Floodplain ecosystems play a key role in soil organic carbon (SOC) storage, as they integrate inputs from both vegetation and deposited sediments. Promoting land uses with low anthropogenic disturbances helps maintain the function of these ecosystems as soil carbon (C) sinks. However, the tipping points along a disturbance gradient where land use transitions generate the largest SOC losses or gains remain unclear, though they are key for effective land use management and climate change mitigation. Because flood events can mobilize and enhance the loss of labile C, determining the main origin of stable SOC, whether from plants or soil microorganisms, is also important to identify the optimal combination of land use, vegetation, and soil type for SOC stabilization in floodplains. We examined how SOC stabilization and origin vary in the floodplain of Lake SaintPierre, Quebec, Canada along an anthropogenic disturbance gradient of six land uses: conventional and improved croplands, temporary and permanent meadows, marshes, and forested swamps. In all 6 land uses at the same elevation, we quantified SOC stocks, mineral-associated organic matter - C (MAOM-C), soil delta 13C and sugar biomarkers. Driven by land use change, variations in plant inputs were associated with changes in topsoil SOC storage. Forested swamps had the highest MAOM-C due to greater plant biomass inputs that increased both microbial- and plant-derived C. Across all land uses, microbial-derived sugars, particularly of fungal origin, were more abundant in MAOM than plant-derived sugars. Notably, MAOM remained below its saturation capacity. Soil C pools, including MAOM-C, and both microbial- and plant-derived C, increased nonlinearly with decreasing disturbance, with a tipping point occurring at the transition from temporary to permanent meadows. Our results highlight the importance of increasing both microbial- and plant-derived C inputs to promote SOC stabilization, while emphasizing the persistent key role of fungal metabolism in floodplain soils.

Methanogenic and methane-oxidizing communities (i.e., the microbial communities involved in methane production and consumption) of the tree phyllosphere remain uncharacterized for most tree species despite increasing evidence of their role in regulating tree methane fluxes. Using 16S rRNA gene sequencing, we studied the methanogenic and methane-oxidizing communities of leaves, wood, and bark of five tree species (Acer saccharinum, Fraxinus nigra, Ulmus americana, Salix nigra, and Populus spp.) growing in the floodplain of Lake St-Pierre (Québec) and assessed their relationships with plant traits. Methane-cycling communities differed primarily between tree tissues (leaf, wood, and bark) but also between tree species according to different traits (e.g., leaf, heartwood and bark pH, leaf humidity). Methanogens were prevalent in wood, while methane-oxidizing taxa were found at higher proportions in leaves and bark. Tissue pH was a particularly important trait modulating methane-cycling community composition and the relative abundance of methanogens and methane-oxidizing taxa in the different phyllosphere compartments. Overall, our study shows that methanogens and methane-oxidizing taxa are prevalent in the phyllosphere of several tree species, suggesting a potential widespread role in the regulation of tree methane fluxes. Better understanding these microbial communities and their drivers can help assess their potential contribution to methane-flux regulation.

ABSTRACT Positive biosphere–climate feedbacks are likely to amplify the Arctic warming, yet major uncertainties persist regarding their magnitude and underlying mechanisms. Vegetation shifts, permafrost thaw, and microbial responses critically influence greenhouse gas emissions, surface albedo, and energy balance, ultimately feeding back to climate change. However, the capacity of emerging and changing Arctic ecosystems to sustain biogeochemical functions remains unclear. Here, we identify ten priority research questions that address vegetation, soil, and hydrological processes most relevant to climate feedback. We argue that simply expanding field observations is insufficient. Instead, monitoring must be strategically designed—aligned with satellite observations and models across scales, structured for causal inference, and integrated within coordinated international ecological networks to ensure comparability and synthesis. Advancing next‐generation, policy‐relevant observation systems will support more accurate projections of Arctic feedbacks, improve the ecological realism of Earth system models, and guide policy‐making. Given the disproportionate role of Arctic ecosystems in the global climate system, developing coordinated and ecologically grounded monitoring strategies is a crucial step toward reducing uncertainties in biosphere–climate interactions.

Abstract. The Soil Moisture Active Passive Level-4 Terrestrial Carbon Flux model (hereafter referred to as the L4C model) provides daily estimates of net ecosystem CO2 exchange (NEE), gross primary production (GPP), and ecosystem respiration (ER) at a global scale. The model is based on direct mechanistic forcing–response relationships between CO2 fluxes and energy proxies (absorbed photosynthetically active radiation and temperature) and moisture proxies (soil moisture and vapor pressure deficit). Although the L4C model aims to provide a representative estimation of the CO2 budget of Arctic and Subarctic (AS) environments, a deeper understanding of carbon cycle processes and targeted refinements are needed to improve its accuracy. In this study, alternative model formulations are proposed for the North American AS regions during the growing season. These formulations are calibrated and evaluated using NEE-derived GPP and ER from 20 eddy covariance towers across western Canada and Alaska, covering the period from 2015 to 2022. Refinements in the representation of energy proxies resulted in greater improvements in model performance than adjustments to moisture proxies. Specifically, implementing a light-response curve in GPP estimation reduced unbiased root mean squared error and bias, while incorporating growing degree days improved correlation. Adjustments to rootzone and surface soil moisture in GPP and ER estimation, respectively, did not yield conclusive performance improvements. Vapor pressure deficit showed limited importance as a driver of GPP in upland tundra and wetlands, whereas it had a stronger impact in taiga forests. Finally, the litterfall scheme used to represent SOC dynamics in the L4C ER model formulation in version 8 demonstrated improved performance relative to version 7. These results highlight opportunities to enhance the accuracy of the L4C model for the North American AS growing season but also underscores the need for further research on ER modeling.

Abstract Anticipating the decomposition of permafrost organic matter (OM) following thaw is critical for improving projections of climate‐carbon feedbacks. Stable carbon (C) and nitrogen (N) isotope profiles provide integrated records of the environmental and biogeochemical processes that shaped OM during its accumulation and may therefore improve our understanding of the factors controlling its future vulnerability to decomposition. Here, we compiled isotope data from 122 soil units across 42 Arctic sites and used Bayesian hierarchical models to test whether relationships between stable isotope signatures and soil C/N ratios varied along a gradient of soil organic carbon (SOC) content. Nitrogen isotope fractionation remained consistent across the SOC gradient, with δ 15 N increasing as C/N decreased. In contrast, carbon isotope fractionation differed markedly between mineral and organic soils. In mineral soils, δ 13 C increased with decreasing C/N, indicating extensive microbial processing and recycling of OM during accumulation. In organic soils, δ 13 C decreased with decomposition, consistent with the preferential preservation of relatively 13 C‐depleted plant‐derived compounds under waterlogged conditions. These contrasting isotope patterns reveal distinct OM accumulation pathways in mineral and organic permafrost soils. Our results provide a revised conceptual framework linking long‐term OM accumulation pathways to the properties of permafrost OM exposed by thaw, improving the interpretation of Arctic soil isotope records and their implications for permafrost carbon vulnerability.

Plant litter decomposition governs how much carbon soils store and emit, yet the microbial traits that shape ecosystem-scale decay remain unresolved. Metagenomes can quantify genes encoding plant cell-wall-degrading enzymes, but it is unclear whether ecosystem differences in decay reflect distinct enzymatic repertoires, and whether these data improve prediction beyond climate and soil properties. We paired standardized green and rooibos tea-bag decomposition assays across 3–24 months with 295 soil metagenomes from 264 global sites. Using 196 European plots for primary inference, we built a stage-resolved catalogue of 17.6 million carbohydrate-active enzyme (CAZyme) genes. Forest microbiomes decomposed tea faster than grasslands, but this was not explained by greater CAZyme family richness. Instead, ecosystems differed in CAZyme abundance, subfamily and protein-sequence variation, and allocation across biochemical stages of plant cell-wall decay, with evidence of ecosystem-specific selection. CAZyme profiles added explanatory power for 24-month mass loss and improved within-ecosystem prediction but generalized poorly across ecosystems and continents. By showing that ecosystem differences in decomposition arise from the stage-specific distribution of shared enzymatic functions rather than their presence alone, this work shifts microbial trait inference beyond gene inventories and provides a mechanistic genomic framework for carbon-cycle modelling within defined environmental limits.

An aerobic, Gram-stain-negative bacterium, designated strain NIPR152T, was isolated from a glacier moss ball (glacier mice; Schistidium agassizii) collected from Ellesmere Island in the Canadian High Arctic. Cells were non-motile, rod-shaped, and formed orange-pigmented colonies. The strain grew at 4–30 °C (optimum, 18 °C), at pH 5.5–8.0 (optimum, pH 7.0–7.5), and tolerated up to 1.0

Floodplains represent major sources of methane, a greenhouse gas with significant warming potential, as high soil humidity favors microbial methane production (methanogenesis) over consumption (methanotrophy). Land use also influences methane dynamics by modulating soil properties that regulate methanogenesis and methanotrophy, with forest soils generally showing the greatest methanotrophic potential. However, how land use shapes microbial functions regulating soil methane fluxes remains largely uncharacterized in floodplains. This study aims to better understand the role of forests relative to grasslands and croplands in floodplain methane budgets by assessing soil methane fluxes and microbial functional potential and activity. We used metagenomics and metatranscriptomics to investigate carbon- and methane-cycling gene abundance and expression across land uses, and how these were linked to variations in soil methane fluxes in the Lake St-Pierre floodplain (Quebec). We demonstrated that forests were long-term methane sinks, exhibiting significantly higher uptake rates than croplands, underscoring the importance to protect forests in floodplain ecosystems. Despite greater methane uptake over years, forest soils exhibited a higher potential for anaerobic metabolisms and could switch to net sources at specific times and sites under high water table conditions. The expression of genes regulating acetate conversion, CO2 reduction, and H2 oxidation explained flux variations, indicating that substrate conversion and electron generation are key steps driving methane production and spatiotemporal variation in soil methane fluxes in floodplains. Overall, our study demonstrates how methane cycling varies between floodplain forests and croplands across temporal scales in response to environmental controls, clarifying their respective roles in climate regulation.

La nature en ville offre de multiples bienfaits tant sur le plan écologique que sociologique, en favorisant la santé physique et mentale. De nombreuses initiatives municipales à travers le monde se penchent ainsi sur l’identification et le maintien de ces services écosystémiques, à l’heure où ces derniers risquent d’être durement touchés par les changements climatiques. Pourtant, peu de recherches se sont concentrées sur les facteurs régulant la combinaison des services rendus par la nature urbaine et des risques auxquels elle est exposée. Notre étude vise donc à quantifier 6 services (perméabilité des sols, régulation de la température de surface, séquestration/stockage de carbone, biodiversité, interactions humaines avec les écosystèmes et cueillette de champignons comestibles) et 5 risques (mortalité des arbres, maladies/dépérissement, dépôts de neige, vulnérabilité aux intempéries et espèces exotiques envahissantes) pour 12 écosystèmes du campus de l’Université du Québec à Trois-Rivières. Les résultats révèlent que les services et les risques varient selon le type d’écosystème. La jeune pinède blanche et les milieux mixtes surannés fournissent les services les plus importants, tandis que la pinède grise et le milieu mixte sont les plus exposés aux risques. Plus les écosystèmes sont âgés et éloignés des perturbations anthropiques, plus ils rendent un grand nombre de services, sans augmenter le niveau moyen des risques. Cette étude fournit une méthodologie facilement ajustable pour évaluer la dynamique spatiale des services offerts et des risques encourus par la nature urbaine et, de ce fait, formuler des recommandations aux gestionnaires.

Growing evidence has shown that, apart from local environmental factors, changes in landscape-level factors by accelerated land-use change can also shape soil pathogenic fungal diversity. However, the global representativeness of such patterns remains unclear. Here, we assess how pathogenic fungal diversity in 511 soil samples worldwide responds to landscape factors, including landscape complexity index based on eight landscape metrics and quantity of different land cover types across six spatial scales (i.e., surrounding landscape, 250 m to 10,000 m radii from the sampling coordinate). We find that while soil variables explain over half of the variance, pathogenic fungal alpha diversity increases with landscape complexity and crop cover proportion, but decreases with grass and tree cover proportion, together explaining 23.4% of the total variance. Landscape factors have weaker impacts on beta diversity, explaining 13.0% of the variance. Across spatial scales, grassland ecosystems exhibit increasingly stronger responses to landscape variables compared to forest ecosystems. Landscape factors have a higher relative contribution to root-associated fungi than leaf/fruit/seed-associated fungi. Our results emphasize the importance of local factors and the complementary role of landscape patterns in shaping global soil pathogenic fungal distributions, highlighting scale-dependent effects across ecosystems and fungal functional groups.

The Arctic is warming faster than anywhere else on Earth, placing tundra ecosystems at the forefront of global climate change. Plant biomass is a fundamental ecosystem attribute that is sensitive to changes in climate, closely tied to ecological function, and crucial for constraining ecosystem carbon dynamics. However, the amount, functional composition, and distribution of plant biomass are only coarsely quantified across the Arctic. Therefore, we developed the first moderate resolution (30 m) maps of live aboveground plant biomass (g m(-2)) and woody plant dominance (%) for the Arctic tundra biome, including the mountainous Oro Arctic. We modeled biomass for the year 2020 using a new synthesis dataset of field biomass harvest measurements, Landsat satellite seasonal synthetic composites, ancillary geospatial data, and machine learning models. Additionally, we quantified pixel-wise uncertainty in biomass predictions using Monte Carlo simulations and validated the models using a robust, spatially blocked and nested cross-validation procedure. Observed plant and woody plant biomass values ranged from 0 to similar to 6000 g m(-2) (mean approximate to 350 g m(-2)), while predicted values ranged from 0 to similar to 4000 g m(-2) (mean approximate to 275 g m(-2)), resulting in model validation root-mean-squared-error (RMSE) approximate to 400 g m(-2) and R-2 approximate to 0.6. Our maps not only capture large-scale patterns of plant biomass and woody plant dominance across the Arctic that are linked to climatic variation (e.g., thawing degree days), but also illustrate how fine-scale patterns are shaped by local surface hydrology, topography, and past disturbance. By providing data on plant biomass across Arctic tundra ecosystems at the highest resolution to date, our maps can significantly advance research and inform decision-making on topics ranging from Arctic vegetation monitoring and wildlife conservation to carbon accounting and land surface modeling.

Freshwater ecosystems are highly biodiverse and provide essential services that support both ecosystem health and economic sustainability. Despite their ecological significance, these ecosystems are disproportionately affected by the global biodiversity crisis. Large river floodplains constitute a fundamental component of freshwater ecosystems, sustaining fish biodiversity, growth, and reproduction. Yet, these floodplains face mounting threats from anthropogenic pressures, including physical modifications and land conversion for agriculture. In this context, there is an urgent need for scalable biomonitoring methods to more effectively assess floodplain ecosystems, which present methodological challenges due to their heterogeneous and dynamic nature. Traditional fish monitoring methods, however, are often invasive, costly, and resource-intensive. In contrast, environmental DNA (eDNA) metabarcoding presents a noninvasive, cost-effective, and scalable alternative. This study compares eDNA metabarcoding and electrofishing for fish community biomonitoring in the floodplain of Lake St. Pierre, the largest floodplain habitat along the St. Lawrence River. We assessed the effectiveness of these methods in monitoring fish community diversity and composition, as well as the influence of floodplain sectors and a gradient of land use from natural wetlands to annual (row) crops. eDNA metabarcoding detected a broader range of species than electrofishing, while both methods consistently identified abundant species. The two methods yielded uncorrelated diversity indices and distinct community compositions. Fish eDNA community composition was strongly associated with floodplain sectors, whereas land use within these sectors had a weaker influence on community diversity and composition. Our findings highlight eDNA metabarcoding as a valuable tool for characterizing broad patterns of fish communities in floodplain ecosystems. This method provides an additional tool to traditional methods for monitoring and conserving threatened floodplain habitats. However, careful consideration of study scale is essential to ensure effective conservation outcomes in these hydrologically dynamic environments.

Peatlands are globally important carbon stores that face increasing threats from human activities and climate change impacts. Comprehensive peatland data are essential for understanding ecosystem responses to these stressors and mapping their past and current characteristics. Current peatland datasets remain limited due to poor representation in global soil mapping initiatives and the absence of a recognized, coordinated central repository for peat depth data. Existing compilations often contain errors, duplicates, and outdated observations, requiring researchers to repeatedly gather and harmonize data on a study-by-study basis. To address these challenges, we present Peat-DBase version 1.0 – a harmonized, quality-controlled global compilation of basal peat depth measurements. Version 1.0 of Peat-DBase comprises 204 902 peat depth measurements from 29 sources spanning 54.933° S to 82.217° N, with a significant proportion of measurements in Atlantic Canada and Scotland due to the inclusion of two particularly large datasets focused on those regions. We supplement the peat study measurements with 94 615 non-peat soil measurements to ensure comprehensive coverage consistent with the relatively low spatial coverage of peatlands globally. Despite the uneven distribution of peat depth measurements, Peat-DBase contains reasonable coverage of the major global peatland complexes in temperate and boreal North America and Europe, portions of Russia, the Amazon and Congo basins, and the Malay Archipelago, though gaps remain in the lower Amazon Basin, Eastern Indonesia, and Eastern Russia. From the current data, peat depths have a median value of 130 cm (IQR: 60–240), although this is influenced by a predominance of measurements in the North Atlantic regions. Peat-DBase's deepest measurement is 2223 cm. While sampling biases and measurement uncertainties exist, Peat-DBase provides an essential foundation for global peatland research. Peat-DBase is under active development and future versions will incorporate additional datasets, information on current peatland status, and improved positional uncertainty quantification. Peat-DBase eliminates the need for overlapping data compilation efforts while identifying critical observational gaps for future research. Peat-DBase is available at https://doi.org/10.5281/zenodo.15530644 (Skye et al., 2025).

The soil microbiome plays a key role in tree growth and nutrient cycling. Although hybrid poplar clones are phylogenetically related, their influence on the soil microbiome may differ due to phenotypic differences, especially their root mean diameter or mass density (RMD). This influence remains relatively unexplored, limiting our understanding of how soil microbial communities contribute to the growth strategies of fast-growing trees. Our objective was to determine how phylogenetically related trees and their root traits influence soil microbial diversity and composition by studying five hybrid poplar clones growing in a plantation located in New Liskeard, Ontario, Canada. We collected soil cores at depths of 0-20 cm (topsoil) and 20-40 cm (subsoil) and analyzed fine root traits and soil bacterial and fungal communities. We found that phylogenetically related hybrid poplars influenced the soil microbiome by shaping the composition of microbial communities. Differences between clones were evident in the relative abundance of soil ectomycorrhizal fungi and Actinobacteriota in both topsoil and subsoil. The increase in relative abundance of ectomycorrhizal fungi was driven by high RMD and root length density (RLD), i.e., high fine root mass and length per unit soil volume. Fine roots with high RLD led to a higher relative abundance of Actinobacteriota in the topsoil, while recalcitrant fine roots (high lignin/nitrogen ratio) promoted their abundance in the subsoil. Thus, root traits are key factors in determining the effects of trees on soil microbiome. The soil microbiome, together with root traits, could be an important aspect to consider in tree selection.

Despite the increasing number of studies investigating tree methane fluxes, the relationships between tree methane fluxes and species traits remain mostly unexplored. We measured leaf and stem methane fluxes of five tree species (Acer saccharinum, Fraxinus nigra, Ulmus americana, Salix nigra, and Populus spp.) in the floodplain of Lake St-Pierre (Québec) and examined how these fluxes vary with species traits (wood density, humidity, pH; leaf water content, pH, stomatal conductance; methanogen and methanotroph relative abundances (RAs) in leaf, wood, and bark). Tree methane fluxes differed among tree species according to traits linked to the transport of soil-produced methane and chemical conditions associated with the regulation of methane-cycling microorganisms. Tree fluxes were correlated positively with heartwood and leaf pH and negatively with their humidity. Stem emissions were positively correlated with methanogen RA in heartwood, and leaf emissions were negatively correlated with the RA of leaf epiphytic methanotrophs, suggesting a contribution of tree microbiota in the regulation of methane fluxes. We demonstrated for the first time that tissue pH may be a particularly important trait influencing tree methane fluxes via the regulation of microbial mechanisms. Species with low tissue pH show potential for methane release mitigation.

Floodplain ecosystems play a key role in soil organic carbon (SOC) storage, as they integrate inputs from both vegetation and sediments. Promoting land uses with low anthropogenic disturbances helps maintain the function of these ecosystems as soil carbon (C) sinks. However, the tipping points along a disturbance gradient where land use transitions generate the largest SOC losses or gains remain unclear, though they are key for effective land use management and climate change mitigation. Because flood events can negatively impact labile C, determining the main origin of stable SOC, whether from plants or soil microorganisms, is also important to identify the optimal combination of land use, vegetation, and soil type for SOC stabilization in floodplains. We examined how SOC storage and stabilization vary in the floodplain of Lake Saint-Pierre, Quebec, Canada along an anthropogenic disturbance gradient of six land uses: conventional and improved croplands, temporary and permanent meadows, marshes, and forested swamps. In all 6 land uses, we quantified SOC stocks, mineral-associated organic matter - C (MAOM-C), particulate organic matter - C (POM-C), soil δ13C and sugars. Land use effects on SOC storage and stabilization were most pronounced in the topsoil (0-10 cm), with forested swamps showing the highest vegetation biomass, SOC, and MAOM-C. Microbe-derived inputs were more abundant in MAOM, whereas plant-derived C dominated POM. SOC gains increased with decreasing disturbance, with a tipping point occurring in the transition from temporary to permanent meadows. Our results highlight the importance of conserving permanent and low-disturbance land uses to promote SOC persistence in floodplain ecosystems and emphasize the central role of microbial metabolism in stabilization processes. ### Competing Interest Statement The authors have declared no competing interest. The QuM-CM-)bec Ministries (Agriculture, Fisheries and Food; Environment, Fight against Climate Change, Fauna and Parks) and the Natural Sciences and Engineering Research Council of Canada (Alliance program), through two multidisciplinary research initiatives: le pM-CM-4le dM-BM-^Rexpertise multidisciplinaire en gestion durable du littoral du Lac Saint-Pierre (2019-2023) and Carbon cycling in QuM-CM-)becM-BM-^Rs wetlands (Carbonique, 2024-2029).

This dataset is for paper "Leaf nitrogen from the perspective of optimal plant function"

Earth harbours an extraordinary plant phenotypic diversity(1) that is at risk from ongoing global changes(2,3). However, it remains unknown how increasing aridity and livestock grazing pressure-two major drivers of global change(4-6)-shape the trait covariation that underlies plant phenotypic diversity(1,7). Here we assessed how covariation among 20 chemical and morphological traits responds to aridity and grazing pressure within global drylands. Our analysis involved 133,769 trait measurements spanning 1,347 observations of 301 perennial plant species surveyed across 326 plots from 6 continents. Crossing an aridity threshold of approximately 0.7 (close to the transition between semi-arid and arid zones) led to an unexpected 88% increase in trait diversity. This threshold appeared in the presence of grazers, and moved toward lower aridity levels with increasing grazing pressure. Moreover, 57% of observed trait diversity occurred only in the most arid and grazed drylands, highlighting the phenotypic uniqueness of these extreme environments. Our work indicates that drylands act as a global reservoir of plant phenotypic diversity and challenge the pervasive view that harsh environmental conditions reduce plant trait diversity(8-10). They also highlight that many alternative strategies may enable plants to cope with increases in environmental stress induced by climate change and land-use intensification.

Plant biomass is a fundamental ecosystem attribute that is sensitive to rapid climatic changes occurring in the Arctic. Nevertheless, measuring plant biomass in the Arctic is logistically challenging and resource intensive. Lack of accessible field data hinders efforts to understand the amount, composition, distribution, and changes in plant biomass in these northern ecosystems. Here, we present The Arctic plant aboveground biomass synthesis dataset , which includes field measurements of lichen, bryophyte, herb, shrub, and/or tree aboveground biomass (g m −2 ) on 2,327 sample plots from 636 field sites in seven countries. We created the synthesis dataset by assembling and harmonizing 32 individual datasets. Aboveground biomass was primarily quantified by harvesting sample plots during mid- to late-summer, though tree and often tall shrub biomass were quantified using surveys and allometric models. Each biomass measurement is associated with metadata including sample date, location, method, data source, and other information. This unique dataset can be leveraged to monitor, map, and model plant biomass across the rapidly warming Arctic.

Aim: Arctic plants survived the Pleistocene glaciations in unglaciated refugia. The number, ages, and locations of these refugia are often unclear. We use high-resolution genomic data from present-day and Little-Ice-Age populations of Arctic Bell-Heather to re-evaluate the biogeography of this species and determine whether it had multiple independent refugia or a single refugium in Beringia. Location: Circumpolar Arctic and Coastal British Columbia (BC) alpine. Taxon: Cassiope tetragona L., subspecies saximontana and tetragona, outgroup C. mertensiana (Ericaceae). Methods: We built genotyping-by-sequencing (GBS) libraries using Cassiope tetragona tissue from 36 Arctic locations, including two similar to 250- to 500-year-old populations collected under glacial ice on Ellesmere Island, Canada. We assembled a de novo GBS reference to call variants. Population structure, genetic diversity and demography were inferred from PCA, ADMIXTURE, fastsimcoal2, SplitsTree, and several population genomics statistics. Results: Population structure analyses identified 4-5 clusters that align with geographic locations. Nucleotide diversity was highest in Beringia and decreased eastwards across Canada. Demographic coalescent analyses dated the following splits with Alaska: BC subspecies saximontana (5 mya), Russia (similar to 1.4 mya), Europe (>200-600 kya), and Greenland (similar to 60 kya). Northern Canada populations appear to have formed during the current interglacial (7-9 kya). Admixture analyses show genetic variants from Alaska appear more frequently in present-day than historic plants on Ellesmere Island. Conclusions: Population and demographic analyses support BC, Alaska, Russia, Europe and Greenland as all having had independent Pleistocene refugia. Northern Canadian populations appear to be founded during the current interglacial with genetic contributions from Alaska, Europe and Greenland. We found evidence, on Ellesmere Island, for continued recent gene flow in the last 250-500 years. These results suggest that a re-analysis of other Arctic species with shallow population structure using higher resolution genomic markers and demographic analyses may help reveal deeper structure and other circumpolar glacial refugia.