Relationship between dissolved organic matter quality and microbial community composition across polar glacial environments

Relationship between dissolved organic matter quality and microbial community composition across... ABSTRACT Vast expanses of Earth’s surface are covered by ice, with microorganisms in these systems affecting local and global biogeochemical cycles. We examined microbial assemblages from habitats fed by glacial meltwater within the McMurdo Dry Valleys, Antarctica and on the west Greenland Ice Sheet (GrIS), evaluating potential physicochemical factors explaining trends in community structure. Microbial assemblages present in the different Antarctic dry valley habitats were dominated by Sphingobacteria andFlavobacteria, while Gammaproteobacteria and Sphingobacteria prevailed in west GrIS supraglacial environments. Microbial assemblages clustered by location (Canada Glacier, Cotton Glacier and west GrIS) and were separated by habitat type (i.e. ice, cryoconite holes, supraglacial lakes, sediment and stream water). Community dissimilarities were strongly correlated with dissolved organic matter (DOM) quality. Microbial meltwater assemblages were most closely associated with different protein-like components of the DOM pool. Microbes in environments with mineral particles (i.e. stream sediments and cryoconite holes) were linked to DOM containing more humic-like fluorescence. Our results demonstrate the establishment of distinct microbial communities within ephemeral glacial meltwater habitats, with DOM-microbe interactions playing an integral role in shaping communities on local and polar spatial scales. Biodiversity, Antarctica, Greenland, microbial, dissolved organic matter INTRODUCTION Glaciers and ice sheets represent a distinct biome, dominated by diverse and biologically active microorganisms (Anesio and Laybourn-Parry 2012; Boetius et al.2015; Anesio et al.2017), that play a pivotalrole in regional and global biogeochemical cycles. These icy environments are subject to relatively similar environmental conditions, including low temperature, frequent freeze-thaw cycles, seasonally low water availability, desiccation and nutrient limitation, but vary in the degree of surface melt and geochemical composition. As such, the presence of distinct microbial communities in environmentally similar glacial meltwater habitats (Cameron et al.2012, 2015, 2016; Stibal et al.2015; Lutz et al.2016; Uetake et al.2016) offers intriguing questions about the factors influencing trends in community structure and function and the impacts of microorganisms on geochemical processes. While regional differences across glacial meltwater communities have been frequently attributed to local aeolian deposition, catchment geology, and physical and geographical variables (Cameron et al.2015, 2016; Cook et al.2015; Musilova et al.2015; Stibal et al.2015; Lutz et al.2016), these factors may, indeed, have only a limited effect on community composition. Findings from a growing number of investigations favor the development of habitat-specific ice surface communities, particularly those adapted to specific ecological niches (Mueller and Pollard 2004; Edwards et al.2013; Musilova et al.2015; Gawor et al.2016; Franzetti et al.2017). Dissolved organic matter (DOM) on ice sheets originates from a variety of allochthonous (i.e. terrestrial, marine or black carbon) and autochthonous (i.e. in situ production) sources (Stibal et al.2010; Stubbins et al.2012; Legrand et al.2013; Antony et al.2014; Hood et al.2015; Smith et al.2017), both of which have been shown to vary in reactivity and may be highly bioavailable to resident microbial communities (Hood et al.2009; Antony et al.2017; Smith et al.2017). Many studies have targeted DOM dynamics on the surface of ice sheets (Anesio et al.2009; Hodson et al.2010; Cook et al.2012; Stibal et al.2012; Hood et al.2015), and both deposition and in situ production have been linked to the biogeochemical evolution of DOM in supraglacial waters (Hood et al.2009; Stubbins et al.2012; Telling et al.2012; Antony et al.2014; Musilova et al.2017; Smith et al.2017). Rather than being passive receivers of allochthonous DOM, recent studies indicate that there are tight interactions between DOM production and composition within bacterial communities in glacial surface habitats (Smith et al.2016, 2017; Musilova et al.2017). Additionally, changes in microbial community structure have been correlated with dissolved organic carbon (DOC) concentrations and the presence of distinct algal populations in Arctic snow and on glaciers, further attesting to the importance of interactions between organic carbon and microbial communities (Lutz et al.2016, 2017). In fact, major components of glacial carbon dynamics are functionally linked to trophic interactions between photo- and heterotrophs during algal blooms, and within cryoconite granules and biofilms (Edward et al.2011; Takeuchi et al.2014; Anesio et al.2017; Lutz et al.2017; Musilova et al.2017; Smith et al.2017). However, the mechanism by which the composition of DOM influences bacterial utilization and ultimately community composition is largely unknown. Globally, glacial meltwater habitats are dominated by cyanobacteria,Proteobacteria, Actinobacteria, Bacteriodetes, Acidobacteria, Chloroflexi and Planctomycetes, with regional differences at lower levels of taxonomic resolution (Boetius et al.2015). Given the habitat-specific selective pressures of cryo-environments, we hypothesize that bacterial assemblages are shaped by geochemistry, which provides the available energy sources necessary to support life. To test this scenario, glacial meltwater habitats from the Earth’s polar regions were investigated. Meltwater samples were collected from two glaciers in the McMurdo Dry Valleys, Antarctica, (stream water, stream sediments, glacial ice, snow, and cryoconite holes) and from the western margin of the Greenland Ice Sheet (GrIS) (supraglacial lakes and cryoconite holes). Illumina MiSeq 16S rRNA gene amplicons were analyzed to assess differences in community structure across and within these environments. Geochemical parameters, including major ions, nutrients, DOC concentration, and fluorescing DOM characteristics were evaluated to infer their influence on microbial community composition. The primary objective of this study was to link trends in bacterial biodiversity to DOM quality across glacial systems, thus enabling better predictions regarding their function in the context of a changing environment. METHODS Field site description Sampling was conducted in the McMurdo Dry Valleys, Antarctica and at the western margin of the GrIS (Fig. 1). The McMurdo Dry Valleys of Antarctica comprise the largest ice-free area of the Antarctic continent and are classified as polar deserts with <10 cm of precipitation per year. Mean annual temperatures range between −15°C and −30°C. This barren region experiences katabatic winds originating in the continental interior, which are important in controlling local climate, physical weathering and material transport. Contrary to expectation, contributions by marine aerosols are minor (Bottos et al.2014; Bowman and Deming 2017). The two Antarctic sampling sites were comprised of glacially fed Antarctic stream ecosystems ∼70 km apart: (i) The Cotton Glacier (CG) supraglacial stream (77°07’S, 161°40’E) is ∼16 km long, consisting of a network of braided channels cut through the ice, ultimately terminating into McMurdo Sound. The CG stream receives a large amount of sedimentary deposits of unknown origin (fluvial, glacial or aeolian) from surrounding areas (Foreman et al.2013). Sediments are primarily located on the lateral margins of the stream, and herein referred to as parafluvial sediments. CG stream water (n = 10), parafluvial sediments (n = 11), snow (n = 1) and surrounding ice (n = 5) were sampled during austral summer field seasons 2009–2012. (ii) The Canada Stream (77°61’S, 162°59’E) represents a proglacial stream in the Taylor Valley and flows over a constant layer of coarse sediments. Samples were collected in 2009–2011 and included: stream water (n = 5), stream sediments (n = 3), as well as samples from its meltwater source, the adjacent Canada Glacier [glacial ice (n = 4), snow (n = 1) and cryoconite holes (n = 4)]. Figure 1. View largeDownload slide Sampling locations in (A) the western margin of the Greenland Ice Sheet and (B) the McMurdo Dry Valleys, Antarctica. Figure 1. View largeDownload slide Sampling locations in (A) the western margin of the Greenland Ice Sheet and (B) the McMurdo Dry Valleys, Antarctica. Supraglacial features in west Greenland were collected from the Paakitsoq region northeast of Jakobshavn Isbrae (69°34′N, 49°48′W). This location is ∼15 km from the ice sheet margin with an ice elevation of ∼962 m above sea level (Banwell et al.2012). Samples were collected during summer 2011, including cryoconite hole (n = 9) and supraglacial lake samples (n = 5) from 'Lake Half Moon' and 'Lake Ponting' (Tedesco et al.2013). More information on sampling regime is provided below and in Table S1 (Supporting Information). Sample collection Temperature, pH, conductivity, dissolved oxygen, and total dissolved solids in the two Antarctic streams were measured on site using a Manta Sub2 probe (Eureka Water Probes Inc., U.S.A.). A malfunction of the probe obstructed in situ measurements in west GrIS supraglacial environments. Antarctic stream water samples (4 L) for geochemical analyses were collected in acid-washed, deionized water-rinsed (6×) fluorinated carboys. Samples for carbon analyses were collected from all sites in combusted (450°C for 5 hr) 125 mL amber bottles. Sediment (20–30 g) and freshly fallen snow samples were aseptically collected with sterile spatulas. Ice cores (diameter: 10 cm and length: 100 cm) and frozen cryoconite holes were removed with an ethanol-cleaned ice corer (Kovacs, U.S.A.). Samples were stored in sterile cryovials or whirl-pack bags and processed/analyzed within 12 hr post-collection. Cored samples were decontaminated by removing ∼2 mm of outer ice with sterile blades using a modification of (Christner et al.2005). Snow samples were melted at 4°C. Meltwater samples (0.5–1 L) for microbial community analysis from glacial streams, and ice and snow melt were filtered onto sterile 0.2 µm, 47 mm Supor®-200 membranes (Pall Corporation, U.S.A.) under low pressure (<7 psi). Filters, submerged in TES buffer (100 mM Tris, 100 mM EDTA, and 2% SDS), and sediments were flash frozen in liquid nitrogen, and stored at −80°C until DNA extraction. Samples for microbial community analysis from supraglacial lakes on the west GrIS were collected on polyethersulfone 0.2 mm Sterivex filters (MilliporeSigma, U.S.A.) with a hand pump. Between 0.5–1 L of sample water was passed through the filters. Sterivex cartridges were filled with RNAlater® (Invitrogen, U.S.A.) and sealed with Luer–Lok caps. Filters were frozen at −20°C and stored on ice for transport. Upon arrival at Montana State University, samples were stored at −80°C until further processing. Water chemistry Water samples for non-purgeable DOC, macronutrients, anions/cations, particulate organic carbon and particulate organic nitrogen were collected and analyzed following the protocols of the McMurdo Dry Valleys LTER (Foreman et al.2013). Geochemical data were statistically analyzed to identify differences across samples, using a general linear model of analysis of variance (ANOVA) in version 17, Minitab Inc., U.S.A. Fluorescence spectroscopy DOM is comprised of a heterogeneous mixture of organic compounds, with the analysis of fluorescent components proven to be useful in assessing its chemical quality (Cory and McKnight 2005; Fellman et al.2010). Samples were filtered through combusted GF/F filters into acid-washed, pre-combusted amber bottles. Samples were analyzed for UV-absorbance (190–1100 nm) using an Agilent 8453 UV-spectrophotometer in a 1 cm quartz cuvette. Samples with absorbance >0.3 at 254 nm were diluted with Milli-Q water to prevent inner-filter effects (Miller and McKnight 2010). Excitation emission matrices (EEMs) were collected with a Horiba Fluoromax-4 Fluorometer, equipped with a Xenon light source and a 1 cm pathlength quartz cuvette (Foreman et al.2013). Post-processing was completed in MATLAB to generate EEMs corrected for Raman scattering and blank water subtraction (McKnight et al.2001; Lawaetz and Stedmon 2009). Due to large difference in characteristics of the fluorescent fraction of DOM between the Antarctic and Arctic samples (e.g. dominance of B1 fluorescence in the Greenland samples) and the lack of samples with intermediate characteristics, it was not possible to validate a PARAFAC model (Bro 1997) for this dataset. In light of these restrictions we quantified the EEMs spectra based on characteristic fluorescence peaks (Coble 1996; Fellmanet al.2010). To statistically compare differences in DOM composition from EEMs data, the relative proportion of individual fluorophores (B1, B2, T1, T2, A and C) was calculated for each sample. Fluorescence intensity (F.I.) values were normalized across different sample locations and summed based on DOM classifications (Coble 1996) (Table S2, Supporting Information). Significant differences between the means of specific DOM fluorophores were determined with an ANOVA general linear model followed by Tukey’s multiple-comparison test (version 17, Minitab Inc., U.S.A.). DNA extraction and PCR amplification The PowerSoil DNA or PowerWater Sterivex kits (Qiagen, U.S.A.) were used to extract genomic DNA from sediment and filters following manufacturer recommendations. Extracted genomic DNA was quantified using the Qubit DNA Kit (Invitrogen, U.S.A.). The V3/V4 regions of the 16S rRNA gene were amplified in 50 µl PCR reactions containing 100–200 pg of extracted genomic DNA, 0.1 µM of each primer, a 1X final concentration of Bulls Eye PREMIUM Taq 2X Mix (Midwest Scientific, U.S.A.) and nuclease free water. The primer design consisted of Illumina adaptor sequences followed by either the universal 341F 5′-acactctttccctacacgacgctcttccgatctCCTACGGGNGGCWGCAG-3′ or 805R 5′-gtgactggagttcagacgtgtgctcttccgatctGACTACHVGGGTATCTAATCC-3′. PCR was performed in an Eppendorf Mastercycler pro S. The amplification protocol consisted of an initial denaturation at 95°C for 3 min, followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, extension at 72°C for 30 s and a final extension step at 72°C for 5 min. Negative PCR controls without DNA template and DNA extractions from filter- and extraction-blanks were amplified to monitor contamination (n = 4) (Salter et al.2014; Nguyen et al.2015). Sequencing and analysis Paired end, 250 bp sequencing was performed on an Illumina MiSeq at the University of Wisconsin-Madison Biotechnology Center. Forward and reverse sequences were joined with the Quantitative Insights Into Microbial Ecology toolkit version 1.9.0 (Caporaso et al.2010). Contigs were subsequently analyzed with the Mothur platform v.1.34.4 (Schloss et al.2009). For sequence quality refinement, sequences containing ambiguous bases, homopolymers > 8 bases and an average quality score below 30 over a 50 bp window were excluded from further analysis (Schloss et al.2011). The maximum sequence length was 467 bp, and sequences shorter than 450 bp were removed. Processed sequences were aligned against the SILVA Gold database in Mothur. Chimeric sequences were removed using UCHIME (Edgar et al.2011) with the SILVA Gold database (Pruesse et al.2007) and a second chimera check using sequences from the present study as a database. Sequences were classified with a Bayesian method (Wang et al.2007) using the Mothur formatted version of the RDP classifier. Operational taxonomic units (OTUs) were defined at ≥97% 16S rRNA sequence identity. Any OTU found in the blank samples were removed from corresponding sequence libraries. Sequences were deposited in the NCBI SRA database under the accession number SRP076570. Statistical analysis of sequencing samples Precipitation events are rare in polar deserts; thus, with only single snow samples collected, these were excluded from subsequent statistical analysis. OTU richness using Chao1 (Chao 1984), Inverse Simpson (Simpson 1949) and the non-parametric multivariate analysis of OTU similarities (ANOSIM; Clarke and Warwick 2001) approach were calculated within Mothur. ANOSIM produces the test statistic R. Large R values (close to 1) indicate that two samples are highly dissimilar, while small R values (close to zero) signify little or no dissimilarity. The most dominant 200 OTUs (89% of the entire sequence library) were subjected to similarity percentage (SIMPER) analysis using PAST version 3.10 (Hammer et al.2001). A canonical correspondence analysis (CCA) was performed on square root transformed OTU profiles within Canoco v4.5 (Microcomputer Power Inc., U.S.A.). Within Canoco, detrended CCA was used to obtain the length of the gradient among OTUs. A gradient length > 4 supported a unimodal method for comparison between all ecosystem components. Differences in microbial communities among sites were calculated with non-metric multidimensional scaling (MDS) ordination, using a Bray Curtis dissimilarity matrix created from OTU tables and refined by the Kruskal’s Stress test. RESULTS Stream water biogeochemistry Both Antarctic streams had low ionic strength (mean 25.2 µS cm−1) and near neutral pH (mean 6.8). Of importance were the significantly higher concentrations of NH4+ (5.1± 0.7µg L−1; ANOVA, DF = 14, F = 44.68, P < 0.001), NO3− (50.5± 24.0µg L−1; ANOVA, DF = 14, F = 13.59, P = 0.004), PO42− (1.6± 0.2µg L−1; ANOVA, DF = 14, F = 9.32, P = 0.012) and particulate phosphorous (10.6± 7.8µg L−1; ANOVA, DF = 14, F = 9.98, P = 0.003) within the CG stream compared to the Canada Stream (Table S3, Supporting Information). Fluorescence in the humic-like regions of the EEMs was apparent in all samples; however, only the fluorescence spectra of Canada Stream water DOM resolved in distinct humic fluorophores. DOM from all other Antarctic and west GrIS environments lacked these discrete regions of humic fluorophores and were characterized by protein-like fluorescence (ANOVA, DF = 55, F = 40.44, P < 0.001) (Fig. 2; Table S4, Supporting Information). No significant difference between any of the Antarctic environments was found in the T1 region (ANOVA, DF = 55, F = 3.53, P > 0.216). DOM in cryo-habitats from the west GrIS lacked humic-like fluorophores (A and C peaks) and showed significantly different protein-like peaks B1 (ANOVA, DF = 55, F = 16.71, P < 0.001) and B2 (ANOVA, DF = 55, F = 18.04, P < 0.020) when compared to all Antarctic environments. Figure 2. View largeDownload slide Change in the relative abundance of fluorescent intensity (F.I.) of fluorescent regions commonly observed in natural waters (Coble 1996) for different environmental locations. Figure 2. View largeDownload slide Change in the relative abundance of fluorescent intensity (F.I.) of fluorescent regions commonly observed in natural waters (Coble 1996) for different environmental locations. Microbial assemblages and biodiversity Raw Illumina MiSeq sequences consisted of 76,026± 15,449 (average ± S.D.) reads per sample. Quality refined sequence libraries contained on average 8,913± 5,640 reads, clustered into 2,156 OTUs across all samples. On average, all samples from Antarctic environments were dominated by Sphingobacteria (48%± 6%), Flavobacteria (20%± 11%), Betaproteobacteria (8% ± 3%) and Gammaproteobacteria (4%± 2%). Supraglacial aquatic habitats from the west GrIS were primarily dominated by Gammaproteobacteria (66%± 11%),Sphingobacteria (24%± 8%) and Betaproteobacteria (3%± 2%) (Fig. 3). CG snow was dominated by Gammaproteobacteria (82%), which differed from the dominance of Betaproteobacteria (46%) and Bacilli (38%) in Canada snow. The observed number of OTUs across all sampled habitats showed similar patterns when compared to the predicted alpha diversity indices (Fig. S1, Supporting Information). Based on Chao richness, the greatest diversity was found in the Canada Stream water (524.0–535.6; 95% CI), followed by the CG stream water (395.3–411.4; 95% CI). Species richness for Antarctic stream sediments, ice and cryoconite hole communities was between 116.4 and 258.6 (95% CI). The lowest species richness was calculated for west GrIS supraglacial lakes (54.8–64.0; 95% CI) and cryoconite holes (52.9–62.2; 95% CI), and confirmed by the Inverse Simpson index (2.7–3.2; 95% CI). In the different geographic locations (CG, Canada Glacier, and the west GrIS), 43.5%, 53.1% and 46.9% of the OTUs were unique. Antarctic environments shared 29.3% of the OTUs while only 3.1% of the OTUs were present across all glacial settings. MDS of Bray–Curtis indices of the OTU profiles clustered the three glacial locations into distinct clades (Fig. 4). Within each clade, sample types clustered into their own representative group, with the exception of one CG sediment sample that grouped more closely with the Canada Glacier samples. Overlap was found between CG ice and CG stream samples. Figure 3. View largeDownload slide Relative abundance of MiSeq 16S rRNA gene sequences representing the distribution of microbial assemblages with individual samples grouped together based on sampling location. Percentages are reported at the class taxonomic level. Figure 3. View largeDownload slide Relative abundance of MiSeq 16S rRNA gene sequences representing the distribution of microbial assemblages with individual samples grouped together based on sampling location. Percentages are reported at the class taxonomic level. Figure 4. View largeDownload slide Multidimensional scaling plot of Bray–Curtis indices on the relative abundance of OTU (97% sequence similarity) profiles for different environmental locations. Dashed lines delineate the different sampled environmental locations (Canada Glacier, Cotton Glacier and Greenland Ice Sheet). Figure 4. View largeDownload slide Multidimensional scaling plot of Bray–Curtis indices on the relative abundance of OTU (97% sequence similarity) profiles for different environmental locations. Dashed lines delineate the different sampled environmental locations (Canada Glacier, Cotton Glacier and Greenland Ice Sheet). OTU specific differences between environments from the same location and the same environment from different locations were determined by ANOSIM (Table S6, Supporting Information). Overall, differences were evident across sites (R = 0.722, P < 0.001) and sample type (R = 0.530, P < 0.001), with statistical evidence for OTU dissimilarities found for 11 out of 14 pairwise comparisons (R ≥ 0.559, P ≤ 0.035). Only microbial assemblages from Canada Glacier cryoconite holes and Canada Glacier ice (R = 0.198, P = 0.081) and west GrIS cryoconite hole and lake habitats (R = 0.112, P = 0.156) were found to be more similar. Moderate community overlap was also determined between CG ice and the CG stream (R = 0.410, P = 0.011). Markedly, OTU profiles found in similar environments (i.e. ice, stream, stream sediments, and cryoconite holes) from different locations were highly dissimilar (R ≥ 0.559 and P ≤ 0.012). With a focus on the 200 most abundant OTUs in each sample, dissimilarities were largely attributed to four families (i.e. Cytophagaceae, Flavobacteriaceae, Chitinophagaceae, and Sphingobacteriaceae) within the phylum Bacteroidetes (SIMPER; Table S5, Supporting Information). Combined with the unclassified microbial realm (7.7%–13.9%), these groups accounted for 82.8% of the dissimilarities between microbial assemblages in ice from CG and Canada Glaciers, 72.1% between microbes associated with sediments in the CG and Canada streams, and 87.7% between bacterioplankton in the CG and Canada streams. These four families also explained 50.9% of the dissimilarities between bacterial assemblages inhabiting cryoconite holes on the Canada Glacier and the west GrIS, with Pseudomonadaceae contributing another 15.2%. Relationship of environmental variables to microbial biodiversity An initial CCA between CG stream and Canada Stream water OTU profiles and the physicochemical dataset was performed to determine whether environmental variables had an effect on microbial assemblage composition. However, variables tested (temperature, pH, dissolved oxygen, conductivity, major cations and anions, and DOC concentrations) were statistically insignificant in explaining assemblage composition. To determine the effect of DOM quality on microbial assemblages in glacial environments, CCA analysis included the F.I. for different DOM fluorophores and OTU profiles (Fig. 5). The established model explained 86.9% (P = 0.004) of the total variation across all OTU profiles, with 77.9% of the total model variance described by the first three axes (all eigenvalues >0.16). The first axis explained 51.2% of the variance (P = 0.002). Microbial assemblages from the west GrIS clustered closely together and were best explained by the protein-like component B1. Canada Stream microbial assemblages were clearly described by the presence of humic-like fractions of DOM. Phylotype composition from Canada Stream sediment, Canada Glacier cryoconite holes, and CG sediment samples were moderately influenced by humic- and protein-like DOM constituents. Based on the model, OTUs found in the CG stream, CG ice and Canada Glacier ice were best explained by protein-like B2 and T2 fluorophores. Figure 5. View largeDownload slide Canonical correspondence analysis of the influence of fluorescent OM variables on the relative abundance of square root transformed OTU (97% sequence similarity) profiles for sampled glacial environments. Different colored circles represent relative fluorescence intensities for fluorophores commonly observed in natural waters (Coble 1996) from each sample. Fluorescence from protein-like OM components are represented as B1 and B2 (tyrosine-like) and T1 and T2 (tryptophan-like). Fluorescent humic-like OM components are represented by A and C. Figure 5. View largeDownload slide Canonical correspondence analysis of the influence of fluorescent OM variables on the relative abundance of square root transformed OTU (97% sequence similarity) profiles for sampled glacial environments. Different colored circles represent relative fluorescence intensities for fluorophores commonly observed in natural waters (Coble 1996) from each sample. Fluorescence from protein-like OM components are represented as B1 and B2 (tyrosine-like) and T1 and T2 (tryptophan-like). Fluorescent humic-like OM components are represented by A and C. DISCUSSION Phyla identified across all samples in this study were typical of glacial ecosystems (Boetius et al.2015). Individual microbial assemblages clustered based on location (i.e. CG, Canada Glacier, and west GrIS) and habitat type (i.e. stream water, sediments, ice, cryoconite holes, and supraglacial lakes), with pronounced dissimilarities apparent between the same habitat types from different locations (Fig. 4; Table S6, Supporting Information). The lack of significant differences between ice and cryoconite hole communities from the Canada Glacier (ANOSIM; R=0.198, P = 0.081) and the moderate overlap between ice and stream microbial consortia from the CG (ANOSIM; R=0.410, P = 0.011) indicate that ice-trapped microbes contributed to the evolution of glacial meltwater communities (Musilova et al.2015; Cameron et al.2016). Following this trajectory, we found that meltwater selected for more habitat-specific microbial assemblages (Table S5, Supporting Information; Musilova et al.2015). Even without an active land bridge, polar ice sheets are under constant exchange with both near and distant habitats (Pearce et al.2009; Price et al.2009; Bottos et al.2014; Bowman and Deming 2017), introducing organisms that must establish themselves among existing communities in order to survive. Nonetheless, dominant microbial phylotypes in snow on polar ice sheets (i.e. Firmicutes, β- and γ-Proteobacteria; this study Fig. 3; Boetius et al.2015; Cameron et al.2015; Musilova et al.2015; Antony et al.2016) do not prevail in glacial ice surface communities. In fact, decreased meltwater dynamics within ice-margin habitats promotes microbial community establishment (Fountain et al.2004; Musilova et al.2015; Stibal et al.2015; Uetake et al.2016), dominated by taxa that were originally rare members in aeolian deposits (i.e. Bacteroidetes, Actinobacteria and Proteobacteria) (Edwards et al.2013; Hell et al.2013; Musilova et al.2015; Stibal et al.2015; Cameron et al.2016; Franzetti et al.2017). Relevant to the discussion on community establishment and stability in glacial meltwater habitats, our study provides evidence that microbial assemblages from samples within the same local environment exhibit a high degree of dissimilarity in compartments along the hydrologic flow path (i.e. ice, stream, stream sediments, and cryoconite holes). Differences in the fluorescent signatures of DOM between glaciers or habitats on the same glacier have been attributed to deposition and the microbial loop (Barker et al.2006; Pautler et al.2013; Musilova et al.2017). Similar to previous reports of glacial meltwater habitats (i.e. supraglacial ponds and streams, cryoconite holes, ice, and subglacial outflow, Bhatia et al.2010; Dubnick et al.2010; Pautler et al.2012, 2013; Barker et al.2013; Foreman et al.2013; Feng et al.2016; D’Andrilli et al., 2017; Smith et al. 2018), DOM fluorescence in the present study was dominated by protein-like components, indicative of microbially produced DOM. Despite the predominance of protein-like components, there was little to no overlap in the overall DOM composition across investigated habitats. In contrast to the terrestrial Canada Stream, the CG supraglacial stream was largely devoid of benthic substrates and microbial mats, features reflective of the overall DOM composition. In the CG stream, DOM is primarily derived from the exudation of labile DOM produced by planktonic microorganisms (Smith et al.2017), does not accumulate inter-seasonally (SanClements et al.2016), and appears to be less complex than from other natural environments (Foreman et al.2013; this study). Conversely, the presence of both protein- and humic-like fluorescence in the Canada stream DOM is typical of Antarctic streams with inter-annual DOM accumulation by microbial mats (McKnight and Tate 1997), characteristics also expressed in habitats shaped by the deposition of terrestrial mineral particles (i.e. CG parafluvial sediments and cryoconite). Striking was the lack of discrete regions of detectable fluorescent humic components in the west GrIS DOM. Increased aeolian deposition and microbial modifications of allochthonous material, including material from wildfires (Khan et al.2017), onto the GrIS was expected to contribute to chemically more heterogeneous DOM (Stibal et al.2010; Pautler et al.2013; Musilova et al.2017). However, the distribution and composition of DOM may vary, depending on the location, distance to the ice sheet margin, season (Stibal et al.2010; Langford et al.2010; Musilova et al.2017), and trajectories of air masses (Khan et al.2017). In each habitat, microbial assemblages were associated with either specific protein-like or humic-like fluorophores, with the calculated correlations between DOM components and OTUs of heterotrophic bacteria suggestive of contrasting taxonomic DOM preferences (Fig. 5). An in-depth analysis of the CG stream DOM determined it to be microbially derived and compositionally labile, with almost half of the DOM comprised of lipid-, protein- and carbohydrate-like molecular constituents (Smith et al.2018). As highlighted by substrate utilization assays (Sanyal et al.2018), Antarctic isolates from supraglacial meltwaters preferentially utilize these compounds. Specifically, diverse lineages within the Bacteroidetes, the most dominant members in all dry valley samples, are well known for degrading high molecular weight organic matter (i.e. proteins and carbohydrates), and grow attached to particles or algal cells (Thomas et al.2011; Fernández-Gomez et al.2013). Corroborating evidence for phylogenetic discrimination in carbon preference and algal-Bacteroidetes arrangements in glacial meltwater habitats has recently been reported for cryoconite holes on the Canada Glacier and CG supraglacial stream, Antarctica (Smith et al.2016). Incubation assays revealed that stream water DOM, dominated by protein-like fluorophores (designations B and T), was directly recycled byBacteroidetes in close spatial organization within Oscillatoria biofilms (Smith et al.2016). It was found that heterotrophic production by Bacteroidetes and Betaproteobacteria within the CG stream relied on freshly derived DOM, while there was no evidence of algal exudate uptake by Alphaproteobacteria (Smith et al.2017). Such differences in carbon preferences are consistent with the current understanding that—unlike Alphaproteobacteria—Bacteroidetes and Betaproteobacteria are capable of utilizing complex carbon sources of varying concentrations (Eiler et al.2003; Thomas et al.2011). In turn, the degradation of complex organic material by Bacteroidetes provides by-products directly targeted by Alpha- and Gammaproteobacteria (Williams et al.2013). Similarly, studies performed by Lutz and others described the production of taxon-specific metabolites by snow and ice algae (Lutz et al.2016, 2017). Associated with different algal blooms were specific taxonomic groups within the Bacteroidetes and Proteobacteria, attesting not only to DOM-microbe interactions, but also to the linkage between DOM exudates/composition and resident phylogeny (Lutz et al.2016, 2017). In contrast with previously reported findings on the microbial community in Greenlandic supraglacial habitats (Stibal et al.2015; Musilova et al.2015; Cameron et al.2012, 2016), in this studyGammaproteobacteria dominated the cryoconite hole and lake communities from the west GrIS. This class of Proteobacteria is generally known to be comprised of fast-growing opportunists (Pinhassi and Berman 2003; Pernthaler and Amann 2005). Further, Gammaproteobacteria have a competitive advantage in phosphate-limited environments (Ho et al.2017), characteristics typical for supraglacial habits (Mindl et al.2007; Stibal and Tranter 2007; Stibal et al.2009). Thus, in combination with extensive drainage of supraglacial habitats in the Paakitsoq region of west GrIS during the summer of 2011 (Banwell et al.2012), the dynamic surface environment may have been a selective force, giving rise to the dominance of opportunistic Gammaproteobacteria lineages. It is well documented in temperate aquatic environments that microbial community composition and underlying metabolic strategies are driven by the chemical composition of DOM (McCarren et al.2010; Koehler et al.2012; Hansell 2013; Buchan et al.2014; Moran et al.2016). Intricate substrate dependencies on specific DOM fluorophores have been shown for major taxonomic groups within many aquatic communities (e.g. Romera-Castillo et al.2011; Sarmento et al.2013; Amaral et al.2016; Blanchet et al.2017), with production and consumption of fluorescent DOM varying on a species level (Fukuzaki et al.2014; Fox et al.2017; Goto et al.2017). Evidence presented here establishes causality between DOM substrate composition and microbial community composition (Fig. 5) in glacial environments. Therefore, changes in DOM composition may result in dramatic shifts in the functional response of these glacial meltwater microbial systems, with subsequent consequences for downstream environments through alterations in the geochemical composition of fluxes. CONCLUSIONS While it is acknowledged that icy ecosystems possess diverse and metabolically active organisms (Boetius et al.2015), understanding the interactions between bacteria and glacial DOM remains an important challenge. Here we demonstrate that DOM composition has a major effect on glacial surface microbial communities, and by extension, influences ecosystem function. Additionally, our data contribute to the growing evidence that supraglacial habitats harbor spatially variable communities governed by local environmental factors (Cameron et al.2012, 2016; Edwards et al.2014; Stibal et al.2015). Undeniably, the global cryosphere is shrinking (Stocker et al.2013) and icy environments need to be understood within the context of climate change. Changes in microbial and functional diversity may have a profound impact on supraglacial DOM cycling, likely changing predicted quantities of carbon discharge (Anesio et al.2009; Cook et al.2012; Hodson et al.2015) and DOM bioavailability to downstream environments (Bhatia et al.2013; Lawson et al.2014; Musilova et al.2017; Smith et al.2017). SUPPLEMENTARY DATA Supplementary data are available at FEMSEC online. ACKNOWLEDGEMENTS The authors wish to thank the members of the Cotton Glacier and Greenland field teams, the Antarctic and Arctic support contractors and the pilots and crew of Petroleum Helicopters Inc. Any opinions, findings, or conclusions expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation (NSF). FUNDING This work was supported by the National Science Foundation [NSF ANT-0838970, ANT-1141978, and DGE-0654336], and through a NASA Earth and Science Space Fellowship. Conflict of interest. None declared. REFERENCES Amaral V, Graeber D, Calliari D et al.   Strong linkages between DOM optical properties and main clades of aquatic bacteria. Limnol Oceanogr . 2016; 61: 906– 18. Google Scholar CrossRef Search ADS   Anesio AM, Hodson AJ, Fritz A et al.   High microbial activity on glaciers: importance to the global carbon cycle. Global Change Biol . 2009; 15: 955– 60. Google Scholar CrossRef Search ADS   Anesio AM, Laybourn-Parry J. Glaciers and ice sheets as a biome. Trends Ecol Evol . 2012; 27: 219– 25. Google Scholar CrossRef Search ADS PubMed  Anesio AM, Lutz S, Chrismas NAM et al.   The microbiome of glaciers and ice sheets. NPJ Biofilms Microbiomes . 2017; 3: 10. Google Scholar CrossRef Search ADS PubMed  Antony R, Grannas AM, Willoughby AS et al.   Origin and sources of dissolved organic matter in snow on the East Antarctic ice sheet. Environ Sci Technol . 2014; 48: 6151– 9. Google Scholar CrossRef Search ADS PubMed  Antony R, Sanyal A, Kapse N et al.   Microbial communities associated with Antarctic snow pack and their biogeochemical implications. Microbiol Res . 2016; 192: 192– 202. Google Scholar CrossRef Search ADS PubMed  Antony R, Willoughby AS, Grannas AM et al.   Molecular insights on dissolved organic matter transformation by supraglacial microbial communities. Environ Sci Technol . 2017; 51: 4328– 37. Google Scholar CrossRef Search ADS PubMed  Banwell AF, Arnold NS, Willis IC et al.   Modeling supraglacial water routing and lake filling on the Greenland Ice Sheet. J Geophys Res Earth Surf . 2012; 117, F04012. Barker JD, Dubnick A, Lyons WB, et al.   Changes in dissolved organic matter (DOM) fluorescence in proglacial Antarctic streams. AAAR . 2013; 45: 305– 17. Google Scholar CrossRef Search ADS   Barker JD, Sharp MJ, Fitzsimons SJ et al.   Abundance and dynamics of dissolved organic carbon in glacier systems. AAAR . 2006; 38: 163– 72. Google Scholar CrossRef Search ADS   Bhatia MP, Das SB, Longnecker K et al.   Molecular characterization of dissolved organic matter associated with the Greenland ice sheet. Geochim Cosmochim Acta . 2010; 74: 3768– 84. Google Scholar CrossRef Search ADS   Bhatia MP, Das SB, Xu L et al.   Organic carbon export from the Greenland ice sheet. Geochim Cosmochim Acta . 2013; 109: 329– 44. Google Scholar CrossRef Search ADS   Blanchet M, Pringault O, Panagiotopoulos C et al.   When riverine dissolved organic matter (DOM) meets labile DOM in coastal waters: changes in bacterial community activity and composition. Aquat Sci . 2017; 79: 27– 43. Google Scholar CrossRef Search ADS   Boetius A, Anesio AM, Deming JW et al.   Microbial ecology of the cryosphere: sea ice and glacial habitats. Nat Rev Microbiol . 2015; 13: 677– 90. Google Scholar CrossRef Search ADS PubMed  Bottos EM, Woo AC, Zawar-Reza P et al.   Airborne bacterial populations above desert soils of the McMurdo Dry Valleys, Antarctica. Microbiol Ecol . 2014; 67: 120– 8. Google Scholar CrossRef Search ADS   Bowman JS, Deming JW. Wind-driven distribution of bacteria in coastal Antarctica: evidence from the Ross Sea region. Polar Biol . 2017; 40: 25– 35. Google Scholar CrossRef Search ADS   Bro R. PARAFAC. Tutorial and applications. Chemom Intell Lab Syst . 1997; 38: 149– 71. Google Scholar CrossRef Search ADS   Buchan A, LeCleir GR, Gulvik CA et al.   Master recyclers: features and functions of bacteria associated with phytoplankton blooms. Nat Rev Microbiol . 2014; 12: 686– 98. Google Scholar CrossRef Search ADS PubMed  Cameron KA, Hagedorn B, Dieser M et al.   Diversity and potential sources of microbiota associated with snow on western portions of the Greenland Ice Sheet. Environ Microbiol . 2015; 17: 594– 609. Google Scholar CrossRef Search ADS PubMed  Cameron KA, Hodson AJ, Osborn AM. Structure and diversity of bacterial, eukaryotic and archaeal communities in glacial cryoconite holes from the Arctic and the Antarctic. FEMS Microbiol Ecol . 2012; 82: 254– 67. Google Scholar CrossRef Search ADS PubMed  Cameron KA, Stibal M, Zarsky JD et al.   Supraglacial bacterial community structures vary across the Greenland ice sheet. FEMS Microbiol Ecol . 2016; 92: fiv164. Google Scholar CrossRef Search ADS PubMed  Caporaso JG, Kuczynski J, Stombaugh J et al.   QIIME allows analysis of high-throughput community sequencing data. Nat Methods . 2010; 7: 335– 6. Google Scholar CrossRef Search ADS PubMed  Chao A. Nonparametric estimation of the number of classes in a population. Scand J Stat . 1984; 265– 70. Christner BC, Mikucki JA, Foreman CM et al.   Glacial ice cores: a model system for developing extraterrestrial decontamination protocols. Icarus . 2005; 174: 572– 84. Google Scholar CrossRef Search ADS   Clarke KR, Warwick RM. Change in marine communities: an approach to statistical analysis and interpretation , 2nd edition. Primer-E Ltd, Plymouth UK 2001. Coble PG. Characterization of marine and terrestrial DOM in seawater using excitation-emission matrix spectroscopy. Mar Chem . 1996; 51: 325– 46. Google Scholar CrossRef Search ADS   Cook J, Edwards A, Hubbard A. Biocryomorphology: integrating microbial processes with ice surface hydrology, topography, and roughness. Front Earth Sci . 2015; 3: 78. Google Scholar CrossRef Search ADS   Cook JM, Hodson AJ, Anesio AM. An improved estimate of microbially mediated carbon fluxes from the Greenland ice sheet. J Glaciol . 2012; 58: 1098– 1108. Google Scholar CrossRef Search ADS   Cory RM, McKnight DM. Fluorescence spectroscopy reveals ubiquitous presence of oxidized and reduced quinones in dissolved organic matter. Environ Sci Technol . 2005; 39: 8142– 9. Google Scholar CrossRef Search ADS PubMed  D’Andrilli J, Smith HJ, Dieser M, Foreman CM. Climate driven carbon and microbial signatures through the last ice age. Geochem Persp Let . 2017; 4: 29– 34. Google Scholar CrossRef Search ADS   Dubnick A, Barker J, Sharp M et al.   Characterization of dissolved organic matter (DOM) from glacial environments using total fluorescence spectroscopy and parallel factor analysis. Ann Glaciol . 2010; 51: 111– 22. Google Scholar CrossRef Search ADS   Edgar RC, Haas BJ, Clemente JC et al.   UCHIME improves sensitivity and speed of chimera detection. Bioinformatics . 2011; 27: 2194– 2200. Google Scholar CrossRef Search ADS PubMed  Edwards A, Anesio AM, Rassner SM et al.   Possible interactions between bacterial diversity, microbial activity and supraglacial hydrology of cryoconite holes in Svalbard. ISME J . 2011; 5: 150. Google Scholar CrossRef Search ADS PubMed  Edwards A, Mur LAJ, Girdwood SE et al.   Coupled cryoconite ecosystem structure-function relationships are revealed by comparing bacterial communities in alpine and Arctic glaciers. FEMS Microbiol Ecol . 2014; 89: 222– 37. Google Scholar CrossRef Search ADS PubMed  Edwards A, Rassner SM, Anesio AM et al.   Contrasts between the cryoconite and ice-marginal bacterial communities of Svalbard glaciers. Polar Res . 2013; 32: 19468. Google Scholar CrossRef Search ADS   Eiler A, Langenheder S, Bertilsson S et al.   Heterotrophic bacterial growth efficiency and community structure at different natural organic carbon concentrations. Appl Environ Microbiol . 2003; 69: 3701– 9. Google Scholar CrossRef Search ADS PubMed  Fellman JB, Hood E, Spencer RGM. Fluorescence spectroscopy opens new windows into dissolved organic matter dynamics in freshwater ecosystems: a review. Limnol Oceanogr . 2010; 55: 2452– 62. Google Scholar CrossRef Search ADS   Feng L, Xu J, Kang S et al.   Chemical composition of microbe-derived dissolved organic matter in cryoconite in Tibetan Plateau glaciers: insights from Fourier transform ion cyclotron resonance mass spectrometry analysis. Environ Sci Technol . 2016; 50: 13215– 23. Google Scholar CrossRef Search ADS PubMed  Fernández-Gomez B, Richter M, Schüler M et al.   Ecology of marine Bacteroidetes: a comparative genomics approach. ISME J . 2013; 7: 1026. Google Scholar CrossRef Search ADS PubMed  Foreman CM, Cory RM, Morris CE et al.   Microbial growth under humic-free conditions in a supraglacial stream system on the Cotton Glacier, Antarctica. Environ Res Lett . 2013; 8: 035022. Google Scholar CrossRef Search ADS   Fountain AG, Tranter M, Nylen TH et al.   Evolution of cryoconite holes and their contribution to meltwater runoff from glaciers in the McMurdo Dry Valleys, Antarctica. J Glaciol . 2004; 50: 35– 45. Google Scholar CrossRef Search ADS   Fox BG, Thorn RM, Anesio AM et al.   The in situ bacterial production of fluorescent organic matter; an investigation at a species level. Water Res . 2017; 125: 350– 9. Google Scholar CrossRef Search ADS PubMed  Franzetti A, Navarra F, Tagliaferri I et al.   Potential sources of bacteria colonizing the cryoconite of an Alpine glacier. PloS One . 2017; 12: e0174786. Google Scholar CrossRef Search ADS PubMed  Fukuzaki K, Imai I, Fukushima K et al.   Fluorescent characteristics of dissolved organic matter produced by bloom-forming coastal phytoplankton. J Plankton Res . 2014; 36: 685– 94. Google Scholar CrossRef Search ADS   Gawor J, Grzesiak J, Sasin-Kurowska J et al.   Evidence of adaptation, niche separation and microevolution within the genus Polaromonas on Arctic and Antarctic glacial surfaces. Extremophiles . 2016; 20: 403– 13. Google Scholar CrossRef Search ADS PubMed  Goto S, Tada Y, Suzuki K et al.   Production and reutilization of fluorescent dissolved organic matter by a marine bacterial strain, Alteromonas macleodii. Front Microbiol . 2017; 8: 507 Google Scholar CrossRef Search ADS PubMed  Hammer Ø, Harper DA, Ryan PD. PAST: Paleontological Statistics Software Package for education and data analysis. Palaeontol Electron 4 . 2001. Hansell DA. Recalcitrant dissolved organic carbon fractions. Ann Rev Mar Sci . 2013; 5: 421– 45. Google Scholar CrossRef Search ADS PubMed  Hell K, Edwards A, Zarsky J et al.   The dynamic bacterial communities of a melting high Arctic glacier snowpack. ISME J . 2013; 7: 1814. Google Scholar CrossRef Search ADS PubMed  Ho A, Di Lonardo DP, Bodelier PL. Revisiting life strategy concepts in environmental microbial ecology. FEMS Microbiol Ecol . 2017; 93: 10. Hodson A, Bøggild C, Hanna E et al.   The cryoconite ecosystem on the Greenland ice sheet. Ann Glaciol . 2010; 51: 123– 9. Google Scholar CrossRef Search ADS   Hodson A, Brock B, Pearce D et al.   Cryospheric ecosystems: a synthesis of snowpack and glacial research. Environ Res Lett . 2015; 10: 110201. Google Scholar CrossRef Search ADS   Hood E, Battin TJ, Fellman J, O’Neel S et al.   Storage and release of organic carbon from glaciers and ice sheets. Nat Geosci . 2015; 8: 91– 6. Google Scholar CrossRef Search ADS   Hood E, Fellman J, Spencer RG et al.   Glaciers as a source of ancient and labile organic matter to the marine environment. Nature . 2009; 462: 1044– 7. Google Scholar CrossRef Search ADS PubMed  Khan AL, Wagner S, Jaffé R et al.   Dissolved black carbon in the global cryosphere: concentrations and chemical signatures. Sci Invest Rep . 2017; 44: 6226– 34. Koehler B, Wachenfeldt E, Kothawala D et al.   Reactivity continuum of dissolved organic carbon decomposition in lake water. J Geophys Res-Biogeo . 2012; 117: G01024. Google Scholar CrossRef Search ADS   Langford H, Hodson A, Banwart S et al.   The microstructure and biogeochemistry of Arctic cryoconite granules. Ann Glaciol . 2010; 51: 87– 94. Google Scholar CrossRef Search ADS   Lawaetz AJ, Stedmon CA. Fluorescence intensity calibration using the Raman scatter peak of water. Appl Spectrosc . 2009; 63: 936– 40. Google Scholar CrossRef Search ADS PubMed  Lawson EC, Wadham JL, Tranter M et al.   Greenland Ice Sheet exports labile organic carbon to the Arctic oceans. Biogeosci . 2014; 11: 4015– 28. Google Scholar CrossRef Search ADS   Legrand M, Preunkert S, Jourdain B et al.   Water-soluble organic carbon in snow and ice deposited at Alpine, Greenland, and Antarctic sites: a critical review of available data and their atmospheric relevance. Clim Past . 2013; 9: 2195. Google Scholar CrossRef Search ADS   Lutz S, Anesio AM, Raiswell R et al.   The biogeography of red snow microbiomes and their role in melting arctic glaciers. Nat Commun . 2016; 7: 11968. Google Scholar CrossRef Search ADS PubMed  Lutz S, Anesio AM, Edwards A et al.   Linking microbial diversity and functionality of arctic glacial surface habitats. Environ Microbiol . 2017; 19: 551– 65. Google Scholar CrossRef Search ADS PubMed  McCarren J, Becker JW, Repeta DJ et al.   Microbial community transcriptomes reveal microbes and metabolic pathways associated with dissolved organic matter turnover in the sea. PNAS USA . 2010; 107: 16420– 7. Google Scholar CrossRef Search ADS PubMed  McKnight DM, Boyer EW, Westerhoff PK. Spectrofluorometric characterization of dissolved organic matter for indication of precursor organic material and aromaticity. Limnol Oceanogr . 2001; 46: 38– 48. Google Scholar CrossRef Search ADS   McKnight DM, Tate CM. Canada Stream: a glacial meltwater stream in Taylor Valley, South Victoria Land, Antarctica. J North Am Benthol Soc . 1997; 16: 14. Google Scholar CrossRef Search ADS   Miller MP, McKnight DM. Comparison of seasonal changes in fluorescent dissolved organic matter among aquatic lake and stream sites in the Green Lakes Valley. J Geophys Res-Biogeosci . 2010; 115: G00F12. Google Scholar CrossRef Search ADS   Mindl B, Anesio AM, Meirer K et al.   Factors influencing bacterial dynamics along a transect from supraglacial runoff to proglacial lakes of a high Arctic glacier. FEMS Microbiol Ecol . 2007; 59: 307– 17. Google Scholar CrossRef Search ADS PubMed  Moran MA, Kujawinski EB, Stubbins A, et al.   Deciphering ocean carbon in a changing world. PNAS . 2016; 113: 3143– 51. Google Scholar CrossRef Search ADS PubMed  Mueller DR, Pollard WH. Gradient analysis of cryoconite ecosystems from two polar glaciers. Polar Biol . 2004; 27: 66– 74. Google Scholar CrossRef Search ADS   Musilova M, Tranter M, Bennett SA et al.   Stable microbial community composition on the Greenland Ice Sheet. Front Microbiol . 2015; 6: 193. Google Scholar CrossRef Search ADS PubMed  Musilova M, Tranter M, Wadham J et al.   Microbially driven export of labile organic carbon from the Greenland ice sheet. Nat Geosci . 2017; 10: 360– 5. Google Scholar CrossRef Search ADS   Nguyen NH, Smith D, Peay K et al.   Parsing ecological signal from noise in next generation amplicon sequencing. New Phytol Lett . 2015; 205: 1389– 93. Google Scholar CrossRef Search ADS   Pautler BG, Dubnick A, Sharp MJ et al.   Comparison of cryoconite organic matter composition from Arctic and Antarctic glaciers at the molecular-level. Geochim Cosmochim Acta . 2013; 104: 1– 18. Google Scholar CrossRef Search ADS   Pautler BG, Woods GC, Dubnick A et al.   Molecular characterization of dissolved organic matter in glacial ice: coupling natural abundance 1H NMR and fluorescence spectroscopy. Environ Sci Technol . 2012; 46: 3753– 61. Google Scholar CrossRef Search ADS PubMed  Pearce DA, Bridge PD, Hughes KA et al.   Microorganisms in the atmosphere over Antarctica. FEMS Microbiol Ecol . 2009; 69: 143– 57. Google Scholar CrossRef Search ADS PubMed  Pernthaler J, Amann R. Fate of heterotrophic microbes in pelagic habitats: focus on populations. Microbiol Mol Biol Rev . 2005; 69: 440– 61. Google Scholar CrossRef Search ADS PubMed  Pinhassi J, Berman T. Differential growth response of colony-forming α-and γ-proteobacteria in dilution culture and nutrient addition experiments from Lake Kinneret (Israel), the eastern Mediterranean Sea, and the Gulf of Eilat. Appl Environ Microb . 2003; 69: 199– 211. Google Scholar CrossRef Search ADS   Price PB, Rohde RA, Bay RC. Fluxes of microbes, organic aerosols, dust, sea-salt Na ions, non-sea-salt Ca ions, and methanesulfonate onto Greenland and Antarctic ice. Biogeosci . 2009; 6: 479– 86. Google Scholar CrossRef Search ADS   Pruesse E, Quast C, Knittel K et al.   SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res . 2007; 35: 7188– 96. Google Scholar CrossRef Search ADS PubMed  Romera-Castillo C, Sarmento H, Álvarez-Salgado XA et al.   Net production and consumption of fluorescent colored dissolved organic matter by natural bacterial assemblages growing on marine phytoplankton exudates. Appl Environ Microb . 2011; 77: 7490– 8. Google Scholar CrossRef Search ADS   Salter SJ, Cox MJ, Turek EM et al.   Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol . 2014; 12: 87. Google Scholar CrossRef Search ADS PubMed  SanClements MD, Smith HJ, Foreman CM et al.   Biogeophysical properties of an expansive Antarctic supraglacial stream. Antarct Sci . 2016; 29: 33– 44. Google Scholar CrossRef Search ADS   Sanyal A, Antony R, Samui G et al.   Microbial communities and their potential for degradation of dissolved organic carbon in cryoconite hole environments of Himalaya and Antarctica. Microbiol Res . 2018, 208: 32– 42. Google Scholar CrossRef Search ADS PubMed  Sarmento H, Romera-Castillo C, Lindh M et al.   Phytoplankton species‐specific release of dissolved free amino acids and their selective consumption by bacteria. Limnol Oceanogr . 2013; 58: 1123– 35. Google Scholar CrossRef Search ADS   Schloss PD, Westcott SL, Ryabin T et al.   Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol . 2009; 75: 7537– 41. Google Scholar CrossRef Search ADS PubMed  Schloss PD, Gevers D, Westcott SL. Reducing the effects of PCR amplification and sequencing artifacts on 16S rRNA-based studies. PLoS One . 2011; 6: e27310. Google Scholar CrossRef Search ADS PubMed  Simpson EH. Measurement of diversity. Nature . 1949; 163: 688. Google Scholar CrossRef Search ADS   Smith HJ, Foster RA, McKnight DM et al.   Microbial formation of labile organic carbon in Antarctic glacial environments. Nat Geosci . 2017; 10: 356– 9. Google Scholar CrossRef Search ADS   Smith HJ, Schmit A, Foster R et al.   Biofilms on glacial surfaces: hotspots for biological activity. NPJ Biofilms Microbiomes . 2016; 2: 16008. Google Scholar CrossRef Search ADS PubMed  Smith HJ, Tigges M, D’Andrilli J et al.   Dynamic processing of DOM: insight from exometabolomics, fluorescence spectroscopy and mass spectrometry. Limnol Oceanogr Lett . 2018; 3: 225– 35. Google Scholar CrossRef Search ADS   Stibal, M, Anesio, AM, Blues, CJD. Phosphatase activity and organic phosphorus turnover on a high Arctic glacier. Biogeosci . 2009; 6: 913– 22. Google Scholar CrossRef Search ADS   Stibal M, Lawson EC, Lis GP et al.   Organic matter content and quality in supraglacial debris across the ablation zone of the Greenland ice sheet. Ann Glaciol . 2010; 51: 1– 8. Google Scholar CrossRef Search ADS   Stibal M, Sabacka M, Zarsky J. Biological processes on glacier and ice sheet surfaces. Nat Geosci . 2012; 5: 771– 4. Google Scholar CrossRef Search ADS   Stibal M, Schostag M, Cameron KA et al.   Different bulk and active bacterial communities in cryoconite from the margin and interior of the Greenland ice sheet. Environ Microbiol Rep . 2015; 7: 293– 300. Google Scholar CrossRef Search ADS PubMed  Stibal M, Tranter M. Laboratory investigation of inorganic carbon uptake by cryoconite debris from Werenskioldbreen, Svalbard. J Geophys Res . 2007; 11: G04S33. Stocker TF, Qin D, Plattner GK et al.   IPCC, 2013: climate change 2013: the physical science basis . Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change, Cambridge University Press, Cambridge, U.K., 2013: 1535. Stubbins A, Hood E, Raymond PA et al.   Anthropogenic aerosols as a source of ancient dissolved organic matter in glaciers. Nat Geosci . 2012; 5: 198– 201. Google Scholar CrossRef Search ADS   Takeuchi N, Nagatsukan N, Uetake J et al.   Spatial variations in impurities (cryoconite) on glaciers in northwest Greenland. Bull Glaciol Res . 2014; 32: 85– 94. Google Scholar CrossRef Search ADS   Tedesco M, Willis IC, Hoffman MJ et al.   Ice dynamic response to two modes of surface lake drainage on the Greenland ice sheet. Environ Res Lett . 2013; 8: 034007. Google Scholar CrossRef Search ADS   Telling J, Anesio AM, Tranter M et al.   Controls on the autochthonous production and respiration of organic matter in cryoconite holes on high Arctic glaciers. J Geophys Res . 2012; 117: G01017. Google Scholar CrossRef Search ADS   Thomas F, Hehemann JH, Rebuffet E et al.   Environmental and gut bacteroidetes: the food connection. Front Microbiol . 2011; 30; 2: 93. Uetake J, Tanaka S, Segawa T et al.   Microbial community variation in cryoconite granules on Qaanaaq Glacier, NW Greenland. FEMS Microbiol Ecol . 2016; 92: fiw127. Google Scholar CrossRef Search ADS PubMed  Wang Q, Garrity GM, Tiedje JM et al.   Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol . 2007; 73: 5261– 7. Google Scholar CrossRef Search ADS PubMed  Williams TJ, Wilkins D, Long E et al.   The role of planktonic Flavobacteria in processing algal organic matter in coastal East Antarctica revealed using metagenomics and metaproteomics. Environ Microbiol . 2013; 15: 1302– 17. Google Scholar CrossRef Search ADS PubMed  © FEMS 2018. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Microbiology Ecology Oxford University Press

Relationship between dissolved organic matter quality and microbial community composition across polar glacial environments

Loading next page...
 
/lp/ou_press/relationship-between-dissolved-organic-matter-quality-and-microbial-300A4xtXBW
Publisher
Blackwell
Copyright
© FEMS 2018.
ISSN
0168-6496
eISSN
1574-6941
D.O.I.
10.1093/femsec/fiy090
Publisher site
See Article on Publisher Site

Abstract

ABSTRACT Vast expanses of Earth’s surface are covered by ice, with microorganisms in these systems affecting local and global biogeochemical cycles. We examined microbial assemblages from habitats fed by glacial meltwater within the McMurdo Dry Valleys, Antarctica and on the west Greenland Ice Sheet (GrIS), evaluating potential physicochemical factors explaining trends in community structure. Microbial assemblages present in the different Antarctic dry valley habitats were dominated by Sphingobacteria andFlavobacteria, while Gammaproteobacteria and Sphingobacteria prevailed in west GrIS supraglacial environments. Microbial assemblages clustered by location (Canada Glacier, Cotton Glacier and west GrIS) and were separated by habitat type (i.e. ice, cryoconite holes, supraglacial lakes, sediment and stream water). Community dissimilarities were strongly correlated with dissolved organic matter (DOM) quality. Microbial meltwater assemblages were most closely associated with different protein-like components of the DOM pool. Microbes in environments with mineral particles (i.e. stream sediments and cryoconite holes) were linked to DOM containing more humic-like fluorescence. Our results demonstrate the establishment of distinct microbial communities within ephemeral glacial meltwater habitats, with DOM-microbe interactions playing an integral role in shaping communities on local and polar spatial scales. Biodiversity, Antarctica, Greenland, microbial, dissolved organic matter INTRODUCTION Glaciers and ice sheets represent a distinct biome, dominated by diverse and biologically active microorganisms (Anesio and Laybourn-Parry 2012; Boetius et al.2015; Anesio et al.2017), that play a pivotalrole in regional and global biogeochemical cycles. These icy environments are subject to relatively similar environmental conditions, including low temperature, frequent freeze-thaw cycles, seasonally low water availability, desiccation and nutrient limitation, but vary in the degree of surface melt and geochemical composition. As such, the presence of distinct microbial communities in environmentally similar glacial meltwater habitats (Cameron et al.2012, 2015, 2016; Stibal et al.2015; Lutz et al.2016; Uetake et al.2016) offers intriguing questions about the factors influencing trends in community structure and function and the impacts of microorganisms on geochemical processes. While regional differences across glacial meltwater communities have been frequently attributed to local aeolian deposition, catchment geology, and physical and geographical variables (Cameron et al.2015, 2016; Cook et al.2015; Musilova et al.2015; Stibal et al.2015; Lutz et al.2016), these factors may, indeed, have only a limited effect on community composition. Findings from a growing number of investigations favor the development of habitat-specific ice surface communities, particularly those adapted to specific ecological niches (Mueller and Pollard 2004; Edwards et al.2013; Musilova et al.2015; Gawor et al.2016; Franzetti et al.2017). Dissolved organic matter (DOM) on ice sheets originates from a variety of allochthonous (i.e. terrestrial, marine or black carbon) and autochthonous (i.e. in situ production) sources (Stibal et al.2010; Stubbins et al.2012; Legrand et al.2013; Antony et al.2014; Hood et al.2015; Smith et al.2017), both of which have been shown to vary in reactivity and may be highly bioavailable to resident microbial communities (Hood et al.2009; Antony et al.2017; Smith et al.2017). Many studies have targeted DOM dynamics on the surface of ice sheets (Anesio et al.2009; Hodson et al.2010; Cook et al.2012; Stibal et al.2012; Hood et al.2015), and both deposition and in situ production have been linked to the biogeochemical evolution of DOM in supraglacial waters (Hood et al.2009; Stubbins et al.2012; Telling et al.2012; Antony et al.2014; Musilova et al.2017; Smith et al.2017). Rather than being passive receivers of allochthonous DOM, recent studies indicate that there are tight interactions between DOM production and composition within bacterial communities in glacial surface habitats (Smith et al.2016, 2017; Musilova et al.2017). Additionally, changes in microbial community structure have been correlated with dissolved organic carbon (DOC) concentrations and the presence of distinct algal populations in Arctic snow and on glaciers, further attesting to the importance of interactions between organic carbon and microbial communities (Lutz et al.2016, 2017). In fact, major components of glacial carbon dynamics are functionally linked to trophic interactions between photo- and heterotrophs during algal blooms, and within cryoconite granules and biofilms (Edward et al.2011; Takeuchi et al.2014; Anesio et al.2017; Lutz et al.2017; Musilova et al.2017; Smith et al.2017). However, the mechanism by which the composition of DOM influences bacterial utilization and ultimately community composition is largely unknown. Globally, glacial meltwater habitats are dominated by cyanobacteria,Proteobacteria, Actinobacteria, Bacteriodetes, Acidobacteria, Chloroflexi and Planctomycetes, with regional differences at lower levels of taxonomic resolution (Boetius et al.2015). Given the habitat-specific selective pressures of cryo-environments, we hypothesize that bacterial assemblages are shaped by geochemistry, which provides the available energy sources necessary to support life. To test this scenario, glacial meltwater habitats from the Earth’s polar regions were investigated. Meltwater samples were collected from two glaciers in the McMurdo Dry Valleys, Antarctica, (stream water, stream sediments, glacial ice, snow, and cryoconite holes) and from the western margin of the Greenland Ice Sheet (GrIS) (supraglacial lakes and cryoconite holes). Illumina MiSeq 16S rRNA gene amplicons were analyzed to assess differences in community structure across and within these environments. Geochemical parameters, including major ions, nutrients, DOC concentration, and fluorescing DOM characteristics were evaluated to infer their influence on microbial community composition. The primary objective of this study was to link trends in bacterial biodiversity to DOM quality across glacial systems, thus enabling better predictions regarding their function in the context of a changing environment. METHODS Field site description Sampling was conducted in the McMurdo Dry Valleys, Antarctica and at the western margin of the GrIS (Fig. 1). The McMurdo Dry Valleys of Antarctica comprise the largest ice-free area of the Antarctic continent and are classified as polar deserts with <10 cm of precipitation per year. Mean annual temperatures range between −15°C and −30°C. This barren region experiences katabatic winds originating in the continental interior, which are important in controlling local climate, physical weathering and material transport. Contrary to expectation, contributions by marine aerosols are minor (Bottos et al.2014; Bowman and Deming 2017). The two Antarctic sampling sites were comprised of glacially fed Antarctic stream ecosystems ∼70 km apart: (i) The Cotton Glacier (CG) supraglacial stream (77°07’S, 161°40’E) is ∼16 km long, consisting of a network of braided channels cut through the ice, ultimately terminating into McMurdo Sound. The CG stream receives a large amount of sedimentary deposits of unknown origin (fluvial, glacial or aeolian) from surrounding areas (Foreman et al.2013). Sediments are primarily located on the lateral margins of the stream, and herein referred to as parafluvial sediments. CG stream water (n = 10), parafluvial sediments (n = 11), snow (n = 1) and surrounding ice (n = 5) were sampled during austral summer field seasons 2009–2012. (ii) The Canada Stream (77°61’S, 162°59’E) represents a proglacial stream in the Taylor Valley and flows over a constant layer of coarse sediments. Samples were collected in 2009–2011 and included: stream water (n = 5), stream sediments (n = 3), as well as samples from its meltwater source, the adjacent Canada Glacier [glacial ice (n = 4), snow (n = 1) and cryoconite holes (n = 4)]. Figure 1. View largeDownload slide Sampling locations in (A) the western margin of the Greenland Ice Sheet and (B) the McMurdo Dry Valleys, Antarctica. Figure 1. View largeDownload slide Sampling locations in (A) the western margin of the Greenland Ice Sheet and (B) the McMurdo Dry Valleys, Antarctica. Supraglacial features in west Greenland were collected from the Paakitsoq region northeast of Jakobshavn Isbrae (69°34′N, 49°48′W). This location is ∼15 km from the ice sheet margin with an ice elevation of ∼962 m above sea level (Banwell et al.2012). Samples were collected during summer 2011, including cryoconite hole (n = 9) and supraglacial lake samples (n = 5) from 'Lake Half Moon' and 'Lake Ponting' (Tedesco et al.2013). More information on sampling regime is provided below and in Table S1 (Supporting Information). Sample collection Temperature, pH, conductivity, dissolved oxygen, and total dissolved solids in the two Antarctic streams were measured on site using a Manta Sub2 probe (Eureka Water Probes Inc., U.S.A.). A malfunction of the probe obstructed in situ measurements in west GrIS supraglacial environments. Antarctic stream water samples (4 L) for geochemical analyses were collected in acid-washed, deionized water-rinsed (6×) fluorinated carboys. Samples for carbon analyses were collected from all sites in combusted (450°C for 5 hr) 125 mL amber bottles. Sediment (20–30 g) and freshly fallen snow samples were aseptically collected with sterile spatulas. Ice cores (diameter: 10 cm and length: 100 cm) and frozen cryoconite holes were removed with an ethanol-cleaned ice corer (Kovacs, U.S.A.). Samples were stored in sterile cryovials or whirl-pack bags and processed/analyzed within 12 hr post-collection. Cored samples were decontaminated by removing ∼2 mm of outer ice with sterile blades using a modification of (Christner et al.2005). Snow samples were melted at 4°C. Meltwater samples (0.5–1 L) for microbial community analysis from glacial streams, and ice and snow melt were filtered onto sterile 0.2 µm, 47 mm Supor®-200 membranes (Pall Corporation, U.S.A.) under low pressure (<7 psi). Filters, submerged in TES buffer (100 mM Tris, 100 mM EDTA, and 2% SDS), and sediments were flash frozen in liquid nitrogen, and stored at −80°C until DNA extraction. Samples for microbial community analysis from supraglacial lakes on the west GrIS were collected on polyethersulfone 0.2 mm Sterivex filters (MilliporeSigma, U.S.A.) with a hand pump. Between 0.5–1 L of sample water was passed through the filters. Sterivex cartridges were filled with RNAlater® (Invitrogen, U.S.A.) and sealed with Luer–Lok caps. Filters were frozen at −20°C and stored on ice for transport. Upon arrival at Montana State University, samples were stored at −80°C until further processing. Water chemistry Water samples for non-purgeable DOC, macronutrients, anions/cations, particulate organic carbon and particulate organic nitrogen were collected and analyzed following the protocols of the McMurdo Dry Valleys LTER (Foreman et al.2013). Geochemical data were statistically analyzed to identify differences across samples, using a general linear model of analysis of variance (ANOVA) in version 17, Minitab Inc., U.S.A. Fluorescence spectroscopy DOM is comprised of a heterogeneous mixture of organic compounds, with the analysis of fluorescent components proven to be useful in assessing its chemical quality (Cory and McKnight 2005; Fellman et al.2010). Samples were filtered through combusted GF/F filters into acid-washed, pre-combusted amber bottles. Samples were analyzed for UV-absorbance (190–1100 nm) using an Agilent 8453 UV-spectrophotometer in a 1 cm quartz cuvette. Samples with absorbance >0.3 at 254 nm were diluted with Milli-Q water to prevent inner-filter effects (Miller and McKnight 2010). Excitation emission matrices (EEMs) were collected with a Horiba Fluoromax-4 Fluorometer, equipped with a Xenon light source and a 1 cm pathlength quartz cuvette (Foreman et al.2013). Post-processing was completed in MATLAB to generate EEMs corrected for Raman scattering and blank water subtraction (McKnight et al.2001; Lawaetz and Stedmon 2009). Due to large difference in characteristics of the fluorescent fraction of DOM between the Antarctic and Arctic samples (e.g. dominance of B1 fluorescence in the Greenland samples) and the lack of samples with intermediate characteristics, it was not possible to validate a PARAFAC model (Bro 1997) for this dataset. In light of these restrictions we quantified the EEMs spectra based on characteristic fluorescence peaks (Coble 1996; Fellmanet al.2010). To statistically compare differences in DOM composition from EEMs data, the relative proportion of individual fluorophores (B1, B2, T1, T2, A and C) was calculated for each sample. Fluorescence intensity (F.I.) values were normalized across different sample locations and summed based on DOM classifications (Coble 1996) (Table S2, Supporting Information). Significant differences between the means of specific DOM fluorophores were determined with an ANOVA general linear model followed by Tukey’s multiple-comparison test (version 17, Minitab Inc., U.S.A.). DNA extraction and PCR amplification The PowerSoil DNA or PowerWater Sterivex kits (Qiagen, U.S.A.) were used to extract genomic DNA from sediment and filters following manufacturer recommendations. Extracted genomic DNA was quantified using the Qubit DNA Kit (Invitrogen, U.S.A.). The V3/V4 regions of the 16S rRNA gene were amplified in 50 µl PCR reactions containing 100–200 pg of extracted genomic DNA, 0.1 µM of each primer, a 1X final concentration of Bulls Eye PREMIUM Taq 2X Mix (Midwest Scientific, U.S.A.) and nuclease free water. The primer design consisted of Illumina adaptor sequences followed by either the universal 341F 5′-acactctttccctacacgacgctcttccgatctCCTACGGGNGGCWGCAG-3′ or 805R 5′-gtgactggagttcagacgtgtgctcttccgatctGACTACHVGGGTATCTAATCC-3′. PCR was performed in an Eppendorf Mastercycler pro S. The amplification protocol consisted of an initial denaturation at 95°C for 3 min, followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, extension at 72°C for 30 s and a final extension step at 72°C for 5 min. Negative PCR controls without DNA template and DNA extractions from filter- and extraction-blanks were amplified to monitor contamination (n = 4) (Salter et al.2014; Nguyen et al.2015). Sequencing and analysis Paired end, 250 bp sequencing was performed on an Illumina MiSeq at the University of Wisconsin-Madison Biotechnology Center. Forward and reverse sequences were joined with the Quantitative Insights Into Microbial Ecology toolkit version 1.9.0 (Caporaso et al.2010). Contigs were subsequently analyzed with the Mothur platform v.1.34.4 (Schloss et al.2009). For sequence quality refinement, sequences containing ambiguous bases, homopolymers > 8 bases and an average quality score below 30 over a 50 bp window were excluded from further analysis (Schloss et al.2011). The maximum sequence length was 467 bp, and sequences shorter than 450 bp were removed. Processed sequences were aligned against the SILVA Gold database in Mothur. Chimeric sequences were removed using UCHIME (Edgar et al.2011) with the SILVA Gold database (Pruesse et al.2007) and a second chimera check using sequences from the present study as a database. Sequences were classified with a Bayesian method (Wang et al.2007) using the Mothur formatted version of the RDP classifier. Operational taxonomic units (OTUs) were defined at ≥97% 16S rRNA sequence identity. Any OTU found in the blank samples were removed from corresponding sequence libraries. Sequences were deposited in the NCBI SRA database under the accession number SRP076570. Statistical analysis of sequencing samples Precipitation events are rare in polar deserts; thus, with only single snow samples collected, these were excluded from subsequent statistical analysis. OTU richness using Chao1 (Chao 1984), Inverse Simpson (Simpson 1949) and the non-parametric multivariate analysis of OTU similarities (ANOSIM; Clarke and Warwick 2001) approach were calculated within Mothur. ANOSIM produces the test statistic R. Large R values (close to 1) indicate that two samples are highly dissimilar, while small R values (close to zero) signify little or no dissimilarity. The most dominant 200 OTUs (89% of the entire sequence library) were subjected to similarity percentage (SIMPER) analysis using PAST version 3.10 (Hammer et al.2001). A canonical correspondence analysis (CCA) was performed on square root transformed OTU profiles within Canoco v4.5 (Microcomputer Power Inc., U.S.A.). Within Canoco, detrended CCA was used to obtain the length of the gradient among OTUs. A gradient length > 4 supported a unimodal method for comparison between all ecosystem components. Differences in microbial communities among sites were calculated with non-metric multidimensional scaling (MDS) ordination, using a Bray Curtis dissimilarity matrix created from OTU tables and refined by the Kruskal’s Stress test. RESULTS Stream water biogeochemistry Both Antarctic streams had low ionic strength (mean 25.2 µS cm−1) and near neutral pH (mean 6.8). Of importance were the significantly higher concentrations of NH4+ (5.1± 0.7µg L−1; ANOVA, DF = 14, F = 44.68, P < 0.001), NO3− (50.5± 24.0µg L−1; ANOVA, DF = 14, F = 13.59, P = 0.004), PO42− (1.6± 0.2µg L−1; ANOVA, DF = 14, F = 9.32, P = 0.012) and particulate phosphorous (10.6± 7.8µg L−1; ANOVA, DF = 14, F = 9.98, P = 0.003) within the CG stream compared to the Canada Stream (Table S3, Supporting Information). Fluorescence in the humic-like regions of the EEMs was apparent in all samples; however, only the fluorescence spectra of Canada Stream water DOM resolved in distinct humic fluorophores. DOM from all other Antarctic and west GrIS environments lacked these discrete regions of humic fluorophores and were characterized by protein-like fluorescence (ANOVA, DF = 55, F = 40.44, P < 0.001) (Fig. 2; Table S4, Supporting Information). No significant difference between any of the Antarctic environments was found in the T1 region (ANOVA, DF = 55, F = 3.53, P > 0.216). DOM in cryo-habitats from the west GrIS lacked humic-like fluorophores (A and C peaks) and showed significantly different protein-like peaks B1 (ANOVA, DF = 55, F = 16.71, P < 0.001) and B2 (ANOVA, DF = 55, F = 18.04, P < 0.020) when compared to all Antarctic environments. Figure 2. View largeDownload slide Change in the relative abundance of fluorescent intensity (F.I.) of fluorescent regions commonly observed in natural waters (Coble 1996) for different environmental locations. Figure 2. View largeDownload slide Change in the relative abundance of fluorescent intensity (F.I.) of fluorescent regions commonly observed in natural waters (Coble 1996) for different environmental locations. Microbial assemblages and biodiversity Raw Illumina MiSeq sequences consisted of 76,026± 15,449 (average ± S.D.) reads per sample. Quality refined sequence libraries contained on average 8,913± 5,640 reads, clustered into 2,156 OTUs across all samples. On average, all samples from Antarctic environments were dominated by Sphingobacteria (48%± 6%), Flavobacteria (20%± 11%), Betaproteobacteria (8% ± 3%) and Gammaproteobacteria (4%± 2%). Supraglacial aquatic habitats from the west GrIS were primarily dominated by Gammaproteobacteria (66%± 11%),Sphingobacteria (24%± 8%) and Betaproteobacteria (3%± 2%) (Fig. 3). CG snow was dominated by Gammaproteobacteria (82%), which differed from the dominance of Betaproteobacteria (46%) and Bacilli (38%) in Canada snow. The observed number of OTUs across all sampled habitats showed similar patterns when compared to the predicted alpha diversity indices (Fig. S1, Supporting Information). Based on Chao richness, the greatest diversity was found in the Canada Stream water (524.0–535.6; 95% CI), followed by the CG stream water (395.3–411.4; 95% CI). Species richness for Antarctic stream sediments, ice and cryoconite hole communities was between 116.4 and 258.6 (95% CI). The lowest species richness was calculated for west GrIS supraglacial lakes (54.8–64.0; 95% CI) and cryoconite holes (52.9–62.2; 95% CI), and confirmed by the Inverse Simpson index (2.7–3.2; 95% CI). In the different geographic locations (CG, Canada Glacier, and the west GrIS), 43.5%, 53.1% and 46.9% of the OTUs were unique. Antarctic environments shared 29.3% of the OTUs while only 3.1% of the OTUs were present across all glacial settings. MDS of Bray–Curtis indices of the OTU profiles clustered the three glacial locations into distinct clades (Fig. 4). Within each clade, sample types clustered into their own representative group, with the exception of one CG sediment sample that grouped more closely with the Canada Glacier samples. Overlap was found between CG ice and CG stream samples. Figure 3. View largeDownload slide Relative abundance of MiSeq 16S rRNA gene sequences representing the distribution of microbial assemblages with individual samples grouped together based on sampling location. Percentages are reported at the class taxonomic level. Figure 3. View largeDownload slide Relative abundance of MiSeq 16S rRNA gene sequences representing the distribution of microbial assemblages with individual samples grouped together based on sampling location. Percentages are reported at the class taxonomic level. Figure 4. View largeDownload slide Multidimensional scaling plot of Bray–Curtis indices on the relative abundance of OTU (97% sequence similarity) profiles for different environmental locations. Dashed lines delineate the different sampled environmental locations (Canada Glacier, Cotton Glacier and Greenland Ice Sheet). Figure 4. View largeDownload slide Multidimensional scaling plot of Bray–Curtis indices on the relative abundance of OTU (97% sequence similarity) profiles for different environmental locations. Dashed lines delineate the different sampled environmental locations (Canada Glacier, Cotton Glacier and Greenland Ice Sheet). OTU specific differences between environments from the same location and the same environment from different locations were determined by ANOSIM (Table S6, Supporting Information). Overall, differences were evident across sites (R = 0.722, P < 0.001) and sample type (R = 0.530, P < 0.001), with statistical evidence for OTU dissimilarities found for 11 out of 14 pairwise comparisons (R ≥ 0.559, P ≤ 0.035). Only microbial assemblages from Canada Glacier cryoconite holes and Canada Glacier ice (R = 0.198, P = 0.081) and west GrIS cryoconite hole and lake habitats (R = 0.112, P = 0.156) were found to be more similar. Moderate community overlap was also determined between CG ice and the CG stream (R = 0.410, P = 0.011). Markedly, OTU profiles found in similar environments (i.e. ice, stream, stream sediments, and cryoconite holes) from different locations were highly dissimilar (R ≥ 0.559 and P ≤ 0.012). With a focus on the 200 most abundant OTUs in each sample, dissimilarities were largely attributed to four families (i.e. Cytophagaceae, Flavobacteriaceae, Chitinophagaceae, and Sphingobacteriaceae) within the phylum Bacteroidetes (SIMPER; Table S5, Supporting Information). Combined with the unclassified microbial realm (7.7%–13.9%), these groups accounted for 82.8% of the dissimilarities between microbial assemblages in ice from CG and Canada Glaciers, 72.1% between microbes associated with sediments in the CG and Canada streams, and 87.7% between bacterioplankton in the CG and Canada streams. These four families also explained 50.9% of the dissimilarities between bacterial assemblages inhabiting cryoconite holes on the Canada Glacier and the west GrIS, with Pseudomonadaceae contributing another 15.2%. Relationship of environmental variables to microbial biodiversity An initial CCA between CG stream and Canada Stream water OTU profiles and the physicochemical dataset was performed to determine whether environmental variables had an effect on microbial assemblage composition. However, variables tested (temperature, pH, dissolved oxygen, conductivity, major cations and anions, and DOC concentrations) were statistically insignificant in explaining assemblage composition. To determine the effect of DOM quality on microbial assemblages in glacial environments, CCA analysis included the F.I. for different DOM fluorophores and OTU profiles (Fig. 5). The established model explained 86.9% (P = 0.004) of the total variation across all OTU profiles, with 77.9% of the total model variance described by the first three axes (all eigenvalues >0.16). The first axis explained 51.2% of the variance (P = 0.002). Microbial assemblages from the west GrIS clustered closely together and were best explained by the protein-like component B1. Canada Stream microbial assemblages were clearly described by the presence of humic-like fractions of DOM. Phylotype composition from Canada Stream sediment, Canada Glacier cryoconite holes, and CG sediment samples were moderately influenced by humic- and protein-like DOM constituents. Based on the model, OTUs found in the CG stream, CG ice and Canada Glacier ice were best explained by protein-like B2 and T2 fluorophores. Figure 5. View largeDownload slide Canonical correspondence analysis of the influence of fluorescent OM variables on the relative abundance of square root transformed OTU (97% sequence similarity) profiles for sampled glacial environments. Different colored circles represent relative fluorescence intensities for fluorophores commonly observed in natural waters (Coble 1996) from each sample. Fluorescence from protein-like OM components are represented as B1 and B2 (tyrosine-like) and T1 and T2 (tryptophan-like). Fluorescent humic-like OM components are represented by A and C. Figure 5. View largeDownload slide Canonical correspondence analysis of the influence of fluorescent OM variables on the relative abundance of square root transformed OTU (97% sequence similarity) profiles for sampled glacial environments. Different colored circles represent relative fluorescence intensities for fluorophores commonly observed in natural waters (Coble 1996) from each sample. Fluorescence from protein-like OM components are represented as B1 and B2 (tyrosine-like) and T1 and T2 (tryptophan-like). Fluorescent humic-like OM components are represented by A and C. DISCUSSION Phyla identified across all samples in this study were typical of glacial ecosystems (Boetius et al.2015). Individual microbial assemblages clustered based on location (i.e. CG, Canada Glacier, and west GrIS) and habitat type (i.e. stream water, sediments, ice, cryoconite holes, and supraglacial lakes), with pronounced dissimilarities apparent between the same habitat types from different locations (Fig. 4; Table S6, Supporting Information). The lack of significant differences between ice and cryoconite hole communities from the Canada Glacier (ANOSIM; R=0.198, P = 0.081) and the moderate overlap between ice and stream microbial consortia from the CG (ANOSIM; R=0.410, P = 0.011) indicate that ice-trapped microbes contributed to the evolution of glacial meltwater communities (Musilova et al.2015; Cameron et al.2016). Following this trajectory, we found that meltwater selected for more habitat-specific microbial assemblages (Table S5, Supporting Information; Musilova et al.2015). Even without an active land bridge, polar ice sheets are under constant exchange with both near and distant habitats (Pearce et al.2009; Price et al.2009; Bottos et al.2014; Bowman and Deming 2017), introducing organisms that must establish themselves among existing communities in order to survive. Nonetheless, dominant microbial phylotypes in snow on polar ice sheets (i.e. Firmicutes, β- and γ-Proteobacteria; this study Fig. 3; Boetius et al.2015; Cameron et al.2015; Musilova et al.2015; Antony et al.2016) do not prevail in glacial ice surface communities. In fact, decreased meltwater dynamics within ice-margin habitats promotes microbial community establishment (Fountain et al.2004; Musilova et al.2015; Stibal et al.2015; Uetake et al.2016), dominated by taxa that were originally rare members in aeolian deposits (i.e. Bacteroidetes, Actinobacteria and Proteobacteria) (Edwards et al.2013; Hell et al.2013; Musilova et al.2015; Stibal et al.2015; Cameron et al.2016; Franzetti et al.2017). Relevant to the discussion on community establishment and stability in glacial meltwater habitats, our study provides evidence that microbial assemblages from samples within the same local environment exhibit a high degree of dissimilarity in compartments along the hydrologic flow path (i.e. ice, stream, stream sediments, and cryoconite holes). Differences in the fluorescent signatures of DOM between glaciers or habitats on the same glacier have been attributed to deposition and the microbial loop (Barker et al.2006; Pautler et al.2013; Musilova et al.2017). Similar to previous reports of glacial meltwater habitats (i.e. supraglacial ponds and streams, cryoconite holes, ice, and subglacial outflow, Bhatia et al.2010; Dubnick et al.2010; Pautler et al.2012, 2013; Barker et al.2013; Foreman et al.2013; Feng et al.2016; D’Andrilli et al., 2017; Smith et al. 2018), DOM fluorescence in the present study was dominated by protein-like components, indicative of microbially produced DOM. Despite the predominance of protein-like components, there was little to no overlap in the overall DOM composition across investigated habitats. In contrast to the terrestrial Canada Stream, the CG supraglacial stream was largely devoid of benthic substrates and microbial mats, features reflective of the overall DOM composition. In the CG stream, DOM is primarily derived from the exudation of labile DOM produced by planktonic microorganisms (Smith et al.2017), does not accumulate inter-seasonally (SanClements et al.2016), and appears to be less complex than from other natural environments (Foreman et al.2013; this study). Conversely, the presence of both protein- and humic-like fluorescence in the Canada stream DOM is typical of Antarctic streams with inter-annual DOM accumulation by microbial mats (McKnight and Tate 1997), characteristics also expressed in habitats shaped by the deposition of terrestrial mineral particles (i.e. CG parafluvial sediments and cryoconite). Striking was the lack of discrete regions of detectable fluorescent humic components in the west GrIS DOM. Increased aeolian deposition and microbial modifications of allochthonous material, including material from wildfires (Khan et al.2017), onto the GrIS was expected to contribute to chemically more heterogeneous DOM (Stibal et al.2010; Pautler et al.2013; Musilova et al.2017). However, the distribution and composition of DOM may vary, depending on the location, distance to the ice sheet margin, season (Stibal et al.2010; Langford et al.2010; Musilova et al.2017), and trajectories of air masses (Khan et al.2017). In each habitat, microbial assemblages were associated with either specific protein-like or humic-like fluorophores, with the calculated correlations between DOM components and OTUs of heterotrophic bacteria suggestive of contrasting taxonomic DOM preferences (Fig. 5). An in-depth analysis of the CG stream DOM determined it to be microbially derived and compositionally labile, with almost half of the DOM comprised of lipid-, protein- and carbohydrate-like molecular constituents (Smith et al.2018). As highlighted by substrate utilization assays (Sanyal et al.2018), Antarctic isolates from supraglacial meltwaters preferentially utilize these compounds. Specifically, diverse lineages within the Bacteroidetes, the most dominant members in all dry valley samples, are well known for degrading high molecular weight organic matter (i.e. proteins and carbohydrates), and grow attached to particles or algal cells (Thomas et al.2011; Fernández-Gomez et al.2013). Corroborating evidence for phylogenetic discrimination in carbon preference and algal-Bacteroidetes arrangements in glacial meltwater habitats has recently been reported for cryoconite holes on the Canada Glacier and CG supraglacial stream, Antarctica (Smith et al.2016). Incubation assays revealed that stream water DOM, dominated by protein-like fluorophores (designations B and T), was directly recycled byBacteroidetes in close spatial organization within Oscillatoria biofilms (Smith et al.2016). It was found that heterotrophic production by Bacteroidetes and Betaproteobacteria within the CG stream relied on freshly derived DOM, while there was no evidence of algal exudate uptake by Alphaproteobacteria (Smith et al.2017). Such differences in carbon preferences are consistent with the current understanding that—unlike Alphaproteobacteria—Bacteroidetes and Betaproteobacteria are capable of utilizing complex carbon sources of varying concentrations (Eiler et al.2003; Thomas et al.2011). In turn, the degradation of complex organic material by Bacteroidetes provides by-products directly targeted by Alpha- and Gammaproteobacteria (Williams et al.2013). Similarly, studies performed by Lutz and others described the production of taxon-specific metabolites by snow and ice algae (Lutz et al.2016, 2017). Associated with different algal blooms were specific taxonomic groups within the Bacteroidetes and Proteobacteria, attesting not only to DOM-microbe interactions, but also to the linkage between DOM exudates/composition and resident phylogeny (Lutz et al.2016, 2017). In contrast with previously reported findings on the microbial community in Greenlandic supraglacial habitats (Stibal et al.2015; Musilova et al.2015; Cameron et al.2012, 2016), in this studyGammaproteobacteria dominated the cryoconite hole and lake communities from the west GrIS. This class of Proteobacteria is generally known to be comprised of fast-growing opportunists (Pinhassi and Berman 2003; Pernthaler and Amann 2005). Further, Gammaproteobacteria have a competitive advantage in phosphate-limited environments (Ho et al.2017), characteristics typical for supraglacial habits (Mindl et al.2007; Stibal and Tranter 2007; Stibal et al.2009). Thus, in combination with extensive drainage of supraglacial habitats in the Paakitsoq region of west GrIS during the summer of 2011 (Banwell et al.2012), the dynamic surface environment may have been a selective force, giving rise to the dominance of opportunistic Gammaproteobacteria lineages. It is well documented in temperate aquatic environments that microbial community composition and underlying metabolic strategies are driven by the chemical composition of DOM (McCarren et al.2010; Koehler et al.2012; Hansell 2013; Buchan et al.2014; Moran et al.2016). Intricate substrate dependencies on specific DOM fluorophores have been shown for major taxonomic groups within many aquatic communities (e.g. Romera-Castillo et al.2011; Sarmento et al.2013; Amaral et al.2016; Blanchet et al.2017), with production and consumption of fluorescent DOM varying on a species level (Fukuzaki et al.2014; Fox et al.2017; Goto et al.2017). Evidence presented here establishes causality between DOM substrate composition and microbial community composition (Fig. 5) in glacial environments. Therefore, changes in DOM composition may result in dramatic shifts in the functional response of these glacial meltwater microbial systems, with subsequent consequences for downstream environments through alterations in the geochemical composition of fluxes. CONCLUSIONS While it is acknowledged that icy ecosystems possess diverse and metabolically active organisms (Boetius et al.2015), understanding the interactions between bacteria and glacial DOM remains an important challenge. Here we demonstrate that DOM composition has a major effect on glacial surface microbial communities, and by extension, influences ecosystem function. Additionally, our data contribute to the growing evidence that supraglacial habitats harbor spatially variable communities governed by local environmental factors (Cameron et al.2012, 2016; Edwards et al.2014; Stibal et al.2015). Undeniably, the global cryosphere is shrinking (Stocker et al.2013) and icy environments need to be understood within the context of climate change. Changes in microbial and functional diversity may have a profound impact on supraglacial DOM cycling, likely changing predicted quantities of carbon discharge (Anesio et al.2009; Cook et al.2012; Hodson et al.2015) and DOM bioavailability to downstream environments (Bhatia et al.2013; Lawson et al.2014; Musilova et al.2017; Smith et al.2017). SUPPLEMENTARY DATA Supplementary data are available at FEMSEC online. ACKNOWLEDGEMENTS The authors wish to thank the members of the Cotton Glacier and Greenland field teams, the Antarctic and Arctic support contractors and the pilots and crew of Petroleum Helicopters Inc. Any opinions, findings, or conclusions expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation (NSF). FUNDING This work was supported by the National Science Foundation [NSF ANT-0838970, ANT-1141978, and DGE-0654336], and through a NASA Earth and Science Space Fellowship. Conflict of interest. None declared. REFERENCES Amaral V, Graeber D, Calliari D et al.   Strong linkages between DOM optical properties and main clades of aquatic bacteria. Limnol Oceanogr . 2016; 61: 906– 18. Google Scholar CrossRef Search ADS   Anesio AM, Hodson AJ, Fritz A et al.   High microbial activity on glaciers: importance to the global carbon cycle. Global Change Biol . 2009; 15: 955– 60. Google Scholar CrossRef Search ADS   Anesio AM, Laybourn-Parry J. Glaciers and ice sheets as a biome. Trends Ecol Evol . 2012; 27: 219– 25. Google Scholar CrossRef Search ADS PubMed  Anesio AM, Lutz S, Chrismas NAM et al.   The microbiome of glaciers and ice sheets. NPJ Biofilms Microbiomes . 2017; 3: 10. Google Scholar CrossRef Search ADS PubMed  Antony R, Grannas AM, Willoughby AS et al.   Origin and sources of dissolved organic matter in snow on the East Antarctic ice sheet. Environ Sci Technol . 2014; 48: 6151– 9. Google Scholar CrossRef Search ADS PubMed  Antony R, Sanyal A, Kapse N et al.   Microbial communities associated with Antarctic snow pack and their biogeochemical implications. Microbiol Res . 2016; 192: 192– 202. Google Scholar CrossRef Search ADS PubMed  Antony R, Willoughby AS, Grannas AM et al.   Molecular insights on dissolved organic matter transformation by supraglacial microbial communities. Environ Sci Technol . 2017; 51: 4328– 37. Google Scholar CrossRef Search ADS PubMed  Banwell AF, Arnold NS, Willis IC et al.   Modeling supraglacial water routing and lake filling on the Greenland Ice Sheet. J Geophys Res Earth Surf . 2012; 117, F04012. Barker JD, Dubnick A, Lyons WB, et al.   Changes in dissolved organic matter (DOM) fluorescence in proglacial Antarctic streams. AAAR . 2013; 45: 305– 17. Google Scholar CrossRef Search ADS   Barker JD, Sharp MJ, Fitzsimons SJ et al.   Abundance and dynamics of dissolved organic carbon in glacier systems. AAAR . 2006; 38: 163– 72. Google Scholar CrossRef Search ADS   Bhatia MP, Das SB, Longnecker K et al.   Molecular characterization of dissolved organic matter associated with the Greenland ice sheet. Geochim Cosmochim Acta . 2010; 74: 3768– 84. Google Scholar CrossRef Search ADS   Bhatia MP, Das SB, Xu L et al.   Organic carbon export from the Greenland ice sheet. Geochim Cosmochim Acta . 2013; 109: 329– 44. Google Scholar CrossRef Search ADS   Blanchet M, Pringault O, Panagiotopoulos C et al.   When riverine dissolved organic matter (DOM) meets labile DOM in coastal waters: changes in bacterial community activity and composition. Aquat Sci . 2017; 79: 27– 43. Google Scholar CrossRef Search ADS   Boetius A, Anesio AM, Deming JW et al.   Microbial ecology of the cryosphere: sea ice and glacial habitats. Nat Rev Microbiol . 2015; 13: 677– 90. Google Scholar CrossRef Search ADS PubMed  Bottos EM, Woo AC, Zawar-Reza P et al.   Airborne bacterial populations above desert soils of the McMurdo Dry Valleys, Antarctica. Microbiol Ecol . 2014; 67: 120– 8. Google Scholar CrossRef Search ADS   Bowman JS, Deming JW. Wind-driven distribution of bacteria in coastal Antarctica: evidence from the Ross Sea region. Polar Biol . 2017; 40: 25– 35. Google Scholar CrossRef Search ADS   Bro R. PARAFAC. Tutorial and applications. Chemom Intell Lab Syst . 1997; 38: 149– 71. Google Scholar CrossRef Search ADS   Buchan A, LeCleir GR, Gulvik CA et al.   Master recyclers: features and functions of bacteria associated with phytoplankton blooms. Nat Rev Microbiol . 2014; 12: 686– 98. Google Scholar CrossRef Search ADS PubMed  Cameron KA, Hagedorn B, Dieser M et al.   Diversity and potential sources of microbiota associated with snow on western portions of the Greenland Ice Sheet. Environ Microbiol . 2015; 17: 594– 609. Google Scholar CrossRef Search ADS PubMed  Cameron KA, Hodson AJ, Osborn AM. Structure and diversity of bacterial, eukaryotic and archaeal communities in glacial cryoconite holes from the Arctic and the Antarctic. FEMS Microbiol Ecol . 2012; 82: 254– 67. Google Scholar CrossRef Search ADS PubMed  Cameron KA, Stibal M, Zarsky JD et al.   Supraglacial bacterial community structures vary across the Greenland ice sheet. FEMS Microbiol Ecol . 2016; 92: fiv164. Google Scholar CrossRef Search ADS PubMed  Caporaso JG, Kuczynski J, Stombaugh J et al.   QIIME allows analysis of high-throughput community sequencing data. Nat Methods . 2010; 7: 335– 6. Google Scholar CrossRef Search ADS PubMed  Chao A. Nonparametric estimation of the number of classes in a population. Scand J Stat . 1984; 265– 70. Christner BC, Mikucki JA, Foreman CM et al.   Glacial ice cores: a model system for developing extraterrestrial decontamination protocols. Icarus . 2005; 174: 572– 84. Google Scholar CrossRef Search ADS   Clarke KR, Warwick RM. Change in marine communities: an approach to statistical analysis and interpretation , 2nd edition. Primer-E Ltd, Plymouth UK 2001. Coble PG. Characterization of marine and terrestrial DOM in seawater using excitation-emission matrix spectroscopy. Mar Chem . 1996; 51: 325– 46. Google Scholar CrossRef Search ADS   Cook J, Edwards A, Hubbard A. Biocryomorphology: integrating microbial processes with ice surface hydrology, topography, and roughness. Front Earth Sci . 2015; 3: 78. Google Scholar CrossRef Search ADS   Cook JM, Hodson AJ, Anesio AM. An improved estimate of microbially mediated carbon fluxes from the Greenland ice sheet. J Glaciol . 2012; 58: 1098– 1108. Google Scholar CrossRef Search ADS   Cory RM, McKnight DM. Fluorescence spectroscopy reveals ubiquitous presence of oxidized and reduced quinones in dissolved organic matter. Environ Sci Technol . 2005; 39: 8142– 9. Google Scholar CrossRef Search ADS PubMed  D’Andrilli J, Smith HJ, Dieser M, Foreman CM. Climate driven carbon and microbial signatures through the last ice age. Geochem Persp Let . 2017; 4: 29– 34. Google Scholar CrossRef Search ADS   Dubnick A, Barker J, Sharp M et al.   Characterization of dissolved organic matter (DOM) from glacial environments using total fluorescence spectroscopy and parallel factor analysis. Ann Glaciol . 2010; 51: 111– 22. Google Scholar CrossRef Search ADS   Edgar RC, Haas BJ, Clemente JC et al.   UCHIME improves sensitivity and speed of chimera detection. Bioinformatics . 2011; 27: 2194– 2200. Google Scholar CrossRef Search ADS PubMed  Edwards A, Anesio AM, Rassner SM et al.   Possible interactions between bacterial diversity, microbial activity and supraglacial hydrology of cryoconite holes in Svalbard. ISME J . 2011; 5: 150. Google Scholar CrossRef Search ADS PubMed  Edwards A, Mur LAJ, Girdwood SE et al.   Coupled cryoconite ecosystem structure-function relationships are revealed by comparing bacterial communities in alpine and Arctic glaciers. FEMS Microbiol Ecol . 2014; 89: 222– 37. Google Scholar CrossRef Search ADS PubMed  Edwards A, Rassner SM, Anesio AM et al.   Contrasts between the cryoconite and ice-marginal bacterial communities of Svalbard glaciers. Polar Res . 2013; 32: 19468. Google Scholar CrossRef Search ADS   Eiler A, Langenheder S, Bertilsson S et al.   Heterotrophic bacterial growth efficiency and community structure at different natural organic carbon concentrations. Appl Environ Microbiol . 2003; 69: 3701– 9. Google Scholar CrossRef Search ADS PubMed  Fellman JB, Hood E, Spencer RGM. Fluorescence spectroscopy opens new windows into dissolved organic matter dynamics in freshwater ecosystems: a review. Limnol Oceanogr . 2010; 55: 2452– 62. Google Scholar CrossRef Search ADS   Feng L, Xu J, Kang S et al.   Chemical composition of microbe-derived dissolved organic matter in cryoconite in Tibetan Plateau glaciers: insights from Fourier transform ion cyclotron resonance mass spectrometry analysis. Environ Sci Technol . 2016; 50: 13215– 23. Google Scholar CrossRef Search ADS PubMed  Fernández-Gomez B, Richter M, Schüler M et al.   Ecology of marine Bacteroidetes: a comparative genomics approach. ISME J . 2013; 7: 1026. Google Scholar CrossRef Search ADS PubMed  Foreman CM, Cory RM, Morris CE et al.   Microbial growth under humic-free conditions in a supraglacial stream system on the Cotton Glacier, Antarctica. Environ Res Lett . 2013; 8: 035022. Google Scholar CrossRef Search ADS   Fountain AG, Tranter M, Nylen TH et al.   Evolution of cryoconite holes and their contribution to meltwater runoff from glaciers in the McMurdo Dry Valleys, Antarctica. J Glaciol . 2004; 50: 35– 45. Google Scholar CrossRef Search ADS   Fox BG, Thorn RM, Anesio AM et al.   The in situ bacterial production of fluorescent organic matter; an investigation at a species level. Water Res . 2017; 125: 350– 9. Google Scholar CrossRef Search ADS PubMed  Franzetti A, Navarra F, Tagliaferri I et al.   Potential sources of bacteria colonizing the cryoconite of an Alpine glacier. PloS One . 2017; 12: e0174786. Google Scholar CrossRef Search ADS PubMed  Fukuzaki K, Imai I, Fukushima K et al.   Fluorescent characteristics of dissolved organic matter produced by bloom-forming coastal phytoplankton. J Plankton Res . 2014; 36: 685– 94. Google Scholar CrossRef Search ADS   Gawor J, Grzesiak J, Sasin-Kurowska J et al.   Evidence of adaptation, niche separation and microevolution within the genus Polaromonas on Arctic and Antarctic glacial surfaces. Extremophiles . 2016; 20: 403– 13. Google Scholar CrossRef Search ADS PubMed  Goto S, Tada Y, Suzuki K et al.   Production and reutilization of fluorescent dissolved organic matter by a marine bacterial strain, Alteromonas macleodii. Front Microbiol . 2017; 8: 507 Google Scholar CrossRef Search ADS PubMed  Hammer Ø, Harper DA, Ryan PD. PAST: Paleontological Statistics Software Package for education and data analysis. Palaeontol Electron 4 . 2001. Hansell DA. Recalcitrant dissolved organic carbon fractions. Ann Rev Mar Sci . 2013; 5: 421– 45. Google Scholar CrossRef Search ADS PubMed  Hell K, Edwards A, Zarsky J et al.   The dynamic bacterial communities of a melting high Arctic glacier snowpack. ISME J . 2013; 7: 1814. Google Scholar CrossRef Search ADS PubMed  Ho A, Di Lonardo DP, Bodelier PL. Revisiting life strategy concepts in environmental microbial ecology. FEMS Microbiol Ecol . 2017; 93: 10. Hodson A, Bøggild C, Hanna E et al.   The cryoconite ecosystem on the Greenland ice sheet. Ann Glaciol . 2010; 51: 123– 9. Google Scholar CrossRef Search ADS   Hodson A, Brock B, Pearce D et al.   Cryospheric ecosystems: a synthesis of snowpack and glacial research. Environ Res Lett . 2015; 10: 110201. Google Scholar CrossRef Search ADS   Hood E, Battin TJ, Fellman J, O’Neel S et al.   Storage and release of organic carbon from glaciers and ice sheets. Nat Geosci . 2015; 8: 91– 6. Google Scholar CrossRef Search ADS   Hood E, Fellman J, Spencer RG et al.   Glaciers as a source of ancient and labile organic matter to the marine environment. Nature . 2009; 462: 1044– 7. Google Scholar CrossRef Search ADS PubMed  Khan AL, Wagner S, Jaffé R et al.   Dissolved black carbon in the global cryosphere: concentrations and chemical signatures. Sci Invest Rep . 2017; 44: 6226– 34. Koehler B, Wachenfeldt E, Kothawala D et al.   Reactivity continuum of dissolved organic carbon decomposition in lake water. J Geophys Res-Biogeo . 2012; 117: G01024. Google Scholar CrossRef Search ADS   Langford H, Hodson A, Banwart S et al.   The microstructure and biogeochemistry of Arctic cryoconite granules. Ann Glaciol . 2010; 51: 87– 94. Google Scholar CrossRef Search ADS   Lawaetz AJ, Stedmon CA. Fluorescence intensity calibration using the Raman scatter peak of water. Appl Spectrosc . 2009; 63: 936– 40. Google Scholar CrossRef Search ADS PubMed  Lawson EC, Wadham JL, Tranter M et al.   Greenland Ice Sheet exports labile organic carbon to the Arctic oceans. Biogeosci . 2014; 11: 4015– 28. Google Scholar CrossRef Search ADS   Legrand M, Preunkert S, Jourdain B et al.   Water-soluble organic carbon in snow and ice deposited at Alpine, Greenland, and Antarctic sites: a critical review of available data and their atmospheric relevance. Clim Past . 2013; 9: 2195. Google Scholar CrossRef Search ADS   Lutz S, Anesio AM, Raiswell R et al.   The biogeography of red snow microbiomes and their role in melting arctic glaciers. Nat Commun . 2016; 7: 11968. Google Scholar CrossRef Search ADS PubMed  Lutz S, Anesio AM, Edwards A et al.   Linking microbial diversity and functionality of arctic glacial surface habitats. Environ Microbiol . 2017; 19: 551– 65. Google Scholar CrossRef Search ADS PubMed  McCarren J, Becker JW, Repeta DJ et al.   Microbial community transcriptomes reveal microbes and metabolic pathways associated with dissolved organic matter turnover in the sea. PNAS USA . 2010; 107: 16420– 7. Google Scholar CrossRef Search ADS PubMed  McKnight DM, Boyer EW, Westerhoff PK. Spectrofluorometric characterization of dissolved organic matter for indication of precursor organic material and aromaticity. Limnol Oceanogr . 2001; 46: 38– 48. Google Scholar CrossRef Search ADS   McKnight DM, Tate CM. Canada Stream: a glacial meltwater stream in Taylor Valley, South Victoria Land, Antarctica. J North Am Benthol Soc . 1997; 16: 14. Google Scholar CrossRef Search ADS   Miller MP, McKnight DM. Comparison of seasonal changes in fluorescent dissolved organic matter among aquatic lake and stream sites in the Green Lakes Valley. J Geophys Res-Biogeosci . 2010; 115: G00F12. Google Scholar CrossRef Search ADS   Mindl B, Anesio AM, Meirer K et al.   Factors influencing bacterial dynamics along a transect from supraglacial runoff to proglacial lakes of a high Arctic glacier. FEMS Microbiol Ecol . 2007; 59: 307– 17. Google Scholar CrossRef Search ADS PubMed  Moran MA, Kujawinski EB, Stubbins A, et al.   Deciphering ocean carbon in a changing world. PNAS . 2016; 113: 3143– 51. Google Scholar CrossRef Search ADS PubMed  Mueller DR, Pollard WH. Gradient analysis of cryoconite ecosystems from two polar glaciers. Polar Biol . 2004; 27: 66– 74. Google Scholar CrossRef Search ADS   Musilova M, Tranter M, Bennett SA et al.   Stable microbial community composition on the Greenland Ice Sheet. Front Microbiol . 2015; 6: 193. Google Scholar CrossRef Search ADS PubMed  Musilova M, Tranter M, Wadham J et al.   Microbially driven export of labile organic carbon from the Greenland ice sheet. Nat Geosci . 2017; 10: 360– 5. Google Scholar CrossRef Search ADS   Nguyen NH, Smith D, Peay K et al.   Parsing ecological signal from noise in next generation amplicon sequencing. New Phytol Lett . 2015; 205: 1389– 93. Google Scholar CrossRef Search ADS   Pautler BG, Dubnick A, Sharp MJ et al.   Comparison of cryoconite organic matter composition from Arctic and Antarctic glaciers at the molecular-level. Geochim Cosmochim Acta . 2013; 104: 1– 18. Google Scholar CrossRef Search ADS   Pautler BG, Woods GC, Dubnick A et al.   Molecular characterization of dissolved organic matter in glacial ice: coupling natural abundance 1H NMR and fluorescence spectroscopy. Environ Sci Technol . 2012; 46: 3753– 61. Google Scholar CrossRef Search ADS PubMed  Pearce DA, Bridge PD, Hughes KA et al.   Microorganisms in the atmosphere over Antarctica. FEMS Microbiol Ecol . 2009; 69: 143– 57. Google Scholar CrossRef Search ADS PubMed  Pernthaler J, Amann R. Fate of heterotrophic microbes in pelagic habitats: focus on populations. Microbiol Mol Biol Rev . 2005; 69: 440– 61. Google Scholar CrossRef Search ADS PubMed  Pinhassi J, Berman T. Differential growth response of colony-forming α-and γ-proteobacteria in dilution culture and nutrient addition experiments from Lake Kinneret (Israel), the eastern Mediterranean Sea, and the Gulf of Eilat. Appl Environ Microb . 2003; 69: 199– 211. Google Scholar CrossRef Search ADS   Price PB, Rohde RA, Bay RC. Fluxes of microbes, organic aerosols, dust, sea-salt Na ions, non-sea-salt Ca ions, and methanesulfonate onto Greenland and Antarctic ice. Biogeosci . 2009; 6: 479– 86. Google Scholar CrossRef Search ADS   Pruesse E, Quast C, Knittel K et al.   SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res . 2007; 35: 7188– 96. Google Scholar CrossRef Search ADS PubMed  Romera-Castillo C, Sarmento H, Álvarez-Salgado XA et al.   Net production and consumption of fluorescent colored dissolved organic matter by natural bacterial assemblages growing on marine phytoplankton exudates. Appl Environ Microb . 2011; 77: 7490– 8. Google Scholar CrossRef Search ADS   Salter SJ, Cox MJ, Turek EM et al.   Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol . 2014; 12: 87. Google Scholar CrossRef Search ADS PubMed  SanClements MD, Smith HJ, Foreman CM et al.   Biogeophysical properties of an expansive Antarctic supraglacial stream. Antarct Sci . 2016; 29: 33– 44. Google Scholar CrossRef Search ADS   Sanyal A, Antony R, Samui G et al.   Microbial communities and their potential for degradation of dissolved organic carbon in cryoconite hole environments of Himalaya and Antarctica. Microbiol Res . 2018, 208: 32– 42. Google Scholar CrossRef Search ADS PubMed  Sarmento H, Romera-Castillo C, Lindh M et al.   Phytoplankton species‐specific release of dissolved free amino acids and their selective consumption by bacteria. Limnol Oceanogr . 2013; 58: 1123– 35. Google Scholar CrossRef Search ADS   Schloss PD, Westcott SL, Ryabin T et al.   Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol . 2009; 75: 7537– 41. Google Scholar CrossRef Search ADS PubMed  Schloss PD, Gevers D, Westcott SL. Reducing the effects of PCR amplification and sequencing artifacts on 16S rRNA-based studies. PLoS One . 2011; 6: e27310. Google Scholar CrossRef Search ADS PubMed  Simpson EH. Measurement of diversity. Nature . 1949; 163: 688. Google Scholar CrossRef Search ADS   Smith HJ, Foster RA, McKnight DM et al.   Microbial formation of labile organic carbon in Antarctic glacial environments. Nat Geosci . 2017; 10: 356– 9. Google Scholar CrossRef Search ADS   Smith HJ, Schmit A, Foster R et al.   Biofilms on glacial surfaces: hotspots for biological activity. NPJ Biofilms Microbiomes . 2016; 2: 16008. Google Scholar CrossRef Search ADS PubMed  Smith HJ, Tigges M, D’Andrilli J et al.   Dynamic processing of DOM: insight from exometabolomics, fluorescence spectroscopy and mass spectrometry. Limnol Oceanogr Lett . 2018; 3: 225– 35. Google Scholar CrossRef Search ADS   Stibal, M, Anesio, AM, Blues, CJD. Phosphatase activity and organic phosphorus turnover on a high Arctic glacier. Biogeosci . 2009; 6: 913– 22. Google Scholar CrossRef Search ADS   Stibal M, Lawson EC, Lis GP et al.   Organic matter content and quality in supraglacial debris across the ablation zone of the Greenland ice sheet. Ann Glaciol . 2010; 51: 1– 8. Google Scholar CrossRef Search ADS   Stibal M, Sabacka M, Zarsky J. Biological processes on glacier and ice sheet surfaces. Nat Geosci . 2012; 5: 771– 4. Google Scholar CrossRef Search ADS   Stibal M, Schostag M, Cameron KA et al.   Different bulk and active bacterial communities in cryoconite from the margin and interior of the Greenland ice sheet. Environ Microbiol Rep . 2015; 7: 293– 300. Google Scholar CrossRef Search ADS PubMed  Stibal M, Tranter M. Laboratory investigation of inorganic carbon uptake by cryoconite debris from Werenskioldbreen, Svalbard. J Geophys Res . 2007; 11: G04S33. Stocker TF, Qin D, Plattner GK et al.   IPCC, 2013: climate change 2013: the physical science basis . Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change, Cambridge University Press, Cambridge, U.K., 2013: 1535. Stubbins A, Hood E, Raymond PA et al.   Anthropogenic aerosols as a source of ancient dissolved organic matter in glaciers. Nat Geosci . 2012; 5: 198– 201. Google Scholar CrossRef Search ADS   Takeuchi N, Nagatsukan N, Uetake J et al.   Spatial variations in impurities (cryoconite) on glaciers in northwest Greenland. Bull Glaciol Res . 2014; 32: 85– 94. Google Scholar CrossRef Search ADS   Tedesco M, Willis IC, Hoffman MJ et al.   Ice dynamic response to two modes of surface lake drainage on the Greenland ice sheet. Environ Res Lett . 2013; 8: 034007. Google Scholar CrossRef Search ADS   Telling J, Anesio AM, Tranter M et al.   Controls on the autochthonous production and respiration of organic matter in cryoconite holes on high Arctic glaciers. J Geophys Res . 2012; 117: G01017. Google Scholar CrossRef Search ADS   Thomas F, Hehemann JH, Rebuffet E et al.   Environmental and gut bacteroidetes: the food connection. Front Microbiol . 2011; 30; 2: 93. Uetake J, Tanaka S, Segawa T et al.   Microbial community variation in cryoconite granules on Qaanaaq Glacier, NW Greenland. FEMS Microbiol Ecol . 2016; 92: fiw127. Google Scholar CrossRef Search ADS PubMed  Wang Q, Garrity GM, Tiedje JM et al.   Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol . 2007; 73: 5261– 7. Google Scholar CrossRef Search ADS PubMed  Williams TJ, Wilkins D, Long E et al.   The role of planktonic Flavobacteria in processing algal organic matter in coastal East Antarctica revealed using metagenomics and metaproteomics. Environ Microbiol . 2013; 15: 1302– 17. Google Scholar CrossRef Search ADS PubMed  © FEMS 2018. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

FEMS Microbiology EcologyOxford University Press

Published: May 14, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off