The researchers collected fresh specimens representing the three genera during macrofungal surveys in Zhashui County and Shangzhou District, both in Shaanxi Province. Basidiomata were photographed and described in the field, including their substrates and habitats. The specimens were dried and deposited in the herbarium of the Institute of Microbiology at Beijing Forestry University.

Microscopic examination included slide preparations in 5% potassium hydroxide, Cotton Blue and Melzer’s reagent. At least 30 basidiospores were measured from each specimen, with sterigmata excluded from basidial measurements and hilar appendages excluded from basidiospore measurements. Genomic DNA was extracted with a CTAB rapid plant genome DNA extraction kit, and the ITS and nLSU regions were amplified and sequenced. The resulting sequences were combined for separate phylogenetic analyses of Hymenochaete, Hyphoderma and Steccherinum.

Maximum-likelihood and Bayesian analyses placed the specimens in three independent lineages within their respective genera. Hymenochaete zhashuiensis, represented by specimen Cui 24846, was closely related to H. tasmanica and had 100% maximum-likelihood bootstrap support and a Bayesian posterior probability of 1.00. The two Hyphoderma specimens, Cui 24841 and Cui 24844, formed an independent clade with 98% maximum-likelihood bootstrap support and a Bayesian posterior probability of 0.99. Three Steccherinum specimens, Cui 24842, Cui 24843 and Cui 24845, formed a distinct lineage with 91% maximum-likelihood support and a Bayesian posterior probability of 0.99, close to S. fissurutum and S. fragrans.

Hymenochaete zhashuiensis has annual, resupinate to effused-reflexed basidiomata measuring 15–20 cm long, 8–12 cm wide and 0.5–3 mm thick. Its hymenial surface is reddish brown to rust-brown when fresh, and its hyphal system is monomitic, with simple-septate generative hyphae. Diagnostic features include abundant dark-brown, thick-walled hymenial setae measuring 66–89 × 7.8–11 µm, branched dendrohyphidia and ellipsoid to broadly ellipsoid basidiospores measuring 4–5.6 × 2.8–3.5 µm. The holotype was collected from a living angiosperm tree in Zhashui County at an elevation of 1,184.0 m.

Hyphoderma brevicystidiatum has white to cream, resupinate and membranaceous basidiomata that are 15–23 cm long, 2.5–4 cm wide and 0.3–1 mm thick. The species has a monomitic hyphal system with clamped generative hyphae and a smooth to slightly grandinioid hymenial surface. Its cylindrical to subcylindrical cystidia measure 24.2–30.5 × 4.2–7.4 µm and are mostly smooth, although some bear crystalline deposits. The basidiospores measure 6.2–10 × 2.7–4.2 µm. The holotype, Cui 24844, was found on a fallen Pinus branch in Shangzhou District at an elevation of 701.6 m; specimen Cui 24841 was collected from a fallen angiosperm branch in Zhashui County.

Steccherinum shangzhouense has annual, resupinate, membranaceous basidiomata that are 13–18 cm long, 1–2 cm wide and up to 120 µm thick. Its hymenial surface is white to cream when fresh, becomes pale gray to grayish upon drying and develops conspicuous cracks. The species has a monomitic hyphal system with simple-septate generative hyphae; cystidia and skeletocystidia were not observed. Its broadly ellipsoid to oblong ellipsoid basidiospores measure 5.2–7.9 × 2.5–5 µm. The holotype, Cui 24845, was collected from a fallen angiosperm branch in Shangzhou District at an elevation of 852 m, with two additional specimens from the district serving as paratypes.

The study was limited to the morphology, DNA data and phylogenetic comparisons available for the collected specimens and selected reference taxa. Its authors described the three species as distinct based on the combined evidence, while noting that research on corticioid-fungus diversity in the Qinling Mountains remains limited. The findings add three named taxa to the documented diversity of wood-inhabiting fungi in the region.

Subject of Research: Taxonomic identification of three new wood-inhabiting corticioid fungal species in China’s Qinling Mountains

Article Title: Three new wood-inhabiting corticioid fungi from the Qinling Mountains, China

Article References: Wan, L.-L., Zhang, L.-S., Dong, J.-H., Zeng, L., Liu, S., & Cui, B.-K. (2026). Three new wood-inhabiting corticioid fungi from the Qinling Mountains, China. MycoKeys, 138, 57-94. https://doi.org/10.3897/mycokeys.138.204396

Image Credits: AI Generated

DOI: 10.3897/mycokeys.138.204396

Keywords: mycology, taxonomy, corticioid fungi, wood-inhabiting fungi, Hymenochaete, Hyphoderma, Steccherinum, phylogenetics, Shaanxi Province, Qinling Mountains, Three, wood-inhabiting

A new study published in the journal Nature Conservation tackles that question by reconstructing the species’ potential distribution across six climatic periods, from the Last Interglacial roughly 120,000 years ago to projected conditions in the 2050s and 2070s. Using an ecological niche modeling approach known as maximum entropy, or MaxEnt, the researchers combined 75 rigorously filtered occurrence records with 19 bioclimatic variables drawn from field surveys and herbarium collections. The result is the first comprehensive climate-driven distribution modeling ever conducted for this species, and it fills a long-standing gap in understanding how this orchid has weathered the Quaternary glacial cycles and what its future may hold.

The modeling workflow began with an extensive data-cleaning effort. The team gathered 229 latitude and longitude records from the Chinese Virtual Herbarium, the Global Biodiversity Information Facility, the Flora of China, and their own fieldwork, then applied a one-square-kilometer buffer across China to eliminate spatial bias from clustered sampling. Redundant and misidentified records were removed, leaving a single occurrence point per buffer zone. This kind of spatial thinning is critical for niche models, which can be misled by dense clusters of records that reflect collector access rather than true ecological preference.

Climate data came from the WorldClim database at a 30-arc-second resolution, generated by the CCSM4 general circulation model developed at the National Center for Atmospheric Research under the Coupled Model Intercomparison Project Phase 5. For future projections, the researchers used the Representative Concentration Pathway 2.6 scenario, a relatively optimistic low-emission trajectory aligned with global mitigation goals. Because multicollinearity among the 19 bioclimatic variables can cause model overfitting, the team screened them with Pearson correlation analysis, removing any factor with a correlation coefficient at or above 0.8. Seven variables survived the screening, and four of them ultimately dominated the model’s predictions.

Those four factors tell a clear story about what this orchid needs. Precipitation of the driest month contributed the most, at 30.6 percent, followed by annual precipitation at 27.0 percent, mean temperature of the warmest quarter at 25.9 percent, and mean diurnal temperature range at 8.5 percent. Together, these four variables accounted for 92 percent of the model’s total contribution. The pattern makes ecological sense: orchids are notoriously sensitive to water availability and seasonal drought, and C. graciliflora inhabits subtropical monsoon zones where rainfall is abundant but strongly seasonal. The response curves revealed surprisingly narrow tolerances. When the probability of occurrence exceeded 0.5, annual precipitation fell between 1,500 and 2,800 millimeters, precipitation of the driest month between 30 and 170 millimeters, mean diurnal range between 6.3 and 8 degrees Celsius, and mean temperature of the warmest quarter between 21.5 and 25 degrees Celsius. These thresholds represent the climatic envelope within which the species can grow and reproduce normally.

The model’s performance was exceptional. The average area under the receiver operating characteristic curve, a standard measure of predictive accuracy, reached 0.989 with a standard deviation of only 0.001, far exceeding the random-prediction benchmark of 0.5. With that level of confidence, the researchers could trace the species’ range through time with meaningful precision. During the Last Interglacial, suitable habitat covered roughly 123,770 square kilometers, concentrated in present-day Chongqing, Yunnan, Guizhou, Hainan, Fujian, and Zhejiang, with a notable concentration in the southern Hengduan Mountains. During the Last Glacial Maximum, those southern Hengduan habitats vanished entirely while suitable areas expanded northward into the Qinling Mountains, the Luoxiao Mountains, and Mount Wuyi.

That shift during the Last Glacial Maximum is the study’s most striking biogeographic finding. It suggests that the Qinling Mountains, Luoxiao Mountains, and Mount Wuyi served as glacial refugia where the orchid survived the coldest millennia of the Pleistocene. The evidence points to two distinct survival strategies operating simultaneously. In the Qinling Mountains, the species appears to have followed a migratory model, retreating northward into northern subtropical transitional forests as ice advanced. At Mount Wuyi, by contrast, suitable habitat persisted continuously through every Quaternary climatic event, consistent with a model of in situ evolution, where populations endured through genetic variation or phenotypic plasticity rather than range shifts. Both patterns have been documented in other Chinese plant species responding to glacial cycles, but this is the first time they have been identified for C. graciliflora specifically.

By the Middle Holocene, roughly 6,000 years ago, the distribution had largely returned to its present-day configuration, with highly suitable areas concentrated along the border of Chongqing and Hubei and in southern Hunan, Fujian, Zhejiang, and Taiwan. Today, the total suitable habitat covers approximately 98,080 square kilometers, about 1.014 percent of China’s land area, centered on the subtropical evergreen broad-leaved forests of central and eastern China. The highly suitable core regions sit in the Jiangxi-Hunan border area and in Fujian and Zhejiang, closely associated with the Luoxiao Mountains and Mount Wuyi, where montane forests between 600 and 2,500 meters elevation provide the humid, acidic, humus-rich soils and shaded understories the orchid requires.

The future projections are sobering. Under the RCP2.6 scenario, total suitable habitat is expected to shrink by approximately 52 percent by the 2050s, dropping from 98,080 to about 46,860 square kilometers, before partially recovering to roughly 69,400 square kilometers by the 2070s. The highly suitable area is projected to collapse by nearly 60 percent, from 29,220 to 11,700 square kilometers, in the 2050s, then rebound to 25,910 square kilometers by the 2070s. The researchers interpret this bimodal decline-and-recovery pattern as evidence of the species’ historical resilience to climate fluctuations. The initial crash likely reflects rapid warming that disrupts the fog-shade equilibrium in core montane habitats, while the subsequent recovery may result from elevational migration as warming pushes subtropical forest communities upward into newly suitable zones. Notably, a highly suitable region is predicted to emerge in the high-altitude areas of Taiwan by the 2050s but disappear again by the 2070s, underscoring how transient even newly available habitat can be under continued warming.

The conservation implications are concrete. The authors recommend prioritizing in situ protection south of the Qinling-Huaihe line, particularly montane forests between 800 and 1,200 meters in the Nanling Mountains and between 1,200 and 1,600 meters on Mount Wuyi, integrating these zones into existing or planned national park systems. For newly identified climatically suitable areas such as Taiwan, small-scale protected zones could facilitate species migration and buffer against human impacts. Ex situ measures, including seed banking, botanical garden cultivation, and assisted reproduction, are recommended where natural habitats are projected to decline, especially given the species’ characteristically low seed set and germination rates. The study also cautions that its focus on bioclimatic variables alone leaves out topography, competition, and other biological factors, and that plant responses to climate change often lag behind model projections. Still, as a baseline for future research on Calanthe and other terrestrial orchids, and as a call to incorporate climate vulnerability into Red List reassessments for a species already listed as near threatened and regulated under CITES, the work offers a template for how phylogeography and conservation planning can converge on the mountains where this resilient orchid has survived everything the past could throw at it.

Subject of Research: Climate-driven distribution modeling and conservation of the Chinese endemic orchid Calanthe graciliflora

Article Title: The past, present, and future distribution of Calanthe graciliflora : implications for conservation and phylogeography

Article References: Guo, P., Lu, A., Zheng, J., Chen, L., Wu, S., Hu, C., Li, M., Huang, W., & Zhai, J. (2025). The past, present, and future distribution of Calanthe graciliflora: implications for conservation and phylogeography. Nature Conservation, 60, 21-38. https://doi.org/10.3897/natureconservation.60.156661

Image Credits: AI Generated

DOI: 10.3897/natureconservation.60.156661

Keywords: Calanthe graciliflora, orchid conservation, MaxEnt, ecological niche modeling, glacial refugia, climate change, Qinling Mountains, Mount Wuyi, phylogeography, Last Glacial Maximum, habitat suitability, subtropical forests

Phytophthora, a genus of oomycetes first established by Anton de Bary in 1876, includes species responsible for devastating agricultural and forestry epidemics worldwide. These soilborne pathogens drive widespread dieback and decline in natural forest ecosystems while simultaneously inflicting severe economic losses on crops. Yet despite their importance, the evolutionary significance of mitochondrial genomic variation within and between species has remained poorly understood, particularly in pristine natural ecosystems where these organisms likely originated and diversified over deep time.

The research team, publishing in IMA Fungus, isolated 143 Phytophthora strains from rhizosphere soils of dominant tree species across the Qinling Mountains using the leaf-baiting method. From these, eleven representative species were selected for whole-genome sequencing on an Illumina NovaSeq 6000 platform. The resulting mitochondrial genomes, assembled and annotated using specialized bioinformatic pipelines, ranged in size from 37,502 base pairs in P. citrophthora to 41,542 base pairs in P. cryptogea, all with a remarkably uniform GC content averaging approximately 22 percent.

What emerged from the analysis was a picture of extreme genomic conservatism punctuated by dramatic structural events. All eleven mitogenomes shared a conserved suite of 34 core protein-coding genes, two ribosomal RNA genes, and between 24 and 25 transfer RNA genes, with no introns disrupting any of the newly sequenced genomes. This rigid gene complement, encompassing ATP synthase subunits, NADH dehydrogenase subunits, ribosomal proteins, and cytochrome oxidase subunits, reflects the essential role these genes play in oxidative phosphorylation and energy metabolism, leaving little room for evolutionary experimentation.

Perhaps the most striking finding came from selective pressure analysis. The ratio of nonsynonymous to synonymous substitutions, known as Ka/Ks, fell significantly below 1 for all 34 core protein-coding genes, ranging from 0.0096 in nad4L to 0.1737 in rpl6. This pattern demonstrates unequivocally that purifying selection is the dominant force shaping Phytophthora mitochondrial genome evolution. Core respiratory chain genes like nad4L, cox1, and atp1 showed exceptionally low ratios, reflecting strict functional conservation, while ribosomal protein genes exhibited slightly higher values, suggesting greater structural flexibility in these components of the translation machinery.

Beyond sequence-level evolution, the researchers uncovered a remarkable pattern of structural organization. By comparing gene arrangements across all eleven species, they classified the mitogenomes into three distinct types. Type I, the conserved type, included five species from Subclade 2c that exhibited completely identical gene orders, reflecting an ancestral state inherited from their common ancestor. Type II, the rearranged type, comprised P. citrophthora and P. himalsilva from Clade 2a, which showed a distinct inversion and translocation of gene blocks. Type III, the inverted type, included four species characterized by large-scale inversions of the genomic region between nad3 and atp1. These three arrangement types mapped cleanly onto phylogenetic clades, suggesting that structural rearrangements serve as stable macro-evolutionary markers of lineage divergence.

To place these genomic patterns in an evolutionary timeframe, the team employed Bayesian relaxed molecular clock analyses calibrated with fossil and biogeographical constraints. The results were staggering: the deep evolutionary split between Phytophthora and its pythiaceous relative Globisporangium ultimum occurred approximately 415 million years ago during the Devonian Period. Major diversification events establishing the core Phytophthora crown clades unfolded across the late Paleozoic to early Mesozoic, roughly 250 to 300 million years ago, while lineage diversification within individual clades progressed throughout the Jurassic and Cretaceous periods.

Notably, the researchers found that mitochondrial and nuclear phylogenetic trees, while largely congruent at major clade levels, exhibited highly supported topological conflicts at specific nodes within Subclade 2c. Both datasets supported mutually exclusive topologies with high confidence, ruling out random error. The researchers attribute this discordance primarily to incomplete lineage sorting during rapid divergence events concentrated in the Miocene, though hybridization and introgression between lineages cannot be excluded. This mito-nuclear discordance highlights the distinct evolutionary histories captured by maternally inherited mitochondrial genomes versus biparentally inherited nuclear markers.

The study also carries significant biogeographical implications. The field recovery of endemic Clade 2 species, including P. acerina, P. citricola, P. citrophthora, P. himalsilva, and P. plurivora, in undisturbed Qinling habitats strongly corroborates East Asia as a primary center of origin and diversification for Clade 2. Conversely, the co-occurrence of introduced pathogens such as P. cactorum and P. multivora points to historic anthropogenic movement, emphasizing the vital role of biogeographically informed evolutionary models in forest health management and global biosecurity.

By integrating comparative mitogenomics, molecular clock calibrations, and field recoveries, this work elevates oomycete mitogenomics from basic structural description to an integrative evolutionary framework. The findings demonstrate that mitochondrial genome structural evolution in Phytophthora is neither a purely passive byproduct of time nor a solely selectionist process, but rather a dynamic continuum driven by the interplay between neutral genomic mechanics and evolutionary selection. As natural forests continue to serve as evolutionary reservoirs for pathogen diversification, this genomic baseline provides crucial foundations for tracing pathogen origins, understanding organellar dynamics, and safeguarding international plant biosecurity in an era of increasing global trade and climate-driven range shifts.

Subject of Research: Mitochondrial genome evolution and phylogeny of forest-associated Phytophthora species in the Qinling Mountains of China

Article Title: Mitochondrial genomic variation, evolutionary dynamics, and phylogeny of forest-associated Phytophthora species in the Qinling Mountains of China

Article References: Sun, Y., Chen, X., Wang, W., Li, Y., Qian, Z., Ma, X., Hieno, A., Kageyama, K., & Li, M. (2026). Mitochondrial genomic variation, evolutionary dynamics, and phylogeny of forest-associated Phytophthora species in the Qinling Mountains of China. IMA Fungus, 17, Article e203360. https://doi.org/10.3897/imafungus.17.203360

Image Credits: AI Generated

DOI: 10.3897/imafungus.17.203360

Keywords: Phytophthora, mitogenomics, oomycetes, Qinling Mountains, forest pathology, molecular clock, purifying selection, phylogenomics, gene rearrangement, plant pathogens, East Asia, species identification

The research, led by Jianqi Zhuang of Sichuan University together with colleagues from institutions across China, set out to answer a question that has long frustrated hazard scientists: how do the rainfall changes projected for a warming world translate into changes in landslide susceptibility on the ground? Susceptibility mapping is a well-established technique, but most existing maps are static snapshots, built from historical records of where landslides have occurred and the terrain conditions that favored them. They say little about how those danger zones might expand, contract, or shift as precipitation patterns evolve. The team’s approach was to fuse two strands of data that rarely meet so thoroughly: a massive inventory of past landslides and a continuous rainfall record stretching from 1980 all the way to 2060, derived from the latest generation of global climate models.

The foundation of the study is its dataset, and it is a formidable one. The researchers assembled 8,373 documented historical landslide locations across the Qinling Mountains. To train machine learning models fairly, they paired each landslide point with an equal number of non-landslide samples drawn from stable terrain, producing a balanced dataset of nearly 17,000 labeled locations. This balance matters more than it might appear. When models are trained on datasets where landslides are vastly outnumbered by safe ground, they tend to underestimate danger, and a growing body of literature has highlighted how sampling strategy can make or break susceptibility predictions. By keeping the classes equal, the team gave their algorithms a level playing field on which to learn what distinguishes failing slopes from stable ones.

Onto each of those locations, the researchers overlaid fifteen conditioning factors: the environmental variables that govern whether a slope can fail. These included topographic attributes such as elevation and slope angle, distances to faults, rivers, and roads, vegetation cover, land use type, landform classification, and, critically, rainfall itself. The analysis of where historical landslides cluster is strikingly consistent. Slopes fail most often at elevations between 500 and 1,000 meters, on gradients of 15 to 30 degrees, within 250 meters of geological faults, within 2,500 meters of rivers, and within 250 meters of roads. In other words, landslides in the Qinling concentrate where tectonic fracturing has weakened the rock, where rivers undercut the toes of slopes, and where road construction has steepened and loaded hillsides, all in places where annual rainfall exceeds 650 millimeters and vegetation cover is sparse or the land has been converted to dryland farming or built development.

To quantify which of these factors matter most, the team first applied frequency ratio analysis, a statistical technique that compares the density of landslides within each class of a factor to the overall density across the landscape. A frequency ratio above one indicates that a class is over-represented among landslide sites and therefore more hazardous. The results pointed to three dominant controls: slope, rainfall, and landform. That triad is physically intuitive. Slope determines the gravitational shear stress acting on hillside material, rainfall controls the pore water pressure that reduces friction along potential failure surfaces, and landform encapsulates the broader geomorphological setting that concentrates or disperses both water and sediment. The fact that rainfall ranks alongside terrain as a first-order control is precisely what makes the study’s forward-looking component so consequential.

That component rests on the Coupled Model Intercomparison Project Phase 6, or CMIP6, the international ensemble of global climate models that underpins much of modern climate projection. The researchers constructed a continuous rainfall series for the Qinling region spanning 1980 to 2060 under the SSP2-4.5 scenario, one of the intermediate emissions pathways that assumes roughly current trajectories of greenhouse gas output continue through mid-century. Using projected future rainfall as an input, they could then re-evaluate susceptibility across the mountains for coming decades rather than for the present alone. The projections show annual rainfall rising over the region, and with it, an expansion of the zones classified as having high to very high landslide susceptibility. Areas that are marginal today, where slopes sit just below the rainfall and saturation thresholds needed for failure, are the most likely to cross into the danger zone as precipitation increases.

The modeling itself was a three-way contest between statistical and machine learning approaches, and the comparison is one of the study’s most instructive contributions. The team used logistic regression, the classical statistical baseline that models landslide probability as a weighted sum of factors; random forest, an ensemble method that aggregates hundreds of decision trees, each trained on random subsets of the data and factors; and a convolutional neural network, a deep learning architecture designed to recognize spatial patterns. Convolutional networks, famously successful in image recognition, excel at capturing local spatial context, effectively learning the texture of hazardous terrain rather than treating each pixel in isolation. That ability to encode the nonlinear, spatially correlated interactions among factors is exactly what landslide susceptibility demands, since a steep slope near a fault and a river is not merely the sum of three independent risks.

The performance metrics tell a clear story. Measured by the area under the receiver operating characteristic curve, or ROC-AUC, the convolutional neural network scored 0.913, compared with 0.865 for random forest and 0.826 for logistic regression. The same ranking held for the precision-recall curve, with AUC values of 0.906, 0.858, and 0.818 respectively, and for accuracy, recall, and F1-score, where the neural network achieved 0.881, 0.874, and 0.880. All three models performed well by the standards of the field, but the deep learning model was consistently superior, confirming that in a geologically and climatically transitional landscape like the Qinling, the added complexity of a convolutional architecture pays for itself. For hazard practitioners, the message is that the choice of model is not a technical footnote; it directly changes which slopes appear on the warning map.

Why does this matter beyond the mountains themselves? Landslides are among the deadliest of climate-driven hazards, responsible for thousands of fatalities worldwide each year, and they disproportionately strike rural communities and infrastructure corridors where warning systems are thin. The Qinling region is crossed by transport links, settlements, and agricultural land that press directly against its unstable mid-elevation slopes. The study’s identification of the specific combinations of terrain, proximity, land use, and rainfall that concentrate historical landslides gives planners a concrete template: which distances to faults and roads to treat as buffer zones, which elevation and slope bands to prioritize for monitoring, and which land use categories to scrutinize when approving new development. The finding that landslides cluster within 250 meters of roads is particularly pointed, since it implicates human engineering as a co-conspirator with climate in shaping future risk.

The broader significance of the work lies in its demonstration of a replicable workflow. By coupling a large, balanced landslide inventory with CMIP6 rainfall projections and benchmarking multiple learning algorithms, the researchers have produced a template that can be applied to other mountain belts facing similar pressures, from the Himalaya to the Alps, where recent studies have also documented intensifying landslide activity under changing climate. The authors note that their findings can inform land-use planning and climate adaptation in the Qinling, aiding efforts to reduce future landslide risks. As global temperatures climb and the hydrological cycle accelerates, the ground beneath some of the world’s most densely populated mountain regions is becoming less forgiving. Maps like these, built on deep learning and forward-looking climate data, may become the difference between communities that see the danger coming and those that do not.

Subject of Research: Machine learning-based landslide susceptibility mapping under future rainfall changes in the Qinling Mountains

Article Title: Landslide susceptibility assessment in the Qinling mountain driven by rainfall changes under future climate scenarios

Article References: Zhuang, J., Ma, Z., Di, B., Heng, S., Chang, L., Dou, J., Guo, X., & Peng, J. (2026). Landslide susceptibility assessment in the Qinling mountain driven by rainfall changes under future climate scenarios. Natural Hazards, 122(21), Article 670. https://doi.org/10.1007/s11069-026-08437-7

Image Credits: AI Generated

DOI: 10.1007/s11069-026-08437-7

Keywords: landslide susceptibility, Qinling Mountains, climate change, CMIP6, rainfall, convolutional neural network, random forest, logistic regression, machine learning, natural hazards, SSP2-4.5, disaster risk

The research, conducted in Foping National Nature Reserve on the southern slopes of the Qinling range, tracked seed production and seed fates across nearly a decade. The team monitored annual seed crops of the three oak species from 2016 to 2021 using seed rain traps suspended beneath the canopies of mature trees, and then ran seed-tagging experiments each autumn to follow what happened to individual acorns after rodents found them. In the springs of 2022 to 2024, they repeated the dispersal experiments to capture the other end of the seasonal cycle, when overwintered seeds face a very different rodent community. Live-trapping surveys showed that the local rodent community was dominated by two species, the Chinese white-bellied rat Niviventer confucianus and the wood mouse Apodemus draco, which together accounted for roughly 80 percent of captures.

The three oaks present rodents with strikingly different packages. Quercus variabilis produces the largest seeds of the trio, averaging about 3 grams fresh weight with a thick pericarp of roughly half a millimeter, relatively high fat content, and the lowest tannin concentration of the three. Quercus glandulifera sits at the opposite extreme: its seeds are small, thin-coated, and loaded with tannins, the bitter defensive compounds that deter many herbivores. Quercus aliena falls in between, though its seeds carry the highest crude protein content. Because rodents must balance the energetic reward of a seed against the time and predator exposure required to open and transport it, these trait differences set the stage for predictable, species-specific foraging decisions.

That predictability showed up clearly in the data. Across almost every autumn of the study, rodents harvested Quercus variabilis seeds fastest, cached them most often, and ate the fewest of them outright. Quercus glandulifera seeds, by contrast, were harvested slowly, consumed in place more frequently, and cached least often. Quercus aliena occupied the middle ground, and although its nutritional profile closely resembled that of Q. glandulifera, rodents still preferred it, apparently on the strength of its larger size. The authors interpret this pattern through the lens of optimal foraging: rodents favor seeds that deliver greater energetic returns per unit of handling time, and thick-shelled seeds are often carried away to safer locations before being opened, because cracking them in the open leaves the animal vulnerable to its own predators.

The study also quantified seed dispersal effectiveness, a framework that combines the quantity of seeds removed with the quality of that dispersal, measured here as the proportion of released seeds that survived the winter after rodent handling. By this measure, Quercus variabilis held a consistent competitive advantage, recording the highest effectiveness in four of the six autumn study years, followed by Q. aliena and then Q. glandulifera. Dispersal distances did not differ significantly among the species, but there was a tendency for the large-seeded Q. variabilis to be carried farther before being eaten or cached, consistent with the optimal caching space model, which predicts that rodents transport higher-value foods to more distant cache sites.

Perhaps the most striking result concerns abundance. The three oaks mast in rough synchrony, producing large seed crops in some years and meager crops in others. Annual seed production varied dramatically: Quercus variabilis peaked at about 21.7 seeds per square meter in 2017 and collapsed to just over one seed per square meter by 2021. When the researchers combined seed crop with rodent availability into a single measure of per capita seed availability, a clear pattern emerged. In years when seeds were abundant relative to the number of rodents, harvest rates slowed markedly—the slowest of the entire study occurred in 2017, the peak mast year. This supports the classic predator satiation hypothesis, which holds that plants overwhelm seed predators by flooding the market, so that a smaller fraction of the crop is consumed.

Yet satiation tells only half the story. In high-availability years, rodents also shifted their behavior after harvesting: a greater proportion of seeds were eaten, both in place and after removal, while a smaller proportion were cached. This aligns with the predator dispersal hypothesis, which proposes that when food is plentiful, rodents have less incentive to invest in long-term storage and instead consume seeds immediately. When seeds were scarce, caching became relatively more attractive as a hedge against future shortage. The authors conclude that the mechanisms underlying both hypotheses can operate at the same time, with satiation governing how quickly seeds are taken and dispersal dynamics governing what rodents do with them once captured.

Seasonality added another layer of complexity. Spring experiments revealed a dramatically different dispersal landscape. Harvest rates were substantially lower than in autumn, and in 2023 the effect was extreme: roughly 15 percent of Quercus variabilis seeds, half of Q. aliena seeds, and 95 percent of Q. glandulifera seeds were still sitting untouched at the release stations fifty days after deployment. The researchers attribute this to reduced rodent abundance in spring following the lean winter months, and to the fact that when all three oak species were simultaneously scarce, rodents switched almost entirely to immediate consumption. No Q. aliena or Q. glandulifera seeds were cached at all in the springs of 2022 or 2023, and overall dispersal effectiveness showed no significant differences among the species in spring, because so few seeds of any kind were cached.

For the oaks, these results carry real evolutionary implications. Quercus variabilis appears to have evolved a winning combination: a large, nutritious seed with enough tannin defense to remain palatable but enough shell to reward caching rather than instant consumption. Quercus glandulifera, with its small, tannin-rich, thin-coated seeds, consistently loses out in the scramble for rodent dispersal services, suggesting it must compensate through other life-history strategies, such as producing seeds in years or microhabitats where competitors are absent, or relying on the simple arithmetic of producing many cheap seeds and accepting high predation. The study’s authors suggest that seed-trait-mediated selection by rodents may modulate the effects of masting itself, shaping indirect seed-seed interactions that ultimately influence how these sympatric species coexist.

The work also underscores how fragile the mutualism between oaks and rodents can be. Scatter-hoarding only translates into forest regeneration when cached seeds are forgotten, and the balance between predation and dispersal tips with every fluctuation in seed crop, rodent population, and season. As climate change alters the frequency and synchrony of mast events across the globe, understanding these trait- and abundance-dependent rules becomes essential for predicting which tree species will regenerate successfully and which will quietly lose their place in the forest. In the Qinling Mountains, at least, the verdict is clear: when it comes to recruiting rodents as gardeners, size, nutrition, and timing are everything.

Subject of Research: Rodent-mediated seed predation and dispersal among sympatric Quercus species in relation to seed traits and mast seeding in the Qinling Mountains, China

Article Title: Trait‐Mediated Seed Predation and Dispersal Are Regulated by Seed Abundance for Sympatric Tree Species in the Qinling Mountains, China

Article References: Wang, J., Han, G., Zhang, B., Han, N., Feng, T., Zhao, J., An, X., Hou, X., & Chang, G. (2026). Trait‐Mediated Seed Predation and Dispersal Are Regulated by Seed Abundance for Sympatric Tree Species in the Qinling Mountains, China. Ecology and Evolution, 16(10), Article e74453. https://doi.org/10.1002/ece3.74453

Image Credits: AI Generated

DOI: 10.1002/ece3.74453

Keywords: seed dispersal, seed predation, rodents, mast seeding, Quercus, Qinling Mountains, seed traits, scatter-hoarding, forest ecology, plant-animal mutualism, seed dispersal effectiveness, China