Spatially consistent microbial biomass and future cellular carbon release from melting Northern Hemisphere glacier surfaces

Earth’s glaciers and ice sheets are losing mass rapidly, increasing meltwater discharge and the delivery of microbes, nutrients, and sediment to downstream environments. Organic carbon (OC) transported by glacial melt impacts microbial productivity, biogeochemical cycles, and ecosystem function, yet the active microbial ecosystems on melting glacier surfaces—particularly within the porous near-surface weathering crust and cryoconite holes—are poorly constrained outside of Greenland. Understanding microbial abundance and export from these supraglacial habitats is essential in the context of 21st-century ‘peak melt,’ as bare-ice areas expand and then contract with glacier retreat. This study asks: what are the typical microbial abundances in supraglacial meltwaters across diverse Northern Hemisphere glaciers, how do hydrological properties relate to these abundances, and what is the projected export of microbial cells and cellular carbon over the coming century?

Prior studies indicate that glacial meltwaters supply bioavailable OC downstream and that microbial communities inhabit a range of cryospheric environments. Reported cell concentrations in other environments often exceed those in supraglacial meltwaters (e.g., ocean surface ~10^5 cells mL^−1, freshwater ~10^6 cells mL^−1, snow ~10^5 cells mL^−1, sea and lake ice ~10^6 cells mL^−1, subglacial waters ~10^5 cells mL^−1). Existing supraglacial assessments suggest 10^3–10^5 cells mL^−1 in Arctic and Alpine settings, with higher abundances reported for surface ice (10^4–10^6 cells mL^−1) and cryoconite hole waters (~10^4 cells mL^−1). Recent work in Greenland’s Dark Zone estimated daily supraglacial microbial carbon release but comparable estimates for other regions were lacking. The weathering crust is recognized as an extensive, hydrologically active, photic habitat connecting supraglacial microbial communities to the drainage network, yet its role in regulating biomass export remains underexplored.

Field campaigns (July 2014–July 2016) sampled 10 sites across Europe, North America, and two western Greenland Ice Sheet sites, focusing on ablation zones during summer melt. Supraglacial meltwater was collected from the weathering crust (depth-integrated auger-hole extractions) and from supraglacial streams. Samples were fixed with paraformaldehyde or glutaraldehyde (2% w/v), stored cold and dark, then frozen. An optimized flow cytometry (FCM) protocol enumerated microbial cells: paired stained (SYBR-Gold 1×) and unstained aliquots were analyzed on a Sony SH-800EC, using gating based on FSC-A/FSC-H (aggregation) and FITC/BSC (microbe identification). Size classes (<1 to >15 µm) were assigned via calibration beads; instrument detection limit was 0.5 µm. Viability discrimination was not possible due to fixation and non-specific staining. In total, 763 weathering crust samples were analyzed for abundance; 142 in-stream samples were likewise analyzed. For 905 records, suspended sediment concentration was estimated via non-stained particle counts in cytograms. Hydrological variables paired with 614 abundance measurements included weathering crust hydraulic conductivity (K) from auger-hole bail-recharge tests with capacitance piezometers, water table depth, water temperature, electrical conductivity (EC), and, at select sites (n>5), stream discharge via salt-dilution gauging or stage-discharge rating. Statistical analyses evaluated relationships between microbial abundance and suspended sediment, K, EC, temperature, and discharge at hemispheric and glacier scales. For century-scale export forecasts, cellular carbon per meltwater volume was derived from the abundance-size distribution using allometric conversions from biovolume to carbon for bacteria and algae. Assumptions included cell length equaling FCM size class median, shape parameters (spherical for <1 µm, rod-shaped with domain-appropriate L:W for larger cells; algae >10 µm with L:W=2.6), and conversion equations M=cV^a with domain-specific constants. Regional upscaling multiplied cellular carbon concentrations by modeled supraglacial runoff. Glacier runoff projections for RGI regions 1–12 (Northern Hemisphere excluding High Mountain Asia) used GloGEM driven by ensembles of CMIP5 GCMs under RCP 2.6/4.5/8.5; Greenland Ice Sheet runoff used the MAR regional model under RCP 4.5/8.5. Extrapolation was limited to regions with comparable glaciological/climatological settings; High Mountain Asia and Southern Hemisphere were excluded. Uncertainties combined the 16th–84th percentile ranges of cell concentrations with modeled runoff uncertainties to propagate low/high export bounds.

• Across 763 weathering crust meltwater samples, mean microbial abundance was 2.2 × 10^4 ± 5.5 × 10^4 cells mL^−1. In-stream meltwaters (n=142) showed 2.2 × 10^4 ± 3.0 × 10^4 cells mL^−1, not significantly different from weathering crust values.
• Abundances are within the same order of magnitude (10^4 cells mL^−1) across all sampled glaciers and both habitat types, aligning with prior reports for Arctic/Alpine supraglacial waters and cryoconite hole waters.
• Regional differences exist: Greenland’s Dark Zone exhibited higher weathering crust concentrations (mean 4.7 × 10^4 cells mL^−1) than several European Alpine and sub-Arctic glaciers (means ~2.9–3.9 × 10^4 cells mL^−1). In-stream contrasts included Fountain Glacier (3.6 × 10^4) vs. Vadrec del Forno (1.5 × 10^4 cells mL^−1).
• Size distribution: modal class 1–2 µm (53.1 ± 8.8% of cells; n=905). Overall, 71.7% ≤2 µm and 86.3% ≤4 µm. Cells >10 µm (likely cyanobacteria/eukaryotic algae) comprised on average 13.5% (~2.9 × 10^3 cells mL^−1).
• Suspended sediment concentration showed the strongest positive correlation with microbial abundance across the dataset, with stronger fits on a glacier-by-glacier basis. Relationships with hydraulic variables (K, discharge) were weak or not significant; EC and temperature relationships were significant but with poor fits (r^2 < 0.04).
• Across five orders of magnitude in K (10^−3 to 10^1 m d^−1) and three orders in Q (10^−3 to 10^0 m^3 s^−1), 97% of samples remained at ~10^4 cells mL^−1, indicating broadly stable supraglacial meltwater abundance.
• Forecasts under RCP 4.5 (next ~80 years) predict mean annual export of 2.9 × 10^22 cells yr^−1, equivalent to 0.65 Mt yr^−1 cellular carbon, from Northern Hemisphere glaciers (RGI 1–12) plus the Greenland Ice Sheet. • Glaciers contribute ~1.8 × 10^22 cells yr^−1 (0.40 Mt C yr^−1); the Greenland Ice Sheet contributes ~1.1 × 10^22 cells yr^−1 (~0.25 Mt C yr^−1). • This cellular carbon equals ~10–15% of Arctic river POC flux (5 ± 1 Mt C yr^−1) and ~0.4% of global biospheric river POC (157 Mt C yr^−1, uncertainty bounds given in cited work).
• Temporal trends: ‘Peak carbon’ export occurs before 2040 (RCP 2.6) and between 2040–2059 (RCP 4.5); no peak within the 21st century under RCP 8.5, with end-of-century exports >2000–2019 levels by ~34%.
• Regional outlooks: Strong declines in mid-latitude/high-altitude glacier regions (Alaska, W Canada/US, Scandinavia, North Asia, Central Europe, Caucasus); increasing or sustained exports from Arctic Canada, Greenland, Svalbard, and the Russian Arctic.
• Source attribution: For glaciers, supraglacial cellular carbon export represents ~56% of prior global glacier POC export estimates (~0.39 of 0.70 Mt C yr^−1), implying surface-derived cells dominate proglacial POC. For the Greenland Ice Sheet, surface cellular carbon export (~0.18 Mt C yr^−1 used for comparison) represents ~20–43% of total proglacial POC (~0.90 Mt C yr^−1), suggesting efficient downstream delivery of supraglacial POC with limited subglacial storage during ablation.

The study addresses key uncertainties about the magnitude and controls of microbial abundance in supraglacial meltwaters and quantifies their projected export as cellular carbon. The broadly consistent abundance of ~10^4 cells mL^−1 across diverse glaciers and hydrological regimes indicates that supraglacial microbial concentrations are relatively stable despite large variations in hydraulic conductivity and discharge. The strongest predictor—suspended sediment concentration—suggests co-mobilization of cells and particulates, potential microbe-sediment associations (e.g., aggregates, particle-attached growth), and nutrient scavenging as mechanisms influencing abundance. Weak relations with K, EC, temperature, and discharge imply that ecological factors (nutrient availability, predation, viral lysis, photodegradation) and glaciological controls (pore structure, retention, filtration, EPS/biofilm adhesion) likely regulate microbial presence and transport within the weathering crust. Upscaled forecasts reveal that supraglacial cellular carbon is a substantial, temporally variable component of POC export from glacierized regions, with ‘peak carbon’ aligning with projected runoff maxima. Regionally, diminishing exports from mid-latitude montane glaciers contrast with increasing contributions from Greenland and high-latitude Arctic regions, portending spatially heterogeneous ecological impacts on downstream aquatic systems. The finding that supraglacial cells constitute the majority of glacier-derived POC (outside major ice sheets) and a sizable fraction for the Greenland Ice Sheet underscores the tight coupling between supraglacial microbial ecosystems and downstream biogeochemistry under ongoing climate warming.

Using a standardized flow cytometry approach across 10 Northern Hemisphere glacier and ice-sheet sites, the study establishes an apparent upper-limit microbial abundance of ~10^4 cells mL^−1 in supraglacial meltwaters, consistent across weathering crust and supraglacial streams. Abundances correlate most strongly with suspended sediment rather than hydraulic properties, indicating that ecological and particle-related processes likely regulate microbial transport. Century-scale projections indicate mean annual exports of 2.9 × 10^22 cells yr^−1 (0.65 Mt C yr^−1) under RCP 4.5, with glaciers contributing ~0.40 Mt C yr^−1 and the Greenland Ice Sheet ~0.25 Mt C yr^−1, and with pronounced regional and temporal variability including ‘peak carbon’ in mid-century under moderate emissions. Supraglacially derived cells are inferred to dominate glacier POC exports and contribute substantially to Greenland Ice Sheet POC, linking surface microbial communities to downstream ecosystem change. Future research should resolve viability and activity fractions, quantify ecological controls (nutrients, biotic interactions), characterize retention and mobility within the weathering crust microstructure, assess seasonal variability, and directly couple supraglacial and proglacial measurements to constrain processing and fate of cellular carbon.

• Methodological constraints: Flow cytometry with fixed samples and SYBR-Gold staining cannot discriminate viable vs. non-viable cells; instrument detection limit of 0.5 µm excludes smaller cells/viruses. Enumeration accuracy is ~84%.
• Assumptions in upscaling: Constant microbial community composition and abundance over ~80 years; equivalence of microbial concentrations in snow and ice melt; all meltwater produced and routed from the surface and efficiently advected to downstream environments; 1 L meltwater = 1 kg. Size and shape assumptions for biovolume-to-carbon conversions may introduce bias.
• Spatial scope: Upscaling limited to RGI regions 1–12 (Northern Hemisphere excluding High Mountain Asia); no extrapolation to High Mountain Asia, Southern Hemisphere, or Antarctic Ice Sheet due to differing surface processes (e.g., debris cover) and lack of observations.
• Temporal coverage: Sampling spans summer ablation seasons over 2014–2016; limited season-long and interannual variability data constrain understanding of seasonal dynamics.
• Hydro-ecological controls unresolved: Weak or inconsistent relationships with hydraulic variables (K, discharge, EC, temperature) leave ecological and glaciological regulators (nutrients, predation, viral lysis, photodegradation, pore structure, retention/filtration, EPS/biofilms) unquantified.
• Export fate: The study assumes efficient downstream delivery of supraglacial cellular carbon; direct paired measurements of surface and proglacial POC are limited, especially outside Greenland.