Single-cell imaging reveals efficient nutrient uptake and growth of microalgae that darken the Greenland Ice Sheet

doi:10.21203/rs.3.rs-5199834/v1

Blooms of dark pigmented microalgae accelerate glacier and ice sheet melting by reducing the surface albedo. However, the role of nutrient availability in regulating their growth remains poorly understood. We studied glacier ice algae on the Greenland Ice Sheet, providing the first single-cell based data on their carbon:nitrogen:phosphorus (C:N:P) ratios and assimilation of dissolved inorganic carbon (DIC) and DIN following various nutrient amendments. The single-cell dataset revealed high C:N and C:P atomic ratios in the algal biomass and the presence of intracellular P storage. Assimilation of DIC by the algae was not enhanced by ammonium, nitrate, or phosphate addition. Our combined results demonstrate that glacier ice algae can optimise nutrient uptake, facilitating the potential colonization of ablating ice sheet surfaces without the need for additional nutrient inputs. This adaptive strategy becomes particularly significant as climate warming accelerates the expansion of melt areas on the Greenland Ice Sheet.

Earth and environmental sciences/Biogeochemistry

Biological sciences/Ecology/Biogeochemistry

Biological sciences/Microbiology/Environmental microbiology/Water microbiology

Biological sciences/Ecology/Stable isotope analysis

Biological sciences/Ecology/Freshwater ecology

glacier ice algae

Greenland Ice Sheet

nutrient limitation

productivity

carbon assimilation

nitrogen assimilation

stoichiometry

HR-SIMS

The ablation zones of glacier and ice sheet surfaces are hotspots for microbial life[^1–5]. Dark-pigmented glacier ice algae (Ancylonema spp.) are the main primary producers on bare ice surfaces[^4,6]. They form extensive blooms during the summer melt season[^7–9], lowering the ice surface albedo and accelerating ice melt as a consequence[^8–12]. Algal blooms on the western margin of the Greenland Ice Sheet have been shown to contribute, on average, 10 to 13% to the surface ice melt[⁹]. The Greenland Ice Sheet has become the single largest contributor to global barystatic sea-level rise[^13–16], and its ablation zones are expanding due to climate warming[¹⁷], exposing more of the ice surface and expanding the potential habitat area for glacier ice algae. However, we still do not know the triggers and controls of algal bloom progression throughout the summer ablation season and the causes of inter-annual variations in bloom extent[^6,8,18–20]. Understanding the mechanisms of algal bloom formation is critical to predicting the progression of blooms on melting bare ice surfaces[^6,19,21,22], and the contribution of the melting of the Greenland Ice Sheet to global sea level rise.

The highly oligotrophic conditions of the ice sheet’s ablation zone may limit the growth and expansion of algal blooms on ice surfaces. Supraglacial environments are characterized by low concentrations of inorganic dissolved macro-nutrients with NH₄⁺ and NO₃^-concentrations typically below 1 µM, and aqueous PO₄^3-below 0.1 µM[^23–25]. Besides algae, microbial communities on glacier surfaces often include protists such as ciliates and dinoflagellates, along with fungi, bacteria, and archaea[^26–31], which compete for and drive the cycling of nutrients. McCutcheon et al.[²⁴] found that phosphate can limit algal productivity and highlighted the positive association between phosphorus-bearing minerals, such as hydroxylapatite, and the accumulation of algal biomass. A potential relationship between minerals and algal growth has also been highlighted by Stibal et al.[³²], who documented a positive correlation between dust loading and algal abundance.

The glacier ice algae inhabiting bare ice surfaces are likely to have evolved strategies to partially compensate for the low macro-nutrient concentrations typical of their habitats. These strategies may include the maintenance of a high carbon:nitrogen:phosphorus (C:N:P) biomass ratio, a high nutrient uptake efficiency and/or intracellular nutrient storage capabilities. However, the nutrient content and uptake rates for cryophilic algae and, in particular, glacier ice algae remain unknown, hampering our mechanistic understanding of potential nutrient limitations on algal growth. Previous analysis have focussed solely on bulk analyses of particulate organic matter (POM) glacier ice algal-colonised ice[^24,33]. The data revealed a wide range of C:N:P ratios, spanning from 690:48:1 to 2615:196:1. The measured bulk C:P[^24,33], and sometimes also bulk C:N ratios[³³] in both glacier ice algae and snow algae dominated POM samples from across the Arctic[³⁴], were much higher than the Redfield C:N:P ratio of 116:16:1 commonly observed in POM of marine ecosystems[³⁵], suggesting either a relatively low macro-nutrient requirement for glacier supraglacial algae or potential limitations in P and N. However, the bulk C:N:P ratios of POM collected from ice surface samples are unlikely to accurately reflect glacier ice algae biomass stoichiometries due to differences in contributions from atmospheric deposition-derived organic matter, necromass, extrapolymeric substances, and other microorganisms (e.g. bacteria, other eukaryotic algae, fungi) in the POM filter fraction. Hence, bulk C:N:P ratios derived from POM of surface ice will invariably span a large range. Only single cell-specific measurements can accurately determine the C:N:P ratios of glacier ice algae, and single cell-specific activity and nutrient uptake assessments help constrain potential kinetic limitations imposed by nutrient availability on algal growth. Thus, targeted and cell-specific measurements are crucial for comprehending the nutrient demands driving algal bloom progression and their future growth dynamics in glacier ecosystems.

In this study, we quantified the C:N:P ratio, dissolved inorganic carbon (DIC) assimilation, NO₃^-and NH₄⁺ assimilation, and growth rates of single glacier ice algal cells (Ancylonema spp.) on the Greenland Ice Sheet under both unamended and nutrient-amendment conditions. Our aim was to gain insights into the physiological responses of this key species to varying levels of nutrient availability. In addition to single-cell analyses, we also used bulk stable isotope biogeochemical rate measurements to quantify the C and N turnover and elemental composition of the microbial community on the Greenland Ice Sheet, which we further characterised by 16S and 18S rRNA gene amplicon sequencing. Together, our findings demonstrate that glacier ice algae are well adapted to the oligotrophic conditions of the Greenland ice sheet, and exhibit no significant productivity response to external nutrient additions. This suggests that as melting exposes new ice surfaces, they can be readily colonized without nutrient limitations hindering algal growth. Our study refines the understanding of how nutrient availability influences glacier ice algal bloom development and highlights the yields insights into the role of algal cells in primary production and nutrient cycling on glaciers and ice sheets.

Physico-chemical conditions on the Greenland Ice Sheet

Dark snow-free surface ice with visibly high concentrations of particulates and algal biomass was sampled from the southern tip of the Greenland Ice Sheet (Figure 1a) to assess the in situ microbial community composition, bulk and single-cell elemental ratios, and to perform incubations (Fig. 1b) to assess bulk and single-cell C-fixation and inorganic N-assimilation. The ice samples were allowed to melt for ~36 hrs at an ambient air temperature of ~4 °C under in situ light conditions, with 18 hrs of daylight. The initial dissolved inorganic nutrient concentrations in this ice melt were 0.08 µM for NH₄⁺, 0.05 µM for NO₃^- and ~0.01 µM for PO₄^3- (> LOD but <LOQ of 0.02 µM) (Table 1). Dissolved organic nitrogen and phosphorous (DON and DOP), were present at concentrations ~5 and ~7 times higher than the inorganic nutrients.

Table 1. Nutrient and base cation concentrations, along with the algal community composition, in the initial surface glacier melt sample from the Greenland Ice Sheet (prior to incubations). Values are reported as means, with standard deviations (SD) where sample size was >1. Sample sizes are indicated (n).

Parameter		Value		SD	n
Nutrient concentrations in situ	NH₄⁺	0.08			1
	NO₃^-	0.05			1
	NO₂^-	<0.01			1
	PO₄^3-	~0.01 (<LOQ)			1
	DON	0.65			1
	DOP	0.1			1
	Ca²⁺	0.66			1
	Mg²⁺	0.36			1
	Na⁺	1.54			1
	K⁺	0.22			1

Community composition in situ	A. nordenskiöldii filaments (filaments ml^-1)	3,880	±	150	2
	A. nordenskiöldii chain length (number of cells)	~3 (1-18)			2
	A. nordenskiöldii abundance (cells ml^-1)	10,700	±	260	2
	A. alaskanum abundance (cells ml^-1)	5,580	±	1,520	2
	Abundance snow algae (cells ml^-1)	2,110	±	210	2

Supraglacial community composition

The bare ice community sampled for the incubation experiments had a mean glacier ice algal (phylum Phragmoplastophyta) abundance of 16.2 ± 1.2 x 10^3 cells ml^-1 based on microscopic analyses, of which ~66% were filamentous Ancylonema cf. nordenskiöldii and ~34% were unicellular A. cf. alaskanum (Table 1). Red-coloured snow algal cysts (phylum Chlorophyta), Chlamydomonas spp., had an abundance of 2.1 ± 0.2 x 10^3 cells ml^-1, and represented 12% of the total eukaryotic algal cells. In total, 76 amplicon sequence variants (ASVs) were found within the 18S rRNA gene amplicon data. This data confirmed the dominance of glacier ice algae among the eukaryotic community. The phylum Phragmoplastophyta had the highest relative abundance (70%; Supplementary Figure 1a) among the eukaryotes, and was comprised solely of Mesotaeniaceae, while fungi contributed ~ 19% of eukaryotes, with the phyla Basidiomycota, Ascomycota, and Chytridiomycota (Supplementary Figure 1a) dominating. Microscopic observations confirmed the presence of pigmented glacier ice algae with parasitic infections by Chytridiomycota (Supplementary Figure 2). Other eukaryotic phyla found at lower relative abundances were Chlorophyta (9%), Cercozoa (1%), and Ciliophora (<1%). The bacterial community (determined by 16S rRNA gene amplicon sequencing) was dominated by Bacteroidia (37%), followed by Actinobacteria (26%), and Alphaproteobacteria (14%). Cyanobacteria represented 4% of the bacterial community. Overall, 71 bacterial ASVs were found.

In situ glacier ice algae C:N:P ratios

The elemental mapping of single glacier ice algal cells in the fresh ice melt yielded a mean in situ C:N biomass ratio of 19 ± 2.9 and a C:P ratio of 509 ± 149, exceeding the Redfield C:N (6.6) and C:P ratio (116) four- and three-fold, respectively[³⁵] (Table 2). The mean glacier algal N:P ratio was 26 ± 5, higher than the Redfield N:P stoichiometry of 16. Overall, we observed a high variability in cellular C:P, N:P, and C:N stoichiometries. Notably, the elemental mapping revealed the presence of small (<1 µm) P-rich inclusions inside individual glacier ice algae cells (Figure 2, white arrows in P elemental map).

C-fixation, N-assimilation, and the effect of nutrient additions in bulk samples and single algal cells

Stable isotope incubations were performed under in situ conditions on the ice surface (Fig 1b) to quantify the activity (based on DIC and DIN uptake) of both single glacier ice algae cells (Figure 3; Figure 4d,e; Table 2) and bulk microbial community (Figure 4a,b; Table 2). All incubations received ¹³C-DIC to assess photoautotrophic C fixation, either with no nutrient amendment (control), or addition of ¹⁵N-NH₄⁺, ¹⁵N-NO₃^-, PO₄^3- or combined ¹⁵N-NH₄⁺+PO₄^3-. Following isotope amendments, samples were taken after 6 (T1) and 30 hours (T2) of incubation. The incorporation rate of the C or N isotopes was calculated based on the change in isotopic composition in the biomass of algal cells and bulk POM over the incubation period. The DIC concentrations measured at T1 from the control bottle was 275 µM, which is far higher than 44-70 µM Andrews et al.^[36] or 15 µM reported in Yallop et al.[¹²]. The latter is closer to the value expected in dilute glacier ice melt in equilibrium with the atmosphere (Supplementary Note 1). The high value measured at T1 might suggest that net heterotrophic activity and/or photooxidation occurred during the incubations. The high initial DIC concentrations of our experiments were likely impacted by the process of ice melting, but these values mitigate potential DIC limitation during the incubations.

Bulk and single cell isotope incorporation measurements were generally obtained for T1 and T2 of the incubation period, except for single cell analyses, where T1 measurements were only performed for the control and ¹⁵N-NH₄⁺ treatments. In total, 244 glacier algal cells were analysed using high-resolution secondary ion mass spectrometry (HR-SIMS), with 24 of the 244 imaged cells (~10%) not exhibiting any DIC fixation (Figure 4d,e). The growth and assimilation data of algal cells described here and in Table 2 are based solely on the active fraction of the population (those showing DICassimilation), while a comprehensive overview of the HR-SIMS data, including inactive cells, is provided in Supplementary Table 1.

Nutrient additions did not stimulate bulk or single-cell C-based growth rates (Figure 4). Rather, the ¹⁵N-NH₄⁺, ¹⁵N-NO₃^-, and combined ¹⁵N-NH₄⁺+PO₄^3- additions resulted in a significantly decreased bulk C-based growth rate (p=0.02, df = 4, with growth rates of 0.36 ± 0.03, 0.29 ± 0.05, 0.24 ± 0.04 day^-1, respectively) compared to the control (which received only ¹³C-DIC; 0.63 ± 0.03 day^-1; Figure 4a). Similarly, the mean C-based growth of single glacier ice algal cells also did not show any stimulation upon nutrient addition, but their C-based growth was significantly decreased in the PO₄^3- and the ¹⁵N-NH₄⁺+PO₄^3- treatments (0.20 ± 0.11, 0.23 ± 0.16 day^-1, respectively; p=1.56e^-9, df = 4) compared to the control (0.47 ± 0.24 day^-1) (Figure 4d). The mean C-based growth rates for single glacier ice algal cells in the control treatment, with values of 0.65 ± 0.36 at T1 and 0.47 ± 0.24 day^-1at T2, correspond to mean C-based doubling times of ~2 days (Figure 4d; Table 2).

The bulk and single-cell N-based growth rates were all higher at T1 compared to T2, which is consistent with the rapid depletion of the N-tracers in solution: <37% of NO₃^- in the ¹⁵N-NO₃^- treatment and <3% of the NH₄⁺in the ¹⁵N-NH₄⁺treatment remained after the first 6 hrs of incubation (T1) (Supplementary Figure 4a,b). There was no statistically significant difference in N-based growth rates between the ¹⁵N-NH₄⁺,¹⁵N-NO₃^-, and ¹⁵N-NH₄⁺+PO₄^3- additions in both bulk and single-cell measurements when compared at the same timepoint (Figure 4b,e). The bulk N-based growth rates for both NO₃^- andNH₄⁺were 0.60 day^-1at T1, with assimilations of 44.8 and 32.4 µmol N L^-1 day^-1, respectively (difference in values is likely due to inhomogeneous biomass distribution; assimilation normalised to biomass: 28.5 and 28.6 µmol N mg^-1 N_POMday^-1, respectively) (Figure 4b, Table 2). The bulk N-based growth rates were only 0.2 day^-1 for T2 for both NO₃^- and NH₄⁺. The N-based growth rates of the ¹⁵N-NH₄⁺and ¹⁵N-NO₃^- treatments were similar also for single glacier ice algal cells: both 0.07 day^-1at T2 (Figure 4e, Table 2). The bulk POC:PON ratio in the nitrogen-spiked treatments decreased between T0 and T1 compared to the control and consistently remained lower, in line with the rapid depletion of the nitrogen tracers in solution, indicating that the observed nitrogen depletion was due to biological uptake (Figure 4c).

Glacier ice algal cells assimilated Cin excess of N and generally in excess of the Redfield C:N ratio (Figure 5). We found that the C-fixation of the single algal cells continued at high rates, despite the measurable N-assimilation slowing down (due to the depletion of the ¹⁵N-tracers) at later time points, with mean C:N assimilation ratios of ~46 at T₁ and ~84-113 at T₂. There was no significant correlation between cell volume and C-based growth rates (Supplementary Figure 5). Overall, we observed a high variability in C-fixation and N-assimilation rates between glacier ice algal cells (Figure 4d,e and 5).

Contribution of glacier ice algae to bulk C and N uptake

Using the single-cell rate measurements, along with data on the abundance and bulk uptake of C or N uptake from DIC or NH₄⁺, we estimated the glacier ice algae contribution to the total C and N uptake in bulk POM. These calculations are sensitive to variability in biomass distribution among the incubation bottles, differences in single-cell activity rates, and variability in algal cell abundance between bottles and timepoints (Supplementary Note 2). The bulk POM measurements represent all particulate matter retained on filters (3 and 0.2 µm), meaning they include not only glacier ice algae but also other organisms, such as dispersed cryoconite material containing cyanobacteria and organic matter (SEM images in Supplementary Figure 6). Glacier ice algal assimilation accounted for approximately 7±6% to 15±12%, of the ¹³C from DIC recovered in POM, while glacier ice algae accounted between ~3±2% to 8±6% of the ¹⁵N from NH₄⁺ recovered in POM (Supplementary Tables 2 and 3).

Table 2. Bulk and single-cell elemental compositions and activity rates of a surface ice community on the Greenland Ice Sheet. Data are shown as means with standard deviations (SD). The number of replicates (n) for the bulk parameter correspond to the analysed replicate samples and for the single-cell data to the number of analysed cells (n). Single-cell assimilation and growth data represent estimates of the active fraction of the algal population. Ranges in ¹³C-DIC,¹⁵N-NO₃^- and¹⁵N-NH₄⁺assimilation rates correspond to different incubation lengths (T1 or T2). For single cell measurements of the entire population (active and inactive cells), see Supplementary Table 1.

Parameter		Mean		SD	n
Bulk particulate organic C and N contents in situ	POC (µmol C L^-1)	2,849_(T0)	±	1,015_(T0)	3
	PON (µmol N L^-1)	142_(T0)	±	56_(T0)	3
Bulk particulate C and N uptake and growth	C assimilation (µmol C L^-1 day^-1) ^a	448_(T2)-1,879_(T1)	±	101_(T2)	4
	C assimilation (µmol C mg^-1 C_POM day^-1)^a	28.2_(T2)-33.9_(T1)	±	1.68_(T2)	4
	C-based growth rate (day^-1)^a	0.62_(T1)-0.63_(T2)	±	0.03_(T2)	4
	C-based doubling time (days) ^a	1.6_(T1)-1.63_(T2)	±	0.09_(T2)	4
	NH₄⁺assimilation (µmol N L^-1 day^-1)^b	9.07_(T2)-32.4_(T1)	±	1.03_(T2)	4
	NO₃^-assimilation (µmol N L^-1 day^-1)^b	9.59_(T2)-44.8_(T1)	±	2.40_(T2)	4
	NH₄⁺assimilation (µmol N mg^-1 N_POMday^-1)^b,d	8.22_(T2)-28.6_(T1)	±	1.41_(T2)	4
	NO₃^-assimilation (µmol N mg^-1 N_POMday^-1) ^b	8.0_(T2)-28.5_(T1)	±	1.35_(T2)	4
	N-based growth rate (NH₄⁺)(day^-1)^b	0.18_(T2)-0.60_(T1)	±	0.03_(T2)	4
	N-based growth rate (NO₃^-)(day^-1)^b	0.17_(T2)-0.61_(T1)	±	0.03_(T2)	4

Single cell elemental composition in situ	C:N atomic ratio	19_(T0)	±	2.9_(T0)	48
	C:P atomic ratio	509_(T0)	±	149_(T0)	48
	N:P atomic ratio	26_(T0)	±	5_(T0)	48
Single-cell C and N assimilation and growth	¹³C assimilation (DIC)(pmol C cell^-1 day^-1) ^a	4.6_(T2)-9.7_(T1)	±	3.46_(T2)-5.5_(T1)	48
(active population)	¹⁵N assimilation (NH₄⁺)(fmol N cell^-1 day^-1) ^b	37.7_(T2)-183_(T1)	±	23.2_(T2)-137_(T1)	45
	¹⁵N assimilation (NO₃^-)(fmol N cell^-1 day^-1)^b	63.5_(T2)	±	64.1_(T2)	32
	¹³C-based growth (DIC)(day^-1) ^a	0.47_(T2)-0.66_(T1)	±	0.24_(T2)-0.37_(T1)	48
	¹⁵N-based growth (NH₄⁺)(day^-1)^b	0.07_(T2)-0.18_(T1)	±	0.03_(T2)-0.05_(T1)	45
	¹⁵N-based growth (NO₃^-)(day^-1)^b	0.07_(T2)	±	0.03_(T2)	32
	¹³C-derived population doubling (days)(DIC) ^a	2.12_(T2)-1.53_(T1)	±	1.06_(T2)-0.84_(T1)	48
	¹⁵N-derived population doubling (days)(NH₄⁺) ^b	14.3_(T2)-5.55_(T1)	±	5.9_(T2)-1.7_(T1)	45

^aMeasurements derived from the control treatment.

^bMeasurements derived from either the ¹⁵NH₄⁺ or ¹⁵NO₃^--amended treatment.

Estimates for single-cells derived at T₂are derived from one replicate bottle of the respective treatment.

We studied the activity, nutrient storage and uptake, and the growth response to nutrient availability of the microbial population of the southern margin of the Greenland Ice Sheet. We documented how the variability in stoichiometric ratios and uptake of C and nutrients in single glacier ice algae cells compare to bulk measurements of POM in surface ice samples. We found an active glacier ice algal community with a mean C-based doubling time of ~ 2±4 days (Table 2), which is comparable to previously measured primary production-based doubling times on the Greenland Ice Sheet[³⁷]. The C-based growth and C-assimilation rates for single cells of glacier ice algae are of the same order of magnitude as those for autotrophic marine diatoms or dinoflagellates[^{38, 39}]. Our nutrient-addition experiments demonstrated that increased concentrations of ¹⁵N-NH₄⁺, ¹⁵N-NO₃⁻, PO₄³⁻, and ¹⁵N-NH₄⁺+PO₄³⁻ did not enhance C-based growth of the bulk community and single algal cells over a 30-hour incubation period (Fig. 4). Instead, we found that C-based growth was equal or even lower under the tested nutrient-loading scenarios. We interpret the lack of growth stimulation following nutrient additions as indicative of sufficient nutrient availability under the in situ conditions. The dissolved inorganic nutrient concentrations of our study site at the time of sampling (Table 1) were comparable or lower than those reported in Holland et al.[²⁵] from the Greenland Ice Sheet. The absence of nutrient stimulation affecting glacier ice algal productivity in our study aligns with the findings by McCutcheon et al.[²⁴], where albeit proposing phosphorus to be a major control on algal growth, the increase in maximum rates of electron transport (a proxy for phtotosynthesis) upon PO₄³⁻ additions occurred only after 5 days of incubation. The effect of PO₄³⁻ addition was, hence, only evident while incubating in a closed system and without an additional supply of nutrients, e.g. by surface ice melt[⁴⁰] or atmospheric deposition[^41–44]. In contrast to previous studies, we chose a shorter incubation time to minimise “bottle effects” and substrate transfer (cross-feeding) between microorganisms. Nevertheless, potential negative impacts on growth due to the incubation in a closed bottle and/or high nutrient loading cannot be excluded. The slower C-based growth under the nutrient-loading scenarios could reflect a sensitivity of the microbial community to high nutrient loads since they are adapted to the highly oligotrophic conditions of glacial surfaces[⁴⁵].

We demonstrated a rapid uptake of inorganic N sources by the microbial community, occurring within just a few hours. This was reflected by the decreasing NH₄⁺ and NO₃⁻ concentrations during incubations (Supplementary Fig. 4a,b), a decrease of POC:PON ratio in the treatments receiving ¹⁵N-NH₄⁺ and ¹⁵N-NO₃⁻ relative to the control (Fig. 4c), and substantial ¹⁵N labelling of algal cells in ¹⁵N-NH₄⁺ and ¹⁵N-NO₃⁻ amended incubations (Figs. 3 and 4e). Rapid inorganic N assimilation by algal cells is typical of oligotrophic systems[⁴⁶] and likely reflects the ability of glacier ice algae to maximise nutrient uptake and store excess N when available, even when they are not N-limited. Indeed, low in situ nutrient concentrations are often associated with high turnover rates in aquatic environments⁴⁷]. Consistent with microbial utilization and recycling of N, DON increased by a factor of 3 in the N-amended treatments compared to the control treatment during the first 6 hrs of incubation (Supplementary Fig. 4c). In addition to possible N-storage, our SEM elemental data showed that glacier ice algae store phosphorus granules (Fig. 2, P elemental map), possibly as polyphosphate, which is a common biological phosphorus storage mechanism in algae as well as in all other domains of life[^48–52]. Such P storage abilities have been reported, for example, in Arctic Cylindrocystis strains (Zygnematophycaea)[⁵³]. Excess storage of nutrients can sustain the metabolic requirements and possibly even allow for growth of glacier ice algae at later times towards the end of the melt season when nutrient concentrations may decrease.

We show that the low cellular N and P content relative to C reflects the overall low nutrient requirements of glacier ice algae, instead of this being a potential sign of nutrient limitation. This is demonstrated by the observation that the algal cells that had a relatively high in situ C:N and C:P biomass content (mean of 509:26:1; Figs. 2 and 4d,e; Table 2) did not show signs of nutrient limitation during the short-term incubations upon nutrient addition. The average stoichiometric ratios of glacier ice algae clearly exceed the marine derived Redfield ratio (C:N = 6.6, C:P of 116:1) and reflect the very different growth conditions in the dilute, oligotrophic ice melt habitat, which requires markedly different physiological adaptations for glacier ice algae relative to marine eukaryotic autotrophs. This deviation from Redfield ratio is in line with Williamson et al.[³³] and Lutz et al.[⁵⁴] assessments, using bulk material (POM) C:N:P ratios. Note, that in contrast to the single-cell analysis conducted here, bulk POM may also capture other organisms (e.g., other eukaryotes, fungal biomass, dispersed cryoconite with cyanobacteria), partially degraded necro-mass and organic matter. Thus, the single cell stoichiometric measurements provide a more direct measure of the elemental ratios for the ice algal community, and provide insights into their nutrient retention processes.

We suggest that freshly fixed C is being transferred from primary producers to the heterotrophic community via cross-feeding. This is supported by the estimated contribution of active algal cells, of only ~ 7 ± 6 to 15 ± 12%, to the total bulk C uptake recovered in bulk POM from ¹³C-DIC. Despite heterogeneous biomass distribution between incubation bottles and a large variability in single cell activity, the estimated contribution of active algal biomass to bulk C uptake is small and gives first insights into C cycling within glacier microbial food webs. Other autotrophic taxa comprised only a small fraction of the total autotrophic community (cyanobacteria: 4% of bacterial ASVs; snow algae: ~10% of algal counts), and it is therefore nevertheless likely that glacier ice algae were the main primary producers in our incubations. We thus deduce that a large fraction of the DIC assimilated by glacier ice algae was rapidly released as DOC (e.g., as exopolymeric substances that can be retained on the filters and contributing to the POM), which could be assimilated by the microbial community. We hypothesize that some freshly fixed C from glacier ice algae was also transferred to Chytridiomycota during infection, as these parasitic fungi rely entirely on autotrophic C[⁵⁵]. Chytridiomycota formed a significant fraction of the eukaryotic ASVs in our samples (Supplementary Fig. 1a and 2), as further confirmed by microscopy showing fungal hyphae (Fig. 1c) and numerous algal cells with signs of parasitic infections (Supplementary Fig. 2e-g). Taken together, glacier ice algae may play an important role in rapidly transferring organic C to the microbial food web.

The combined single-cell elemental and isotopic imaging revealed that glacier ice algal elemental composition and their C- and N-based growth rates were highly variable (Figs. 3, 4d,e and 5). Phenotypic intrapopulation variability could also be seen microscopically, as individual cells differed in size, cell division stage or pigmentation (Fig. 1c, Supplementary Fig. 2). The striking variability and DIC and DIN uptake rates in the algal cells (Fig. 4d,e and 5) may explain at least some of the observed variability in the single-cell elemental ratios. Further, different activity modes within the glacier ice algal population, different growth stages of individual cells[⁵⁶], and/or a selective allocation of C, N or P to cell maintenance, including biomolecule replacements and/or repair[^{57, 58}], and/or micro-scale variations in nutrient availability, may all impact on the C:N:P ratio of individual cells. The increasing DIC:DIN assimilation ratios observed throughout the incubation period (Fig. 5) likely reflects a temporal decoupling of primary productivity from DIN uptake in the algal cells. Such decoupling might reflect the adaptations of glacier ice algae to the conditions typical of the summer ablation season: high light, low nutrient availability and continuous DIC supply[⁵⁹]. Intrapopulation variability of microorganisms is ubiquitous in nature[⁶⁰] and variable activity or phenotypic modes in glacier ice algal populations are likely critical for adapting to environmental gradients over time. From our measurements, we also provide the first assessment of the active/alive (90%) and non-active/dead (10%) fraction of the glacier ice algae population (Fig. 4). The fraction of active cells may depend on factors such as the season, location, rate of dead cell degradation, and infections by parasitic fungi Chytridiomycota[^61–63]. Single-cell analyses thus provide key insights into the adaptive responses and microbial interactions of individual microbes.

As our climate warms, the Greenland Ice Sheet faces prolonged summer ice melt, which may extend the duration and magnitude of surface algal blooms. New bare ice surfaces may also be colonized by algae if sufficient nutrients are available to support their growth[^{24, 25, 33}]. However, direct measurements of nutrient uptake and growth of glacial microbial communities have been lacking, limiting our understanding of their nutrient requirements in these oligotrophic glacier environments. We address this gap with the first measurements of dual DIC and NH₄⁺ or NO₃⁻ assimilation, as well as the elemental composition of both the supraglacial community and individual glacier ice algal cells. Our findings suggest that the growth of glacier ice algae and the autotrophic microbial community is not limited by nutrient availability under the in situ conditions of our sampling site. Glacier ice algae efficiently assimilate available DIN sources, are able to store excess P intracellularly, and exhibit variable and elevated C:N:P biomass ratios (mean of 509:26:1) compared to Redfield stoichiometries. These findings underscore the optimised metabolic adaptations to low nutrient levels in situ and their potential to grow on new emerging bare ice surfaces using the allochthonous nutrients supplied by atmospheric deposition[^41–44], surface ice melt[⁴⁰] or N₂-fixation[^{41, 64}]. Hence, algal surface blooms could occur in these newly exposed bare regions, which would result in albedo reduction and enhanced melting, and thus constitute a climate feedback.

Study area and ice sampling

The supraglacial algal community was sampled near the SW tip of the Greenland Ice Sheet (61°05'8708"N, 46°50'9442"W), close to the PROMICE station QAS-M (61°05'54.7"N, 46°50'01.0"W), at an elevation of 680 m. On July 12, 2020, surface ice was collected by scraping off the top ~2 cm, which were placed into two 5 L Whirl-pack bags (Nasco, USA). The two bags were closed by wrapping the bag top over itself several times, likely sealing the bag from exchange of gases with the atmosphere. The ice was allowed to melt under in situ light conditions at an ambient air temperature of ~4 °C. The ice took ~36 hrs to melt completely. The two bags of melted ice were combined into a single Whirl-pack bag, homogenized, subsampled and used for the incubations, as described below.

Incubation experiment: C-fixation and N-uptake

We performed stable isotope incubation experiments to measure autotrophic C-fixation from ¹³C-DIC and N-assimilation from ¹⁵NO₃^- and ¹⁵NH₄⁺ of the supraglacial community. Additionally, we tested the effect of combined ¹⁵NH₄⁺ and PO₄^3-, and only PO₄^3- (see Supplementary Figure 3 for a graphical overview of the set-up). The melted surface ice, without any amendments, represents the T0 time point of our incubation experiment. The homogenized meltwater was then distributed into five 1 L blue cap Schott bottles, in which four different treatments and one control were prepared. C-fixation by algae was traced by adding 30 µmol L^-1 ¹³C-labelled bicarbonate (¹³C-NaHCO₃, ≥98 ¹³C atom%; Sigma-Aldrich) to all treatments and control. N-assimilation was traced by adding ¹⁵N-labelled ammonium sulfate (¹⁵N-(NH₄)₂SO₄) and ¹⁵N-labelled sodium nitrate (¹⁵N-NaNO₃) to separate treatments (both ≥98 ¹⁵N atom%, Sigma-Aldrich) at ~10 µM final concentration. The effect of PO₄^3- availability on C and N uptake was assessed by adding potassium di-hydrogen phosphate at a final concentration of 10 µM to separate ¹³C-DIC-only and ¹³C-DIC+¹⁵N-NH₄⁺ treatments. Once the tracers and nutrients were added, the liquid was gently homogenised and then further distributed from the 1L Schott bottles into triplicated 250 mL serum bottles, closed with butyl rubber stoppers and aluminium crimps, leaving a 10 mL air head-space. These bottles were then incubated under in situ conditions on the ice surface for ~30 hrs, where they received a total amount of shortwave radiation of 346 W m^-2[⁶⁵].

Subsamples were taken at T0 from the melted glacier ice without nutrient or tracer addition, T1 (~6 hrs incubation time since nutrient or tracer addition, from the 1 L Schott bottles) and at T2(~30 hrs incubation time, from the triplicate serum bottles). The following subsamples were taken: 1) for measurements of the atom% of the DIC pool after ¹³C-DIC addition (T1 and T2 time points), a liquid sample was collected with a syringe without headspace and bubble formation into 5.9 ml exetainers (Labco, Wales, UK), containing 100 µL saturated ZnCl₂solution to stop the biological activity. The exetainers were stored in the dark at 4 °C until analysis; 2) For bulk C-fixation and N-assimilation measurements, the sample (145-193 mL) was filtered onto pre-combusted (450 °C) glass fibre filters (GF/F nominal pore size of 0.7 µm; Whatman, Maidstone, UK) and stored in plastic dishes at -80 °C until analysis. All bottles and laboratory equipment, such as filtration towers and forceps, were cleaned by soaking in 5 % HCl overnight, followed by soaking and rinsing in Milli-Q. A filter rosette with one filter unit per treatment was used to avoid potential cross-contamination of isotopically labelled material; 3) For single-cell analyses by HR-SIMS (collected at T0 and T2, and for the control and ¹⁵N-NH₄⁺ treatments at T1) and SEM-EDS (collected at T0), 5 mL subsamples were collected and fixed with 2% EM-grade paraformaldehyde (PFA; EMS, USA) for 24 hrs at 4 °C. The fixed cells were then filtered onto 3 µm pore size gold–coated polycarbonate filters (25 mm diameter; GTTP, Merck Millipore, Eschborn, Germany), washed three times with ~10 mL of 0.2 µm filtered glacier stream water and stored at -20 °C; 4) Samples for dissolved inorganic and organic nutrient measurements (collected at T0, T1, T2) were collected by filtering 30 mL through 0.2 µm PES filters (25 mm diameter, Merck Millipore) with a polypropylene syringe into pre-washed 30 mL HDPE Nalgene bottles. To avoid any contamination for ultra-trace ion analysis, the bottles and caps were previously soaked in 5% HCl overnight and thereafter soaked in fresh Milli-Q water (Millipore, USA) for three days, with Milli-Q water replacement every day[⁶⁶]. Once the nutrient samples were taken, the bottles were stored frozen at -20 °C until analysis; 5) For microscopy and cell counts (collected at T0), 2 mL of the sample liquid was preserved in duplicates in 2.5% EM-grade Glutaraldehyde (EMS, USA) and stored in the dark at 4°C; 6) For amplicon sequencing (collected at T0), 500 mL of the melted ice surface sample was filtered onto a sterile, 0.2 µm cellulose nitrate filter (Thermo Scientific Nalgene), which was preserved in a sterile cryotube, flash-frozen and transported to the home laboratory in a cryo-shipper. The filter was stored at -80 °C until nucleic acid extraction.

Quantification of ¹³C-DIC, ¹⁵N-NH₄⁺ and ¹⁵N-NO₃^- atom% and atom%excess

The abundance of heavy stable isotope tracers, expressed in percentage ('atom%'), depends on the concentration of the added heavy isotope and its dilution with the naturally occurring isotopes of the same compound (DIC, NH₄⁺, NO₃^-). The concentration of the heavy isotope corrected for the naturally occurring heavy isotope already present in the sample before tracer addition is termed ‘atom% excess’. For determining the ¹³C-atom% of DIC, 2 ml subsamples from each ZnCl₂ fixed exetainer were injected into helium-flushed exetainers and acidified with phosphoric acid following Torres et al. 2005[⁶⁷] to convert all DIC to CO_2.Headspace subsamples were injected into GC-IRMS (isoprime precision, precision ± 0.1‰ for ¹³C-standards of 0-100 nM). The ¹³C-atom% of DIC in the ambient water was calculated from the measured concentrations of ¹³C-CO₂ and ¹²C-CO₂. Since the ¹³C-atom% of DIC between T1 and T2 decreased slightly (means of 4.4 to 3.6 atom% excess for all treatments), we used the mean value between T1 and T2 for each of the respective treatments for the T2 ¹³C-label incorporation calculations (3.9 ¹³C-atom% excess). For determining the ¹⁵N-atom% of NH₄⁺or NO₃^-, we used their added concentrations (10 µM with ≥98 ¹⁵N-atom%) and corrected it for the dilution with naturally occurring NH₄⁺ or NO₃^- at T0 (0.078 and 0.05 µM, respectively, with 0.36 ¹⁵N-atom% natural abundance), yielding 98 and 97 atom% excess for ¹⁵N-NH₄⁺ and ¹⁵N-NO₃, respectively.

Isotopic analyses of bulk particulate organic matter

The C and N contents and the isotopic composition of bulk particulate organic matter (at T0, T1 and T2) were determined from the particulate material collected on GF/F filters, which were dried at 60 °C, decalcified overnight under 37% HCl fumes in a desiccator and again dried again at 60 °C. One-quarter of each filter was packed into tin capsules and analysed by an elemental analyser (Thermo Flash EA 1112) coupled to a continuous-flow Thermo Delta Plus XP isotope ratio mass spectrometer; Thermo Finnigan, Dreieich, Germany) (EA-IRMS) at the Max-Planck-Institute for Marine Microbiology (MPIMM), Germany. Caffeine was used as a standard for isotope ratio monitoring and C and N quantifications. The limit of detection (LOD) for isotopic enrichment was 1.078 ¹³C-atom% and 0.365 ¹⁵N-atom%.

Single-cell elemental ratios and HR-SIMS analyses

Single-cell elemental ratios were obtained at the T0 timepoint using scanning electron microscopy (SEM, Quanta FEG 250, Thermo Fisher Scientific) coupled to energy-dispersive X-ray spectroscopy (EDS, Bruker Nano GmbH)[^68–73], at the MPIMM. To avoid charging effects through the presence of large numbers of minerals, cells had to be transferred from the GTTP filters (3 µm) onto filters with a thicker gold coating prior to analysis (25 mm, 0.8 µm pore size, 40/20 nm coating; APC, Eschborn Germany). This was done by adding one drop of Milli-Q onto the filter surface with algal cells, placing the new filter piece with thicker gold coating onto a drop of water, freezing both filters together for 2 minutes and once frozen, removing the old filter by peeling it off. This procedure transferred substantial amounts of the original filter material onto the new filter surface without the need to scrape off any cells. Additionally, filters were gently rinsed with Milli-Q to remove some minerals/sediment grains. For morphological and autofluorescence-based identification of algal cells, the gold-coated filters were cut into sections (approx. 5x5 mm) and areas of interest were marked and imaged using a laser micro-dissection (LMD) microscope (6000 B, Leica) prior to SEM-EDS measurements. The EDS system is equipped with two QUANTAX XFlash 6/30 (Bruker Nano GmbH, Germany) detectors. The detector area is 30 mm² and the detectors have an energy resolution at Mn K α line of <123 eV, allowing for the quantification of light elements. An NBS SRM 1155 ANSI 316 stainless steel standard was used to check the performance of the EDS system. 10 kV was used as a minimum accelerating voltage to analyse the sample for all major elements contained in the algal cells, also restricting the penetration depth to around 2 µm (demonstrated for cyanobacterial filaments in Schoffelen et al.[⁷³]), so reducing any potential signal from the filter surface. The analysis of the elemental content of algal cells was performed using the standardless P/B-ZAF method (Quantax 400 software, version 1.9; Bruker), suitable for samples with topography and allowing for measurements of light to heavy elements. Further details on the data processing can be found in Khachikyan et al.[⁶⁹]. Cells which were too thin for a robust signal were excluded from data processing by manually inspecting the obtained spectra. Single cell relative C:P, C:N, and N:P atomic ratios were determined from the measurements in atom%, while the data in mass% was used to calculate the absolute elemental content of algal cells (see next section).

The T0 sample, one replicate per treatment of the T2 timepoint, and additionally the T1 of the control and the ¹⁵N-NH₄⁺ amended treatments (due to rapid NH₄⁺cycling), were used for HR-SIMS analysis. The pre-imaged filter pieces from SEM-EDS analysis and additional filter pieces were mounted on a glass slide and coated with a 5 nm layer of gold prior to HR-SIMS analyses. Single-cell ¹⁵N and ¹³C assimilation rates of algal cells were determined by HR-SIMS (IMS 1280, CAMECA, Gennevilliers, France) at the Natural History Museum in Stockholm, Sweden. Areas of interest were pre-sputtered with a primary Cs⁺ ion beam of 3 nA for 240s over an area of 80 x 80 µm and then analysed with a 100 pA beam over 70 x 70 µm at a spot size of 1 µm for 60 cycles. The HR-SIMS images (256 x 256 pixel) were recorded for ¹²C¹⁵N⁻, ¹³C¹⁴N⁻ and ¹²C¹⁴N⁻ ions with a peak-switching routine at a mass resolving power of 12,000 (M/ΔM) using a low-noise ion-counting electron multiplier. The detection limit was < 0.01 counts per second (cps). For integration times of 60 s (¹²C¹⁴N⁻), 300s (¹²C¹⁵N⁻) and 120s (¹³C¹⁴N⁻) over 60 cycles, a run was expected to have total background count lower than 0.6, 3 and 1.2, respectively, not requiring any baseline correction. For the 256 x 256 pixel resolution, this approximates to background levels of 1e^-5, 5e^-5 and 2e^-5 cps pixel^-1, respectively. Images were processed using the CAMECA WinImage2 software. Secondary ion images were drift-corrected and accumulated for each measurement and the detector dead time, electronically gated at 44 ns, was processed on each pixel. Regions of interest (ROIs) were manually drawn around the algal cells. The ¹³C/(¹³C+¹²C) and¹⁵N/(¹⁵N+¹⁴N) ratios were subsequently calculated as means for each ROI. Unlabelled (natural abundance) glacier ice algae cells from non-incubated samples were also measured (n=29) and mean isotope fractions (0.0037±0.00006 and 0.0111±0.00016 for ¹⁵N and ¹³C, respectively) were subtracted from the labelled samples to obtain ‘excess’ isotope fractions of the biomass. A. alaskanum and A. nordenskioeldii are grouped together as glacier ice algae within this study, due to their taxonomically close relationship and partial size overlap[⁷⁴], which challenged an unambiguous species identification from microscopic images obtained for the filtered cells. We acknowledge that the fixation of the algal cells with PFA after incubations for HR-SIMS analysis may result in a decrease of ¹³C-enrichement of ca. 4-8%, and, to a lesser extent ¹⁵N-enrichment[^75–77]. However, this effect is likely considerably smaller than the differences observed in the C- and N-based growth rates between single-cell (fixed with PFA) and bulk (preserved by freezing) measurements in our study. We therefore chose not to apply any corrections to the measured enrichment values of the single cell analyses. Cells were considered as enriched/active if their mean ¹³C/(¹³C+¹²C) enrichment exceeded the mean observed natural abundance value + 3x the standard deviation of unlabelled control cells[⁷⁸](1.15 ¹³C-atom% for glacier ice algae).

Cellular biovolume, dry weight, and absolute elemental content of glacier ice algal cells

Cell dimensions were obtained from HR-SIMS images using ImageJ. Biovolumes were subsequently calculated by assuming cylindrical shapes for glacier ice algae after Hillebrand et al.[⁷⁹]. Cellular dry weights (pg cell^-1) were calculated by multiplying the algal biovolumes (mean of 1414±873 µm^-3 for all imaged algal cells, n=244) by the glacier ice algal-specific buoyant density of 1160 kg m⁻³[⁸⁰] and a mean dry fraction of 0.28 (obtained from C. vulgaris[⁸¹]). Absolute elemental contents of glacier ice algal cells (pg element cell^-1) were determined by multiplying the median mass fraction of C, N or P in glacier ice algal cells (0.72, 0.04 and 0.04 C, N and P, respectively, derived from SEM-EDS, Supplementary Table 5) by the cellular dry weights (pg cell^-1)[⁶⁹].

C- and N-assimilation rates determined by EA-IRMS and HR-SIMS

Bulk C-assimilation rates were calculated using the following equation[⁸²]:

where ¹³C-atom% excess_POC represents the ¹³C-atom% of incubated POC minus its natural abundance atom%, POC refers to the biomass concentration (µmol C L^-1), ¹³C-atom% excess_DIC represents the ¹³C-atom% in DIC minus its natural abundance atom% and Δt represents the incubation period (in days, T0-T1 or T0-T2). We assume that the ¹³C-assimilation rates correspond to net photosynthesis, as any ¹³C fixed during the incubation (1.1 days including ~6 hours of twilight) may have partially been respired again, which would not be measured by HR-SIMS.

Bulk N-assimilation rates from NH₄⁺ or NO₃^-were calculated analogously from the ¹⁵N-atom% of incubated PON minus its natural abundance atom%, the corresponding PON concentration of the sample (µmol N L^-1), the ¹⁵N-atom% of NH₄⁺ or NO₃^- present in the incubation water minus their natural abundance atom%, and the incubation period, as described above.

Single-cell specific C-fixation rates were calculated according to the following equation[⁸²]:

where ¹³C-atom% excess_cellrepresents the ¹³C-atom% of single algal cells minus their natural abundance atom%, C_cell represents the mean C content of single algal cells (pmol C cell^-1, calculated as described above), ¹³C-atom% excess_DICrepresents the ¹³C-atom% in DIC minus its natural abundance atom% and Δt represents the incubation period (in days, T0-T1 or T0-T2).

Single-cell specific N-assimilation rates from NH₄⁺ or NO₃^-were calculated analogously from the ¹⁵N-atom% of single algal cells minus the natural abundance atom%, the ¹⁵N-atom% of NH₄⁺ or NO₃^-present in the incubation water minus the natural abundance atom%, the corresponding mean N content of single algal cells (pmol N cell^-1, calculated as described above), and the incubation period, as described above.

C-and N-based growth rates

Growth rates based on ¹³C-DIC, ¹⁵NH₄ or ¹⁵NO₃ isotope uptake were calculated for the bulk community (EA-IRMS measurements) or single algal cells (HR-SIMS measurements). C-based growth rates (day^-1) were calculated following Martínez-Pérez et al.[⁷¹], based on Montoya et al.[⁸²]:

where ¹³C-atom% excess_DIC represents the ¹³C-atom% in DIC minus its natural abundance atom%, ¹³C-atom% excess_POCthe ¹³C-atom% of incubated POC (of either bulk or single cell biomass) minus its natural abundance atom%, and Δt representing the incubation period (in days, T0-T1 or T0-T2.

N-based growth rates were calculated analogously from the ¹⁵N-atom% excess of either NH₄⁺ or NO₃^- in the incubation water, the ¹⁵N-atom% excess of PON of either bulk or single cell biomass, and the incubation period, as described previously. The C or N-based growth rates assume exponential growth[⁷¹] and that all newly incorporated ¹³C or ¹⁵N are due to biomass increase[⁸³], e.g. a growth rate of 1 day^-1 means that cells double their C or N content once per day and, thus, divide once. The obtained growth rate estimates are independent of the biomass[⁸²]. A fraction of assimilated ¹³C or ¹⁵N may be allocated to C- or N-storage, recycling or replacing of cell components without net per cell growth. However, as this fraction is unknown, we do not consider it in our calculations. See Polerecky et al.[⁵⁷] and Halbach[⁵⁸] for more details on assumptions for isotope uptake calculations. Population doubling times were calculated as 1/growth rate.

Glacier ice algae contribution to bulk C- and N-uptake

Similar to previous studies[^84,85], we estimated the relative contribution by active glacier ice algae to the total bulk C and N uptake (originating from ¹³C-DIC or ¹⁵NH₄⁺) for the different timepoints:

where assimilation_cell is the mean assimilation rate of active glacier ice algae of the respective substrate (pmol element cell^-1 day^-1), N_cellis the mean abundance of the active glacier ice algae (cells L^-1) and assimilation_bulkrepresents the assimilation rates of the bulk community of the corresponding time point (µmol element L^-1 day^-1). The active glacier ice algal cell numbers are derived from algal counts at T0, corrected for the active population fraction based on SIMS measurements of C fixation (90% active cells). Biomass distribution between incubation bottles was variable due to rapid sinking of particulate material, thus, the large uncertainty associated with the parameter of assimilation_bulk contributes to the uncertainty of relative contribution by glacier ice algae. To account for varying biomass between bottles and the potential varying algal abundance, we also performed the calculations using the algal abundance corrected by the fractional change in POC concentrations between T0 and T1, as well as T0 and T2. This revealed a consistently low contribution (7-12% for C from DIC and 3-4% for N from NH₄⁺; Supplementary Tables 2 and 3). Uncertainties in the contribution of the glacier ice algal community assimilation to total assimilation derive from the combined uncertainties of each variable, following the laws of error propagation (Supplementary Note 2).

Dissolved nutrient analysis

Dissolved NO₃^-, NO₂^-, NH₄⁺, and PO₄^3- concentrations from T0, T1 and T2 were analysed on a Metrohm Ion chromatography system (883 Basic IC Plus and 919 Autosampler Plus) at Uppsala University, Sweden. The IC was equipped with a peristaltic pump to enable full loop injections (400 µl) to decrease the LOD and limit of quantifications (LOQ)[⁶⁶]. Sample tubes were stored with a lid in the autosampler to avoid contamination with N from air. LOD’s and LOQ’s were determined as 3 x and 10 x the standard deviation (STDEC) of the lowest nutrient concentrations from standards, according to the EPA procedure for method detection limit[⁸⁶]. LOD’s were 0.011, 0.008, 0.005 and 0.004 µM and LOQ’s were 0.022, 0.027, 0.018 and 0.007 µM, for NO₂^-, NO₃^-, PO₄^3-,and NH₄⁺, respectively. The corresponding mean precisions were ±3, ±8, ±5 and ±3% and accuracy -8, -12, -4 and -1%, for NO₂^-, NO₃^-, PO₄^3-,and NH₄⁺, respectively, as determined from a comparison of QC standards with 0.043, 0.026, 0.015, and 0.015 µM levels.

Total dissolved nitrogen (TDN) was analysed on a Shimadzu TNM (Tokyo, Japan). DON was calculated as DON=TDN – DIN, where DIN is (NO₃^- + NH₄⁺ + NO₂^-). Total dissolved phosphorus (TDP) was analysed by the molybdenum blue method after digestion with potassium persulfate and autoclaving at 121 °C for 60 min. DOP was then calculated as DOP=TDP - PO₄^3-. There was insufficient liquid left from the first collected sample (T0) and the reported concentration was measured from a sampling location close to the experimental site, but three days later. The LODs were 0.83 and 0.03 µM and LOQs 0.03 and 0.07 µM for TDN and TDP, respectively. The accuracy for TDN was 12% and precision 5% with a standard of 0.03 µM N.

The T0/in situ samples for analysis of Na²⁺, Mg²⁺, K⁺ and Ca²⁺ were acidified using Aristar HNO₃. The analyses of major, minor and trace element analyses was carried out with an inductively-coupled plasma mass spectrometer (ICP-MS; Thermo Fisher iCAPQc). The precision of the analyses was between 1-5% and LOD’s for Na²⁺, Mg²⁺, K⁺ and Ca²⁺ were 0.2, 0.03, 0.65 and 0.46 µg L^-1, respectively. The ICP-MS analyses was conducted by Stephen Reid at the University of Leeds, UK.

Community composition and algal abundance

The algal abundance and community composition at T0 were microscopically characterised from the glutaraldehyde preserved samples and algal cells counted on a B/W FlowCAM™ II (Fluid Imaging Technologies, Maine, USA) using a 100 µm x 2 mm flow cell, a 10 x objective and the automated-imaging mode. A minimum of 760 total algal cells per sample were counted. Algal cells were subsequently taxonomically identified, using the VisualSpreadSheet (VISP). Additional pictures of the supraglacial community were taken from unfixed, fresh sample material using a Nikon Eclipse Ti microscope. Images of the fresh unfixed and fixed algal cells were screened for signs of fungal infections (Supplementary Figure 3).

Amplicon sequencing was performed to determine the microbial composition of the sample prior to incubation. DNA extraction was performed using the DNeasy PowerSoil Pro Kit (Qiagen) according to the manufacturer’s protocol. Thereafter, DNA concentration was measured on a Qubit 3.0 (Invitrogen) with the broad-range dsDNA kit (Invitrogen). Amplification was performed for the bacterial 16S rRNA gene using Bakt_341F (5’- CCTACGGGNGGCWGCAG-3‘) and Bakt_805R (5’- GACTACHVGGGTATCTAATCC-3‘)[⁸⁷] and for the 18S rRNA gene using 528F (5’- GCGGTAATTCCAGCTCCAA-3‘) and 706R (5’-AATCCRAGAATTTCACCTCT-3‘)[⁸⁸]. The amplicon library was built in a two-step PCR. Each reaction of the first PCRs contained 12.5 μL of 2x PCRBIO Ultra Mix (PCR Biosystems), 0.5 μL of forward and reverse primer from a 10 μM stock, 0.5 μL of bovine serum albumin (BSA) to a final concentration of 0.025 mg mL^-1, 0.6 μL of sterile water and 5 μL of template DNA. For the first PCR, conditions were as follows: at 95°C for 2 min, followed by 38 cycles of 95 °C for 15 s, 55 °C for 15 sec, 72 °C for 40 sec, with a final extension performed at 72 ˚C for 4 min. An electrophoresis 1% agarose gel was run with PCR products before proceeding. Samples were subsequently indexed in a second PCR. In the second PCR, 5 µl of product from the first PCR was used as template to add indexes and sequencing adaptors in a reaction consisting of 12.5 μl of 2x PCRBIO Ultra Mix (PCR Biosystems), 2 μl of each index primer (P5/P7), and DNase free water to a final volume of 28 µl. For the second PCR, conditions were as follows: pre-incubation at 98 °C for 1 min, followed by 13 cycles of 98 °C for 10 sec, 55 °C for 20 sec, and 72 °C for 40 sec, and ending with a final step at 72 °C for 5 min. The final PCR products were purified with 15 µl HighPrep PCR magnetic beads (MagBio Genomics Inc. Gaithersburg, Maryland, US) according to the manufacturer's instructions and eluted in 27 µl TE buffer. Aliquots of the PCR products were run on a 1.5% agarose gel and checked under UV light. Concentrations of the amplified and purified DNA samples were measured on a Qubit 2.0 fluorometer (Invitrogen, Eugene, Oregon, US). The samples were then equimolarly pooled, and this final library was sequenced on an Illumina MiSeq using the V2 kit (Illumina Inc. SanDiego, California, US) resulting in 2×250 bp reads.

Data analysis

Statistical analysis and plotting was undertaken in R[⁸⁹]. The non-parametric Kruskal-Wallis t-test was used to explore the similarity of data for individual treatments generated by HR-SIMS and EA-IRMS, followed by a -hoc test of multiple comparisons using the Fisher's least significant difference criterium and Holm’s p-adjustment method. Data were considered significantly different at p<0.05. Inactive cells (i.e. those with no significant ¹³C-DIC incorporation) were excluded from statistical tests involving cell activity. Results are presented as mean ± standard deviation. The 16S and 18S rRNA gene amplicons were pre-processed and analysed using the DADA2 R package[⁹⁰] for ASVs. Taxonomic assignment was made using the SILVA (V148) rRNA gene database[⁹¹]. Detailed documentation of the pipelines, including parameter setups, is available in Trivedi et al.[⁹²]. Results were visualized using the phyloseq v1.36.0[⁹³] and ggplot2 v3.3.5 R packages[⁹⁴]. Classes and phyla with <1% mean relative abundance were grouped under “Others” for 16S and 18S rRNA gene data representation, respectively.

Data availability

All data analysed in this study are included in this article and its Supplementary Information. The amplicon sequencing data are available at SUB14452981 under BioProject ID PRJNA1113015.

Acknowledgements

The presented work is part of the project DeepPurple which has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant Agreement No. 856416). Alexandre Anesio and Martin Hansen received support from the Aarhus University Research Foundation (grant numbers AUFF-T-2017-FLS-7-4 and AUFF-2018), Katharina Kitzinger, Sten Littmann and Marcel Kuypers from the Max Planck Society. Liane G Benning and Rey Mourot were also supported through funding from The Helmholtz Recruiting Initiative (award no. I-044-16-01). James A Bradley was supported by the CNRS Chaires de Professeur Junior (CPJ) and the Alexander von Humboldt Foundation. We would like to thank Swantje Lilienthal for their help during the SEM-EDS imaging and Gabriele Klockgether, Wiebke Mohr and Gaute Lavik for fruitful discussions. We would like to thank Marie Bolander Jensen for organising the analysis of the amplicon sequencing and Christoffer Bergvall for analysing the nutrient samples. We acknowledge NordSIMS-Vegacenter for the provision of facilities and experimental support and we thank Kerstin Lindén and Heejin Jeon for their assistance. NordSIMS-Vegacenter is funded by the Swedish Research Council as a national research infrastructure (Dnr. 2021-00276) and is further supported by the Swedish Museum of Natural History and the University of Iceland. Data from the Programme for Monitoring of the Greenland Ice Sheet (PROMICE) were provided by the Geological Survey of Denmark and Greenland (GEUS) at http://www.promice.dk.

Author contributions

LH drafted the manuscript, collected the samples, and analysed the data. KK helped in study design and sampling protocols, conducted the IRMS analysis and helped with data interpretation. AA helped in study design, sample collection and data interpretation. MJW conducted the HR-SIMS analysis and helped during data analysis and interpretations. SL conducted the SEM-EDS analysis, and its data analysis and helped with their interpretations. MH helped during the study design and data interpretations. MMMK helped during the study design and provided funding for SEM-EDS and IRMS analyses. LGB helped for study design and data interpretations. RM helped with sample collection, in-field processing as well as DNA extraction and analysis of the sequencing data. MO and JB helped in the analysis of stable isotope data and its interpretations. MT helped with data interpretation. LEJ supervised DNA library building, sequenced the amplicon libraries and helped during their analysis. MT, LGB and AA obtained the funding for the Deep Purple ERC Synergy project. All co-authors contributed to the drafting of this manuscript.

Smith, H. J. et al. Biofilms on glacial surfaces: hotspots for biological activity. NPJ Biofilms Microbiomes 57, 10–13 (2016).
Anesio, A. M. & Laybourn-Parry, J. Glaciers and ice sheets as a biome. Trends Ecol Evol 27, 219–225 (2012).
Hoham, R. W. & Remias, D. Snow and Glacial Algae: A Review. Journal of Phycology vol. 56 264–282 Preprint at https://doi.org/10.1111/jpy.12952 (2020).
Anesio, A. M., Lutz, S., Chrismas, N. A. M. & Benning, L. G. The microbiome of glaciers and ice sheets. NPJ Biofilms Microbiomes 3, 0–1 (2017).
Margesin, R. & Collins, T. Microbial ecology of the cryosphere (glacial and permafrost habitats): current knowledge. Appl Microbiol Biotechnol 103, 2537–2549 (2019).
Williamson, C. J. et al. Glacier Algae: A Dark Past and a Darker Future. Front Microbiol 10, (2019).
Lutz, S., McCutcheon, J., McQuaid, J. B. & Benning, L. G. The diversity of ice algal communities on the Greenland Ice Sheet as revealed by oligotyping. Microb Genom 4, 1–10 (2018).
Williamson, C. J. et al. Ice algal bloom development on the surface of the Greenland Ice Sheet. FEMS Microbiol Ecol 94, 1–10 (2018).
Cook, J. M. et al. Glacier algae accelerate melt rates on the south-western Greenland Ice Sheet. Cryosphere 14, 309–330 (2020).
Di Mauro, B. et al. Glacier algae foster ice-albedo feedback in the European Alps. Sci Rep 10, 1–9 (2020).
Stibal, M. et al. Algae Drive Enhanced Darkening of Bare Ice on the Greenland Ice Sheet. Geophys Res Lett 44, 11,463-11,471 (2017).
Yallop, M. L. et al. Photophysiology and albedo-changing potential of the ice algal community on the surface of the Greenland ice sheet. ISME J 6, 2302–2313 (2012).
IPCC. IPCC Special Report on the Ocean and Cryosphere in a Changing Climate. Intergovernmental Panel on Climate Change (2019).
van den Broeke, M. et al. Greenland Ice Sheet Surface Mass Loss: Recent Developments in Observation and Modeling. Curr Clim Change Rep 3, 345–356 (2017).
Van Den Broeke, M. R. et al. On the recent contribution of the Greenland ice sheet to sea level change. Cryosphere 10, 1933–1946 (2016).
Fettweis, X. et al. Estimating the Greenland ice sheet surface mass balance contribution to future sea level rise using the regional atmospheric climate model MAR. Cryosphere 7, 469–489 (2013).
Ryan, J. C. et al. Greenland Ice Sheet surface melt amplified by snowline migration and bare ice exposure. Sci Adv 5, (2019).
Halbach, L. et al. Dark ice in a warming world : advances and challenges in the study of Greenland Ice Sheet ’ s biological darkening. 1–6 (2023).
Tedstone, A. J. et al. Algal growth and weathering crust state drive variability in western Greenland Ice Sheet ice albedo. Cryosphere 14, 521–538 (2020).
Winkel, M. et al. Seasonality of Glacial Snow and Ice Microbial Communities. Front Microbiol 13, (2022).
Cook, J. M. et al. Glacier algae accelerate melt rates on the south-western Greenland Ice Sheet. Cryosphere 14, 309–330 (2020).
Williamson, C. J. et al. Algal photophysiology drives darkening and melt of the Greenland Ice Sheet. Proceedings of the National Academy of Sciences 201918412 (2020) doi:10.1073/pnas.1918412117.
Stibal, M. et al. Environmental Controls on Microbial Abundance and Activity on the Greenland Ice Sheet: A Multivariate Analysis Approach. Microb Ecol 63, 74–84 (2012).
McCutcheon, J. et al. Mineral phosphorus drives glacier algal blooms on the Greenland Ice Sheet. Nat Commun 12, 1–11 (2021).
Holland, A. T. et al. Dissolved organic nutrients dominate melting surface ice of the Dark Zone (Greenland Ice Sheet). Biogeosciences 16, 3283–3296 (2019).
Perini, L. et al. Darkening of the Greenland Ice Sheet: Fungal Abundance and Diversity Are Associated With Algal Bloom. Front Microbiol 10, 557 (2019).
Lutz, S., Anesio, A. M., Edwards, A. & Benning, L. G. Linking microbial diversity and functionality of arctic glacial surface habitats. Environ Microbiol 19, 551–565 (2017).
Maccario, L., Vogel, T. M. & Larose, C. Potential drivers of microbial community structure and function in Arctic spring snow. Front Microbiol 5, 413 (2014).
Larose, C. et al. Microbial sequences retrieved from environmental samples from seasonal Arctic snow and meltwater from Svalbard, Norway. Extremophiles 14, 205–212 (2010).
Bellas, C. M., Anesio, A. M. & Barker, G. Analysis of virus genomes from glacial environments reveals novel virus groups with unusual host interactions. Front Microbiol 6, 656 (2015).
Zawierucha, K. et al. A hole in the nematosphere: tardigrades and rotifers dominate the cryoconite hole environment, whereas nematodes are missing. J Zool 313, 18–36 (2021).
Stibal, M. et al. Microbial abundance in surface ice on the Greenland Ice Sheet. Front Microbiol 6, 1–12 (2015).
Williamson, C. J. et al. Macro-Nutrient Stoichiometry of Glacier Algae From the Southwestern Margin of the Greenland Ice Sheet. Front Plant Sci 12, 1–8 (2021).
Lutz, S., Anesio, A. M., Edwards, A. & Benning, L. G. Linking microbial diversity and functionality of arctic glacial surface habitats. Environ Microbiol 19, 551–565 (2017).
Redfield, A. C. The biological control of chemical factors in the environment. American Scientist vol. 46 230 Preprint at https://doi.org/10.2307/27827150 (1958).
Andrews, M. G., Jacobson, A. D., Osburn, M. R. & Flynn, T. M. Dissolved Carbon Dynamics in Meltwaters From the Russell Glacier, Greenland Ice Sheet. J Geophys Res Biogeosci 123, 2922–2940 (2018).
Williamson, C. J. et al. Ice algal bloom development on the surface of the Greenland Ice Sheet. FEMS Microbiol Ecol 94, 1–10 (2018).
Klawonn, I. et al. Untangling hidden nutrient dynamics: rapid ammonium cycling and single-cell ammonium assimilation in marine plankton communities. ISME Journal 13, 1960–1974 (2019).
Olofsson, M. et al. Nitrate and ammonium fluxes to diatoms and dinoflagellates at a single cell level in mixed field communities in the sea. Sci Rep 9, 1–12 (2019).
Holland, A. T., Williamson, C. J., Tedstone, A. J., Anesio, A. M. & Tranter, M. Dissolved Nitrogen Speciation and Concentration During Spring Thaw in the Greenland Ice Sheet Dark Zone: Evidence for Microbial Activity. Front Earth Sci (Lausanne) 10, 1–12 (2022).
Telling, J. et al. Microbial nitrogen cycling on the Greenland Ice Sheet. Biogeosciences 9, 2431–2442 (2012).
Keiding, K. & Heidam, N. Z. Observations on acidity and ions in East Greenland precipitation. Tellus B: Chemical and Physical Meteorology 38, 345–352 (1986).
Davidson, C. I., Chu, L., Grimm, T. C., Nasta, M. A. & Qamoos, M. P. Wet and dry deposition of trace elements onto the Greenland ice sheet. Atmospheric Environment (1967) 15, 1429–1437 (1981).
Hodson, A. et al. Glacial ecosystems. Concepts & Synthesis 78, 1920–1931 (2008).
Jensen, M. B. et al. The dark art of cultivating glacier ice algae. Bot Lett (2023) doi:10.1080/23818107.2023.2248235.
Lindemann, C., Fiksen, Ø., Andersen, K. H. & Aksnes, D. L. Scaling laws in phytoplankton nutrient uptake affinity. Front Mar Sci 3, 26 (2016).
Olofsson, M., Power, M. E., Stahl, D. A., Vadeboncoeur, Y. & Brett, M. T. Cryptic constituents: The paradox of high flux–low concentration components of aquatic ecosystems. Water (Switzerland) 13, (2021).
Rao, N. N., Gómez-García, M. R. & Kornberg, A. Inorganic Polyphosphate: Essential for Growth and Survival. http://dx.doi.org/10.1146/annurev.biochem.77.083007.093039 78, 605–647 (2009).
Sanz-Luque, E., Bhaya, D. & Grossman, A. R. Polyphosphate: A Multifunctional Metabolite in Cyanobacteria and Algae. Front Plant Sci 11, 938 (2020).
Karl, D. M. & Björkman, K. M. Dynamics of DOP. Biogeochemistry of Marine Dissolved Organic Matter 249–366 (2002) doi:10.1016/b978-012323841-2/50008-7.
Kornberg, A. Inorganic polyphosphate: Toward making a forgotten polymer unforgettable. Journal of Bacteriology vol. 177 491–496 Preprint at https://doi.org/10.1128/jb.177.3.491-496.1995 (1995).
Cliff, A. et al. Polyphosphate synthesis is an evolutionarily ancient phosphorus storage strategy in microalgae. Algal Res 73, 103161 (2023).
Barcytė, D., Pilátová, J., Mojzeš, P. & Nedbalová, L. The Arctic Cylindrocystis (Zygnematophyceae, Streptophyta) Green Algae are Genetically and Morphologically Diverse and Exhibit Effective Accumulation of Polyphosphate. J Phycol 56, 217–232 (2020).
Lutz, S., Anesio, A. M., Edwards, A. & Benning, L. G. Linking microbial diversity and functionality of arctic glacial surface habitats. Environ Microbiol 19, 551–565 (2017).
Klawonn, I. et al. Characterizing the “fungal shunt”: Parasitic fungi on diatoms affect carbon flow and bacterial communities in aquatic microbial food webs. Proc Natl Acad Sci U S A 118, 1–11 (2021).
Olofsson, M. et al. High single-cell diversity in carbon and nitrogen assimilations by a chain-forming diatom across a century. Environ Microbiol 21, 142–151 (2019).
Polerecky, L. et al. Calculation and Interpretation of Substrate Assimilation Rates in Microbial Cells Based on Isotopic Composition Data Obtained by nanoSIMS. Front Microbiol 12, 3657 (2021).
Halbach, L. Nutrient requirements and pigment signatures of glacial algae on the Greenland Ice Sheet. (Aarhus University, 2022).
Williamson, C. J. et al. Macro-Nutrient Stoichiometry of Glacier Algae From the Southwestern Margin of the Greenland Ice Sheet. Front Plant Sci 12, 1–8 (2021).
Cooper, G. A., Liu, M., Peña, J. & West, S. A. The evolution of mechanisms to produce phenotypic heterogeneity in microorganisms. Nature Communications 2022 13:1 13, 1–13 (2022).
Kobayashi, K., Takeuchi, N. & Kagami, M. Distribution of parasitic chytrids of glacier algae in Alaska; Cryoconite holes as a hotspot of chytrid infection. (2022) doi:10.21203/RS.3.RS-2189377/V1.
Perini, L. et al. Interactions of Fungi and Algae from the Greenland Ice Sheet. Microb Ecol (2022) doi:10.1007/s00248-022-02033-5.
Perini, L. et al. Darkening of the Greenland Ice Sheet: Fungal Abundance and Diversity Are Associated With Algal Bloom. Front Microbiol 10, 557 (2019).
Telling, J. et al. Nitrogen fixation on Arctic glaciers, Svalbard. J Geophys Res Biogeosci 116, 2–9 (2011).
Fausto, R. S. et al. Programme for Monitoring of the Greenland Ice Sheet (PROMICE) automatic weather station data. Earth Syst Sci Data 13, 3819–3845 (2021).
Rodriguez, E. S. et al. Capillary ion chromatography with on-column focusing for ultra-trace analysis of methanesulfonate and inorganic anions in limited volume Antarctic ice core samples. J Chromatogr A 1409, 182–188 (2015).
Torres, M. E., Mix, A. C. & Rugh, W. D. Precise δ13C analysis of dissolved inorganic carbon in naturalwaters using automated headspace sampling and continuous-flowmass spectrometry. Limnol Oceanogr Methods 3, 349–360 (2005).
Heldal, M., Norland, S. & Tumyr, O. X-ray microanalytic method for measurement of dry matter and elemental content of individual bacteria. Appl Environ Microbiol 50, 1251–1257 (1985).
Khachikyan, A. et al. Direct Cell Mass Measurements Expand the Role of Small Microorganisms in Nature. Appl Environ Microbiol 85, 1–1 (2019).
Khan, A. L., Dierssen, H. M., Scambos, T. A., Höfer, J. & Cordero, R. R. Spectral characterization , radiative forcing and pigment content of coastal Antarctic snow algae : approaches to spectrally discriminate red and green communities and their impact on snowmelt. 133–148 (2021).
Martínez-Pérez, C. et al. The small unicellular diazotrophic symbiont, UCYN-A, is a key player in the marine nitrogen cycle. (2016) doi:10.1038/NMICROBIOL.2016.163.
Meador, T. B. et al. Carbon recycling efficiency and phosphate turnover by marine nitrifying archaea. Sci Adv 6, (2020).
Schoffelen, N. J. et al. Single-cell imaging of phosphorus uptake shows that key harmful algae rely on different phosphorus sources for growth. Sci Rep 8, 1–13 (2018).
Procházková, L., Řezanka, T., Nedbalová, L. & Remias, D. Unicellular versus filamentous: The glacial alga ancylonema alaskana comb. et stat. nov. and its ecophysiological relatedness to ancylonema nordenskioeldii (zygnematophyceae, streptophyta). Microorganisms 9, (2021).
Woebken, D. et al. Revisiting N2 fixation in Guerrero Negro intertidal microbial mats with a functional single-cell approach. The ISME Journal 2015 9:2 9, 485–496 (2014).
Meyer, N. R., Fortney, J. L. & Dekas, A. E. NanoSIMS sample preparation decreases isotope enrichment: magnitude, variability and implications for single-cell rates of microbial activity. Environ Microbiol 23, 81–98 (2021).
Musat, N., Musat, F., Weber, P. K. & Pett-Ridge, J. Tracking microbial interactions with NanoSIMS. Curr Opin Biotechnol 41, 114–121 (2016).
Harding, K. J. et al. Cell-specific measurements show nitrogen fixation by particle-attached putative non-cyanobacterial diazotrophs in the North Pacific Subtropical Gyre. Nat Commun 13, 1–10 (2022).
Hillebrand, H., Dürselen, C. D., Kirschtel, D., Pollingher, U. & Zohary, T. Biovolume calculation for pelagic and benthic microalgae. J Phycol 35, 403–424 (1999).
Chevrollier, L.-A. et al. Light absorption and albedo reduction by pigmented microalgae on snow and ice. Journal of Glaciology 1–9 (2022) doi:10.1017/jog.2022.64.
Healey, F. P. Physiological indicators of nutrient deficiency in algae. Fish. Mar. Serv. Res. Dev. Tech. Rep. 585, (1975).
Montoya, J. P., Voss, M., Kahler, P. & Capone, D. G. A Simple , High-Precision , High-Sensitivity Tracer Assay for N ( inf2 ) Fixation . These include : A Simple , High-Precision , High-Sensitivity Tracer Assay for N 2 Fixation. Appl Environ Microbiol 62, 986–993 (1996).
Kitzinger, K. et al. Single cell analyses reveal contrasting life strategies of the two main nitrifiers in the ocean. Nat Commun 11, (2020).
Martínez-Pérez, C. et al. The small unicellular diazotrophic symbiont, UCYN-A, is a key player in the marine nitrogen cycle. (2016) doi:10.1038/NMICROBIOL.2016.163.
Klawonn, I. et al. Untangling hidden nutrient dynamics: rapid ammonium cycling and single-cell ammonium assimilation in marine plankton communities. ISME Journal 13, 1960–1974 (2019).
US Environmental Protection Agency. Definition and Procedure for the Determination of the Method Detection Limit—Revision 1.11. EPA 821-R-16-006 (2016).
Herlemann, D. P. R. et al. Transitions in bacterial communities along the 2000 km salinity gradient of the Baltic Sea. The ISME Journal 2011 5:10 5, 1571–1579 (2011).
Cheung, M. K., Au, C. H., Chu, K. H., Kwan, H. S. & Wong, C. K. Composition and genetic diversity of picoeukaryotes in subtropical coastal waters as revealed by 454 pyrosequencing. ISME J 4, 1053–1059 (2010).
Team, Rs. RStudio: Integrated Development for R. Preprint at (2020).
Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods 13, 581–583 (2016).
Quast, C. et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res 41, (2013).
Trivedi, C. B. et al. DNA/RNA Preservation in Glacial Snow and Ice Samples. Front Microbiol 13, (2022).
Mcmurdie, P. J. & Holmes, S. phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data. doi:10.1371/journal.pone.0061217.
Wickham H. Ggplot2: Elegant Graphics for Data Analysis. (Springer-Verlag, New York, 2016). doi:https://ggplot2.tidyverse.org.

There is NO Competing Interest.

SIHalbachetal.docx

Single-cell imaging reveals efficient nutrient uptake and growth of microalgae that darken the Greenland Ice Sheet

Status:

Version 1

Abstract

Figures

Introduction

Results

Physico-chemical conditions on the Greenland Ice Sheet

Supraglacial community composition

In situ glacier ice algae C:N:P ratios

C-fixation, N-assimilation, and the effect of nutrient additions in bulk samples and single algal cells

Discussion

Conclusions

Methods