Environmental Research Letters

Background: Surgical care centralization in the U.S. delays access and increases carbon emissions. Global targets suggest patients live within 2-hours of a surgical facility. This study quantifies the environmental impact of travel for cataract surgery in rural Michigan and models the potential emissions reductions from decentralizing surgical and follow-up services. Methods: A retrospective, cross-sectional study analyzed electronic medical records from a rural Michigan ophthalmology practice (March-November 2023). We calculated travel distances using population-weighted centroids and estimated emissions using U.S. Department of Energy vehicle data. A k-means clustering model optimized additional facility placement, and a gradient analysis identified optimal numbers for decentralization points, for emissions reductions. Results: The 920 patients traveled a median of 55.45 km (IQR: 43.33-88.20 km) for surgery and 55.07 km (IQR: 43.54-87.82 km) for follow-up visits, generating Total Surgical Access Emissions (TSAE) of 57,168 kgCO2; (median of 59.20 kgCO2; IQR: 32.31-81.87) under the centralized model. The k-means decentralization model and gradient analysis identified 7 hospitals and 9 clinics, respectively, as the optimal expansion points, reducing emissions by 34.07% (19,475 kgCO2 saved) and 39.52% (22,590 kgCO2; saved). The Surgical Access Carbon Impact (SACI) model demonstrated that achieving two-hour access to clinic services reduced excess emissions by 54.7%. Sensitivity analyses using fuel-efficient vehicles (Toyota Prius and Tesla Model 3) or reducing follow-up visit frequency reduced emissions by 54.03% (30,888 kgCO2) and 25.83% (14,768 kgCO2), respectively. Conclusion: Decentralizing surgical services in rural U.S. settings could cut travel-related emissions by up to 40%, significantly reducing healthcare-related carbon footprints while improving timely access to care. The SACI metric provides a novel framework for integrating environmental sustainability into U.S. health policy and service planning

Tropical dryland ecosystems are highly biodiverse and fragmented and are experiencing significant anthropogenic and climatic changes. With increasing extremes in temperature and precipitation, coupled with significant alteration, these ecosystems are at greater risk of increased exposure and vulnerability to climatic change; however, little work has quantified the climatic shifts occurring within these ecosystems globally. Here, we aim to fill this gap by using the ERA-5 reanalysis and CHIRPS precipitation data to quantify changes in essential climatic variables in tropical drylands since 2000. Overall, we find that regional pressures differ, with tropical dry forests, savannas, and shrublands becoming hotter and drier in the Neotropics and parts of the Afrotropics and Australasia. By contrast, the tropical dry forests in the Indomalayan, Oceania, and Nearctic are experiencing hotter and wetter conditions. Globally, though, these ecosystems are experiencing more change than the global average, suggesting they may be approaching tipping points in their resilience, ultimately shrinking the area where they can survive.

Dengue fever risk and burden has increased globally in the past decade, with record-breaking outbreaks driving high case numbers, outbreaks increasing in existing transmission suitable regions, and occurring in new locations. A combination of global change processes, including climate change, have provided the environmental backdrop for introductions and resurgences of mosquito-transmitted dengue virus. Understanding shifts in exposure risk is integral to public health preparedness. This study provides global mapping of the thermal suitability of dengue transmission for CMIP6 climate scenarios, across a range of general circulation models (GCMs), and we created spatially explicit demographic projections of transmission risk using year-matched RCP-SSP frameworks for demographic and emissions scenarios. Globally, poleward shifts in projected distributions of suitability for transmission for both Ae. aegypti and Ae. albopictus suitability are shown in both the near term (2030s) and longer term (2050). Under a 'middle of the road' climate scenario (CMIP6 SSP2-4.5), regions in Africa and Asia are the major areas driving increases in year-round (12 months) population at risk (PAR) through 2050, with an anticipated net gain in 932 million people at risk for Ae. aegypti transmission and 24 million for Ae. albopictus, which includes multiple regions losing areas of year-round suitability as temperatures exceed the higher thermal boundary for transmission. In contrast, the estimated net increase in PAR for one or more months of transmission suitability at a global scale by 2050 is 3.29 billion people for Ae. aegypti transmission and 3.30 billion for Ae. albopictus transmission. This snapshot of a 'middle-of-the-road' combination of climate and demographic driven increases in potential dengue transmission exposure emphasizes the importance of both expanding suitability in new areas, and growing populations in areas approaching and becoming exposed year-round. Globalization, urbanization, and shipping will continue to provide the potential for introductions into newly suitable areas as season lengths increase, sparking outbreaks in unexposed populations. This is compounded and becomes ever more probable as the number of people and places at year-round risk also increases. This project provides all global gridded outputs for onward mapping and reuse, to add to the toolkit to anticipate and prepare for prevention and response to dengue in a changing world.

ContextArtificial light at night (ALAN) has profound impacts on individual organisms and entire communities. Still, humans tend to underestimate the true biological (spatial) footprint of ALAN, in part because of our limited sensitivity to light compared to other organisms. ObjectivesWe sought to demonstrate how far ALAN can reach into dark spaces at levels that can impact organismal behavior and physiology using a fundamental physical law, the inverse square law. MethodsWe created a spatially explicit model of light spread on real landscapes, parameterized using increasingly available landscape-scale vegetation data to account for attenuation through forests and blocking by topographic relief. ResultsLighting types common in rural areas can produce biologically important effects more than 1 kilometer from the source, and effects of large lights might stretch 3 kilometers or more. The footprint of a light is determined by the complex and multidimensional interaction between characteristics of the light itself and the environment. For example, attenuation through a dense forest might decrease the footprint of a light more than 90% compared to the same light on a grassland. In complex environments, even small changes in light placement and characteristics can lead to large changes in the biological footprint of the light. ConclusionsDesigners and land stewards must account for lighting type, brightness, directionality, and reflected light to create ecologically responsible lighting. Vertical vegetation and topography strongly influence the propagation of biologically detrimental light, and environmental context is vital when planning and installing lights to minimize the biological impacts.

Biochar amendments are increasingly applied to improve soil physical functioning and support carbon dioxide removal, but their effects on intrinsic soil thermal properties remain poorly characterised. We conducted the first global systematic meta-analysis of 19 independent studies, 231 control-biochar comparisons, and 529 property-specific effect sizes to test how biochar changes soil heat transfer and storage. Biochar reduced thermal conductivity by 17.6% (95% CI, -22.7 to -12.2), thermal diffusivity by 11.0% (-14.5 to -7.3), and volumetric heat capacity by 8.3% (-12.3 to -4.1). Gravimetric heat capacity showed no significant overall response (+3.3%; -7.6 to 15.4) but was supported by fewer studies. Negative responses were directionally consistent for thermal conductivity, diffusivity, and volumetric heat capacity. Moderator analyses showed that responses were most consistently associated with post-application bulk density and changes in bulk density, while application rate modulated response magnitude and soil texture constrained context dependence. Co-variation among thermal conductivity, thermal diffusivity, and volumetric heat capacity matched expected physical dependencies, indicating coordinated structural reorganisation rather than independent shifts in isolated parameters. These estimates describe intrinsic conductive and storage properties; field-scale soil temperature responses may also be modified by albedo, evaporation, vegetation, and surface energy balance. Improved integration of soil thermal measurements with moisture dynamics, structural changes, and carbon cycling is essential to accurately represent biochar effects in soil and land-surface models.

Nepal's rangelands provide multiple benefits, including support for pastoral livelihoods and alpine biodiversity, regulation of water and soil nutrients, and sequestering carbon. Climate change and anthropogenic pressures are altering these rangelands, leading to vegetation and biodiversity change. However, national-scale assessments of rangeland change are limited in Nepal. This study quantified rangeland changes at multiple spatial scales and assessed the climatic and non-climatic drivers of rangeland change. About 80.7% of Nepal's high-altitude rangeland (> 2,000m) outside protected areas showed no significant change. Among areas exhibiting significant annual maximum NDVI trends, 383,281 ha (18.6%) showed positive and 14,702 ha (0.7%) showed negative trends, corresponding the ratio of increase in vegetation greenness and decline in vegetation greenness to 26:1. Climate predicted positive trends covered 627,184 ha (30.5%), whereas residual trends caused by non-climatic drivers covered 94,656 ha (4.6%). Climate induced negative trends covered 47,609 ha (2.3%) while residual trends were observed in 6,260 ha (0.3%). Negative trend pixels were concentrated mainly within the 3,000 to 5,000 m elevation band, with Karnali Province recording the highest proportional climate predicted decline in vegetation greenness (3.4%). At the municipality scale, rangeland change showed no significant relationship with grazing pressure derived from gridded livestock data, suggesting that grazing pressure alone did not explain the non-climatic vegetation signal. These spatially explicit, nationally consistent results identify where rangeland change is occurring and help distinguish climatic and non-climatic drivers of rangeland vegetation change, providing evidence to support targeted rangeland management under Nepal's federal governance structure.

Native vegetation protection is a key strategy for delivering both biodiversity and climate benefits, and protected areas have been widely adopted to keep tropical biomes standing. Yet deforestation is driven by interrelated environmental and social factors, and the effectiveness of protected areas varies considerably across space. Here, we evaluated the impact of 802 protected and conserved areas in the Brazilian Amazon on preventing vegetation loss and avoiding carbon emissions over the past 40 years using statistical matching to address the location bias of protection. We found that protected areas were effective throughout the study period, reducing the probability of deforestation per km2 by an average of 0.5 percentage points per year. While the Amazon biome lost 14% of its native vegetation between 1986 and 2024, protected areas prevented the deforestation of 290,436 km2, nine times their actual internal loss. They also stored 45,336 Mt of carbon in 2016 (61% of the Amazon stock) and prevented the emission of 7,300 Mt of CO2 by 2024. Deforestation inside the areas and remoteness reduced their impact, while areas that were initially more preserved were more effective. Area size and age had no influence over impact once we analyzed the amount of avoided deforestation per size and age. Impact also varied across Brazilian states, highlighting the role of regional context. All three protection categories (conservation units, indigenous lands, and quilombola territories) showed a positive mean impact, indicating that each, in aggregate, contributes to reducing deforestation. These findings provide robust evidence of the substantial role of Amazonian protected areas in habitat conservation and climate mitigation, while underscoring that this contribution remains undervalued. We advocate for strategically expanding protection to areas of greatest potential impact, and for securing adequate funding to ensure protected areas can fulfill that potential.

Abstract Background: Despite considerable reductions in under-five mortality during the Millennium Development Goal era, progress towards Sustainable Development Goal (SDG) 3.2 remains uneven across Africa. Identifying countries at greatest risk of missing the target is essential for prioritizing interventions and resource allocation. Methods: A Bayesian spatial forecasting ecological study was conducted using 2024 country-level data from 49 African countries obtained from UNICEF. Spatial dependence was assessed using Global Moran's I and Local Indicators of Spatial Association. Bayesian structured additive regression models with Gaussian, Gamma, and Exponential likelihoods were fitted using Integrated Nested Laplace Approximation (INLA) and compared using the Deviance Information Criterion (DIC), Watanabe-Akaike Information Criterion (WAIC), and conditional predictive ordinates. Posterior exceedance probabilities were estimated, an SDG Failure Index (SFI) and a Priority Intervention Index (PII) were developed, and Bayesian posterior predictive simulations were performed to estimate country-specific probabilities of attaining SDG 3.2 by 2030. Results: Significant spatial clustering of under-five mortality was observed with (Moran's I = 0.355, p < 0.001), and hotspots in Benin, Cameroon, and Nigeria. The Gamma model provided the best fit (DIC = 114.92; WAIC = 111.71). Diarrhoea was the only significant predictor (posterior mean=0.030; 95% credible interval: 0.004-0.056). Twenty-three countries (46.9%) were classified as high risk, whereas only five (10.2%) had achieved SDG 3.2. West Africa recorded the highest mean mortality (7.05%) and North Africa the lowest (1.64%). Bayesian projections indicated that only five countries were likely to achieve SDG 3.2 by 2030, while 41 (83.7%) were unlikely to do so. Conclusion: Considerable geographical inequalities in under-five mortality persist across Africa, and most countries remain off-track for achieving SDG 3.2 by 2030. The integration of exceedance probability mapping, the SDG Failure Index, the Priority Intervention Index, and Bayesian probability forecasting provides a practical framework for monitoring progress and prioritizing countries requiring accelerated action towards achieving SDG 3.2.

The impact of weather on the health of Americans and the American health system is substantial. Using available health and economic data, we developed a data-driven scenario that describes a compounded heat emergency in an archetypal community in the United States. We then characterize the potential human and economic costs of such a heat emergency to demonstrate the widespread impact on health outcomes, health systems, and society.

Rift Valley fever (RVF) is a mosquito-borne disease that can cause severe illness and death in both humans and livestock. Since 2016, Uganda has experienced recurrent but localized RVF outbreaks concentrated in the countrys southwestern region. The ecological drivers of this emergence remain unclear, as outbreaks have occurred throughout the year and show little association with meteorological patterns. We evaluated whether crop cultivation, particularly banana cultivation, is associated with RVF outbreak occurrence after controlling for likely confounders. We conducted a longitudinal study of human-inhabited 5 x 5 km grid cells across southwestern Uganda from 2016-2024. Annual Sentinel-2 satellite imagery composites were used to classify land cover into banana, coffee, ground crops, and non-crop categories, and the proportion of each land type was calculated for every grid-cell year. Because land cover proportions are compositional, isometric log-ratio transformations were used to estimate the independent effects of each land type. Confounding was addressed through propensity weighting, and crop substitution effects were estimated using g-computation. Banana land cover was the only land type consistently associated with increased RVF outbreak likelihood. In grid-cell years with low baseline banana cover, a 10-percentage point substitution from other land classes into banana was associated with a 1.64-fold increase in the odds of an RVF outbreak (95% CI: 1.17-2.29). In a simplified banana-only model, each 10-percentage point increase in banana cover was associated with a 1.21-fold increase in outbreak odds (95% CI: 1.02-1.43). Holding banana cover constant, substitutions among coffee, ground crop, and non-crop land showed weak or null associations. These findings suggest that banana cultivation may be an important ecological feature influencing RVF transmission dynamics and outbreak risk in southwestern Uganda. Author SummaryRift Valley fever (RVF) is a mosquito-borne disease that affects both humans and livestock and has caused repeated outbreaks in southwestern Uganda since 2016. While rainfall and flooding are often linked to RVF outbreaks elsewhere, Ugandas recent outbreaks have occurred across seasons and are not well explained by weather patterns alone. We investigated whether agricultural land use could help explain where outbreaks occur. Using satellite imagery from 2016-2024, we measured the amount of banana cultivation, coffee cultivation, ground crops, and non-crop land across southwestern Uganda and evaluated their association with RVF outbreak occurrence. We found that areas with greater banana cultivation were consistently more likely to experience RVF outbreaks, even after accounting for environmental and demographic factors. In contrast, coffee, ground crops, and non-crop land showed little evidence of an independent association with outbreak risk. These findings suggest that banana cultivation may create ecological conditions that favor RVF transmission. Rather than indicating that bananas themselves cause disease, the results point to banana-growing landscapes as potential environments where interactions among mosquitoes, livestock, and humans may increase transmission opportunities. Understanding these local ecological drivers could help improve surveillance, risk assessment, and prevention strategies for RVF in Uganda and other endemic regions.

Aims: Species' distributions are determined by the interplay between ecological niche and dispersal ability, constrained by biogeographical barriers. Bald-headed uakaris (Cacajao spp.) are highly specialized primates often associated with seasonally flooded forests. In this study, we used ecological niche models to assess changes in habitat suitability and geographic distribution of uakari species under future scenarios. Location: Western Amazonia. Methods: We integrated ecological niche models, current deforestation data, and dispersal ability to estimate habitat suitability under two Shared Socioeconomic Pathway (SSP) scenarios: intermediate (SSP2-4.5) and very high (SSP5-8.5) greenhouse gas (GHG) emissions. Results: Our models project shifts in suitable conditions for all species. Three of the five species are projected to experience substantial reductions ([&ge;]62%) in suitable habitat conditions within their current ranges by 2050 under both future scenarios. Across the western Amazonia, up to 219,189 km2 and 211,276 km2 of land are projected to be unsuitable within the uakari ranges under the intermediate and very high emissions scenarios, respectively. This is particularly relevant for C. calvus, C. rubicundus, and C. ucayalii. At the species level, the uakaris may lose between 343 km2 and 84,531 km2 of their ranges in the intermediate scenario and 858 km2 and 76,216 km2 in the very high scenario. Shifts in suitability due to climate change are expected to vary from 6 to 191 km in the intermediate scenario and from 5 to 168 km in the very high scenario. Furthermore, the uakaris may lose between 0.5% and 8% of their current ranges due to deforestation in all scenarios. Main conclusions: Our findings reveal a high sensitivity of the uakaris to climate change impacts. It is projected that all species may experience contractions in the suitable areas and spatial suitability within their ranges by 2050, underscoring climate change as a relevant threat to these taxa.

Conventional mosquito surveillance typically relies on contemporaneous data, making it challenging to anticipate future vector surges. To support proactive vector management, this study evaluates a multi-model forecasting framework designed to generate probabilistic 1- to 4-week-ahead forecasts of Aedes aegypti relative abundance per trap night. The framework was validated using multi-year surveillance data across four US jurisdictions spanning varied environments (from subtropical to temperate and arid). We found that an ensemble approach aggregating statistical and machine learning models generally achieved the best performance across all locations and forecast horizons. Relative forecast performance improved as the forecast horizon extended from 1 to 4 weeks ahead. The most challenging data to forecast were primarily restricted to low mosquito activity periods or atypical population peaks with unusual timing or magnitude. While full integration into routine vector management workflows represents a long-term process requiring operational adaptation, this work advances forecasting research and establishes a baseline for translating these approaches into real-time applications for public health authorities, with downstream effects in mitigating the risks of mosquito-borne diseases.

AO_SCPLOWBSTRACTC_SCPLOWMountain gradients facilitate our understanding of species range limits, competition dynamics, stress-resilience trade-offs, and determinants of vegetation zone boundaries. Forest compositional models often use altitude as the main predictor, a proxy for temperature that is defensible where floristic transitions are gradual and climate relationships are linear. However, mountains with distinct assemblages, representing tropical gradients or areas with complex biogeographic history, require a modeling framework that reflects non-linear dynamics or interactions between environmental factors, including outlier events (rather than mean conditions). Our study system encompasses both tropical and temperate forests along a broad ([~]3000 m) altitudinal gradient, positioned within a narrow latitudinal band (< 1{degrees}) and composed of mature, continuous forest in the Bhutan Himalaya. To represent the breadth of climatic conditions experienced over a trees lifetime, we used a Bayesian modeling paradigm and integrated multi-generational field knowledge to develop a priori hypotheses and informed priors, with consideration of monsoon seasonality and possible ecophysiological thresholds. Our approach followed three stages (the Pattern, the Mechanism, the Test). Specifically, we interpolated microclimate data and derived custom metrics based on thermodynamics, propagating uncertainty into subsequent models to test whether climate posteriors outperformed altitude in explaining growth form partitioning. For spatial patterns, we identified six distinct vegetation zones (encompassing 145 species from 57 families), with a mid-gradient peak in richness at the tropical-temperate transition zone, and convergence of deciduousness at either end of the gradient. For individual growth forms, abundance was tied to different ecological mechanisms, explained by adaptations to climatic stressors and competition trade-offs. For instance, evergreen broad-leaved dominance was linked to ephemeral cloud immersion, whereas tropical deciduous species were affiliated with higher vapor pressure deficit at lower altitudes. Most importantly, compositional (between-group) models showed that the interaction between frost events and fog probability (air saturation prior to the dry season) governed growth form partitioning more than any single factor; temperate deciduous species, confined to a narrow altitudinal band, exemplified this finding. Our methodological approach is transferable to other data-sparse mountain systems, and our results highlight the vulnerability of unique habitat types and montane endemics under climate change scenarios that alter the fog-frost dynamics. Second abstract in DzongkhaTo see the second abstract in Dzongkha, the official language of Bhutan, please visit our Zenodo site: https://doi.org/10.5281/zenodo.19081441.

Tropical dry evergreen forests (TDEFs) are a unique and highly threatened forest type of the dry tropics. Their restoration could be strengthened if native species demonstrate carbon sequestration comparable to widely used non-native trees. We assessed biodiversity and carbon sequestration in a restored TDEF in India, developed over 50 years from a largely barren landscape. The site now supports high woody-plant diversity, with 91 native species across 34 families. Aboveground biomass (AGB) averaged 66.91 +/- 41.2 Mg/ha comparable to seasonally dry tropical forests globally. Although native species were planted more recently and are shorter than non-natives, they contributed 23.86 +/- 23.4 Mg/ha to AGB and show potential for future increases in basal area. Given their comparable wood densities and capacity to attain similar heights, native species are predicted to sequester carbon at levels similar to non-natives in the long term. AGB was unrelated to species diversity. Overall, native TDEF species can achieve carbon storage while maintaining ecological integrity.

The introduction of salt iodization is associated with improved health and socioeconomic outcomes, but is not yet universally adopted and not always sustained. Using a quasi-experimental event study with difference-in-differences over space and time, we quantify the impacts of iodine deficiency in utero and infancy on childhood mortality and later academic achievement in Ethiopia, comparing cohorts born just before and after the May 1998 border closure that interrupted access to iodized salt. Rural children with fewer months of early-life exposure to iodized salt scored lower on standardized secondary-school exams, especially in districts with low environmental iodine, with excess deaths emerging in infancy and persisting through early childhood. These findings reveal the long-term benefits of salt iodization for health and education, especially for people with low intake of iodine from environmental sources.

Quagga mussels (Dreissena rostriformis bugensis) are ecosystem engineers that can alter nutrient cycling, benthic-pelagic coupling, and food-web structure in deep lakes. Although their invasion trajectories are well documented in the Laurentian Great Lakes in North America, depth-specific population dynamics remain poorly resolved in recently invaded European perialpine lakes. We analyzed five annual lake-wide surveys (2021-2025) from 54 stations spanning 2.4-253 m depth in Lake Constance to quantify changes in quagga mussel density, biomass, and shell-length distribution. Contrary to expectations of lake-wide exponential growth, shallow-water populations (< 20 m) showed no significant increase during the study period and appear to have reached carrying capacity before monitoring began. In contrast, densities increased monotonically at intermediate depths (40-125 m), indicating ongoing expansion into deeper strata. Mean shell length declined with depth, and size distributions in shallow waters shifted toward larger individuals, consistent with a transition from active recruitment to somatic growth of established mussels. Compared with the Laurentian Great Lakes, Lake Constance already has substantially higher shallow-water biomass, whereas deeper invasion trajectories are broadly similar. These results show that quagga mussel invasion in deep European lakes can combine rapid littoral saturation with slower profundal expansion, complicating direct transfer of predictions from the Great Lakes. Continued depth-stratified monitoring will be essential for anticipating future ecosystem effects in perialpine lakes.

The Southern Hemisphere is home to extensive forests of giant kelp (Macrocystis pyrifera), including in Argentina and the southern islands of Tierra del Fuego, which has been proposed as a potential climate refugium. This study presents the first regional time series of M. pyrifera canopy dynamics in Argentina using Landsat satellite imagery from 1985 to 2023. The forests analyzed support 247.61 km{superscript 2} of emergent canopy and are situated in the coastal waters of Argentina and a small portion of Chilean islands, with 4%, 28%, and 68% in the Chubut, Santa Cruz, and Tierra del Fuego A.e.I.A.S, respectively. The small portion of Chilean Islands are included as part of the Tierra del Fuego province analyses. Range limits were scrutinized, in part, using expert knowledge and multisatellite comparisons. Linear regression shows that between 1998 and 2023, 7.4% of kelp sites exhibited a significant trend in annual canopy area, with all observed significant trends in the positive direction. Partitioning by province boundaries, linear regression produces significant positive increases in kelp canopy area across all three provinces, although reassessment when longer temporal continuity is also warranted, where available. Observed seawater nitrate concentrations were high throughout the region (7-23 {micro}mol L-{superscript 1}), suggesting that nitrate availability was not a primary driver of canopy variability. However, positive relationships between kelp canopy and the Antarctic Oscillation suggest that regional climate variability--which alters sea surface temperature and other oceanographic conditions--may be exerting a strong influence on kelp dynamics in this region. These findings document relative stability of kelp forest area in Argentina over the most recent two and a half decades and provide preliminary evidence supporting possible increases in kelp area for the region.

Background: The effect of extreme temperatures on miscarriage is not well understood. Even less understood is the gestational period most vulnerable to extreme temperature exposure, as early miscarriages are often missed in incident datasets. We employ a birth-rate based approach to infer the risk of miscarriage in response to extreme temperature exposure by gestational week. Methods: We conducted a population-based ecological study using birth registry data from the 7,948 municipalities of Italy between 2013 and 2024 (4.5 million births). The analyses were stratified by five climatically coherent macro-regions (Ecoregions). To infer unreported pregnancy losses, we regressed birth rates dated from the last menstrual period against weekly temperatures across gestational weeks 3-21, accounting for temporal seasonality and spatial heterogeneities. Findings: Exposure to heat (mean weekly temperature of 30.4 degree/C) during gestational weeks 3-4 was associated with a reduction of birth rates of 1.62 (0.71 - 2.51)%, and of 1.91 (0.92 - 2.88)% to mean weekly temperature of 1.6 degree/C. Whilst heat was found to be harmful during gestational weeks 3-4 and 18-21, cold spells were found to be consistently harmful from the 3th up to the 12th week, depending on the Ecoregion. Interpretation: Pregnancies are vulnerable to extreme temperatures during the post-conceptual period and the second trimester. The findings underscore the need for a pre-conceptual cohort to clarify the mechanisms of loss, and urge public health action to protect pregnancies from the beginning of gestation.

Antarctic benthic organisms are highly diverse and play a critical role in the Southern Ocean ecosystem. Despite decades of sampling, vast areas of the Antarctic continental shelf remain biologically unsurveyed due to logistical and financial constraints, limiting baseline knowledge essential for effective conservation planning. Species distribution models (SDMs) allow biodiversity to be inferred in the absence of biological data by linking benthic community patterns to environmental predictors. However, the resolution of the environmental predictors, particularly bathymetry, varies significantly between regions, casting doubt about how reliably SDMs can be used to predict into regions where only coarse-resolution data are available. Here, we show that SDMs trained on high-resolution data underestimate Antarctic benthic morphospecies richness by up to 18% when applied to aggregated coarse-resolution environmental data (and up to 50% when using satellite-derived ETOPO bathymetry). Using six systematically degraded versions of high-resolution multibeam bathymetry and annotated seafloor imagery across three Antarctic regions, we evaluate SDM performance both with and without additional environmental variables. High-resolution bathymetry captures terrain complexity most effectively, but we find that the spatial distribution of richness hotspots and the median richness per cell remain consistent, provided models are applied at the same resolution at which they were trained. Our results suggest that while high-resolution bathymetry may enhance local predictions, coarse-resolution data may be more robust for regional-scale predictions, such as those used for Antarctic shelf-wide spatial planning.

Summary Background: Spatial variation in heat-related mortality remains poorly understood, particularly at fine geographical scales. We conducted a nationwide small-area study to examine the association between spatial variation in heat-related mortality and environmental, demographic, health, and socio-economic factors. Methods: We obtained daily all-cause mortality data for people aged [&ge;] 65 years during the summers of 2011-2023 and linked them with municipality-level daily temperature estimates from the ERA5-Land reanalysis dataset. We applied a two-stage Bayesian hierarchical model to estimate small-area heat-related mortality and assess the contribution of community characteristics to spatial variability. Findings: Heat-related mortality showed marked geographical differences, with the highest rates in southern and southeastern Italy. Across municipalities, the relative risk at the 90th temperature percentile, relative to the minimum mortality temperature, ranged from 1.06 to 1.33. The heat-attributable fraction exceeded 6% in several southern municipalities, while excess mortality surpassed 8 deaths per 1,000 inhabitants in parts of the Po Valley, Tuscany, Apulia, and Sicily. National heat-attributable mortality peaked in 2022, with an estimated 17,828 deaths (95% credible intervals: 17,339, 18,285) among older adults. Municipalities with higher average temperatures, less green space, higher obesity prevalence, and more residents aged [&ge;] 85 years had higher heat-related mortality. Educational attainment and employment were among the strongest modifiers of spatial variation. Interpretation: Our findings highlight substantial small-area differences in heat-related mortality across Italy and identify socio-economic deprivation as a key determinant of vulnerability. Heat is likely to disproportionately affect disadvantaged communities, reinforcing the need for adaptation strategies addressing social inequality. Funding: Imperial College Research Fellowship; Italian Ministry of Health PNC (CUP J55I22004450001); NIHR Imperial Biomedical Research Center (BRC NIHR203323).