Bacterial community structure and functional profiling of high Arctic fjord sediments

Kongsfjorden, an Arctic fjord is significantly affected by the glacier melt and Atlantification, both the processes driven by accelerated warming in the Arctic. This has lead to changes in primary production, carbon pool and microbial communities, especially that in the sediment. In this study, we have examined the bacterial community structure of surface (0–2 cm) and subsurface (3–9 cm) sediments of Kongsfjorden using the high throughput sequencing analysis. Results revealed that bacterial community structure of Kongsfjorden sediments were dominated by phylum Proteobacteria followed by Bacteroidetes and Epsilonbacteraeota. While α- and γ-Proteobacterial class were dominant in surface sediments; δ-Proteobacteria were found to be predominant in subsurface sediments. The bacterial community structure in the surface and subsurface sediments showed significant variations (p ≤ 0.05). Total organic carbon could be one of the major parameters controlling the bacterial diversity in the surface and subsurface sediments. Functional prediction analysis indicated that the bacterial community could be involved in the degradation of complex organic compounds such as glycans, glycosaminoglycans, polycyclic aromatic hydrocarbons and also in the biosynthesis of secondary metabolites.


Introduction
Marine sediment depicts most complex microbial habitats on earth. Benthic bacterial communities in the ocean play a significant role in remineralization of organic matter (Ravenschlag et al. 2000). Marine microorganisms hydrolyse high molecular weight organic matter to sufficiently small constituents for their cellular intake and also make them easily available for higher trophic levels. Bacteria and archaea are the dominant microbial communities in marine sediments. These sediment communities are impacted by several physical and chemical parameters (Nguyen and Landfald 2015;Jorgensen et al. 2012) and can be sensitive to environmental changes and are also affected by geographic distance and ocean currents (Hamdan et al. 2013;Xiong et al. 2014). Hence microbial community composition analysis is important to understand the benthic ecosystem processes and thereby to study about the effect of climate change.
Kongsfjorden, located on the northwest coast of Spitsbergen in the Svalbard Archipelago is a tide water glacial fjord in the Arctic. Atlantic water inflow as well as several glaciers influence the water column in the Kongsfjorden (Hop et al. 2002). The fjord is now identified as an ideal environment to study the impacts of climate change in the Arctic, indicated by the decrease in ice cover and a rise in melting. Physical gradients of salinity and temperature induced by the glacier melting, which generate turbidity and also sedimentation along the fjord system (Hop et al. 2002;Kotwicki et al. 2004) that ultimately affects the distribution of pelagic and benthic organisms (Kotwicki et al. 2004;Piquet et al. 2014). At the head of the fjord, tidal glaciers discharge fresh water and suspended loads, which generate steep environmental gradients (Wȩsławski et al. 2000), biomass and diversity of inner fjord benthic community are also affected (Hop et al. 2002). Thus, marine sediment microbial structure and function are greatly influenced by the changes in the physicochemical properties of water such as temperature, nutrient availability, oxygen saturation, hydrodynamics and light patterns, being linked to climate change.
Bacterial diversity in an environment is usually discovered by both culture dependent and independent techniques. Culture dependent approaches enable to characterize the cultivated bacteria and their ecological and physiological role. Nevertheless, through culture dependent approaches only a small part (< 1%) of the bacteria can be cultivated (Cole et al. 2014). Culture independent methods detect the bacterial diversity which can be cultured as well as that are yet to be cultured. Previous studies have mentioned that there are complex microbial groups present in marine sediments, although most of the sequences were distantly related to the culturable bacteria. Moreover, presence of a significant fraction of bacteria that do not belong to any known taxonomic group indicates the presence of unique species which warrants further study of bacterial communities in the marine sediments (Xuezheng et al. 2014;Zhang et al. 2015;Sinha et al. 2019). Culture independent methods, besides providing insights in to the uncultured diversity, also help us to access the functional genes and biochemical pathways within these uncultured microorganisms (Simon and Daniel 2011).
Advancement of culture independent methods has revolutionised the studies of bacterial communities in Arctic fjords. In Arctic sediments, various culture independent bacterial diversity studies have been conducted Zeng et al. 2011Zeng et al. , 2017Fang et al. 2019). The primary objective of this study was to explore the bacterial diversity of surface (0-2 cm) and subsurface (3-9 cm) sediments from various stations of Kongsfjorden, Arctic. Metagenomic analysis of V 3 -V 4 amplicon regions of 16 S rRNA genes were carried out in this study by using Illumina HiSeq 2500 platform. This study allows broad understanding of bacterial diversity in surface and subsurface sediments of Kongsfjorden.

Site description and sample collection
Kongsfjorden is situated on the western side of Spitsbergen in Svalbard archipelago (79° N, 12° E), which is influenced by both Arctic and Atlantic water. The sediment samples were collected on board research vessel R V Teisten from four locations of the Kongsfjorden (Fig. 1) using a Van Veen's grab during Arctic expedition in October 2016. Using a sterile glass corer of one-inch diameter, surface (0-2 cm) and subsurface sediments (3-9 cm) from four stations were subsampled aseptically from the undisturbed sediments collected in the grab. Physico-chemical variables of the sediment samples are presented in Table 1. The sediment samples were frozen at − 80 °C and brought to the laboratory at Cochin University of Science and Technology (CUSAT) for metagenomic analysis.

Geochemical analysis
Geochemical characteristics of eight sediment samples from four stations were carried out. Total carbon, nitrogen and sulphur analysis of sediments using CHNS analyzer (ELE-MENTAR Vario EL111) and analysis of trace elements by inductively coupled plasma mass spectrometry (Icap RQ ICP-MS) were done at Sophisticated Test and Instrumentation Centre (STIC), Cochin University of Science and Technology (CUSAT). Total organic carbon was detected by using Walkley-Black titration method (Walkley and Black 1947). Statistical analysis (ANOVA) was conducted to find out the significant relationship between environmental parameters.

DNA extraction and PCR amplification
As per manufacturer's instructions, total genomic DNA of eight sediment samples from four different stations was extracted using Ultra Clean soil DNA isolation kit (MO BIO, CA, USA). DNA concentration was determined using Qubit™ 4 Fluorometer and quality of the isolated DNA samples were checked by agarose gel electrophoresis using 1% agarose gel. Isolated DNA from the surface sediment samples (0-2 cm) and sub-surface sediment samples (3-9 cm) of four stations were pooled separately and sent for 16 S V 3 -V 4 amplicon sequencing (Macrogen Inc, Republic of Korea). The universal primer set Pro341F/Pro805R was used for amplification of V 3 -V 4 regions of 16 S rRNA (Takahashi et al. 2014).

S rRNA library preparation and sequencing
In order to generate amplicon library, a unique pair of barcoded primers with Illumina adaptors were used. By using Qubit 2.0 Fluorometer and Agilent Bioanalyzer 2100 system quality of the libraries was analysed. Library was loaded on a flow cell for the cluster generation, these fragments are occupied on the surface of oligos which are complementary to the adapters in the library. Discrete clonal clusters were generated by using the fragments through bridge amplification. Subsequent to cluster generation, sequencing of templates was performed on an Illumina HiSeq 2500 platform (2 × 300 bp size). Afterwards, DNA sequence data was submitted in the NCBI Sequencing Read Archive with accession number PRJNA678587.

Quality filtering, operational taxonomic units (OTUs) picking and annotation
After sequencing on an Illumina HiSeq platform raw fastq reads were created. The 16S rRNA primers and barcodes were removed to produce paired-end reads using in-house PERL script. Quality of the reads was detected using the FastQC program. Trimmed paired-end reads with Phred score quality > Q20 was taken for consensus generation. The paired-end reads were combined using FLASH tool (latest version. 1.2.11). By using UCLUST program, operational taxonomic units (OTUs) were generated by clustering the merged reads on the basis of their sequence similarity (similarity cut off = 0.97). In this step, singleton OTUs were removed. Then UCHIME method is applied for chimera removal using VSEARCH tool. The OTUs were processed using QIIME programme to generate a characteristic sequence for every OTU, then the sequence is aligned into SILVA rRNA 132 rRNA reference database. By using QIIME software OTUs depicting chloroplasts and mitochondrial sequences were cleared (Caporaso et al. 2010).

Diversity analysis
For the analysis of alpha diversity, OTU numbers were rarefied. Shannon, Chao1 and observed species metrices were used to determine alpha diversity indices. Venny 2.1 software was used to create a Venn diagram to indicate the number of OTUs in surface and subsurface sediments and the common OTUs shared between the surface and subsurface sediments. ANOVA was performed to find out the difference in bacterial community structure in the surface and subsurface samples.

Prediction of functional profiles
Prediction of functional gene profiles of bacterial community composition in the surface and subsurface sediments were generated on the basis of 16 S rRNA gene sequences data using PICRUSt2 (Langille et al. 2013;Douglas et al. 2020). Greengenes database was used to compile the data for PICRUSt analysis. By using the database, OTUs were picked on the basis of 97% sequence similarity. By using KEGG Orthology (KO), estimate of the number of predicted gene families were obtained from metagenomic predictions. These counts were introduced into KEGG pathways and relative abundance for KO and KEGG pathways were calculated. NSTI (nearest sequenced taxon index) values were used to analyse the accuracy of functional prediction.

Characteristics of sediment samples
The sampling sites and geochemical characteristics of the 8 sediment samples (both surface and subsurface sediments) from four different stations were analysed and summarized (Table 1). Total carbon and total organic carbon in the surface sediments varied from 2.17 to 3.38% and 0.94 to 2.86% respectively while that in subsurface sediments varied from 1.75 to 3.15% and from 0.39 to 1.68% respectively. Total nitrogen in the surface and subsurface sediment samples ranged from 0.06 to 0.3% and from 0.05 to 0.25% respectively. TS (total sulphur) was detected only in sediments from two stations. Subsurface sediment of station-1 had higher amount of TS (0.2%) than the surface sediments of station-1 and station-2, where both the stations had 0.12% TS. The Mg, P, Ca and Fe content of surface sediment samples were found to vary from 2.62 to 4.95%, 0.20 to 0.49%, 0.84 to 1.45% and 2.10 to 2.81% respectively, while that in subsurface sediments varied from 3.17 to 3.82%, 0. TC, TOC and TN contents were found to be higher in the surface sediment of station-1. Mg (magnesium) and Mn (manganese) content were found to be higher in surface sediment of station-3. Amount of P (phosphorus) was higher in subsurface sediment of station-2. Compared to other stations Ca (calcium) content was lower in surface sediment of station-1 and Fe (iron) content was lower in subsurface sediment of station-4. Amount of Mo (molybdenum) was higher in subsurface sediment of station-1. Amount of Co (cobalt) was found to be comparatively lower in both surface and subsurface sediments of station-1 and higher in both surface and subsurface sediments of station-3. Cu (copper) content was higher in subsurface sediment of station-3. Compared to surface sediments, subsurface sediments had higher content of Zn (zinc). Statistical analyses indicated that there were no significant statistical variation in the environmental parameters (TC, TOC, Fe, Mg, Ca, Mo and TOC:TP ratio) among the four stations of Kongsfjorden (Supplementary Table 1), however, the parameters such as TC, TOC, TN, TC:TN showed significant difference in surface and subsurface sediments (p ≤ 0.05). It clearly indicates that TC, TOC and TN are important parameters controlling the geochemistry and bacterial community structure in the surface and subsurface sediments.

Diversity of bacteria in the sediments of Kongsfjorden
A diverse group of microbial community have been reported from Arctic sediments as well as in the water column. In this study, total of 2129 OTUs from surface sediments and 2203 OTUs from subsurface sediments were obtained. In total 4332 OTU and 158,021 reads were identified among surface and subsurface sediments. The identified OTUs belonged to thirty six known phyla, one unknown phylum, thirteen candidate phyla and an uncultured phylum. In addition, 156 sequences remained unclassified and were not found in reference databases (Supplementary Table 2). Top 25 phyla with higher relative abundance is shown in Table 2.

Functional potentials of bacterial diversity
Functional diversity of bacterial community composition in surface and subsurface sediments were revealed by PIC-RUSt2 assignment of KO's to KEGG pathways. In surface Fig. 5 The relative abundance of predicted functional genes of microbial communities from surface and subsurface sediments based on KEGG database sediments (V-A), total of 6627 different KO's which containing 88,924,551 counts were predicted. Total of 6549 KO's with 70,697,628 counts were predicted in subsurface sediments (V-B). In surface sediments, these counts were introduced to 164 KEGG pathways and indicate NSTI score of 0.21, which shows reliability of functional prediction. In subsurface sediments, counts were assigned to 160 KEGG pathways and NSTI score of 0.29 was indicated. Functional prediction showed significant variations in the surface and subsurface sediments (p ≤ 0.05). KEGG pathways which having important role in cellular homeostasis such as, carbohydrate metabolism (12.89 and 13.21%), amino acid metabolism (12.66 and 12.57%), energy metabolism (6.3 and 6.21%), replication and repair (5.17 and 5.44%), xenobiotic degradation and metabolism (4.45 and 3.77%), translation (3.02 and 3.34%), biosynthesis of other secondary metabolites (2.07 and 2.15%) and membrane transport (1.58 and 1.71%) were identified in the surface and subsurface sediments respectively (Fig. 5). In carbohydrate metabolism, predicted gene clusters for complex polysaccharide degradation such as β-glucosidase (0.08 and 0.04%) and endoglucanase (0.03%) was observed in surface and subsurface bacterial community. Genes for endo-1,4-beta xylanase, D-Inositol-3-phosphate glycosyl transferase, glycan degradation, glycosaminoglycan degradation, polycyclic aromatic hydrocarbon degradation were also observed in low abundance. Gene clusters for glycogen phosphorylase was observed in bacterial community of subsurface sediments (0.07%). KEGG pathways for xenobiotic degradation consisted of toluene degradation (0.52 and 0.62%), chloroalkane and chloroalkene degradation (0.42 and 0.32%), nitrotoluene degradation (0.35 to 0.61%), caprolactam degradation (0.28 and 0.20%) and benzoate degradation (0.23%); similarly, pathways for biosynthesis of other secondary metabolites indicated by streptomycin biosynthesis (1.51 and 1.58%) were also observed at level 2 in surface and subsurface sediment bacterial community respectively.

Discussion
The deposition and long-term burial of organic carbon in marine sediments has played a significant role in controlling atmospheric oxygen and carbon dioxide concentrations over the past 500 million years (Berner 1982). Carbon burial in polar environments represents the dominant natural mechanism of long-term organic carbon sequestration with respect to climate change . Fjords have been considered as hotspots of organic carbon burial as they received large amounts of carbon through annual productivity cycle and glacial melt. For instance, Kongsfjorden receives 5-10% of the organic carbon from terrestrial origin by melting of glaciers and about 90-95% of the organic carbon through primary production primarily fuelled by intrusion of Atlantic water (Kuliński et al. 2014;Buchholz and Wiencke 2016) reported that macroalgae, in particular Kelps contribute significantly to the pool of organic carbon in the Kongsfjorden. The sampling was conducted at the onset of winter; as the season is ideal for studying the microbial communities associated with organic carbon degradation in the Kongsfjorden. We found that spatially there was no significant difference in the TC and TOC content among the stations. Further, Sinha et al. (2017a, b) have reported that seasonal variability in community is much more profound than spatial variability within a particular season. However, it is pertinent to note that vertical variability in distribution of organic carbon between the surface and subsurface sediment showed significant variation (p ≤ 0.05).
Considering the above, the strategy adopted in this study with respect to sampling, extraction of DNA and sequencing could be adequate to delineate the diversity of bacteria in the surface and subsurface sediments during the onset of winter. We pooled the DNA sample in the surface stations together and subsurface stations together for bacterial diversity analysis because no significant variation in the organic matter content in the different stations. Our study delineated the observations that organic matter content may have influences the structuring the bacterial communities in the surface and subsurface sediments of Kongsfjorden. The functional profiling indicated that the bacteria were involved in the various remineralization processes of organic carbon such as glycan degradation, glycosaminoglycan degradation, polycyclic aromatic hydrocarbon degradation, etc.
In this study, bacterial community structure of surface and subsurface sediments of Kongsfjorden revealed through Illumina sequencing is consistent with several previous studies on bacterial community pattern of Arctic sediments (Ravenschlag et al. 1999(Ravenschlag et al. , 2001Li et al. 2009Li et al. , 2015Teske et al. 2011;Xuezheng et al. 2014;Zhang et al. 2014;Zeng et al. 2017;Fang et al. 2019). The dominant phyla identified in our study was known to be widely seen in deep sea sediments (Zhang et al. 2015;Sinha et al. 2019), Crustal fluids of a deep sea hydrothermal field (Kato et al. 2013) and deep sea vents (Sylvan et al. 2012). Studies of Tian et al. (2009) disclosed that the 16 S rRNA gene clone sequences retrieved from Kongsfjorden sediments having similarity to the bacterial sequences obtained from several hydrothermal vent sites. This could be very likely due to the seepage of deep ocean sediments into fjord basins by the action of bottom water currents.
In our study, Proteobacteria was the most dominant phyla comprising 39.87 and 33.25% of the total bacterial community structure in the surface and subsurface sediments respectively. Within this phylum, γ-Proteobacteria and δ-Proteobacteria were the two main classes encountered. In the surface sediments, the γ-Proteobacteria became Table 3 List of bacterial groups reported in diverse locations in several studies similar to the dominant bacterial groups identified in this study

Dominant bacteria and reported study locations References
Proteobacteria were most abundant followed by Epsilonbacteraeota, Bacteroidetes, Chloroflexi, Verrucomicrobia, Planctomycetes, Acidobacteria, Fusobacteria and Actinobacteria. In the phylum Proteobacteria, Gammaproteobacteria was more abundant in surface sediments followed by Deltaproteobacteria; Deltaproteobacteria exhibited higher abundance than Gammaproteobacteria in subsurface sediments. Genera Sulfurovum and Sulfurimonas found to be dominant in surface and subsurface sediments  Bano and Hollibaugh (2002), Sinha et al. (2017a, b) Firmicutes the largest group (18.09%) followed by δ-Proteobacteria (17.99%), whereas the α-Proteobacteria contributed only 3.72%. In the subsurface sediments, δ-Proteobacteria was the dominant class of phylum Proteobacteria (24.58%), whereas the γ-Proteobacteria made up 7.76%. In contrast, to the abundance of γ and δ-Proteobacteria, abundance of α-Proteobacteria (0.88%) was lower in subsurface sediments. Similar findings was also reported in a study of Teske et al. (2011) on microbial community composition of surface and subsurface sediments from a fjord of Svalbard by using 16 S rRNA gene clone library analysis. They have observed dominance of γ-Proteobacteria in surface sediments and δ-Proteobacteria in subsurface sediments. Several studies have revealed that the Proteobacteria was the most cosmopolitan group in surface sediments (Ravenschlag et al. 2001;) and γ-Proteobacteria was the most significant group of bacteria in majority of the marine sediments (Li et al. 1999;Oni et al. 2015). Similar observations were also reported in sediments of diverse locations as shown in Table 3. In several studies, δ-Proteobacteria comprise a significant fraction of bacterial communities in marine surface sediments (Park et al. 2011;Bienhold et al. 2016), a finding similar to our observations, where we have got significant fraction of δ-Proteobacteria in surface sediments (17.99%). However, the dominance of δ-Proteobacteria was slightly less than γ-Proteobacteria (18.09%) in surface sediments. The γ and δ-Proteobacteria are the predominant group in various sediments of diverse locations listed in Table 3.
Other dominant groups of bacteria detected in our study comprise members of phyla Epsilonbacteraeota, Bacteroidetes, Chloroflexi, Verrucomicrobia, Planctomycetes, Acidobacteria, Fusobacteria, Actinobacteria, Spirochaetes, Kiritimatiellaeota, Firmicutes and Latescibacteria (Supplementary Table 10). Similar bacterial phyla reported in other studies is detailed in Table 4. The microbial community composition has slight differences in surface and subsurface sediments. Bacteroidetes was second most dominant phyla (23.47%) in surficial sediments, while in subsurface sediments, Epsilonbacteraeota (32.09%) was the second most dominant. In surface sediments phylum Epsilonbacteraeota decreased to 7.65% and in subsurface sediments phylum Bacteroidetes decreased to 8.61%. Phylum Chloroflexi, Acidobacteria, Actinobacteria, Latescibacteria and Planctomycetes were more dominant in subsurface sediments and predominance of phylum Verrucomicrobia, Fusobacteria and Firmicutes were also have been observed in surface sediments. In our study, Chloroflexi was more abundant in subsurface sediments which is agreement with the study of Teske et al. (2011). Several molecular methods and studies have discovered that the Actinobacteria constitute a small fraction in marine bacterial communities, however perform significant ecological function in marine environments (Venter et al. 2004;Ward and Bora 2006;Polymenakou et al. 2009). Phylum Planctomycetes is represented by a broad distribution in most of the marine sediments (Fuerst 1995;Polymenakou et al. 2009;Oni et al. 2015) and involved in the global carbon and nitrogen cycles (Glöckner et al. 2003). We found that Verrucomicrobia more abundant in surface sediments (3.40%) in our study. Several studies also reported phylum Bacteroidetes, Epsilonbacteraeota, Chloroflexi, Acidobacteria, Actinobacteria, Planctomycetes, Verrucomicrobia and Firmicutes at diverse locations listed in Table 3.
In our study, genus Lutibacter of family Flavobacteriaceae under the phylum Bacteroidetes was the more dominant OTU in surface sediments. Several studies have reported the phylum have been consistently enriched in phytoplankton blooms (Kirchman 2002;Abell and Bowman 2005;Teeling et al. 2012). Bacteroidetes are extensively distributed in marine environments, have the capacity to degrade particulate organic matter in the ocean and also play a significant role in ocean carbon cycling (Riemann et al. 2000;Kirchman 2002;Bauer et al. 2006). They are being adapted to the degradation of polymeric substances in ocean (González et al. 2008) mainly polysaccharides and proteins (Cottrell and Kirchman 2000). We have found that, genus Sulfurovum of family Sulfurovaceae under the phylum Epsilonbacteraeota was found to be more dominant OTU in subsurface sediments. Same OTU was seem to be second most dominant in surface sediments. In contrast, OTU of phylum Epsilonbacteraeota more dominant in subsurface than surface sediments. OTU of family Desulfobulbaceae belonging to δ-Proteobacteria was found to be the second most dominant in subsurface sediments. At the genus level, Lutibacter (6.43%) and Marinifilum (3.63%) of phylum Bacteroidetes, Sulfurovum (3.63%) and Sulfurimonas (2.10%) belonging to phylum Epsilonbacteraeota, Psychromonas (2.80%) and Woeseia (2.52%) of phylum γ-Proteobacteria were found to be dominant in surface sediments. Comparatively, genus Sulfurovum (24.51%) and Sulfurimonas (7.38%) were found to be more predominant in subsurface sediments than in surface sediments. SEEP-SRB1 (3.25%) of phylum δ-Proteobacteria dominant in subsurface sediment. Genera Sulfurimonas and Sulfurovum in this study   α-Proteobacteria were also more detected in bottom water Zeng et al. (2009) depicted a considerable portion of bacteria in the sediments of Kongsfjorden, which is agreeing with the study of Zeng et al. (2017). Regarding sulfur cycling, Epsilon-Proteobacteria play a significant role in marine environment (Campbell et al. 2006). Several studies reported that the bacterial members of δ, γ and ε-Proteobacteria mediate sulfide reduction and oxidation Nakagawa et al. 2004;Campbell et al. 2006;Muyzer and Stams 2008). Sulfate reducer phylotypes belonging to δ-Proteobacteria have been reported in Arctic sediment (Ravenschlag et al. 2001;Zeng et al. 2011). Moreover δ-Proteobacteria sulfate reducer, the γ-Proteobacteria sulfate oxidizer phylotypes were also discovered in Arctic sediments (Ravenschlag et al. 1999). Phylotypes belonging to ε-Proteobacteria, were affiliated with sulfur oxidizing genera of Sulfurovum and Sulfurimonas and have been reported earlier (Table 3). Diverse of bacterial communities harbouring in surface and subsurface sediments in Kongsfjorden have a potential role in biogeochemical cycling. In this study, it was found that a significant fraction of retrieved sequences from both surface and subsurface sediments do not belong to any taxonomic division and they are categorized into unknown and uncultured bacteria. Similar findings have been reported in various studies (Polymenakou et al. 2009), the presence of uncultivated bacterial diversity in polar sediments was also reported by several molecular studies (Ravenschlag et al. 1999;Bowman and McCuaig 2003). Our studies revealed, only slight differences in bacterial community composition of surface and subsurface sediments. However, there were distinct differences in the predominance of bacterial communities between surface and subsurface sediments. When compared with previous reports, our results also indicated that Proteobacteria were dominant in the Arctic marine sediment and within this phylum, γ-Proteobacteria were dominant in surface and δ-Proteobacteria were seemed to be dominant in subsurface sediments of Kongsfjorden. The higher relative abundance of γ and α-Proteobacteria, Bacteroidetes and Verrucomicrobia in the surface sediments may be due to the higher quantity of phytoplankton derived organic matter in surficial sediments. They have a significant function in carbon cycle in the ocean. In marine sediments, bacterial abundance and community composition control by the abundance and quality of organic carbon present (Orcutt et al. 2011). Mostly, easily degradable organic matter is used by the microorganisms in surface sediments (Cowie and Hedges 1994;Wakeham et al. 1997). Microorganisms accumulated on the surface sediment from the water column are progressively buried into deeper layers as sedimentary particulate matter deposits on the surface of the sediment. Certain selection processes filter out the bacterial community during burial and producing subsurface and deeper layers below the subsurface, that are populated by a rare group of bacteria in the surface sediment (Jochum et al. 2017). Persisting populations such as Chloroflexi can adapt into the deep biosphere which was also present in the subsurface sediment below the bioturbation zone (Petro et al. 2019). Sulfate reduction is a significant bacterial process in marine sediment involved in nearly 50% remineralization of total organic carbon (Canfield et al. 1993). We have identified the predominance of sulfate reducing δ-Proteobacteria, sulfur oxidizing Epsilonbacteraeota, and γ-Proteobacteria in both surface and subsurface sediments. In this study, Sulfurovum and Sulfurimonas have important role in sulfur oxidation and genera SEEP-SRB1 of class δ-Proteobacteria which are involved in sulfate reduction found to be predominant in subsurface sediment.
Our studies revealed that the phylum γ, α-Proteobacteria and Bacteroidetes were more predominant in surface sediments. Surface sediments contain higher proportions of labile algal derived organic matter which could support these bacterial groups. Several researchers reported their marked presence in marine waters and sediments for the degradation of algal derived organic matter (Teeling et al. 2012;Miyatake et al. 2014;Ruff et al. 2014;Landa et al. 2014). Studies of Cottrell and Kirchman (2000) reported that, rather than utilizing polymers γ-Proteobacteria found to be more adapted to monomers. PICRUSt analysis revealed that the genes in the bacterial community structure in surface and subsurface sediments were involved in carbohydrate and amino acid metabolism in higher abundance. Predicted gene clusters for xenobiotic degradation and metabolism, biosynthesis of secondary metabolites, and genes for membrane transport were also observed. In carbohydrate metabolism, genes responsible for degradation complex polysaccharide represented endoglucanase, β-glucosidase, glycogen phosphorylase was observed and this indicates the important ecological function in carbon cycling by the sediment bacterial community of Kongsfjorden, Arctic.

Method of study
Major findings References DNA clone library analysis Predominance of bacteria related to the sulfur cycle, δ and γ-Proteobacteria were dominated in the permanently cold marine sediments of Arctic Ravenschlag et al. (1999)

Conclusions
High throughput sequencing analysis disclosed a highly diverse bacterial community composition in the surface and subsurface sediments of Kongsfjorden. Proteobacteria were found to be the predominant phylum in both surface and subsurface sediments. The γ, δ-Proteobacteria, and Epsilonbacteraeota which are having a potential role in the carbon remineralization have appeared in both surface and subsurface sediments with relatively higher abundance. Analysis of functional profiles indicated that the bacterial community have the potential to degrade complex organic matter such as glycans, glycosaminoglycans, polycyclic aromatic hydrocarbons etc. The present study thus demonstrates that the significant variability in the community composition of bacteria between the surface and subsurface sediments during the onset of winter in Kongsfjorden could be considerably influenced by the availability of total carbon and organic carbon in the sediments.