Mountains, China

Article Title: Assessing macrofungal diversity to identify conservation priority areas: a case study from the southern section of the Hengduan Mountains, China

Article References: Liu, S.-L., Liu, D.-M., Shen, S., Wang, X.-W., Li, Q.-Z., Yu, J., Li, X.-C., & Zhou, L.-W. (2026). Assessing macrofungal diversity to identify conservation priority areas: a case study from the southern section of the Hengduan Mountains, China. IMA Fungus, 17, Article e168781. https://doi.org/10.3897/imafungus.17.168781

Image Credits: AI Generated

DOI: 10.3897/imafungus.17.168781

Keywords: macrofungi, fungal conservation, Hengduan Mountains, biodiversity hotspot, phylogenetics, community assembly, nature reserves, DNA barcoding, elevation gradient, IMA Fungus, Assessing, macrofungal