Introduction
The Pacific sector of the Southern Ocean, and especially the Amundsen Sea, are the least studied oceanic regions in the world (Griffiths Reference Griffiths2010). Severe ice conditions year-round and the geographic remoteness make sampling in this area very difficult. The biodiversity of the Amundsen Sea, especially of the coastal and shelf areas, is almost unknown (Kaiser et al. Reference Kaiser, Barnes, Sands and Brandt2009). Recently, scientists began to highlight the diversity and distribution of isopods and phytoplankton in this isolated region (Kaiser et al. Reference Kaiser, Barnes, Sands and Brandt2009, Fragoso & Smith Reference Fragoso and Smith2012). Gravalosa et al. (Reference Gravalosa, Flores, Sierro and Gersonde2008) concentrated on the distribution of coccolithophores and showed that their dispersion is restricted north of the Polar Front. Fragoso & Smith (Reference Fragoso and Smith2012) focused their study areas near the coast and delivered an overview of the phytoplankton assemblages in this area. They revealed diatom dominated assemblages in offshore areas of the Amundsen Sea. However, they used pigment based and microscopic analysis and thereby, the taxonomical resolution was not very detailed. So far, no comprehensive survey of the whole eukaryotic protist spectrum in the Amundsen Sea exists.
In the course of the controversially conducted debate about the “everything is everywhere” hypothesis (Lachance Reference Lachance2004), many studies focused on the biogeography of protists (Finlay Reference Finlay2002, Finlay & Fenchel Reference Finlay and Fenchel2004). Our recent study, focusing on the distribution of eukaryotic protists along a transect from New Zealand to the coast of Antarctica, revealed distinct biogeographical patterns, defined by the oceanic fronts (Wolf et al. unpublished). These patterns were driven by strong environmental gradients and included different large-scale water masses. To complement our knowledge about the biogeography of protists in the Pacific sector of the Southern Ocean, their distribution has to be highlighted on a smaller, more regional scale. Narrow environmental differences within a large-scale water mass have to be investigated.
Most investigations of eukaryotic protist composition and distribution in the Southern Ocean mainly used traditional microscopic and pigment extraction based methods (Ishikawa et al. Reference Ishikawa, Wright, van den Enden, Davidson and Marchant2002, Wright et al. Reference Wright, Ishikawa, Marchant, Davidson, van den Enden and Nash2009). However, microscopic surveys have difficulties in identifying small cells and pigment analysis only targets autotrophic cells. Here, molecular tools are advantageous. Few investigations in the Southern Ocean used molecular approaches, such as denaturing gradient gel electrophoresis (DGGE) (Gast et al. Reference Gast, Dennett and Caron2004) or 18S rRNA gene cloning and sequencing (Lopez-Garcia et al. Reference Lopez-Garcia, Rodriguez-Valera, Pedros-Alio and Moreira2001). The automated ribosomal intergenic spacer analysis (ARISA) approach provides a quick overview of the diversity and facilitates the comparison of different samples. It is well established for investigations of prokaryotic diversity (Smith et al. Reference Smith, Barrett, Tusnady, Rejto and Cary2010), and we successfully implemented it for the analysis of eukaryotic phytoplankton diversity. The newly emerging 454-pyrosequencing approach (e.g. of the V4 region of the 18S rRNA gene) allows assessing microbial communities with high-resolution, based on sufficient deep taxon sampling (Margulies et al. Reference Margulies, Egholm and Altman2005, Stoeck et al. Reference Stoeck, Bass, Nebel, Christen, Jones, Breiner and Richards2010), regardless of cell size and nutrition.
The objective of this study is to determine the composition of late summer eukaryotic protist assemblages in the Amundsen Sea, south of the southern boundary of the Antarctic Circumpolar Current. We used state of the art molecular approaches, such as ARISA and 454-pyrosequencing, and high-performance liquid chromatography (HPLC). Furthermore, we want to assess the impact of different environmental conditions on the biogeography of protists, within this oceanic region. The pigment and ARISA analysis provide an overview of differences in structure and diversity of the whole investigated area and the 454-pyrosequencing of selected samples gives more detailed information about the species composition, dominant representatives and the distribution of the rare biosphere (phylotypes with an abundance <1% of total sequences) in the observed area.
Materials and methods
Location and sampling
A total of 34 surface-water samples were taken on a regular basis (c. every 40 km) during the RV Polarstern cruise ANT XXVI/3 between 12 February and 22 March 2010 in the Amundsen Sea (Fig. 1a) along a west–east transect from the eastern Ross Sea to the western Bellingshausen Sea (within 71.06–74.39°S and 160.27–101.58°W). All surface water samples were collected using the ship pumping system (membrane pump), located at the bow at 8 m depth below the surface. For the determination of pigments, 4 l samples were immediately filtered onto 25 mm Whatman GF/F filters and stored at -80°C until further analysis in the laboratory. For molecular analysis, 2 l samples were immediately fractionated, by filtering them on Isopore Membrane Filters (Millipore, USA) with a pore size of 10 μm, 3 μm and 0.2 μm. Filters were stored at -80°C until analysis in the laboratory.
Fig. 1 Study area and environmental conditions. a. Location of surface water samples and water depth. b. Surface water temperature. c. Surface water salinity. d. Ice coverage.
Pigment analysis (HPLC)
Samples were measured using a Waters HPLC-system, equipped with an auto sampler (717 plus), pump (600), PDA (2996), a fluorescence detector (2475) and EMPOWER software. For analytical preparation, 50 μl internal standard (canthaxanthin) and 1.5 ml acetone were added to each filter sample and then homogenized for 20 sec in a Precellys® tissue homogenizer. After centrifugation, the supernatant liquid was filtered through a 0.2 μm PTFE filter (Rotilabo) and placed in Safe-Lock Tubes (Eppendorf, Germany). An aliquot (100 μl) was transferred to the auto sampler (4°C). Just prior to analysis, the sample was premixed with 1 M ammonium acetate solution in the ratio 1:1 (v/v) in the auto sampler and injected onto the HPLC-system. The pigments were analysed by reverse-phase HPLC, using a VARIAN Microsorb-MV3 C8 column (4.6 x 100 mm) and HPLC-grade solvents (Merck, Germany). Solvent A consisted of 70% methanol and 30% 1 M ammonium acetate and solvent B contained 100% methanol. The gradient was modified after Barlow et al. (Reference Barlow, Cummings and Gibb1997). Eluting pigments were detected by absorbance (440 nm) and fluorescence (Ex: 410 nm, Em: > 600 nm). Pigments were identified by comparing their retention times with those of pure standards and algal extracts. Additional confirmation for each pigment was done by comparing their absorbance spectra between 390 and 750 nm with the library of the standards. Pigment concentrations were quantified based on peak areas of external standards, which were spectrophotometrically calibrated using extinction coefficients published by Bidigare (Reference Bidigare1991) and Jeffrey et al. (Reference Jeffrey, Mantoura and Bjornland1997). For correction of experimental losses and volume changes, the concentrations of the pigments were normalized to the internal standard canthaxanthin. Phytoplankton group composition was calculated applying the CHEMTAX program and input ratios of Mackey et al. (Reference Mackey, Mackey, Higgins and Wright1996). To estimate the various size classes of the phytoplankton, the following groups were combined: prasinophytes and pelagophytes for picoplankton (<2 μm), haptophytes and cryptophytes for nanoplankton (2–20 μm), and dinoflagellates and diatoms for microplankton (20–200 μm). Their respective contribution to total biomass is based on their CHEMTAX derived chlorophyll a (chl a) concentration.
DNA extraction
The DNA was extracted with the E.Z.N.A.TM SP Plant DNA Kit (Omega Bio-Tek, USA). At the beginning, the filters were incubated with lysis buffer. All further steps were performed as described in the manufacturer's instructions. At the end, the DNA was eluted in 60 μl of elution buffer and the extracts were stored at -20°C until further analysis. DNA concentration was measured with a NanoDrop 1000 (Thermo Fisher Scientific, USA) (average DNA concentration: 23 ng μl-1).
PCR amplification, ARISA
An equal volume of extracted DNA of each size fraction (>10 μm, 3–10 μm and 0.2–3 μm) from each sample was pooled. The ITS1 (internal transcribed spacer) region was amplified in triplicates using the primer-set 1528F (5′-GTA GGT GAA CCT GCA GAA GGA TCA-3′) (modified after Medlin et al. (Reference Medlin, Elwood, Stickel and Sogin1988)) and ITS2 (5′-GCT GCG TTC TTC ATC GAT GC-3′) (White et al. Reference White, Bruns, Lee and Taylor1990). The 1528F primer was labelled at the 5′-end with the dye 6-FAM (6-carboxyfluorescein). The PCR (polymerase chain reaction) mixtures contained 1 μl of DNA extract, 1 x HotMaster Taq Buffer containing 2.5 mM Mg2+ (5 Prime, USA), 0.8 mM dNTP-mix (Eppendorf, Germany), 0.2 μM of each Primer and 0.4 U of HotMaster Taq DNA polymerase (5 Prime, USA) in a final volume of 20 μl. Reactions were carried out in a Mastercycler (Eppendorf, Germany) under the following conditions: an initial denaturation at 94°C for 3 min, 35 cycles of denaturation at 94°C for 45 sec, annealing at 55°C for 1 min and extension at 72°C for 3 min, and a final extension at 72°C for 10 min. PCR fragments were separated by capillary electrophoresis on an ABI Prism 310 Genetic Analyser (Applied Biosystems, USA).
PCR amplification, 454-pyrosequencing
Seven samples were sequenced (Table I). For each fraction of a sample, we amplified c. 670 base pair (bp) fragments of the 18S rRNA gene, containing the highly variable V4-region, using the primer-set 528F (5′-GCG GTA ATT CCA GCT CCA A-3′) and 1055R (5′-ACG GCC ATG CAC CAC CAC CCA T-3′) (modified after Elwood et al. (Reference Elwood, Olsen and Sogin1985)). The PCR mixtures were composed as described previously for ARISA. Reaction conditions were as follows: an initial denaturation at 94°C for 3 min, 30 cycles of denaturation at 94°C for 45 sec, annealing at 59°C for 1 min and extension at 72°C for 3 min, and a final extension at 72°C for 10 min. An equal volume of PCR reaction of each size fraction from each sample was pooled and purified with the MinElute PCR purification kit (Qiagen, Germany) following the manufacturer's instructions. Pyrosequencing was performed on a Genome Sequencer FLX system (Roche, Germany) by GATC Biotech AG (Germany).
Table I Summary of recovered 454-pyrosequencing reads, quality filtering and number of OTUs (operational taxonomic units). Samples are arranged from west–east.
*reads with a minimum length of 300 bp and a maximum length of 670 bp.
**abundant OTU = number of reads ≥1% of total reads, otherwise it is rare.
Data analysis, ARISA
Electropherograms were analysed using the GeneMapper Software v4.0 (Applied Biosystems, USA). Peaks with a size smaller than 50 bp (corresponding to primer and primer dimer peaks) were removed from the dataset. To remove the background noise and to get sample-by-binned-OTU (operational taxonomic unit) tables, the data were binned using the binning scripts, according to Ramette (Reference Ramette2009), for R (R Development Core Team 2008). The resulting sample-by-binned-OTU tables were transformed into presence/absence matrices and the distances between the samples were calculated, using the Jaccard index implemented in the R package vegan (Oksanen et al. Reference Oksanen, Blanchet, Kindt, Legendre, O'Hara, Simpson, Solymos, Stevens and Wagner2011), which was also used in the following steps. MetaMDS (maximum random starts of 300) plots were computed. Clusters were determined using the hclust function in R. To test, whether the resulting clusters differ significantly, an ANOSIM was performed. A Euclidean distance matrix with the normalized environmental parameters was calculated. The correlation between the ARISA distance matrix and the environmental distance matrix was tested with a Mantel test (10 000 permutations), implemented in the R package ade4 (Dray & Dufour Reference Dray and Dufour2007). A principal component analysis (PCA) with the environmental parameters and the HPLC size fractions was performed (R package ade4).
Data analysis, 454-pyrosequencing
Raw sequence reads were processed to obtain high quality reads. The forward primer 528F, used for the sequencing, attaches c. 25 bp upstream of the V4 region, which has in general a length of c. 230 bp (Nickrent & Sargent Reference Nickrent and Sargent1991). Reads with a length under 300 bp were excluded from further analysis to assure inclusion of the whole hyper variable V4 region in the analysis and to get rid of short reads. Unusually long reads that were greater than the expected amplicon size (>670 bp) and reads with more than one uncertain base (N) were removed. Remaining reads were checked for chimeric sequences with the software UCHIME 4.2.40 (Edgar et al. Reference Edgar, Haas, Clemente, Quince and Knight2011) and all reads considered being chimeric were excluded from further analysis. The high quality reads of all samples were clustered into OTUs at the 97% similarity level using the software Lasergene 10 (DNASTAR, USA). Subsequently, reads not starting with the forward primer were manually removed. Consensus sequences of each OTU were generated, which reduced the amount of sequences to operate with and attenuated the influence of sequencing errors and uncertain bases. The 97% similarity level has been shown to be the most suitable to reproduce original eukaryotic diversity (Behnke et al. Reference Behnke, Engel, Christen, Nebel, Klein and Stoeck2011) and has the effect of bracing most of the sequencing errors (Kunin et al. Reference Kunin, Engelbrektson, Ochman and Hugenholtz2010). Furthermore, known intragenomic SSU polymorphism levels can range to 2.9% in dinoflagellate species (Miranda et al. Reference Miranda, Zhuang, Zhang and Lin2012). Operational taxonomic units comprised of only one sequence (singletons) were removed. The consensus sequences were aligned into a reference alignment obtained from SILVA (see below) using the software HMMER 2.3.2 (Eddy Reference Eddy2011). Subsequently, taxonomical affiliation was determined by placing the consensus sequences into a reference tree, containing about 1200 high quality sequences of Eukarya from the SILVA reference database (SSU Ref 108), using the software pplacer 1.0 (Matsen et al. Reference Matsen, Kodner and Armbrust2010). The compiled reference database is available on request in ARB-format. OTUs assigned to fungi and metazoans were excluded from further analysis. Rarefaction curves were computed using the freeware program Analytic Rarefaction 1.3. The dataset generated in this study has been deposited at GenBank's Short Read Archive (SRA) under Accession No. SRA057133.
Results
Environmental conditions
The investigated area showed a very heterogeneous setting, in terms of water depth, surface temperature, surface salinity and ice coverage (Fig. 1a–d). Samples 33–46 and 49–57 were lying offshore, with water depths (Fig. 1a) from 1969 m (sample 57) to 4334 m (sample 34). Samples 47 and 48 (polynya) and 58–71 (Pine Island Bay) were inshore, with water depths from 398 m (sample 69) to 714 (sample 48). Sample 60 was lying over the continental slope and showed a greater depth (1447 m).
The surface water temperature ranged between -1.63°C (sample 47) and -0.24°C (sample 60) (Fig. 1b). Hence, the temperature only varied weakly, but showed significantly higher values at the most eastern sample sites (60–62), located at the transition to the Bellingshausen Sea.
Surface water salinity (Fig. 1c) showed values between 32.35 PSU (sample 57) and 33.51 PSU (sample 34). In general, salinity was higher in the western part of the transect and declined eastwards. Among the eastern samples, sample 69 showed a very high salinity (33.32 PSU).
Most samples were located near the ice-edge (Fig. 1d) with no ice. At samples 45–48, we crossed an ice field to reach a polynya, with a high spatial variability of the ice cover (5–50%). Samples 57 and 58 were taken in an ice field and showed an ice coverage of 10–50%. Sample 69 was obtained in a region with 100% ice cover.
Structure/diversity overview
We used a combination of HPLC and ARISA to assess the impact of different environmental conditions on the structure/diversity of the plankton assemblages in the sub-polar region.
High performance liquid chromatography
Total chl a concentrations (Fig. 2a) along the entire transect ranged between 0.11 μg l-1 (sample 36) and 9.58 μg l-1 (sample 70). In general, the highest chl a concentrations occurred in samples lying inshore. The chl a concentrations in these areas always exceeded 1 μg l-1. However, the majority of samples (21) showed chl a concentrations lower than 0.5 μg l-1. All these samples, except for sample 68, were taken offshore.
Fig. 2 Total chlorophyll a (chl a) concentration and size class distribution of total chl a based on CHEMTAX identification of the various algae classes. a. Total chl a. b. Proportion of picoeukaryotes. c. Proportion of nanoeukaryotes. d. Proportion of microeukaryotes.
The contribution of the three size classes (picoeukaryotes (0.2–2 μm), nanoeukaryotes (2–20 μm) and microeukaryotes (>20 μm)) to total chl a showed that picoeukaryotes (Fig. 2b) did not significantly contribute to phytoplankton biomass throughout the entire transect. The highest contribution of picoeukaryotes occurred in samples 47 (4%), 48 (3.2%), 54 (6.5%) and 62 (7.7%), mainly samples lying inshore. In all other samples, picoeukaryotes did not exceed a contribution of 1.9%. In general, nanoeukaryotes showed the highest contribution to total chl a in offshore samples (Fig. 2c). In these areas, they contributed up to 58.5% (sample 36). In offshore samples, they accounted for 33% ± 10% of chl a on average, whereas in inshore samples they only accounted for 25% ± 15% on average. Sample 68, with a contribution of nanoeukaryotes of 63%, presented as an outlier, just as for total chl a concentration. The lowest contribution of nanoeukaryotes (c. 14%) was shown by the two polynya samples (samples 47 and 48). Microeukaryotes were always the dominant size class (Fig. 2d), except for samples 33, 35, 36 and 68, where nanoeukaryotes were dominant. Microeukaryotes contributed 35.8–84.5% to total chl a, in which the highest values occurred generally in inshore samples (except sample 68). In these areas, they contributed 73% ± 15% on average, whereas in offshore ocean samples they accounted for 66% ± 11% on average.
Automated ribosomal intergenic spacer analysis
The fragment length analysis of the ITS1 region of all 34 surface water samples resulted in 97 different fragments with a length of 50–432 bp, of which 16 only occurred in one sample (unique fragments). The number of fragments in each sample was 26 on average, ranging from nine (sample 50) to 49 (sample 68). The ordination analysis based on the ARISA profiles (Fig. 3) clustered the samples in three groups. Group A includes samples 33–42, group B contains samples 44 and 45 and 49–62, and group C includes samples 46–48 and 68–71. The three groups show significantly different ARISA profiles (ANOSIM, R = 0.637, P = 0.001). Groups A and B consist of offshore samples and represent the western and eastern part of the transect, respectively. Group C consists of samples collected inshore. Samples 58–62 fall into group B, although they were located over the shelf.
Fig. 3 MDS plot based on Jaccard distances of all 34 samples, gained via ARISA profiles. Colours of the samples indicate the three groups (red = group A, blue = group B, green = group C).
The ARISA profiles distances are significantly correlated with the distances of environmental conditions profiles (Mantel test, r = 0.142, P = 0.023). Figure 4 shows the PCA of the environmental conditions and the HPLC size fractions with the three ARISA groups plotted in. The two axes are explaining 67% of the total variance. Group C is mainly separated from group A and B by higher microeukaryotic contribution and a higher ice coverage. Group A is primarily separated from group B by higher salinities, lower temperatures, and a higher nanoeukaryotic contribution. Group B shows the highest water temperatures and the highest contribution of picoeukaryotes.
Fig. 4 Principal component analysis of environmental conditions and HPLC size fractions with plotted ARISA groups (A, B and C). Both axes are explaining 67% of the variance (PC1: 39%, PC2: 28%). Group A shows greater water depths, higher salinities, and a higher contribution of nanoeukaryotes. Group B is characterized by lower salinities and the highest picoeukaryotic contribution. Group C shows a higher ice coverage and a high contribution of microeukaryotes. d = axis scaling factor.
Detailed community structure
To obtain detailed taxonomic information about the community, we sequenced seven samples (samples 41, 47, 51, 57, 62, 69 and 70), spanning the entire transect and including all three ARISA groups. Three samples (samples 41, 51 and 57) were taken in open ocean waters and four samples (samples 47, 62, 69 and 70) were taken in inshore waters.
454-pyrosequencing
The summary of recovered 454-pyrosequencing reads is shown in Table I. In total, 278 116 sequence reads were obtained from 454-pyrosequencing, of which 77.1% had an acceptable length (300–670 bp). After the quality filtering, 56.5% of the total reads were left for analysis. The number of analysed reads ranged between 14 219 (sample 47) and 29 241 (sample 57). Subsequent to the clustering, 4044 different OTUs could be observed. The number of OTUs for each sample (Table I) ranged between 893 (sample 47) and 1687 (sample 69), at which only 0.7% (sample 57 and 70) to 1.7% (sample 47) were abundant (number of reads ≥1% of total reads). The proportion of unique OTUs (i.e. OTUs occurring in one sample only) was 36%.
The rarefaction curves (Fig. 5) show that none of the samples demonstrates saturation. However, the stacking of the curves suggests that samples 41, 47 and 51 harboured the lowest diversity.
Fig. 5 Rarefaction analysis for each of the seven sequenced samples based on clustering at the 97% similarity level.
The relative abundance of sequences assigned to major protist groups is shown in Fig. 6. Haptophytes showed a read abundance of 9–17% in offshore samples and 14–37% in inshore samples, except in sample 69, where they accounted for only 3%. Chlorophytes occurred in significant amounts only inshore where they composed 1.7–6.3% of the reads. Sample 69 was again an exception, because here chlorophytes only accounted for 0.6% of the reads. Pelagophytes only occurred in great quantities in one offshore sample (sample 41) with 11.6% of the sequence reads. Diatoms were the dominating group in samples 41, 47, 57, 69 and 70 with a read abundance of 40%, 52%, 44.7%, 48.3% and 40.3%, respectively. In samples 51 and 62, they accounted for 28.2% and 11.3% of the reads, respectively. Labyrinthulids occurred in significant amounts only inshore, in samples 62, 69 and 70, where their read abundance accounted for 2.5%, 2.5% and 6.3%, respectively. The read abundance of the marine stramenopiles (MAST) group comprised 1.4–6.6%, whereas the highest abundance occurred in sample 62. Dinoflagellates dominated the sequence assemblage in sample 51 with 38%. In the other samples, they accounted for 9.5–21.2% of the reads. In general, dinoflagellates showed a higher read abundance in offshore than in inshore samples. The highest read abundance of Syndiniales occurred in sample 57 (12.6%). In the other samples, they accounted for 2.5–10.7% of the reads. Ciliates played a minor role in all sequence assemblages, except in sample 69, where they account for 17.6% of the sequences.
Fig. 6 Relative abundance of sequence reads, gained via 454-pyrosequencing, assigned to major taxonomic groups. Blue encircled samples = “offshore”, green encircled samples = “inshore”, * = 100% ice coverage.
Of the 4044 OTUs, 34 were abundant (i.e. abundance > 1%) in at least one sample. A detailed overview of the relative read abundances of the abundant phylotypes is shown in Fig. 7. Three phylotypes were abundant in all seven samples (Phaeocystis sp. 1, Eucampia sp. and unclassified (unc.) Dinoflagellate 1). They were also among the most abundant phylotypes across the entire transect. The Phaeocystis sp. 1 OTU showed the highest read abundance inshore, in samples 70 and 62, with 21% and 26.7%, respectively. However, in sample 69 it was almost rare, with the lowest read abundance of 1.5%. The chlorophytes, represented by Micromonas sp. and Pyramimonas sp., were only abundant inshore (sample 47 and 62), with 3.7% as highest sequence abundance. We found nine abundant phylotypes among the diatoms. The most abundant was Eucampia sp., with a read abundance up to 23.6% (sample 69). Only in sample 47 and 51, the most abundant diatom phylotype was not Eucampia sp., but Pseudo-nitzschia sp. (13.8%) and Chaetoceros sp. 1 (12.6%), respectively. Pelagomonas sp., belonging to the pelagophytes, showed a high read abundance offshore, in sample 41 (10.4%), whereas it was nearly rare in all other samples. Among the rest of the “other stramenopiles”, the unc. labyrinthulid OTU showed the highest read abundance in sample 70 (4.8%). We found four abundant dinoflagellate phylotypes, of which the unc. Dinoflagellate 1 was the most abundant, with a read abundance ranging from 5.2–19.8%. The highest abundance appeared offshore in sample 51. The other dinoflagellate phylotypes did not exceed a read abundance of 2.1%. Among the abundant Syndiniales phylotypes, the unc. Syndiniales 2 showed the highest read abundance with 3.9% in sample 57. Ciliate phylotypes were only abundant in sample 69, where the unc. Ciliate 1 OTU showed the highest sequence abundance (3.4%). The rare biosphere accounted for 34.2% (sample 47) to 45.8% (sample 62) of all reads.
Fig. 7 Colour-coded matrix plot, illustrating the relative read abundance of abundant OTUs (operational taxonomic units) (abundance ≥1%, at least in one sample) in the seven sequenced samples. White boxes indicate the absence of the respective OTU.
Discussion
Structure/diversity overview and biogeographical patterns
One aim of this study was to determine the structure and diversity of eukaryotic protist assemblages in the Amundsen Sea and to assess the impact of environmental conditions on their biogeographical patterns. We used a combination of pigment analysis (HPLC) and ARISA to get an overview of the structure/diversity and the biogeographical patterns. The resulting ARISA profiles were linked with the environmental conditions.
Previous biogeographical classifications of surface waters are broad and of larger scale (Spalding et al. Reference Spalding, Agostini, Rice and Grant2012). For shelf regions the existing classifications are more detailed (Spalding et al. Reference Spalding, Fox, Halpern, McManus, Molnar, Allen, Davidson, Jorge, Lombana, Lourie, Martin, McManus, Recchia and Robertson2007). In our previous study we confirmed characteristic protistan assemblages for each large-scale water mass in the Southern Ocean (Wolf et al. unpublished). However, it is also important to study more regional patterns, to complement our knowledge about the diversity and biogeography of protists in the Pacific sector of the Southern Ocean.
In general, we observed clear differences of total chl a concentrations between the samples taken offshore and inshore. Inshore, the concentrations always exceeded 1 μg l-1. This is congruent with other studies, which observed higher chl a concentrations in Antarctic shelf and coastal waters than in open oceanic waters (Hashihama et al. Reference Hashihama, Hirawake, Kudoh, Kanda, Furuya, Yamaguchi and Ishimaru2008, Olguin & Alder Reference Olguin and Alder2011). Along the entire transect, sample 70 showed the highest chl a concentration with 9.58 μg l-1, indicating a large phytoplankton bloom in this area. The high chl a value is not surprising, since recent investigations observed chl a concentrations up to 8–14 μg l-1 in the shelf area of the Amundsen Sea (Alderkamp et al. Reference Alderkamp, Mills, van Dijken, Laan, Thuroczy, Gerringa, de Baar, Payne, Visser, Buma and Arrigo2012, Fragoso & Smith Reference Fragoso and Smith2012, Mills et al. Reference Mills, Alderkamp, Thuroczy, van Dijken, Laan, de Baar and Arrigo2012).
The high chl a concentrations we observed above the shelf were accompanied by higher proportions of microeukaryotes. Higher chl a concentrations were often connected with high abundances of larger cells, like diatoms (Ishikawa et al. Reference Ishikawa, Wright, van den Enden, Davidson and Marchant2002). The geomorphology in the shelf areas promotes upwelling and mixing and thus, the nutrient availability in this region is higher, which promotes the build-up of biomass and favours larger cells (Irwin et al. Reference Irwin, Finkel, Schofield and Falkowski2006). Picoeukaryotes were of minor importance throughout the entire transect, which is in contrast to Diez et al. (Reference Diez, Massana, Estrada and Pedros-Alio2004), who found out that cells <5μm can contribute up to 80% to total chl a in Southern Ocean waters. However, they investigated a different area of the Southern Ocean (Drake Passage) and focused on cells <5μm, which include small nanoeukaryotes. In our study, nanoeukaryotes were the counterpart to microeukaryotes in the investigated area. They showed their highest contribution in samples where microeukaryotes were less abundant.
The ARISA profiles generally support the existence of an offshore and an inshore group in the investigated area. The offshore group is split into a western and an eastern part, of which the eastern part was characterized by lower salinities, due to melting ice in this area. Samples 58–62 belong to the second offshore group, although they were taken above the shelf. One explanation could be that these areas are more influenced by open oceanic water. In these areas, Circumpolar Deep Water (CDW) is flowing onto the continental shelf through troughs in the shelf as modified CDW (Alderkamp et al. Reference Alderkamp, Mills, van Dijken, Laan, Thuroczy, Gerringa, de Baar, Payne, Visser, Buma and Arrigo2012) and may influence the surface layer (upwelling). However, it appears more likely that wind is the major determining factor, influencing the direction of the surface currents.
Detailed community structure
This study delivers the first protist diversity overview gained by molecular data. Previous studies used pigment based techniques and therefore lack deeper taxonomical resolution (Alderkamp et al. Reference Alderkamp, Mills, van Dijken, Laan, Thuroczy, Gerringa, de Baar, Payne, Visser, Buma and Arrigo2012, Fragoso & Smith Reference Fragoso and Smith2012, Mills et al. Reference Mills, Alderkamp, Thuroczy, van Dijken, Laan, de Baar and Arrigo2012).
The most prominent taxonomic group across the entire transect were the diatoms. This group was previously observed to dominate in the Amundsen Sea, especially in the sea ice zones (Alderkamp et al. Reference Alderkamp, Mills, van Dijken, Laan, Thuroczy, Gerringa, de Baar, Payne, Visser, Buma and Arrigo2012, Fragoso & Smith Reference Fragoso and Smith2012, Mills et al. Reference Mills, Alderkamp, Thuroczy, van Dijken, Laan, de Baar and Arrigo2012). The most dominant diatom in the Pine Island Bay was Eucampia sp. Garibotti et al. (Reference Garibotti, Vernet, Ferrario, Smith, Ross and Quetin2003) found a large contribution to total diatom biomass of Eucampia antarctica (Castracane) Mangin in Marguerite Bay (Antarctic Peninsula). It seems that the conditions in bays may constitute an optimal environment for Eucampia to grow. In the Amundsen polynya, we found Pseudo-nitzschia sp. as the most dominant diatom, whereas offshore, Chaetoceros sp. was generally the dominant diatom. These two genera were previously reported to dominate in waters around Antarctica (Gomi et al. Reference Gomi, Umeda, Fukuchi and Taniguchi2005).
Sample 62, in contrast, showed a dominance of Phaeocystis sp. A dominance of Phaeocystis antarctica Karsten in several regions of the Amundsen Sea was previously reported (Alderkamp et al. Reference Alderkamp, Mills, van Dijken, Laan, Thuroczy, Gerringa, de Baar, Payne, Visser, Buma and Arrigo2012, Mills et al. Reference Mills, Alderkamp, Thuroczy, van Dijken, Laan, de Baar and Arrigo2012). Arrigo et al. (Reference Arrigo, Robinson, Worthen, Dunbar, DiTullio, VanWoert and Lizotte1999) revealed that Phaeocystis antarctica dominates where waters are deeply mixed, whereas diatoms dominate in highly stratified waters. Hence, the domination of Phaeocystis sp. in sample 62 could be due to more deeply mixed water. Another explanation could be that the succession at the eastern edge of the transect was most advanced (post bloom), due to a longer period free of ice, retraced via Advanced Microwave Scanning Radiometer (AMSR) satellite images (Spreen et al. Reference Spreen, Kaleschke and Heygster2008). In polar waters, after a diatom dominated bloom, Phaeocystis often dominated the post bloom situation (McMinn & Hodgson Reference McMinn and Hodgson1993).
Sample 69 showed the most extreme ice condition with 100%. Here, the read abundance of Phaeocystis was very low. The lack of wind stress, due to the ice coverage, could have caused the water to be highly stratified, and therefore led to a low Phaeocystis abundance (Arrigo et al. Reference Arrigo, Robinson, Worthen, Dunbar, DiTullio, VanWoert and Lizotte1999, Goffart et al. Reference Goffart, Catalano and Hecq2000). Under the ice, ciliates showed their highest read abundance. This corresponds to other observations of the under-ice community structure, which revealed that heterotrophic biomass was dominated by ciliates (Ichinomiya et al. Reference Ichinomiya, Honda, Shimoda, Saito, Odate, Fukuchi and Taniguchi2007).
In contrast to our previous study, focusing on the distribution of eukaryotic protists across the main oceanic fronts of the Southern Ocean (Wolf et al. unpublished), the distribution of OTUs was more even. The proportion of unique OTUs was only half the amount (37%) that it was across the main fronts of the Southern Ocean (76%). This is distinctly visible in the distribution of the abundant biosphere (Fig. 7). There were only a few OTUs, which were not present in all samples (10.1%), whereas in our previous study there were many more (20.4%). Here, in the single large-scale water mass, the observed “rare biosphere” serves as a background population and several species may become abundant when the environmental conditions change. However, in both studies the rarefaction curves suggest that none of the samples have been exhaustively analysed by sequencing. Nevertheless, the rarefaction curves indicate that the highest diversity was observed under the ice (sample 69).
There are some potential biases, which can confound the interpretation of molecular data. The amplification of the different species in an environmental sample can vary and some species might not be captured by the primers used (primer specificity) (Jeon et al. Reference Jeon, Bunge, Leslin, Stoeck, Hong and Epstein2008, Stoeck et al. Reference Stoeck, Bass, Nebel, Christen, Jones, Breiner and Richards2010). The ARISA approach can only serve as an approximate overview of the diversity structure, due to the qualitative character and the identical fragment lengths of some different species (Caron et al. Reference Caron, Countway, Jones, Kim and Schnetzer2012). Additionally, the number of rRNA gene copies depends on the cell size and varies between eukaryotes from one to several hundreds (Zhu et al. Reference Zhu, Massana, Not, Marie and Vaulot2005). Especially the dinoflagellates seem to have more rRNA gene copies than the other taxonomical groups and thus might be overrepresented in molecular sequence data. The placement of sequences gained via 454-pyrosequencing has to be interpreted with care, because the length of only c. 500–600 bp is affecting the robustness. Therefore, we generally did not classify the OTUs beyond the genus level.
In conclusion, we have shown that within a single water mass protist assemblages differed in dimensions and species composition, according to geographical and environmental conditions. There were two major groups, the offshore and the inshore group. Biomass and microeukaryotes contribution to total chl a were highest in the inshore group, whereas in the offshore group the contribution of nanoeukaryotes was the highest across the entire transect. Diatoms were the most prominent protist class, and the diatom species appearing as most abundant differed between the locations. We delivered the first taxon detailed protist diversity overview in the Amundsen Sea during summer.
Acknowledgements
This study was accomplished within the Young Investigator Group PLANKTOSENS (VH-NG-500), funded by the Initiative and Networking Fund of the Helmholtz Association. We thank the captain and crew of the RV Polarstern for their support during the cruise. We are grateful to F. Kilpert and B. Beszteri for their bioinformatical support. We also want to thank A. Schroer, A. Nicolaus and K. Oetjen for technical support in the laboratory and Steven Holland for providing access to the program Analytic Rarefaction 1.3. We would like to acknowledge E.M. Nöthig and K. Kohls for their insightful discussions and comments on this manuscript. We also gratefully acknowledge the constructive comments of the reviewers.
