The analysis examined a deliberately bounded population. Emerging listed food enterprises were operationally defined as companies no more than 15 years past listing, a restriction intended to focus on listed businesses combining growth pressure, financing needs and developing operating routines rather than mature enterprises with established distribution, financing and management systems. The initial roster contained 188 A-share firms in Chinese food industries classified under CSRC codes C13, C14 and C15. After excluding firms with special-treatment status, missing profile or province information, unmatched firm-year records and listing ages above 15 years, 122 eligible companies remained. Public financial statements covering 2015–2023 provided 992 firm-year rows, from which 748 valid resilience observations for 119 firms were reconstructed. The common regression sample contained 732 observations from 2017 through 2023.

Food supply-chain resilience was not observed directly. Instead, the researchers constructed a composite accounting proxy from resistance, recovery and stability in revenue, operating cash flow and net margin. Resistance captured inverse adverse annual movement, recovery represented an accounting rebound following a prior decline, and stability measured inverse volatility over a rolling three-year period. An entropy-weighted combination integrated nine indicators, while separate resistance, recovery and stability measures were used in sensitivity tests. Financial records from 2015 and 2016 supplied the lagged and rolling inputs needed for the first eligible 2017 observations, and missing outcome components were excluded rather than imputed. This construction is transparent and reproducible, but it does not directly register supplier failures, logistics interruptions, cold-chain losses, stockouts, spoilage, food-safety incidents or traceability breakdowns. A recovery score may also reflect mean reversion or an ordinary business cycle rather than a rebound from a verified external disruption.

Regional digital finance was measured with the Peking University aggregate Digital Financial Inclusion Index and matched to each company’s registered province and reporting year. It represented the surrounding financial environment, not direct evidence that a company used digital payments, platform credit, digital procurement, inventory software or other enterprise-level services. This separation defines the study’s proposed regional-to-firm conversion gap: a digitally advanced province may make more financial options available, but a business must still translate those options into supplier settlement, inventory governance, quality assurance, traceability and disruption-response routines. The theoretical relationship is also not necessarily positive. Easier finance could help a company retain suppliers, maintain safety stock, fund inspections or support cold-chain investment. At the same time, it could finance more stock-keeping units, distributor expansion, marketing or larger inventory commitments, potentially increasing spoilage exposure and coordination complexity before operating routines can absorb the expansion.

The model sequence progressively separated broad regional differences from changes within individual firms. The digital-finance coefficient was 0.002 in pooled ordinary least squares and 0.002 in the random-effects model, 0.044 with firm fixed effects, −0.005 with firm and year fixed effects, and 0.059 after province-specific trends were added. This changing pattern did not support a stable positive within-firm relationship. Firm effects removed time-invariant differences such as business model, brand heritage, listing cohort, registered location and product category. Year effects absorbed common national patterns, including nationwide regulatory changes, pandemic disruption and general growth in digital finance. Province-trend specifications addressed differential regional trajectories but sacrificed identifying variation and produced wider intervals. Even so, the fixed-effects design is an associational test: it cannot fully remove time-varying regional confounding, reverse adaptation or geographic measurement error.

Power and equivalence diagnostics showed why the near-zero point estimate could not be treated as proof of no relationship. The standardized digital-finance coefficient was −0.046, while the approximate 80% minimum detectable effect was 1.575 standard deviations of the resilience proxy. Equivalence tests using bounds of 0.05 and 0.10 SCR standard deviations were not significant, with p-values of 0.497 and 0.462, so the researchers did not establish practical equivalence. The confidence interval remained compatible with both modest adverse and modest beneficial associations. The short panel also left relatively little independent variation in province-year indicators after persistent firm differences and common annual patterns were removed. Raw regional digital finance and provincial household consumption expenditure were strongly correlated at 0.823, although their residualized correlation fell to 0.090 after firm and year means were removed, making conditional estimates difficult to estimate precisely.

Neither proposed indirect mechanism was established by the available proxy-based evidence. The association between regional digital finance and the entropy-weighted operational-capability measure was −0.038, with a 95% confidence interval from −0.119 to 0.043. Operational capability’s association with the resilience proxy was 0.009, with an interval from −0.064 to 0.082. The operational measure combined inventory turnover, receivables turnover, asset turnover and the inverse cash-conversion cycle. It was treated as an outcome-oriented proxy for operating efficiency rather than a direct measure of enterprise-resource-planning deployment, supplier-platform integration, replenishment routines, cold-chain monitoring, traceability or disruption protocols. Equal-weight variants were similarly imprecise. A single-indicator test found a positive association for asset turnover, but the study did not treat that result as validation of the complete pathway. Tests of whether regional consumption capacity moderated the digital-finance relationship were also inconclusive: the linear interaction was −0.016, with a 95% interval from −0.052 to 0.019, and the quadratic interaction was 0.006, with an interval from −0.004 to 0.016.

Robustness checks did not reveal a consistent positive pattern across credible definitions and samples. The digital-finance estimate remained close to zero or changed sign when researchers used an equal-weight resilience outcome, isolated its resistance, recovery or stability components, substituted alternative digital-finance dimensions, changed covariance estimators, excluded municipalities, restricted the sample to balanced panels or shifted exposure by one or two years. Province-clustered standard errors yielded an estimate of −0.005 with a 95% interval from −0.118 to 0.108, while Driscoll–Kraay inference produced −0.005 and an interval from −0.131 to 0.121. One-year and two-year lag estimates were −0.084 and 0.060, respectively, and both were imprecise. Four theory-led interactions tested whether baseline size, maturity, operational capability or inventory pressure changed the association. None remained statistically significant after Benjamini–Hochberg or Holm adjustment, making the subgroup analysis exploratory rather than evidence that particular firms respond differently.

The study’s main contribution is diagnostic: it identifies where contextual financial development may fail to become an observable firm-level resilience outcome and specifies the operational evidence needed to examine that conversion more directly. Its findings remain constrained by the annual accounting proxy, registered-province matching, short 2017–2023 estimation period, limited statistical power and sample of recently listed companies. Registered location may not represent the places where a company produces, stores goods, obtains inputs or sells products, and such measurement error could attenuate estimates toward zero. The analysis also cannot observe direct digital-finance adoption or responses to identified supply-chain shocks. Future research would need information on enterprise adoption, plant and warehouse locations, supplier networks, cold-chain performance, product-level demand, traceability and disruption-specific operating responses, along with designs that can better separate contextual exposure from firm-level conversion. Until such evidence is available, the most defensible conclusion is that this panel did not demonstrate a stable benefit from regional digital finance, while also lacking the precision to conclude that meaningful resilience associations are absent.

Subject of Research: Associations between regional digital finance and accounting-based supply-chain resilience among emerging listed food enterprises in China

Article Title: Regional digital finance and food supply-chain management resilience: evidence on the regional-to-firm conversion gap in China’s emerging listed food enterprises

Article References: Zhao, J. (2026). Regional digital finance and food supply-chain management resilience: evidence on the regional-to-firm conversion gap in China’s emerging listed food enterprises. Frontiers in Sustainable Food Systems, 10, Article 1948192. https://doi.org/10.3389/fsufs.2026.1948192

Image Credits: AI Generated

DOI: 10.3389/fsufs.2026.1948192

Keywords: regional digital finance, food supply chains, supply-chain resilience, China, listed food enterprises, fixed-effects analysis, operational capability, regional consumption, statistical power, Regional, digital, finance

Chinese sectoral fossil-energy data and emissions factors from the Intergovernmental Panel on Climate Change were used to measure production-side direct fossil-energy CO2 attributed to export domestic value added. The analysis combined logarithmic mean Divisia index decomposition, fixed-effects models, an external shift-share instrument, value-chain configuration tests and a within–between decomposition. Baseline fixed-effects and instrumental-variable estimates showed a negative overall association between regional value-chain cooperation and absolute production-side direct CO2 emissions.

The relationship remained negative in samples restricted to observations after 2013. Within established partner–industry relationships, deeper regional value-chain integration was significantly associated with lower carbon per unit of exports, while the coefficients for export scale and absolute emissions were statistically insignificant. The results therefore distinguish the carbon performance associated with deeper trade relationships from the separate effect of trade volume.

Value-chain configuration also showed distinct associations. Denser two-way embedment was associated with higher carbon pressure, whereas a more forward-oriented structure was associated with lower carbon. Because the study used observational panel estimates and model-based emissions accounting, these results identify associations rather than establish that regional value-chain integration caused the observed emissions changes.

The authors suggest combining production-efficiency gains with forward-oriented domestic value creation and monitoring scale effects in deeper agri-food production networks.

Subject of Research: Carbon performance of regional value-chain integration in China’s agri-food exports

Article Title: Can regional value-chain integration decouple China's agri-food export growth from production-side direct CO2 attributed to export domestic value added? Evidence from belt and road partner economies

Article References: Tang, K., Yang, W., & Huang, Y. (2026). Can regional value-chain integration decouple China's agri-food export growth from production-side direct CO2 attributed to export domestic value added? Evidence from belt and road partner economies. Frontiers in Sustainable Food Systems, 10, Article 1977325. https://doi.org/10.3389/fsufs.2026.1977325

Image Credits: AI Generated

DOI: 10.3389/fsufs.2026.1977325

Keywords: agri-food exports, carbon intensity, China, regional value chains, Belt and Road, food processing, production emissions, trade networks, regional, value-chain, integration, decouple

The reconstruction is based on fluxes of three combustion-derived anhydrosugars. Levoglucosan, produced mainly during the pyrolysis of plant cellulose, served as the principal tracer of biomass burning, while its isomers mannosan and galactosan, formed principally during hemicellulose pyrolysis, helped characterize the relative composition of combusted material. The researchers standardized the three flux records and averaged them to construct a composite fire activity index, or FAI. They compared this index with elemental ratios, total organic carbon, grain-size measurements and climate-model output to examine whether wildfire changes coincided with local hydrological conditions or broader circulation variability.

The chronology came from a Bayesian age-depth model previously established for the Wodi Co core using accelerator mass spectrometry radiocarbon dating, the IntCal20 calibration curve and the rbacon modelling package. The full 880-centimeter core extends from approximately 5,120 years before present to 69 years after present, while the analyzed upper 810 centimeters cover roughly the past 4,550 years and end at about 62 years after present. The model showed no age reversals, although sediment accumulation rates changed substantially, including relatively rapid deposition around 2,300 to 1,600 years before present. Calculating fluxes rather than relying on concentrations alone was therefore important for separating changes in the amount of burned material from changes in the rate at which sediment accumulated.

Stratigraphically constrained clustering divided the molecular fire record into three principal phases. From about 4,550 to 2,885 years before present, the proxies showed generally low to moderate biomass burning. Mean levoglucosan flux was 1.55 ± 1.06 micrograms per square centimeter per year, mean mannosan flux was 0.92 ± 0.33 micrograms per square centimeter per year and mean galactosan flux was 1.90 ± 1.58 micrograms per square centimeter per year. The FAI averaged −0.63 ± 0.64 during this first phase, indicating activity below the full-record average. The most active and variable period followed, extending from roughly 2,885 to 1,946 years before present. Mean levoglucosan flux rose to 3.66 ± 3.27 micrograms per square centimeter per year, while mannosan and galactosan fluxes and the FAI also reached their highest phase averages.

Large standard deviations in the high-fire phase show that the increase was episodic rather than uniform. Several short fire episodes, including a multi-decadal peak near 2,000 years before present, were embedded within the broader interval. The authors note that the peaks may reflect short-lived enhancements in fire intensity or biomass combustion, although variations in fuel availability, atmospheric transport and sedimentary preservation may also have contributed. Since about 1,946 years before present, mean levoglucosan, mannosan and galactosan fluxes declined to 1.25 ± 1.21, 0.98 ± 0.62 and 1.74 ± 1.35 micrograms per square centimeter per year, respectively. The FAI averaged −0.76 ± 0.82 in this final phase, returning activity to levels broadly comparable with those of the earliest phase.

Several lines of evidence pointed toward drier conditions during the high-fire interval. In the Wodi Co catchment, calcium-to-silicon ratios rose as a whole, a pattern interpreted as greater concentration of lake water and reduced effective moisture, although detrital inputs and carbonate sources may also affect the ratio. Conversely, iron-to-manganese ratios declined, consistent with lower lake levels and more oxic sedimentary conditions. Simulated precipitation for the wettest quarter also remained comparatively low, averaging 321.44 ± 4.95 millimeters during the main fire phase. The FAI was significantly correlated with calcium-to-silicon at Spearman’s rho of 0.24 and inversely correlated with iron-to-manganese at rho of −0.34, with both relationships significant at p less than 0.05. By contrast, total organic carbon, the greater-than-64-micrometer fraction and zircon-to-rubidium ratio were not significantly correlated with the principal fire proxies.

Independent statistical tests identified changes at roughly the same broad times. A multiple structural-break analysis selected two FAI transitions, at approximately 2,910 and 1,950 years before present, closely matching the clustering boundaries. The result remained similar when first-order autoregressive errors were included, indicating that moderate temporal autocorrelation did not erase the principal transitions. The iron-to-manganese record also retained two breakpoints, at approximately 1,740 and 2,910 years before present, whereas the calcium-to-silicon breakpoint at about 2,250 years before present was sensitive to the statistical treatment of temporal dependence. The timing of the main fire-state transitions was therefore more robust than the exact timing of every hydrological change.

The broad climate context points to a weakened Indian summer monsoon and enhanced El Niño-like conditions during the principal fire interval. An Oman Cave speleothem oxygen-isotope record indicated relatively weak monsoonal moisture transport, while a sediment record from El Junco Lake in the tropical Pacific indicated pronounced El Niño-like conditions, including the strongest anomaly in the approximately 4,500-year comparison period. These large-scale indicators broadly coincided with the Wodi Co fire episode, although their principal extremes lagged local changes by about 500 years. The ENSO index was significantly associated with lower simulated dry-season relative humidity at the lake, with rho of −0.47 and p less than 0.05, whereas the Indian summer monsoon showed only a weak relationship with that measure, with rho of −0.26 and p below 0.10. The winter westerly jet was positioned slightly farther north during the high-fire phase, at a mean December-to-February latitude of 30.02 ± 1.05 degrees north, but its position was not significantly related to either the moisture indicators or the fire proxies.

After about 2,000 years before present, the transition was marked by reduced and less variable burning, lower calcium-to-silicon, higher iron-to-manganese and a general increase in simulated wet-season precipitation. Mean precipitation for the wettest quarter was 325.15 ± 5.50 millimeters during the final phase, with an overall increase after about 2,100 years before present. The change was also visible in the mean calcium-to-silicon ratio of 0.76 ± 0.21 and iron-to-manganese ratio of 131.92 ± 53.95. This local shift broadly agrees with hydroclimatic reorganizations described across parts of the Asian monsoon domain, although the response was spatially uneven and the exact breakpoints did not match perfectly among records. The authors interpret the Wodi Co change as a local reorganization of the relationship between moisture and wildfire within the broader “2-kyr shift,” rather than as proof that a single event occurred simultaneously everywhere. After the transition, wildfire activity remained generally low and stable.

The reconstruction offers a long-term example of drought-related drying coinciding with greater biomass burning in a high-elevation environment, but it does not establish that moisture alone caused every fire peak. The levoglucosan-to-mannosan ratio ranged from 0.15 to 6.96 and may indicate changes in the relative contribution of woody or herbaceous fuels, yet overlapping fuel signatures, combustion conditions, atmospheric transport and differential preservation prevent it from identifying fuel types conclusively. Organic-carbon accumulation likewise cannot serve as a direct measure of vegetation productivity because it is affected by aquatic production, catchment delivery, degradation and preservation. The Wodi Co catchment also lacks independent archaeological, pollen-based disturbance or anthropogenic-biomarker records, so the magnitude of past human influence cannot be separated confidently from climatic effects, especially during the late Holocene. The study instead supports a moisture-linked regional explanation for the long-term pattern, while treating the Indian summer monsoon, ENSO and westerly circulation as the larger climatic backdrop rather than independent, immediate causes of individual fire episodes.

Subject of Research: Holocene wildfire variability and regional moisture conditions in alpine northwestern Yunnan

Article Title: Regional moisture conditions and mid- to late-Holocene wildfire variability along the southeastern margin of the Tibetan Plateau

Article References: Yang, P., Gao, Y., Hu, C., Kong, L., Chen, G., Chen, J., Li, R., Ma, J., Fu, Z., & Xie, Z. (2026). Regional moisture conditions and mid- to late-Holocene wildfire variability along the southeastern margin of the Tibetan Plateau. Frontiers in Earth Science, 14, Article 1963785. https://doi.org/10.3389/feart.2026.1963785

Image Credits: AI Generated

DOI: 10.3389/feart.2026.1963785

Keywords: paleofire, lake sediments, Wodi Co, Tibetan Plateau, Holocene, wildfire, moisture variability, Indian summer monsoon, ENSO, levoglucosan, Yunnan, Regional

Most known populations are concentrated in small valleys near Clusone, especially within the Borlezza hydrogeographic basin, with isolated records from Valcanale, Val Nossana and Val Conchetta. The perennial plant grows on dolomitic debris cones, generally on poorly developed and well-drained soils from 450 to 1,250 metres above sea level. Field observations indicated stable subpopulations, with no continuous decline observed or estimated.

Some populations crossed by motocross and mountain-bike trails were threatened by trampling. The assessors also identified habitat modification as debris cones become colonized by shrubs and trees, landslides or cliff collapse associated with erosion, and potentially increasing drought and extreme weather as future threats. Because the species was recently described, the assessors said further information on its range, reproductive biology and ecology was needed to plan in situ and ex situ conservation.

Preliminary experiments at Milan’s Città Studi Botanical Garden indicated low germination rates, although the available evidence did not establish general reproductive success. Campanula bergomensis is not protected by international, national or regional law, but populations in Val Nossana, Val Bielone, Romentareck, Val dei Mulini and Monte Simer fall within two Natura 2000 Special Areas of Conservation. The assessors called for its inclusion in Lombardy Region’s protection list.

Iris hippolyti is a perennial geophyte endemic to the Kokchatau and Karatau remnant low mountains in the eastern Kyzylkum Desert of Uzbekistan. The two mountain areas cover approximately 2.5 km² and reach a maximum elevation of 447 metres above sea level. Until 2018, the species was known from one population at Kokchatau, where an estimated 300 individuals were recorded in 2015; the count was 138 in 2020.

A second subpopulation with fewer than 40 individuals was documented in Karatau in 2022, bringing the assessed global population to 175 individuals. For Iris hippolyti, the assessors calculated an extent of occurrence and area of occupancy of 8 km² each and identified two locations based on the principal threats. They recorded a 54% observed decline in individual numbers over 10 years, continuing declines in habitat extent and quality and mature individuals, and a continuous decline in mature individuals.

Livestock grazing, recreation and associated habitat destruction were documented threats: satellite imagery showed that more than 2,000 m² of habitat had been destroyed for recreational activities near the Zafarabad settlement. Severe late-February frost events, occurring approximately once every five to six years, were also reported to cause significant juvenile mortality. The species was classified as Endangered under criteria B1b(iii, v) + 2b(iii, v) + D.

Iris hippolyti is listed as Category 1, or Critically Endangered, in Uzbekistan’s National Red Book and has been introduced into cultivation at the Tashkent Botanical Garden. The assessors described reintroduction of cultivated individuals as essential and identified reinforced in situ conservation, protected ex situ cultivation, bioclimatic modelling and identification of suitable areas as proposed elements of a conservation strategy. These actions were recommendations, not evidence of recovery.

Stipa aktauensis is a perennial tussock grass distributed across the Kuldzhuktau, Auminzatau, Bukantau, Tamditau, Muruntau, Aktau and Sultan-Uizdag remnant mountains in Uzbekistan’s Kyzylkum Desert. It grows on dry gravelly slopes and rock cracks at altitudes from 400 to 700 metres above sea level and can reach 25–30 cm in height. Livestock grazing was identified as a continuing pressure because it directly consumes plants, degrades habitat, limits regeneration and increases soil erosion.

Mining and energy-related development were also reported, including exploration for light metals and marble in Kuldzhuktau, construction of wind-power stations in Kuldzhuktau and Tamditau, and expanding stone-crushing and gravel production at Sultan-Uizdag. The assessment calculated an extent of occurrence of 32,725 km² but an area of occupancy of 44 km², identified five locations, and recorded continuing declines in range, occupancy, habitat extent and quality, subpopulations and mature individuals. Stipa aktauensis was classified as Endangered under criterion B2ab(i, ii, iii, iv, v).

The assessors reported that earlier cultivation attempts at a Kyzylkum desert station were unsuccessful and that its seeds were not conserved in any germplasm bank. They recommended research and monitoring to clarify reproductive biology, ecology, threats and population trends, together with ex situ measures including establishment of a seed bank.

Subject of Research: Global IUCN Red List assessments of three narrow-range vascular plants in Italy and Uzbekistan

Article Title: Global and Regional IUCN Red List Assessments: 17

Article References: Fenu, G., Caccianiga, M., Khabibullaev, B. S., Mangili, F., Shomurodov, K. F., Valle, B., & Orsenigo, S. (2024). Global and Regional IUCN Red List Assessments: 17. Italian Botanist, 18, 141-153. https://doi.org/10.3897/italianbotanist.18.139285

Image Credits: AI Generated

DOI: 10.3897/italianbotanist.18.139285

Keywords: IUCN Red List, Campanula bergomensis, Iris hippolyti, Stipa aktauensis, Italy, Uzbekistan, Kyzylkum Desert, plant conservation, endangered species, Global, Regional, IUCN

These Red List assessments were based on compiled occurrence records, field observations, population estimates, information on ecology and reproduction, and geospatial calculations. The two Italian taxa were evaluated under criterion B using an extent of occurrence calculated with a minimum convex hull in QGIS 3.28.13 and an area of occupancy calculated with a fixed grid of 2 × 2 km cells. The two Uzbek taxa were evaluated with GeoCAT using the same grid dimensions.

For Alyssum wulfenianum subsp. wulfenianum, the Italian regional assessment estimated an extent of occurrence of 60.2 km² and an area of occupancy of 20 km². The regional population was estimated at about 3,600 mature individuals. This narrow endemic is restricted to the southeastern limestone Alps, principally around Cave del Predil and the Rio del Lago valley in Friuli-Venezia Giulia, with a smaller distribution across the border in Carinthia, Austria.

The subspecies is a perennial that grows on calcareous rocks, screes, debris, and gravel-sand riverbeds at approximately 600–900 m above sea level. It is described as a hyperaccumulator metallophyte associated with high concentrations of zinc and lead and is also reported to accumulate thallium. It occurs where natural contamination or mining has created metal-rich substrates and is considered a poor competitor on non-metalliferous soils. Subpopulations on the slopes of Mount Re were described as stable, whereas those in the Rio del Lago stream bed were highly fluctuating and in decline.

The assessment classified A. wulfenianum subsp. wulfenianum as Endangered under criteria B1ab(ii,iii,iv,v)+2ab(ii,iii,iv,v), reflecting no more than five locations and continuing declines in area of occupancy, habitat extent and quality, locations or subpopulations, and mature individuals. Potential future changes associated with mine remediation, hydrological regulation, gravel extraction, and phytoremediation were distinguished from the observed condition of the populations. Increased recreational use of the Cave del Predil geomineral park and resulting human trampling were also identified as threats. No information on extreme fluctuations was available for this assessment.

For Noccaea cepaeifolia, the Italian assessment estimated an extent of occurrence of 21.5 km² and an area of occupancy of 16 km². Its Italian population was described as a few thousand individuals, with additional small, scattered subpopulations. The narrow endemic occurs in the calcareous southeastern Alps, with its Italian range centered on Cave del Predil and a limited extension into Carinthia. Some subpopulations in former mining areas and on nearby slopes were considered stable, but reductions had been observed in lower-altitude populations along Rio del Lago.

N. cepaeifolia was also classified as Endangered in Italy under criteria B1ab(ii,iii,iv,v)+2ab(ii,iii,iv,v). As with A. wulfenianum subsp. wulfenianum, the assessment cited no more than five locations, continuing declines in geographic and demographic measures, potential habitat alteration from mining remediation and riverbed works, and trampling associated with recreation. The assessment did not report information on extreme fluctuations. Both Italian assessments noted that detailed reproductive information was unavailable, and the authors called for research and monitoring of reproductive biology, environmental influences on reproduction, population trends, and the effects of mining remediation and flooding.

The global assessment of Phlomis nubilans estimated 50,000–100,000 mature individuals from field studies conducted over the preceding 10 years, alongside 71 known occurrence records. Its calculated extent of occurrence was 4,381.61 km² and its area of occupancy was 172 km². The endemic perennial occurs in the Nuratau Mountains, where agricultural development, excessive grazing, geological exploration, mining, and quarrying were identified as pressures. Grazing was described as the main indirect threat because livestock trample young plants and contribute to soil erosion and habitat degradation, although livestock do not eat the species.

Although P. nubilans met the numerical area thresholds associated with criterion B for Endangered, the assessment cited more than 10 locations, more than 10,000 mature individuals, no severe fragmentation, and no known extreme fluctuations. It was therefore classified as Near Threatened. A species-distribution model projected a significant reduction in suitable range under the RCP8.5 scenario as aridification increased; this was a model-based projection rather than an observed loss of area. The demographic structure of the species was described as poorly studied, with available data regarded as contradictory.

Phlomoides anisochila was classified as Endangered globally under criteria B1bc(ii,iii,iv)+B2bc(ii,iii,iv)+C2ab. Field work documented about 1,000 individuals across 34 occurrence records, with small isolated groups commonly containing 10–30, 50–70, or occasionally 100–200 plants. Its extent of occurrence was 214.5 km² and its area of occupancy was 40 km². The assessment recorded extreme fluctuations in mature individuals and estimated fewer than 250 mature plants in the largest subpopulation, supporting the application of criterion C as well as criterion B. Grazing, geological exploration, avalanches, mudflows, and drought-related pressures were identified, while species-distribution modeling projected a reduction in suitable range under the RCP8.5 scenario as drought increased. The authors proposed expanded protection within Nuratau Nature Reserve, regulation of grazing, restrictions on mining and geological exploration, research and monitoring, cultivation methods, reintroduction, and seed banking for the two Uzbek taxa.

Subject of Research: Regional and global IUCN Red List assessments of four threatened plant taxa in Italy and Uzbekistan

Article Title: Global and Regional IUCN Red List Assessments: 18

Article References: Orsenigo, S., Beshko, N. Y., Casolo, V., Castello, M., Khabibullaev, B. S., Petrussa, E., Shomurodov, K. F., Uliana, L. R., & Fenu, G. (2025). Global and Regional IUCN Red List Assessments: 18. Italian Botanist, 19, 185-203. https://doi.org/10.3897/italianbotanist.19.156865

Image Credits: AI Generated

DOI: 10.3897/italianbotanist.19.156865

Keywords: IUCN Red List, plant conservation, Italy, Uzbekistan, metallophytes, Nuratau Mountains, Cave del Predil, endemic plants, habitat degradation, population monitoring, Global, Regional

Allium ravenii, commonly called Raven’s garlic, is a perennial geophyte in the amaryllis family known only from saline plains beside the Eastern Chink of the Ustyurt Plateau in Karakalpakstan, Uzbekistan. Its only known population occupies gypsum or calcareous soils at approximately 150 metres above sea level in an extremely arid zone influenced by the drying of the Aral Sea. The locality is also the species’ type locality. Regional climate analyses cited in the assessment indicate increasing aridity and moisture deficits, particularly during spring, while long-term records show that temperature variation has increased in both the cold and warm halves of the year.

Field research in 2009 and 2010 recorded about 20 A. ravenii individuals in full bloom, but a monitoring expedition in 2015 found only six plants, identified in the assessment as mature individuals. Targeted searches during the growing seasons of 2020 and 2021 found no plants, a result following very dry years, including less than 75 millimetres of precipitation in 2020, and a sharp reduction in spring rainfall. The assessors did not classify the species as extinct because underground portions of several mature individuals were found in 2015, indicating that plants could persist below ground when no shoots were visible. Both the extent of occurrence and the area of occupancy were calculated as 4 square kilometres, and the mature population was estimated at about six individuals.

Several pressures converge on that single population. The Ustyurt Plateau and Eastern Chink serve as nomadic pastures for sheep, goats and camels, while saiga antelope and goitered gazelle also graze there. The plateau contains large natural-gas reserves, and several drilling stations operate near the garlic population. The site lies beside the only road crossing the plateau toward the Aral Sea, where irregular traffic includes tourists and equipment used by geologists; satellite images showed more than ten new roads around the population. A single location, continuing decline, habitat deterioration, population fluctuations and the extreme smallness of the population supported classification under Red List criteria B, C and D. The locality is not included in a reserve or other protected area, although Uzbekistan’s national Red Book lists the species under category 1, meaning it is on the brink of extinction.

Centaurea seguenzae, or Seguenza’s knapweed, presents a different pattern of extreme restriction. The perennial species is endemic to the Peloritani Mountains in northeastern Sicily, where its entire known distribution comprises one population and one stand on the rocky slopes of Capo Tindari near Patti. It grows on wind-exposed coastal cliffs at elevations of up to 300 metres, in lithosols formed from Mesozoic marble and calcium-silicate feldspar. Its habitat supports a community of chasmophytic plants. A rough count found more than 250 mature individuals, although inaccessible cliffs mean that the total may have been underestimated. Direct observations showed no significant decline in population size or the number of mature individuals in recent years.

Despite that absence of a measured abundance decline, the habitat of C. seguenzae is deteriorating. The cliffs are susceptible to natural erosion, detachment of individual blocks and landslides, including a small-to-massive landslide near the Verde lagoon in spring 2009. The invasive cactus Opuntia ficus-indica occupies much of the rocky habitat and restricts the space available to the native knapweed, contributing to an ongoing loss of suitable habitat. Potential fires are an additional threat because the surrounding territory is periodically affected by arson. The species has an extent of occurrence and area of occupancy of 4 square kilometres. Its occurrence at a single location, together with continuing decline in habitat quality, led to a Critically Endangered classification under criterion B. This differs from its previous global assessment of Near Threatened.

Some conservation measures are already in place for C. seguenzae. Its population lies within the Laguna di Oliveri–Tindari Special Area of Conservation, part of the European Natura 2000 network. The species was one of the targets of the SiMaSeed project, funded under the Interreg Italy–Malta 2014–2020 programme. Living plants and seeds were preserved at the seed bank of Catania University, which had stored more than 8,000 seeds under accession number SiMaSeed/CT/19/370. The assessors nevertheless called for further research into reproductive biology and population trends, monitoring of human activities in nearby tourist areas and urgent management of invasive alien plants.

Haloxylon persicum, known as white saksaul or ghada, is a small tree distributed broadly from Egypt to China. Its global Red List category is Least Concern, but the regional assessment for Iraq identifies a much more severe situation. Only two small, fragmented Iraqi populations remain in the southern desert of Al-Muthanna Governorate. The larger extends for about 45 kilometres at Busayyah, roughly 100 kilometres south of Al-Nasiriya, while the second lies near Qusair, approximately 20 kilometres south of Hammar Marsh. Both occupy sparsely vegetated sandy dunes without temporary watercourses, where the main water supply is annual rainfall averaging about 93 millimetres. The tree can live for 15 to 25 years, reproduce vegetatively and tolerate high temperatures and drought.

Iraq’s remaining saksaul population has declined dramatically during the past decade. The latest monitoring counted 1,222 juvenile plants, including 356 near Hammar Marsh and 866 at the other site, but no mature individuals. The assessment links the sharp reduction chiefly to indiscriminate removal. Historically, Bedouin communities harvested branches and stems for fuel, which helped safeguard the trees; more recently, collectors have cut the roots of large adults for hookah smoking. Loggers reportedly severed stems close to the ground and pulled root systems with vehicles, compromising vegetative reproduction. Local residents described mature saksaul as reaching five metres and providing shade for as many as four camels, whereas the surviving juveniles were about 0.4–0.7 metres to approximately 1.6 metres tall.

Grazing by camels, sheep and goats, dust generated by traffic on unpaved roads, recreational camping, off-road vehicle use and three consecutive years of severe drought added further pressure to the Iraqi populations. The area has also experienced severe continental temperature extremes, including 53.8 °C at Basrah Airport from 1949 through March 2022 and −13.8 °C at Samawa in January 2020. With no mature individuals, 1,222 juveniles divided among two subpopulations and a continuing decline, the Iraqi population was classified as Critically Endangered under criteria C and D. No dedicated protection measures for H. persicum are currently in place in Iraq. The assessors recommended monitoring, in situ and ex situ conservation, possible translocation and public awareness work, particularly among young people. They emphasized that conservation is also a social, legal and cultural issue connected to changing tribal customs.

Subject of Research: Global and regional Red List assessments of three threatened plant taxa

Article Title: Global and Regional IUCN Red List Assessments: 14

Article References: Fenu, G., Al-Rammahi, H. M., Cambria, S., Cristaudo, A. E., Khabibullaev, B. S., Mohammad, M. K., Shomurodov, K. F., Tavilla, G., & Orsenigo, S. (2022). Global and Regional IUCN Red List Assessments: 14. Italian Botanist, 14, 81-94. https://doi.org/10.3897/italianbotanist.14.97176

Image Credits: AI Generated

DOI: 10.3897/italianbotanist.14.97176

Keywords: IUCN Red List, Critically Endangered, plant conservation, Raven’s garlic, Seguenza’s knapweed, white saksaul, Uzbekistan, Sicily, Iraq, arid ecosystems, Global, Regional

Oryza rhizomatis, known locally as Uru wee and in English as rhizomatous rice, is a perennial grass endemic to Sri Lanka, where it is restricted to the low-altitude dry zone of the eastern and southwestern parts of the island. Field records place it in only six administrative districts — Puttalam, Anuradhapura, Kurunagala, Hambantota, Monaragala and Ampara — in forests and open tall scrub with grassy clearings, typically in swampy or periodically flooded areas under full sun or partial shade. The dry zone where it grows receives between 850 and 1,900 millimeters of rain each year, and the species depends on the persistence of those wet microhabitats. Unlike cultivated rice, this wild relative propagates both by seed, which can enter dormancy, and vegetatively through underground rhizomes, a growth form that distinguishes it from close relatives in the Oryza officinalis complex.

The new global assessment reveals a species in measurable retreat. Using GeoCAT, the Geospatial Conservation Assessment Tool, the assessors calculated an extent of occurrence of 14,603 square kilometers, down from 20,213 square kilometers two decades earlier, a decline of roughly 28 percent. The area of occupancy, computed on a fixed 2-by-2-kilometer grid, fell from 120 to 76 square kilometers, a 37 percent reduction. Recent field surveys conducted between 2020 and 2022 failed to confirm twelve populations previously reported in Puttalam, Kurunagala, Anuradhapura and Hambantota districts, and observers describe ongoing habitat degradation driven by urbanization and grazing across most localities. On that evidence, the species qualifies as Vulnerable under criterion A2c, which captures an observed decline in area of occupancy, extent of occurrence and habitat quality over the assessment window.

The threats catalogued for the wild rice read like a checklist of pressures on Sri Lanka’s dry-zone landscapes. Housing and road construction consume populations in the Northwestern Province, where some plants grow along roadsides and are disturbed by roadside cleaning. Cattle and goats graze many sites, and elephants were observed browsing populations inside or near forests. Three populations in Puttalam district and one in Anuradhapura district sit near cultivated fields, exposing them to herbicides and pesticides and to the further expansion of agricultural land. The coastal population at Kirinda in Hambantota district was devastated by the 2004 Indian Ocean tsunami, which reduced it to as few as 20 individuals. Finally, because the species depends on swampy or periodically flooded ground, drought and the drying of swamps during the wet season under higher temperatures and reduced precipitation could erode its remaining habitat.

Conservation infrastructure for the wild rice already exists in fragments. Twenty ex situ accessions are stored in the germplasm bank of the International Rice Research Institute and 23 more at the Plant Genetic Resources Center in Peradeniya, Sri Lanka, while the species occurs within Yala, Luhugala and Wilpattu national parks. The first in situ conservation site for wild rice, at Vanathavilluwa in Puttalam district, was established under the Northwestern Province Environmental Statute No. 12 of 1990. The assessors call for research and monitoring to clarify population trends and for new seed collections to broaden the genetic variability held in storage — a notable point for crop scientists, since wild relatives of rice carry genetic diversity that breeders may one day draw upon for traits such as drought tolerance or disease resistance.

Half a world away in latitude but equally exposed, Saxifraga facchinii is a rare perennial hemicryptophyte found only on calcareous and dolomitic rocks of the Italian Dolomites, at elevations of roughly 2,600 to 3,200 meters. It occurs on several massifs — Catinaccio/Rosengarten, Pale di San Martino, Marmolada, Sella, the Sasso Lungo/Langkofel group, Tofane, Conturines and the Odle — but more than half of its occupied habitat lies in small, isolated patches, unambiguously meeting the IUCN definition of severe fragmentation. The plant belongs to high-alpine to nival communities that include endemic companions such as Draba dolomitica, and its most probable seed dispersal strategy is barichory, meaning seeds simply fall close to the parent, limiting the species’ ability to colonize new ground.

The upgraded Endangered category rests on a combination of restricted range, fragmentation and projected climate-driven habitat loss. The assessors calculated an extent of occurrence of 978 square kilometers with QGIS minimum convex hull and an area of occupancy of 196 square kilometers on the 2-by-2-kilometer grid, applying criterion B1ab(iii,iv)+B2ab(iii,iv) because of continuing projected decline in habitat extent and quality and in the number of subpopulations. Climate modeling projects habitat loss to 2070 of an average 75.52 percent under the rcp 4.5 scenario and 87.59 percent under rcp 8.5, with substantial uncertainty ranges. Field evidence already shows the process in motion: subpopulations recorded near the species’ lower elevational limit of about 2,500 meters on the Schlern/Sciliar massif at the start of the twentieth century are now missing, their former sites converted to arid meadows with Sesleria caerulea as expanding low-alpine communities fill in the rocky habitats the saxifrage requires. Additional pressures include ski-area development at high elevations, trampling by hikers and climbers on trails, and collection by rock-garden enthusiasts.

Unlike the Sri Lankan rice, Saxifraga facchinii enjoys no legal protection at international, national or regional levels, although many of its populations grow within Dolomite natural parks such as Paneveggio–Pale di San Martino, Puez Odle, Sciliar–Catinaccio, Fanes Sennes Braies and Dolomiti d’Ampezzo. The assessors recommend a research and monitoring program to illuminate the species’ reproductive biology and population trend, alongside the design of in situ and ex situ conservation measures for possible population reinforcement and assisted colonization — the deliberate movement of populations to sites where they may survive under a changing climate. A specific monitoring program, they note, would help detect emerging threats such as rising temperatures, expanding high-elevation ski resorts or stochastic events before they alter the species’ status. The assessment grew out of a doctoral thesis funded by the Free University of Bozen-Bolzano and took shape at the COST Action 18201 Conserve Plants training school on Red List assessment held in Podgorica in spring 2022, an example of how capacity-building networks feed directly into formal conservation outcomes.

The third assessment shifts the lens from global to regional, examining Helianthemum caput-felis, the cat-head rock rose, in Sardinia. This thermophilous, long-lived half shrub, 35 to 50 centimeters tall, is distributed in disjunct and fragmented populations around the western Mediterranean — the southeastern Iberian Peninsula, the Balearic Islands, Sardinia and northwestern Africa. In Sardinia it is confined to the Sinis Peninsula, where two main populations at Capo Mannu and Su Tingiosu anchor a scatter of smaller groups. The species is entomophilous and mostly self-incompatible, its hermaphrodite flowers lasting only three to four days, and its seeds germinate in autumn at the onset of the rains, with high germination rates after scarification. Globally, the species is categorized as Endangered under criterion E, and it has been listed as Endangered in Europe and Spain, Critically Endangered in Algeria and, until now, in Italy.

The Sardinian picture is unexpectedly brighter. A four-year demographic study across populations in Spain and Sardinia found the global population roughly at equilibrium, and the Sardinian populations show the greatest stability, with a finite growth rate above one, probably because they are structured mainly around large, long-lived individuals. Capo Mannu is in demographic growth and Su Tingiosu stable, apart from losses caused by landslides and collapses of the rocky cliffs. Monitoring spanning more than two decades has detected no continuous decline in extent of occurrence, area of occupancy, habitat quality or mature individuals — both metrics stand at just 16 square kilometers. On that basis the assessors downgraded the regional category to Near Threatened, while adopting the precautionary approach embedded in IUCN guidelines: if the predicted threats materialize, the species’ tiny area of occupancy combined with declining habitat surface would immediately qualify it for Endangered. All Sardinian populations fall within protected areas, including Natura 2000 sites, the Penisola del Sinis – Isola di Mal di Ventre marine protected area and the Montiferru–Sinis regional park, the species is covered by Annex I of the Bern Convention and Annex II of the Habitats Directive, and seeds are banked in European germplasm collections. Successful translocations elsewhere in its range suggest that reintroduction and reinforcement could be widely deployed if local extirpations occur.

Read together, the three assessments demonstrate the machinery of contemporary Red List science at work. Geospatial software converts herbarium records and field waypoints into extent and area of occupancy; fixed grids and criteria A through E standardize how decline, fragmentation and fluctuation are weighed; demographic matrices test whether populations are growing or shrinking; and climate envelopes project where habitat will remain by mid-century. They also show how the same framework can yield opposite trajectories — an upgrade for a climate-squeezed alpine endemic, a downgrade for a Mediterranean shrub whose monitored populations are holding steady — and how categories can change as data improve. For the rhizomatous rice, the Dolomite saxifrage and the Sardinian rock rose alike, the assessments mark not an endpoint but a baseline: the starting point for the monitoring, seed collection and habitat protection that will determine whether the next evaluation records recovery or further decline.

Subject of Research: Global and Regional IUCN Red List Assessments: 15

Article Title: Global and Regional IUCN Red List Assessments: 15

Article References: Orsenigo, S., Calderisi, G., Cogoni, D., Mondoni, A., Nascimbene, J., Rota, F., Sajeevan, T., Wijayasinghe, M., Wellstein, C., & Fenu, G. (2023). Global and Regional IUCN Red List Assessments: 15. Italian Botanist, 15, 177-191. https://doi.org/10.3897/italianbotanist.15.107040

Image Credits: AI Generated

DOI: 10.3897/italianbotanist.15.107040

Keywords: Global, Regional, IUCN, List, Assessments, scientific research

The research, led by David Jenkins and colleagues, tackles the question through beta diversity, the measure of how dissimilar two communities are in their species composition. The logic is elegant: if the clearest possible separation between adjacent communities occurs at a particular spatial resolution, then that resolution represents the empirical size of a community. Maximal beta diversity, in other words, marks the scale at which nature draws its sharpest boundaries between assemblages of species.

To test this, the team assembled point data for ants, birds, diatoms, and trees, four taxa chosen for their radically different life histories, dispersal abilities, and data origins. Ant records spanned two centuries of literature synthesis, totaling nearly 25,000 samples and 699 species. Bird data came from the eBird community science platform, with over 810,000 cleaned reports from January and July between 2010 and 2015, standardized as counts per person-hour. Diatoms were sampled from the top five centimeters of lake sediment across 1,070 lakes in a single year, yielding 1,957 species. Tree data, covering stems wider than five centimeters at breast height, came from the US Forest Service’s Forest Inventory Analysis across four years spaced five years apart, totaling nearly 40,000 samples and 403 species.

The analytical framework organized all georeferenced observations into hexagonal polygons, with cell size defining the spatial grain. At each grain, ranging from one square kilometer to one million square kilometers, the researchers calculated mean pairwise beta diversity between each central hexagon and its immediate neighbors. This neighborhood-based approach, borrowed conceptually from remote sensing research on optimized spatial resolution, controls for sample size across grains because hexagonal grids are constrained to fewer than six neighbors.

A critical methodological decision involved how beta diversity was calculated. The team used a relatively new deviation from a null model based on spatial heterogeneity, called beta NBD for its underlying negative binomial distribution. This measure, developed by Youhua Chen and Fangliang He, represents clustered species patterns more accurately than the older deviation based on spatial homogeneity. The newer measure made only modest adjustments to observed beta diversity and showed no vulnerability to sample size effects, whereas the older measure substantially altered observed values and increased with greater sampling effort.

The results were strikingly consistent. For all four taxa, beta diversity peaked at intermediate, regional spatial grains. Trees showed peak values at roughly 300 square kilometers, remarkably stable across four evaluated years spanning two decades. Winter birds peaked at approximately 500 square kilometers, while summer birds and ants both peaked near 2,000 square kilometers. Diatoms, the microscopic algae inhabiting lake bottoms, showed the largest peak at about 11,000 square kilometers. Hump-shaped quadratic curves fit the data far better than linear models, with Akaike Information Criterion weights exceeding 0.95 in all cases.

These rankings defied the team’s initial expectations. They had predicted that diatoms, confined to discrete lake habitats, would show the smallest community scales, followed by ants, trees, and birds. Instead, diatoms exhibited the largest peak, and trees the smallest. The researchers suggest that diatom community size roughly matched the average density of sampled lakes across the contiguous US, meaning the apparent regional scale may actually reflect the average dispersion of discrete diatom habitats rather than long-range dispersal. For terrestrial taxa inhabiting intergrading habitats, the regional peaks may reflect genuinely regional processes.

The repeatability of the patterns adds considerable weight to the findings. Bird beta diversity scaling was consistent across six consecutive years, though sensitive to seasonality, with winter communities concentrating at smaller spatial scales than summer ones. Tree patterns showed very little within- or among-annual variation at their peak grain. Given that birds and trees differ so enormously in life history and mobility, this commonality suggests that community assembly processes provide a strong and general basis for the observed regional scaling.

The results broadly support the regional community concept advanced by Robert Ricklefs, who argued in 2008 that regional scales should be preferred to local ones to capture the full suite of selective conditions driving diversity. The relatively low beta diversity at fine spatial grains in this study supports Ricklefs’ argument that local communities are one end of spatiotemporal continua rather than empirically distinct entities. The authors suggest that metacommunity ecology may be wise to treat local communities as sample points rather than bounded units, steering the field toward a species-centric approach.

Beyond its conceptual implications, the work bears fundamental consequences for conservation biogeography in the Anthropocene. If ecological communities are regional in scale, then local conservation efforts can best ensure success when coordinated regionally, through corridors and networks among conserved sites. The findings may also help researchers reconsider the appropriate spatial grains to use when studying diversity responses to land use and climate change, suggesting that rasterized study units matching peak beta diversity would obtain maximal resolution. The authors note that similar peaked spatial scaling may exist for other biogeographical patterns, potentially revealing maximal signal amidst noise and better representing natural patterns than maps designed for visual appeal.

Subject of Research: Peak beta diversity occurs at regional spatial grains

Article Title: Peak beta diversity occurs at regional spatial grains

Article References: Ohyama, L., Bogota-Gregory, J. D., & Jenkins, D. G. (2024). Peak beta diversity occurs at regional spatial grains. Frontiers of Biogeography, 17, Article e132675. https://doi.org/10.21425/fob.17.132675

Image Credits: AI Generated

DOI: 10.21425/fob.17.132675

Keywords: Peak, beta, diversity, occurs, regional, spatial, grains, scientific research

The study focused on physeal fractures, injuries that involve the physis, the cartilaginous growth plate found only in children whose skeletons are still developing. These fractures matter enormously. The growth plate is the engine of longitudinal bone growth, and trauma to it can disrupt that process permanently, leading to limb-length discrepancy, angular deformity, or arrested growth if the injury is not recognized and managed correctly. Radiographic assessment is notoriously difficult because cartilage is invisible on plain X-rays, so the fracture line may be subtle, displaced, or entirely occult. Salter-Harris type I injuries, in which the fracture passes through the physis itself, are particularly treacherous, since the only clue may be soft-tissue swelling or a barely perceptible widening of the growth plate.

The research team designed a retrospective, single-center, case-only analysis. They assembled 805 confirmed physeal fractures in children aged zero to twelve years, all sustained during 2025. Each case was evaluated once, in March and April 2026, through a recorded GPT-5 API workflow. The model received anonymized radiographs together with a small, structured package of clinical context: the child’s age, the target anatomical region, and a brief complaint or mechanism of injury. Direct identifiers, accession numbers, exact dates, and burned-in text were removed or masked before submission, in compliance with institutional approval and the Turkish Personal Data Protection Law. The model outputs were generated retrospectively and never influenced the clinical care of any patient.

The reference standard was deliberately conservative. Three orthopedic surgeons, a pediatric orthopedics specialist, an orthopedic trauma specialist, and a senior orthopedic surgeon, each with at least eight years of clinical experience, independently reviewed every radiograph. They then reached consensus with the model outputs concealed, and no radiologist took part. The primary endpoint was a definite or probable fracture call in the correct anatomical region. Importantly, the study did not require the model to explicitly recognize the physis or to assign a correct Salter-Harris classification; simply flagging a fracture in the right place was enough to count as a success. This design choice means the reported sensitivity of 72.5 percent is, if anything, a generous measure of the model’s performance.

Because every case in the cohort was fracture-positive, the study could not estimate specificity, predictive values, overall accuracy, or receiver operating characteristic measures. That is a structural consequence of the case-only design, and the authors are explicit about it. What the design does allow is a clean, unobstructed view of false negatives, the failures that matter most in a screening context. When a model says no and the patient actually has a fracture, the consequences can be serious: delayed presentation, inappropriate discharge, and, in the case of physeal injuries, potential permanent growth disturbance. Understanding where and why those failures cluster is the first step toward mitigating them.

The headline finding was the overall sensitivity of 72.5 percent, with a 95 percent confidence interval of 69.4 to 75.5 percent. But the more interesting story lies in the stratified patterns. In an unadjusted secondary comparison, the false-negative rate was 29.9 percent for Salter-Harris type I and II injuries, compared with 17.9 percent for type III and IV injuries, a risk ratio of 1.67 with a confidence interval of 1.15 to 2.41 and a p-value of 0.004. When type I injuries were excluded entirely, sensitivity barely moved, rising only marginally to 72.8 percent. The pattern is consistent with a known weakness of image-based models: injuries whose radiographic signature is subtle or indirect, such as fractures confined to or passing through the radiolucent growth plate, are harder to detect than injuries with obvious cortical disruption.

Anatomical location also mattered. The descriptive false-negative rates were highest in the elbow, at 36.9 percent, and at the distal tibia, at 31.2 percent. The distal radius, one of the most common sites of pediatric physeal fracture and often the most radiographically obvious, fared better, with a false-negative rate of 20.1 percent. The elbow is a notoriously complex region, with multiple ossification centers that appear and fuse in a predictable sequence as a child grows, and distinguishing a normal ossification center from a fracture fragment requires sophisticated pattern recognition. The distal tibia presents its own challenges, with overlapping bony structures and subtle physeal widening that can evade even experienced eyes.

Age, by contrast, did not significantly change the failure rate. Children aged zero to six years had a false-negative rate of 30.0 percent, while those aged seven to twelve had a rate of 24.9 percent, a difference that was not statistically significant with a p-value of 0.108. This is a hypothesis-generating observation rather than a definitive conclusion, but it suggests that the model’s difficulties are driven more by the nature of the injury and the anatomy involved than by the developmental stage of the skeleton. Younger children do have more cartilage and fewer ossified structures, which might have been expected to make detection harder, yet the difference did not reach statistical significance in this cohort.

The authors are careful to delineate the limits of what their study can claim. It did not assess explicit physeal recognition or Salter-Harris classification, and it cannot establish full diagnostic accuracy, performance relative to clinicians, or clinical utility. There was no comparison group of human readers, so the study cannot say whether the model is better or worse than the clinicians it might assist. The single-center design and the restriction to confirmed fracture cases mean the findings describe failure patterns in a specific population rather than real-world screening performance. Still, the central conclusion stands on its own: under the recorded context-assisted GPT-5 API workflow, more than a quarter of confirmed physeal fractures did not receive a qualifying regional fracture call.

What does this mean for the deployment of multimodal large language models in pediatric imaging? The study is a caution against treating these systems as diagnostic oracles, particularly in the high-stakes domain of growth plate injuries, where a missed fracture can alter a child’s skeletal future. The regional and Salter-Harris stratifications offer actionable intelligence: clinicians and developers now know that type I and II injuries, elbow fractures, and distal tibial fractures are the model’s weakest zones, and that vigilance should be highest exactly where the model is least reliable. As multimodal AI continues its rapid advance into radiology workflows, studies like this one, which map failure patterns with the same rigor usually reserved for successes, will be essential to ensuring that the technology augments human judgment rather than quietly undermining it.

Subject of Research: Artificial intelligence models can miss subtle pediatric growth plate fractures, and a new study maps exactly where the failures cluster.

Article Title: Regional fracture-detection false-negative patterns of a multimodal large language model in confirmed pediatric physeal fractures: a case-only analysis

Article References: Erginoğlu, S. E., Ülgen, N. K., Nazlıgül, A. S., Yiğit, N., & Akkurt, M. O. (2026). Regional fracture-detection false-negative patterns of a multimodal large language model in confirmed pediatric physeal fractures: a case-only analysis. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02771-9

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02771-9

Keywords: Regional, fracture-detection, false-negative, patterns, multimodal, large, language, model, confirmed, pediatric, physeal, fractures

The Rila Mountains, the highest range in Southeast Europe, present a challenging environment for soil modeling due to their significant altitudinal variation and diverse climatic conditions. The study area covers approximately 263,000 hectares, with elevations ranging from below 1,000 meters to the peak of Musala at 2,925 meters. This vertical gradient creates distinct vegetation zones, from oak woodlands in the foothills to alpine meadows at the summit, each influencing soil formation processes differently. The researchers utilized this natural laboratory to test the suitability of SoilGrids for ecosystem accounting, a framework that integrates environmental data with economic analysis to inform sustainable policy decisions. By focusing on a region with limited local soil data, the study highlights the potential of global models to fill critical information gaps in national monitoring programs.

The methodological approach involved comparing SoilGrids predictions with a robust set of independent point observations collected from three distinct sources. These sources included field data from the International Co-operative Programme on Monitoring and Assessment of Air Pollution Effects on Forests, known as ICP-Forest, published scientific data, and new observations conducted specifically for this study. The team analyzed 39 sample plots distributed across different altitudes and soil types, primarily Cambisols, Umbrisols, and Luvisols. To ensure a fair comparison, the researchers focused on the top 30 centimeters of soil, a standard depth for many ecosystem assessments. They evaluated six key parameters: pH, bulk density, coarse fraction, soil organic carbon content, nitrogen content, and soil organic carbon stock. This multi-source validation strategy helps mitigate the biases that might arise from relying on a single dataset.

The results indicate that SoilGrids performs well in capturing the general dynamics of soil parameters across the altitudinal range. There was a strong agreement between the observed and predicted average values for bulk density, coarse fraction, and soil organic carbon content. Specifically, the average measured soil organic carbon stock in the top 30 centimeters was 58.54 tonnes per hectare, while the average predicted value was 55.38 tonnes per hectare. Statistical tests, including paired t-tests and Wilcoxon signed-rank tests, confirmed that there were no significant differences between the field data and model predictions for these parameters. This suggests that SoilGrids can reliably estimate the carbon stored in soils, a crucial metric for understanding the global carbon cycle and assessing the capacity of ecosystems to sequester carbon.

However, the study also revealed notable discrepancies in the prediction of nitrogen content and soil pH. The model predicted nitrogen levels that were almost twice as high as those observed in the field. Similarly, the SoilGrids data indicated soils that were less acidic than those measured in the field, with predicted pH values mostly between 5.7 and 6.6, whereas field observations reported values between 4.5 and 5.8. These differences are statistically significant and suggest that the global model may not fully capture the specific biogeochemical processes occurring in mountainous forest soils. The researchers noted that the SoilGrids model is trained on a dataset dominated by agricultural land and older sampling dates, which may not accurately represent the unique conditions of high-altitude coniferous forests where organic matter decomposition and nutrient cycling differ significantly from agricultural systems.

The discrepancies in pH and nitrogen are likely attributable to the input data used to train the SoilGrids models. The global database contains approximately 240,000 soil profiles, but more than 60% of these were collected between 1960 and 2000. In regions like the Rila Mountains, where soil acidity is driven by the weathering of carbonate-free silicate rocks and the influence of coniferous vegetation, these historical biases can lead to underestimations of acidity. The study highlights that while the model captures broad spatial trends, such as the increase in soil organic carbon with altitude, it may lack the precision required for detailed local assessments of specific chemical properties. This finding underscores the importance of validating global models with local data before using them for high-stakes policy decisions, such as reporting on greenhouse gas emissions under the United Nations Framework Convention on Climate Change.

Despite these limitations, the study concludes that SoilGrids remains a valuable tool for ecosystem reporting, particularly when local data is scarce. The model provides a spatially explicit representation of soil properties that can be integrated into ecosystem accounting frameworks, such as the System of Environmental-Economic Accounting for Ecosystems. By combining SoilGrids data with local knowledge of climate, vegetation, and soil types, researchers can improve the accuracy of their assessments. The study recommends using stratification methods to group soil measurements by climate region and soil type, ensuring that each stratum has sufficient data to minimize sampling error. This approach allows for a more nuanced understanding of soil variability and helps in deriving region-specific emission factors that are essential for accurate national greenhouse gas inventories.

The implications of this research extend beyond Bulgaria, offering insights for other regions with similar ecological characteristics. As countries strive to meet their climate commitments and monitor biodiversity loss, the need for reliable soil data becomes increasingly urgent. SoilGrids offers a cost-effective and accessible solution for generating this data, but its application must be accompanied by careful validation and contextualization. The study demonstrates that while global models can provide a solid foundation for ecosystem assessments, they are not a substitute for local field measurements. By understanding the strengths and weaknesses of these tools, scientists and policymakers can make more informed decisions about land management, conservation, and climate mitigation strategies. The Rila Mountains study serves as a case study for the broader challenge of integrating global data with local realities to achieve a comprehensive understanding of our planet’s soil resources.

Subject of Research: Comparative assessment of SoilGrids system with regional soil data for advancing ecosystem reporting in Bulgaria

Article Title: Comparative assessment of SoilGrids system with regional soil data for advancing ecosystem reporting in Bulgaria

Article References: Stoeva, L., & Zhiyanski, M. (2024). Comparative assessment of SoilGrids system with regional soil data for advancing ecosystem reporting in Bulgaria. One Ecosystem, 9, Article e133091. https://doi.org/10.3897/oneeco.9.e133091

Image Credits: AI Generated

DOI: 10.3897/oneeco.9.e133091

Keywords: Comparative, assessment, SoilGrids, system, regional, soil, data, advancing, ecosystem, reporting, Bulgaria, scientific research