Exploring bacterial diversity in Arctic fjord sediments: a 16S rRNA–based metabarcoding portrait

Abstract

The frosty polar environment houses diverse habitats mostly driven by psychrophilic and psychrotolerant microbes. Along with traditional cultivation methods, next-generation sequencing technologies have become common for exploring microbial communities from various extreme environments. Investigations on glaciers, ice sheets, ponds, lakes, etc. have revealed the existence of numerous microorganisms while details of microbial communities in the Arctic fjords remain incomplete. The current study focuses on understanding the bacterial diversity in two Arctic fjord sediments employing the 16S rRNA gene metabarcoding and its comparison with previous studies from various Arctic habitats. The study revealed that Proteobacteria was the dominant phylum from both the fjord samples followed by Bacteroidetes, Planctomycetes, Firmicutes, Actinobacteria, Cyanobacteria, Chloroflexi and Chlamydiae. A significant proportion of unclassified reads derived from bacteria was also detected. Psychrobacter, Pseudomonas, Acinetobacter, Aeromonas, Photobacterium, Flavobacterium, Gramella and Shewanella were the major genera in both the fjord sediments. The above findings were confirmed by the comparative analysis of fjord metadata with the previously reported (secondary metadata) Arctic samples. This study demonstrated the potential of 16S rRNA gene metabarcoding in resolving bacterial composition and diversity thereby providing new in situ insights into Arctic fjord systems.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-023-01217-6.

Introduction

The cryospheric component of Earth’s biosphere has attained global research focus, as the increasing pace of glacial melts seriously alter the structural composition of the ecosystem, thereby resulting in significant change in the biogeochemical cycles and energy flow within the system [1–7]. Since the polar regions are considered one of the “biological hotspots,” it is essential to understand the diversity, distribution and functional roles of its structural components mainly the unseen microbial drivers. Compared to other oceans in the world, Arctic marine environments remain poorly studied due to the extreme climatic conditions and logistical restrictions, including large travel distances for ocean research vessels.

Fjords represent distinct marine ecosystems in the pan-Arctic region, and an Arctic fjord is characterised by relatively mild temperatures compared to other regions at similar latitudes [8]. Tidal glaciers at the head of the fjord discharge freshwater and suspended loads into the fjords, which create steep environmental gradients in salinity, temperature and sedimentation rates [9]. Kongsfjorden-Krossfjorden system is located on the west coast of Spitsbergen in the Svalbard Archipelago between the latitude 78° 04′ N and lat. 77° 30′ N and longitude 11° 3′ E and long. 13° 6′ E. Both these fjords are fed with freshwater with the presence of various tidewater glaciers, calving glaciers and streams. They are also influenced by Atlantic water intrusion from the West Spitsbergen Current [10, 11], which transports relatively warm saline water masses northwards [12, 13]. As a result, this dynamic twin fjord system acts as a key site for Arctic biodiversity monitoring and is also considered for climate change modelling studies.

The identity and diversity of Arctic bacteria, which play a crucial role in the ecosystem, have been investigated using traditional isolation methods, conventional DNA-based techniques and metabarcoding approaches [14–16]. Environmental DNA (eDNA) metabarcoding is an emerging method for the simultaneous molecular detection of different microbes and evaluate microbial diversity from complex environmental samples. eDNA metabarcoding allows the detection of different microbes at low densities, which are difficult to detect via conventional isolation techniques. Over the past few years, eDNA metabarcoding-based studies have been carried out to explore bacterial populations in complex samples like sediment and soil [17, 18]. Metabarcoding made the metagenomic studies more reachable via targeted sequencing, i.e. sequencing of the chosen amplified regions of genomic DNA like 16S rRNA and 18s rRNA [19]. During recent years, 16S rRNA metabarcoding approaches were implemented to assess polar bacterial community dynamics by revealing their taxonomic composition [20]. Several primer sets for this marker gene have been extensively tested across various sample sets from diverse Arctic habitats such as surface water, sea ice, marine snow, deep ocean basin and deep-sea sediment [21].

Available data regarding Arctic sediment bacterial communities are restricted to lakes, ponds, glaciers, ice sheets, etc. [22–25], and their prevalence in fjord sediments is comparatively less characterised [26]. Recent studies with Illumina-based sequencing provide new insights on the fjord bacterial communities [27–32]. The aim of this study was to explore bacterial communities from Arctic fjord sediments and their comparative analysis with other Arctic-related metadata. In the present study, 16S rRNA metabarcoding and metadata analysis provided the composition of sediment bacterial population of Kongsfjorden and Krossfjorden in Svalbard Archipelago.

Materials and methods

Site description and sample collection

Sediment samples were collected from Kongsfjorden (78°58′526″N and 11°50′13″E) and Krossfjorden (79°08′60″N and 11°44′59″E) using Van Veen’s grab onboard research vessel R V Teisten during the Summer Arctic Expedition (July 2017 and August 2019) by National Centre for Polar and Ocean Research (NCPOR), India (Fig. 1). Approximately, 5 kg sediment (from 290 m depth) was collected from both stations, kept in sterile polythene bags and stored at −80 °C in Arctic station. Samples were later transported to the home laboratory at School of Marine Sciences, Cochin University of Science and Technology (CUSAT), India, for further analyses.

Fig. 1.

Map showing Kongsfjorden and Krossfjorden in Arctic from where sediment collection was done. Map is generated using the Generic Mapping Tools (GMT) packages [33]. Topography and Bathymetry data source are from Ryan et al. [34]

Sediment parameter analysis

The overlying water temperature and salinity were measured using a Waterproof Portable metre (Cyberscan series 600) (Eutech instruments, Thermo Fisher Scientific, USA). pH of the sediment samples was measured using pocket pH metre EcoTestr pH 1 (Eutech Instruments, Thermo Fisher Scientific, USA). Organic carbon content and total organic matter were determined by wet oxidation method [35, 36]. Total carbon, hydrogen, nitrogen and sulphur analysis were done using CHNS analyser (ELEMENTAR Vario EL111), and trace elements were estimated using Inductively Coupled Plasma Atomic Emission Spectrometer (iCAP RQ ICP-MS-Thermo Fisher Scientific, USA) at DST-SAIF, Sophisticated Test and Instrumentation Centre (STIC), CUSAT, Kerala, India.

Pre-processing of sediment samples

Sediment samples (2 kg) from both Kongsfjorden (KGS) and Krossfjorden (KRS) were suspended in 8 L of autoclaved and virus free (0.02 μm membrane filtered) seawater. Tween 20 (1 mL/L) was added to the sediment suspension for maximum dislodgement of bacterial particles. The sediment suspension was subjected to low-speed centrifugation (500 rpm) to remove coarse sediment particles followed by high-speed centrifugation (12,000 rpm) to settle all suspended particles along with bacteria. Metagenomic DNA was extracted from this sediment residue (fine sediment with bacteria) using a conventional protocol with slight modifications [37].

The supernatant from the high-speed centrifugation was subjected to sequential filtration using filter membranes of different pore sizes (11 μm, 1.45 μm, 0.45 μm and 0.22 μm respectively) using a stainless-steel Millipore filtration unit to filter out soil particles and other organisms including bacteria (Supp. Fig. 1). Bacterial fraction entrapped in 1.45 μm, 0.45 μm and 0.22 μm filter membranes were detached by treating these membranes with a mixture of 1.5 mL TEN buffer (100 mM of Tris-HCl, 100 mM of EDTA (pH 8.0), 100 mM of sodium phosphate (pH 8.0) and 1.5 M of NaCl) and 1.5 μL of Tween 20. To facilitate the detachment of microbial cells from filter membranes, 50 mL centrifuge tubes containing filter membranes were vortexed for 10 min. The eluent was collected in 2 mL centrifuge vials and centrifuged (12,000 rpm for 15 min) to pelletise the microbial fraction. This microbial pellet was subjected to metagenomic DNA extraction using the above protocol [37].

Metagenomic DNA extraction

The fine sediment residue and microbial pellet obtained after pre-processing were treated with 500 μL of TEN buffer, 50 μL of 10% cetyl trimethyl ammonium bromide (CTAB), 50 μL of 20% sodium dodecyl sulphate (SDS) and 10 μL of proteinase K (20 mg/mL). This suspension was incubated (55 °C for 2 h) and centrifuged (10,000 rpm for 10 min). An equal volume of chloroform:iso-amyl alcohol (24:1, vol/vol) suspension was added with the obtained supernatant, and centrifuged (10,000 rpm for 20 min). The resulting aqueous phase was precipitated with 60 μL of 3 M sodium acetate and 600 μL of ice-cold iso-propanol and kept at 4 °C overnight. Crude DNA from aqueous phase was pelletised (12,000 rpm for 30 min) and was washed (10,000 rpm for 10 min) with 500 μL of ice-cold 70% (vol/vol) ethanol followed by absolute ethanol. The DNA pellet was air-dried, re-suspended in 20–30 μL TE buffer and stored at −20 °C in a freezer (Sanyo Electric Co. Ltd., Japan). Metagenomic DNA from the sediment residue and bacterial pellet (from supernatant) were pooled and used for further processing.

Quantity estimation and PCR confirmation of extracted metagenomic DNA

Yield of extracted metagenomic DNA from KGS and KRS was estimated in a Qubit Fluorometer (Invitrogen, UK) as per the manufacturer’s instructions. Amplification of bacterial 16S rRNA gene was performed using universal primers viz., 27-F (5′-AGAGTTTGATCMTGGCTCAG-3′) and 1492-R (5′-TACGGYTACCTTGTTACGACTT-3′) [38] for confirming the presence of bacteria in the extracted metagenomic DNA. PCR was performed in a thermal cycler (Applied Biosystems, USA) using PCR programme with an initial denaturation at 95 °C for 5 min, 35 cycles of denaturation at 94 °C for 45 s, annealing at 58 °C for 45 s and extension at 72 °C for 1 min and a final extension at 72 °C for 10 min. The amplicons were separated along with 100 bp ladder (New England Biolabs (NEB)) in 1% agarose-TBE gel (40 mM of Tris-HCl, 20 mM of Boric acid, 1 mM of EDTA, pH 8.0) and visualised under a Gel Documentation system (G:BOX-F³, Syngene, UK).

16S rRNA gene sequencing

After confirming PCR amplification, metagenomic DNA from both stations were sent to Macrogen Inc., Republic of Korea, for metabarcoding. After performing quality control (QC), the V3 and V4 regions of 16S rRNA were amplified using primer sets 341F: 5′-CCTACGGGNGGCWGCAG-3′ and 805R: 5′-GACTACHVGGGTATCTAATCC-3′ [39]. Primer pairs were concatenated with an Illumina overhang adapter sequence at the 5′ end (5′ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG for 341F; 5′ GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG for 805R) for compatibility with Nextera indexing. PCR replicates were generated using KAPA HiFi HotStart ReadyMix (KAPA Biosystems, Wilmington, MA, USA) under the following conditions: initial denaturation at 95 °C for 3 min, 25 cycles of denaturation at 95°C for 30 s, annealing at 55 °C for 30 s, extension at 72 °C for 30 s and final extension at 72 °C for 5 min. PCR amplicons were used to prepare the sequencing library using Herculase II Fusion DNA Polymerase Nextera XT Index Kit v2 (Illumina, San Diego, CA, USA). The Illumina MiSeq platform (Illumina, San Diego, USA) was used to perform 16S rRNA marker-based metabarcoding with paired-end read length of 300 bp. Finally, the sequencing data was converted into raw data for metadata analysis.

Metadata processing, diversity and comparative analysis

Illumina raw sequences were processed and analysed using Metagenomic Rast Server (MG-RAST) version 4.0.3 (http://metagenomics.anl.gov/) [40]. In MG-RAST, raw reads were uploaded as FASTQ files after demultiplexing of paired-end reads. Low-quality regions of uploaded FASTQ files were trimmed using SolexaQA [41]. Artificial duplicate reads (ADRs) were analysed by DRISEE (Duplicate Read Inferred Sequencing Error Estimation) [42]. Bowtie [43] was used for the best alignment of short DNA sequences to a reference genome. Ribosomal RNA identification was done by VSEARCH [44] and the taxonomic diversity of sequenced reads was decoded using “Greengenes 13_5” reference database [45]. Their abundance was analysed using “Representative Hit Classification” approach with an E-value cut-off 1 × 10⁻⁵, minimum sequence identity of 60% and a minimum alignment length of 15 bp. Metadata generated from this study was compared with similar 16S rRNA gene-based metadata (from different Arctic ecosystems) retrieved from MG-RAST (Table 1). Tabular findings from MG-RAST were used as input in MEGAN (MEtaGenome ANalyser) version 6.21.5 [46] and MicrobiomeAnalyst (https://www.microbiomeanalyst.ca/) [47] to generate taxonomic profiles for comparative analysis.

Table 1.

Summary of metagenome secondary data used to compare fjord sediment bacterial communities with other Arctic ecosystems. All metagenomes are publicly available on the MG-RAST server

MG-RAST ID	Sample Name	Sampling material	Biosample ID	Library ID	Location	Coordinates
mgm4906251.3	CUSAT_B7_combined_	Kongsfjorden sediment (KGS)	mgs882986	mgl882988	Kongsfjorden, Svalbard	78.58526, 11.5013
mgm4906280.3	CUSAT_B8_combined_	Krossfjorden sediment (KRS)	mgs882989	mgl882991	Krossfjorden, Svalbard	79.0860, 11.4459
mgm4875703.3	bar60	Glacier meltwater (GMW)	mgs781282	mgl781284	Svalbard, Norway	78.71974, 14.374894
mgm4626065.3	Total_samples	Coastal water (CW)	mgs291513	mgl291515	Kongsfjorden, Svalbard	79.8, 11.94
mgm4469175.3	Catlin-pool-2.IC3.S1.L3	Arctic Ice core sample (ICS)	mgs29262	–	Canada	78.7115, −104.8776667
mgm4729028.3	PF_11a_A	Permafrost soil (PS)	mgs554450	mgl554452	Alaska	65.1, −147.5
mgm4455848.3	951	Arctic desert soil (ADS)	mgs28788	mgl10165	Canada	78.88333, −75.78333
mgm4729014.3	83	Arctic terrestrial soil (ATS)	mgs554198	mgl554200	Svalbard, Norway	78.8967045, 12.06317073
mgm4900961.3	Genomowy_DNA_33_2	Cold sediment (CS)	mgs808520	mgl808522	Spitsbergen, Hornsund, Norway	77.002055, 15.541668
mgm4900963.3	Genomowy_DNA_6_2	Liquid environmental material (LEM)	mgs808526	mgl808528	Spitsbergen, Hornsund, Norway	77.002055, 15.541668
mgm4854933.3	S1_2	Feces of Arctic bird (FAB)	mgs758012	mgl758014	SiriusPasset, Greenland	82.791525, 42.44666667
mgm4646484.3	diet_SD	Feces of Arctic Ground Squirrel (FGS)	mgs322369	mgl322371	USA	61.1925538, −149.815182

Data availability statement

The raw reads are available on the MG-RAST server with accession numbers mgm4906251.3 (KGS) and mgm4906280.3 (KRS). 16S rRNA metabarcoding data from this study have been deposited in the NCBI Sequence Read Archive (SRA) under the accession numbers SRR12762477 (KGS) and SRR12762479 (KRS). Bio project IDs include PRJNA666789 (KGS) and PRJNA666794 (KRS) and Bio sample IDs include SAMN16327582 (KGS) and SAMN16327715 (KRS).

Results

Geochemical properties of the fjord sediments

During sampling, sediment samples were taken from a depth of 294 m from Kongsfjorden and 290 m from Krossfjorden. The water temperature for the fjord sediments were 2.23 °C (KGS) and 4.29 °C (KRS), while salinity was 34.94‰ (KGS) and 34.85‰ (KRS). The highest pH was observed in KRS (8.24), while the pH of KGS was 7.91. The total carbon (TC) and total organic carbon (TOC) content from KGS were found as 3.25% and 2.65%, and that of KRS was 2.94% and 2.4%, respectively. The trace elements/heavy metals such as Mg (2.65%), Fe (0.60%), Ni (24.03 ppm), P (623.69 ppm), Hg (0.22 ppm) and Se (4.32 ppm) were higher in KGS. However, the concentrations of Al (1.75%), K (1.03%), Ca (0.46%), Mn (415.45 ppm), Cu (14.84 ppm), Pb (14.75 ppm), As (7.81 ppm) and Co (11.40 ppm) were found to be higher in KRS (Table 2).

Table 2.

Details of various geochemical parameters of the sediment samples from Kongsfjorden and Krossfjorden

Parameters	Kongsfjorden	Krossfjorden
Depth (m)	294	290
Pressure (dbar)	297.348	293.717
Temperature (°C)	2.2323	4.2913
Salinity (PSU)	34.9451	34.8556
Density (kg/m3)	27.9087	27.6403
Conductivity (S/m)	3.1074	3.275018
pH	7.91	8.24
TN%	0.15	0.14
TS%	0.10	0.06
TH%	0.02	0.01
TC%	3.25	2.94
TOC (%)	2.65	2.4
TOM (%)	4.569	4.137
Trace elements/heavy metals	Kongsfjorden	Krossfjorden
25Mg (%)	2.65	2.50
27 Al (%)	1.44	1.75
39 K (%)	0.71	1.03
44 Ca (%)	0.45	0.46
57 Fe (%)	0.60	0.44
60 Ni (ppm)	24.03	22.15
31 P (ppm)	623.69	617.71
55 Mn (ppm)	385.33	415.45
200 Hg (ppm)	0.22	0.21
63 Cu (ppm)	14.83	14.84
208 Pb (ppm)	14.10	14.75
75 As (ppm)	7.24	7.81
59 Co (ppm)	10.61	11.40
78 Se (ppm)	4.32	3.64

Metagenomic DNA extraction and confirmation of bacterial communities

Yields of the metagenomic DNA extracted from KGS and KRS were 873.63 μg/g and 1675.14 μg/g respectively. Presence of bacterial communities in the metagenomic DNA samples was confirmed by visualising PCR amplicons with 16S rRNA gene primers (Supp. Fig. 2).

Overview of metabarcoding raw data

On 16S rRNA metabarcoding, the raw metadata from KGS contained 107,005 reads with 47,443,156 base pairs at an average length of 443 bps, while that of KRS sample resulted in 100,091 reads with 44,655,791 bp at an average length of 446 bps. Analysis statistics of metadata obtained after MG-RAST analysis is given in Supp. Table 1. Based on the analysis of KGS metadata, 99.95% sequences were represented as bacteria, and remaining sequences were categorised as eukaryota (0.04%) and unclassified sequences (0.01%). The percentage of bacteria, eukaryota and unclassified sequences in the KRS metadata was 99.91%, 0.08% and 0.01%, respectively.

Bacterial diversity and community structure in the fjord sediments

A total of 72,566 and 63,410 bacterial OTUs were detected from KGS and KRS, respectively. From KGS, 397 bacterial species affiliated to 17 phyla, 28 classes, 59 orders, 111 families and 198 genera were found whereas 633 species assigned under 18 phyla, 32 classes, 68 orders, 142 families and 277 genera were observed in KRS. At phylum level, 57% (KGS) and 52% (KRS) of the total bacterial sequences were found as “Unclassified.” Proteobacteria, Planctomycetes, Bacteroidetes and Firmicutes were the most abundant phyla detected from both the fjord sediments. Members from Spirochaetes, Thermodesulfobacteria and Acidobacteria were found exclusively in KGS, while Chlorobi, Deinococcus-Thermus, Fusobacteria and Thermotogae were detected from KRS sample alone (Fig. 2).

Fig. 2.

Circle plot representation of phylum level distribution of bacterial communities in a KGS and b KRS samples. Each phylum represented in specific colour and the size of each circle plot corresponds to the relative abundance of respective phylum

Among 28 bacterial classes detected from KGS, the dominant group gamma- and alpha-proteobacteria altogether contributed 97% of the total bacterial sequences along with unclassified bacterial reads. The Planctomycetacia, followed by Flavobacteria, Bacilli and Delta-proteobacteria, were the other dominant classes in KGS sample. In KRS sample, 52% bacterial reads remained as “Unclassified.” Gamma- and alpha-proteobacteria, Flavobacteria, Planctomycetacia and Actinobacteria together contributed 44% of the total bacterial reads. Classes viz., Aquificae, Dictyoglomia, Beta-proteobacteria, Thermodesulfobacteria, Opitutae, Nitrospira and Mollicutes were reported in low abundance from both sediments. Gloeobacteria, Solibacteres and Spirochaetes were detected from KGS alone while Chlorobia, Deinococci, Fusobacteria, Negativicutes, Spartobacteria and Thermotogae were identified from KRS only (Supp. Fig. 3).

At the order level, Pseudomonadales and Planctomycetales contributed a significant proportion (43%) in KGS, along with unclassified bacterial reads. Other than 52% unclassified bacterial reads, Aeromonadales, Flavobacteriales, Pseudomonadales and Alteromonadales were found to be dominant in KRS (35%) sample. Vibrionales and Aeromonadales (KGS) and Oceanospirillales and Planctomycetales (KRS) were the other dominant orders detected. Chlorobiales, Chromatiales, Deinococcales, Nitrosomonadales, Rubrobacterales, Mycoplasmatales, etc., were found in very low proportion from both sediments. Gloeobacterales, Solibacterales, Spirochaetales and Thermodesulfobacteriales were reported from KGS alone (Fig. 3).

Fig. 3.

Alluvial plot showing bacterial communities (phylum, class and order) in a KGS and b KRS samples. Topmost 30 orders were selected for representation

At family level, Pseudomonadaceae and Moraxellaceae were the dominant families accounting for 33% of the bacterial sequences from KGS sample. In KRS sample, Aeromonadaceae and Flavobacteriaceae were the major families accounting for 22% of the sequences. Rhodospirillaceae, Planctomycetaceae, Vibrionaceae, Bacillaceae, Clostridiaceae, Shewanellaceae, Oceanospirillaceae and Rhodobacteraceae were the other dominant families from both the fjord sediments. Almost 74 (KGS) and 79 (KRS) families were detected in a very low abundance which include Helicobacteraceae, Hydrogenothermaceae, Desulfobulbaceae and Psychromonadaceae from KGS and Pseudoalteromonadaceae, Actinomycetaceae, Geobacteraceae and Pelobacteraceae from KRS. Unclassified reads derived from different bacterial phyla and classes were also detected from both samples in various proportions. Sixteen families viz., Cellulomonadaceae, Chlorobiaceae, Erythrobacteraceae, Rhodothermaceae, etc., were reported from KRS only (Supp. Fig. 4).

In KGS sample, major genera included Pseudomonas, Acinetobacter, Aeromonas, Photobacterium, Thalassospira, Planctomyces, Aliivibrio, Rhodopirellula, Bacillus, Novispirillum and Flavobacterium. In addition, almost 150 genera were detected from KGS in very small proportion which include Oceanimonas, Polaribacter, Geobacillus, Acidimicrobium, Psychromonas, Rhodobacter, etc. (Fig. 4a). Among 277 genera, Aeromonas and Flavobacterium (20%) were the most represented clade from KRS followed by Shewanella, Psychrobacter, Pseudomonas, Gramella, Rhodopirellula and Alcanivorax. Bacterial genera detected from KRS in very small proportion include Lactobacillus, Marinosulfonomonas, Acetobacterium, Pseudoalteromonas, Arthrobacter, Thioalkalivibrio, etc. (Fig. 4b).

Fig. 4.

Genus level distribution of bacterial communities in a KGS and b KRS samples by Cytoscape. Each node in this network is coloured based on taxonomic levels and represents up to genus level. Size of each nodule is proportional to the relative abundance of each bacterial groups

Comparative analysis of fjord bacterial metadata with other Arctic ecosystems

Phyla level distribution of bacteria in different samples with respect to sample texture are shown in principal component analysis (PCA). The first principal coordinate (PC1) of the analysis accounted for 27.2% of the variation in the data. It shows the grouping of soil/sediment and water samples and separation of faecal sample. PC2 accounted for 18.1% of the variance in the microbial communities. Based on the microbial phyla abundance, the same type of samples grouped together and formed two clusters except the faecal sample FGS. Almost all soil/sediment samples (PS, CS, ADS and ATS), including KGS and KRS, grouped to form one cluster. All water samples (ICS, GMW and LEM) were grouped and formed a single cluster along with soil/sediment samples and faecal sample FAB. Water sample CW and faecal sample FGS were lying well separated from the water/sediment cluster (Fig. 5).

Fig. 5.

Principal component analysis (PCA) of different Arctic samples. All bacterial phyla from each sample were selected for plotting PCA. KGS, Kongsfjorden sediment; KRS, Krossfjorden sediment; GMW, glacial meltwater; CW, coastal water; ICS, ice core sample; PS, permafrost soil; ADS, Arctic desert soil; ATS, Arctic terrestrial soil; CS, cold sediment; LEM, liquid environmental material; FAB, faeces of Arctic bird; FGS, faeces of Arctic ground squirrel

Family level analysis showed that the Arctic samples formed two clusters in which KGS and KRS samples grouped with glacial/water samples viz., ICS, GMW and CW samples. Second cluster included faecal samples (FGS and FAB) along with soil/sediment samples (ATS, LEM, CS and PS). In the first cluster, glacial/water samples showed an abundance of cold-active/marine alpha-proteobacterial members (from Cytophagaceae, Flavobacteriaceae, Rhodobacteriaceae, etc.) while fjord samples are dominated with cold-active/halophilic Gamma-proteobacterial members (from Alteromonadaceae, Halomonadaceae, etc.). In the second cluster, both faecal and soil/sediment samples revealed the abundance of pathogenic bacterial members from Parachlamydiaceae, Paenibacillaceae, Micrococcaceae, etc. (Fig. 6).

Fig. 6.

Family level heatmap representation and comparative analysis of bacterial communities from different Arctic samples. Topmost 30 families were represented in this figure using MEGAN 6 software. KGS, Kongsfjorden sediment; KRS, Krossfjorden sediment; GMW, glacial meltwater; CW, coastal water; ICS, ice core sample; PS, permafrost soil; ADS, Arctic desert soil; ATS, Arctic terrestrial soil; CS, cold sediment; LEM, liquid environmental material; FAB, faeces of Arctic bird; FGS, faeces of Arctic ground squirrel

Genera level comparative analysis was done using topmost bacterial genera. Faecal, fjord, soil and sediment samples grouped together and formed two clusters, while CW and ICS stood as two outgroups. CW was dominated by Cytophaga and Psychrobacter, while ICS sample showed the dominance of Pseudomonas, Chryseobacterium, Flavobacterium and Planctomyces. First cluster formed by soil (ADS, PS) and faecal (FAB and FGS) samples are prevailed by members from alpha-, gamma- and delta-proteobacteria. On comparison with metadata from other samples, KGS and KRS grouped with soil/sediment (ATS, LEM, CS and GMW) samples mainly due to the abundance of Aeromonas, Thalassospira, Rhodopirellula, Pirellula, etc. (Fig. 7).

Fig. 7.

Genus level distribution and comparative analysis of bacterial communities from different Arctic samples. Topmost 25 genera were represented in this heatmap. KGS, Kongsfjorden sediment; KRS, Krossfjorden sediment; GMW, glacial meltwater; CW, coastal water; ICS, ice core sample; PS, permafrost soil; ADS, Arctic desert soil; ATS, Arctic terrestrial soil; CS, cold sediment; LEM, liquid environmental material; FAB, faeces of Arctic bird; FGS, faeces of Arctic ground squirrel

Discussion

Kongsfjorden and Krossfjorden are semi-open glacial fjords located on the western side of Spitsbergen and share a common mouth to the open sea. Due to climate change and global warming, these Arctic fjords experience enhanced glacial retreat and meltwater input, influencing the mixing of warm, saline Atlantic water with sediment-rich and less saline glacial meltwater [10, 11]. Microbial diversity in Kongsfjorden and Krossfjorden systems has already led to a shift by decreased salinity and temperature and increased sediment load via these mixing [5, 9, 48]. Change in the temperature of glacial ice has led to the growth of new microbes which are not psychrophiles, thereby altering the microbial community structure [6, 7]. Shift in salinity, temperature and sediment load might be the major determinants of microbial community composition and diversity in Spitsbergen fjords, while freshwater might be the source of the entrance of non-marine species [8]. Increasing ocean temperatures and decreasing sea ice cover suggested that the Arctic-Atlantic boundary of Kongsfjorden and Krossfjorden are in a good position as an indicator system for global warming impacts [8, 12].

Geochemical properties of sediment samples

Hydrological conditions change seasonally in the Arctic fjords and water masses are homogeneous during winter. During summer, warmer surface water masses from the West Spitsbergen Current (WSC) cause thermal-salinity stratification along with the freshwater outflow from the glaciers [49, 50]. The temperature and salinity of the fjord sediments from previous studies varied between 2.2 and 5.9 °C and 33.7-34.9 PSU and were comparable with that of Kongsfjorden and Krossfjorden samples [31, 32, 51]. The highest pH was observed in KRS (8.24), while the pH of KGS was 7.91. The TOC observed in the KGS (2.65) and KRS (2.4) corresponds to the results from previous reports, while KRS properties and their seasonal variability were not much explored earlier [31, 32, 49, 51–53]. In KGS, about 90-95% of the organic carbon was reported to be contributed by primary production and 5-10% of total organic carbon of terrestrial origin from glacial melting [54]. The physico-chemical properties of the fjord sediments are similar and comparable to previous studies [31, 32, 49, 51].

The concentration of major elements did not vary much in the KGS and KRS samples. However, Mg and Fe were marginally higher in KGS than in KRS sample. Al, Ca and K were lower in KGS compared to KRS. Studies also support the presence of more calcite and dolomite in these fjord sediments [55, 56]. Global atmospheric, oceanic and biological cycling of elements causes the transport of trace metals from lower and middle latitudes to various polar habitats [49, 57]. Higher concentrations of Cu, Ni and Zn have been reported from the Arctic terrestrial sediments, which can be related to the abundant soluble rocks like sulfides and carbonates in the glacier regions. The enrichment of these elements might be due to the contact of meltwater with glacial debris [31, 58]. The trace metal concentrations in fjord (KGS and KRS) sediments were comparable to those from earlier studies conducted in Kongsfjorden sediments [31, 32, 49, 59]. The metal concentrations from both the fjord sediments were found to be lower than the baseline values reported from studies by Lu et al. [49].

An increasing trend of Hg and Cd levels was reported from marine birds and mammals in the Canadian Arctic [60, 61]. There have been a number of studies on the distribution of trace metals in the Arctic environments [49, 62, 63]. In sedimentary environments in the Arctic, heavy metals viz., Zn, Cu, Pb, Cd, Co, Ni, Mn, Cr and Hg have been reported in higher concentrations due to anthropogenic processes [64, 65]. In our study, Fe (0.60 ppm), Ni (24.03 ppm) and Se (4.32 ppm) were found to be higher in KGS sample. Similarly, KRS exhibited higher concentrations of Mn (415.45 ppm), Pb (14.75 ppm) and Co (11.40 ppm). Apart from Mn, no significant variations were observed in the trace elemental concentrations of both fjord sediments. Cu, Pb, Hg, As, Co and Se concentrations showed little variation in both samples. Research conducted in coastal sediments of the Arctic has focused on trace metals such as Cd, Cr, Cu, Hg, Pb and Zn [49, 66]. Studies from Arctic concluded that occurrence of trace metals is strongly influenced by physical and chemical parameters of the sediments and anthropogenic processes [49, 64, 66].

Fjord bacterial diversity and their comparative analysis with other Arctic ecosystems

Major bacterial phyla explored from Arctic environments include Proteobacteria, Firmicutes, Bacteroidetes, Planctomycetes, Actinobacteria and Acidobacteria [29, 67, 68]. The dominant bacterial phyla detected from KGS and KRS are comparable to the most abundant phyla reported from various Arctic environments [29, 67–70]. Current study detected 17 and 18 bacterial phyla from the sediments of Kongsfjorden and Krossfjorden, respectively. About half of the reads were put under “unclassified” as with other samples, viz., GMW, CW, PS, ATS, CS and LEM.

Along with a group of unclassified bacterial reads, Proteobacteria was dominant in KGS and KRS samples and was similar to the previous studies from Arctic sediments by both culture-based and conventional culture-independent (16S rRNA/18S rRNA clone library-based) approaches [29, 71–73]. The dominance of Proteobacteria, Firmicutes, Bacteroidetes and Actinobacteria from freshwater as well as marine samples (water, lake sediments and soils, polar desert soils, tundra soils, ice and fjord sediments) of Arctic was also reported [71–74]. The dominance of Proteobacteria and Bacteroidetes observed in both KGS and KRS, along with ICS, PS and ADS samples, is also related to the results from Kongsfjorden sediments [26, 31] and tundra soil [75, 76]. Similar observation was recorded from ice, sediment, soil, meltwater and faecal samples (from comparative analysis) from Arctic.

Gamma-, alpha-, delta- and beta-proteobacteria were found as dominant classes/groups in KGS, similar to earlier studies from Arctic ice and the winter cover (snow and ice) by cultivation or molecular methods [5, 73, 77]. CW, PS and ATS soil samples also supported the above finding with respect to the presence of gamma- and alpha-proteobacteria. In addition, previous reports from sediments of Pacific Arctic Ocean and northern Bering Sea [78–80] showed the dominance of delta- and gamma-proteobacteria which was consistent with our findings in the fjord sediments. Higher abundance of Gamma-proteobacteria was also detected from the fjord samples and was similar to reports from other studies [74, 81, 82]. The increase of Gamma-proteobacteria was also reported in connection with increased dissolved organic matter and phytoplankton bloom dynamics in the summer season [82, 83]. Beta-proteobacteria was found to be highly dominant in glaciers [84] and its presence was noted in GMW sample as well. In contrast, low abundance of Beta-proteobacteria (up to 3% of all proteobacterial reads) is reported as a characteristic of Atlantic water masses [85, 86] and this trend could be observed in CW and PS samples along with KGS and KRS samples. Besides, a low abundance of epsilon-proteobacteria was observed in both the fjord sediments and all other samples. Our analyses indicated that only zeta-proteobacteria (among six Proteobacterial classes) were not reported from both the sediments, which was reliable with earlier investigations [80, 87].

Bacteroidetes, usually found in summer coastal Arctic communities [88, 89], was another dominant group in our fjord sediments where it comprised 0.66% (KGS) and 11.79% (KRS) of total bacterial composition. Its dominance with Proteobacteria in Arctic soils suggests a connection between the dominant bacterial taxa in the fjord sediments and organic matter degradation [81]. Higher levels of Bacteroidetes could be observed in soil or sediment samples such as PS, ADS, ATS and CS. Firmicutes was also reported in the present study, which was in consensus with the earlier reports from Arctic sediments [29, 68, 80]. Faecal samples (FAB and FGS) showed a higher abundance of Firmicutes along with Bacteroidetes and Proteobacteria from the current study, which was supported by earlier reports from Arctic birds and ground squirrels [90–92]. The 16S rRNA gene sequences related to the phyla Chloroflexi and Chlamydiae were also detected from fjord samples, similar to a previous study from Kongsfjorden sediments [93]. However, they were present in low abundances in all analysed Arctic samples in this study.

Though Arctic environments support diverse bacterial communities, comparatively little is known about members of the phylum Actinobacteria by culture-dependent and culture-independent methods [67, 94]. Actinobacteria was found to be abundant in all other Arctic samples except KGS and KRS samples. Dominant Actinobacterial members isolated from Arctic marine sediments include Brevibacterium, Pseudonocardia and Mycobacterium sp. [67]. Actinocorallia and Rhodococcus from KGS and Subtercola and Streptomyces from KRS were noted as dominant actinobacterial members. Pseudonocardia and Mycobacterium sp. may be associated with environmental pollution occurring in the Arctic region [95]. Phylum Acidobacteria was detected in low abundances from fjord sediments and other Arctic samples included in the comparative analysis. In contrast, studies of Alaskan, Canadian and Siberian Arctic soils reported Acidobacteria as a dominant phylum [96, 97].

Members of phylum Verrucomicrobia were detected from fjord samples, and their presence suggested them as freshwater taxa entering the fjord from rivers [30]. PS and CS samples recorded higher levels of Verrucomicrobia than other samples, including fjord sediments. As per previous studies, temperature, water column depth and nitrogen concentration are the main factors influencing taxonomic composition of Verrucomicrobia [98]. Roseibacillus and Luteolibacter were the two most abundant genera detected from Svalbard fjords [99, 100], whereas Prosthecobacter and Rubritalea were the dominant ones from KGS and KRS samples.

Cyanobacteria comprised 0.24% of the total bacterial community in the pan-Arctic tundra soils [69], 4.50% in polar desert soils [68] and 0.13% (KGS) and 0.29% (KRS) in fjord sediments. Cyanobacteria have been detected in many nutrient-poor Arctic soils and have been reported to fix atmospheric nitrogen and carbon dioxide [101, 102]. This pointed out the significant role of Cyanobacteria in nutrient cycling in the nitrogen and/or phosphorus-limited Kongsfjorden polar desert soils [103] and fjord sediments. Bacterial phyla such as Planctomycetes, Cyanobacteria, Chloroflexi, Chlamydiae and Verrucomicrobia with higher relative abundances were reported from polar desert soil [68] than the levels reported from pan-Arctic tundra soil [69].

Planctomycetes was another dominant phylum detected from both fjord samples and other Arctic samples. Planctomyces, Rhodopirellula and Pirellula were the genera detected among which Rhodopirellula and Pirellula are generally associated with giant kelp as epiphytes. These kelp-associated bacterial communities play a significant role in the Arctic carbon metabolism along with Bacteroidetes and Proteobacterial members. From the current study, bacterial phyla viz., Spirochaetes, Thermodesulfobacteria and Acidobacteria were detected from KGS alone, whereas Chlorobi, Deinococcus-Thermus, Fusobacteria and Thermotogae were detected from KRS alone.

Conclusion

16S rRNA gene metabarcoding of KGS and KRS metagenome prevailed the presence of representatives from phyla viz., Proteobacteria, Bacteroidetes, Planctomycetes, Firmicutes, Actinobacteria, Cyanobacteria, Chloroflexi and Chlamydiae. A significant proportion of unclassified bacterial reads was also detected. Class level analysis revealed the dominance of Gamma-proteobacteria from both the fjord samples along with Alpha-proteobacteria, Flavobacteria, Planctomycetacia, etc. Psychrobacter, Pseudomonas, Acinetobacter, Aeromonas, Photobacterium, Flavobacterium, Gramella and Shewanella were the major genera reported from both the fjord sediments. Comparative analysis of KGS and KRS metadata with the earlier reported data (secondary data) of other 10 Arctic samples supported the above findings. The resolution of the fjord metadata highlights the efficiency of 16S rRNA gene metabarcoding for detecting microbes present in fjord sediments.

Abbreviations

KGS

Kongsfjorden sediment

KRS

Krossfjorden sediment

rRNA

Ribosomal RNA

MG-RAST

Metagenomic Rast Server

MEGAN

MEtaGenome ANalyser

Author contribution

JP, JC, JT and RP collected the sediment samples. BK carried out the experiment of the present work with support from SS and JG. MS and ASM assisted the processing of sediment samples. The work was carried out under the supervision of RP. BK wrote the manuscript. SS, RP, JP, CER, JG and MS reviewed and edited the manuscript. All authors have read and approved the manuscript.

Declarations

This article does not contain any studies with human participants or animals performed by any of the authors.

Competing interests

The authors declare no competing interests.