The Fram Strait is an important conduit between the North Atlantic and the central Arctic. The eastern Fram Strait is heavily influenced by the West Spitsbergen Current, which brings warm, salty Atlantic water into the high Arctic (Figure 1) and controls the location of the Polar Front by pushing polar water further north (Tesi et al. Reference Tesi, Muschitiello, Mollenhauer, Miserocchi, Langone, Ceccarelli, Panieri, Chiggiato, Nogarotto, Hefter, Ingrosso, Giglio and Giordano2021). The upper ∼500 m in the eastern Fram Strait are dominated by Atlantic water, with cooler, fresher Arctic water underneath (Beszczynska-Möller et al. Reference Beszczynska-Möller, Fahrbach, Schauer and Hansen2012). In most of the Fram Strait, seafloor depth is 1000–4000 m, and the maximum depth of 5669 m occurs in the Molloy Deep (Klenke and Schenke Reference Klenke and Schenke2002). Near-bottom flow is variable in the eastern Fram Strait, and temperatures of both surface and bottom waters have increased over the last two decades (Soltwedel et al. Reference Soltwedel, Bauerfeind, Bergmann, Bracher, Budaeva, Busch, Cherkasheva, Fahl, Grzelak, Hasemann, Jacob, Kraft, Lalande, Metfies, Nöthig, Meyer, Quéric, Schewe, Włodarska-Kowalczuk and Klages2016; von Appen et al. Reference von Appen, Schauer, Somavilla, Bauerfeind and Beszczynska-Möller2016). The seafloor is characterized by soft sediments comprised primarily of silt, with sparse dropstones (Soltwedel et al. Reference Soltwedel, Jaeckisch, Ritter, Hasemann, Bergmann and Klages2009).

Figure 1.

Study site locations in the Fram Strait, between Greenland and Svalbard. Circle, Senke; diamond, S3; WSC, West Spitsbergen Current.

The pelagic community in the Fram Strait includes Phaeodaria, a widespread group of marine heterotrophic eukaryotic protists within the phylum Cercozoa (Nakamura and Suzuki Reference Nakamura, Suzuki, Ohtsuka, Suzaki, Horiguchi, Suzuki and Not2015). Phylogenetic relationships among cercozoans and sister groups are complex and unresolved, but the Phylum Cercozoa is monophyletic (Burki and Pawlowski Reference Burki and Pawlowski2006; Irwin et al. Reference Irwin, Tikhonenkov, Hehenberger, Mylnikov, Burki and Keeling2019). These zooplanktonic organisms play an important role in ocean carbon flux, in one case accounting for 10% of all organic carbon in mesopelagic sinking particles (Ikenoue et al. Reference Ikenoue, Kimoto, Okazaki, Sato, Honda, Takahashi, Harada and Fujiki2019). Phaeodarian skeletons are important for silica flux to the deep ocean (Takahasi et al. Reference Takahasi, Hurd and Honjo1983).

Abundance and biomass estimates for Phaeodaria are challenging because of the meticulous attention required to collect specimens intact (Stukel et al. Reference Stukel, Biard, Krause and Ohman2018). Recent studies suggest that phaeodarian abundances vary seasonally and interannually and have been previously underestimated (Nakamura and Suzuki Reference Nakamura, Suzuki, Ohtsuka, Suzaki, Horiguchi, Suzuki and Not2015). Within the eastern Arctic, Phaeodaria abundances can range 11–1554 m⁻³ (Ikenoue et al. Reference Ikenoue, Bjørklund, Krabberød, Nishino and Wassmann2023). Lattice structures of phaeodarians have been observed using the Lightframe On-site Keyspecies Identification (LOKI) underwater imaging system in the Fram Strait (Niehoff et al. Reference Niehoff, Hildebrandt, Henke and Bauerfeind2016, Reference Niehoff, Köhler and Hildebrandt2017; Schmid et al. Reference Schmid, Aubry, Grigor and Fortier2016). While imaging can be used for abundance and vertical distribution estimates of Phaeodaria (Biard et al. Reference Biard, Stemmann, Picheral, Mayot, Vandromme, Hauss, Gorsky, Guidi, Kiko and Not2016; Stukel et al. Reference Stukel, Biard, Krause and Ohman2018), specimen collection is necessary for species-level identification.

Here, we report collections of phaeodarians in the eastern Fram Strait, including specimens with 18S sequences that form distinct clusters when compared to publicly available sequences. Out of 465 species included in Haeckel’s (Reference Haeckel1887) monograph and an unknown number of species described since, only 23 phaeodarian species (<5%) had sequences available in a public repository (GenBank, accessed August 2025). Our specimens most likely belong to the >95% of phaeodarian species which have been described morphologically but have not previously been sequenced.

All specimens were collected during expedition PS126 on board R/V Polarstern. Sampling was conducted in the context of the Long-Term Ecological Research observatory HAUSGARTEN, located in the eastern Fram Strait, which has been sampled annually since 2002 to understand climatic influence and ecosystem functioning in this gateway between the North Atlantic and the central Arctic basin (Soltwedel et al. Reference Soltwedel, Bauerfeind, Bergmann, Budaeva, Hoste, Jaeckisch, von Juterzenka, Matthiessen, Mokievsky, Nöthig, Quéric, Sablotny, Sauter, Schewe, Urban-Malinga, Wegner, Włodarska-Kowalczuk and Klages2005, Reference Soltwedel, Bauerfeind, Bergmann, Bracher, Budaeva, Busch, Cherkasheva, Fahl, Grzelak, Hasemann, Jacob, Kraft, Lalande, Metfies, Nöthig, Meyer, Quéric, Schewe, Włodarska-Kowalczuk and Klages2016). A high-volume plankton pump (McLane WTS-LV) was deployed to the seafloor on a free-fall lander. The original target of this sample collection was marine invertebrate larvae, so deployment parameters followed standards for larval collection. The pump intake valve was positioned ∼1 m above the seafloor, and one pump deployment was conducted at each station in this study (Table 1). Station Senke consists of a seafloor depression adjacent to a steep rocky reef with abundant sponges (Meyer et al. Reference Meyer, Soltwedel and Bergmann2014). The lander was deployed at the foot of the rocky reef (Figure 1). Station S3 is a regular HAUSGARTEN sampling station and primarily consists of soft sediments with dispersed dropstones (Soltwedel et al. Reference Soltwedel, Bauerfeind, Bergmann, Budaeva, Hoste, Jaeckisch, von Juterzenka, Matthiessen, Mokievsky, Nöthig, Quéric, Sablotny, Sauter, Schewe, Urban-Malinga, Wegner, Włodarska-Kowalczuk and Klages2005; Taylor et al. Reference Taylor, Krumpen, Soltwedel, Gutt and Bergmann2016).

Table 1.

Collection locations and specimens described in this study

Each lander deployment was ∼24 hours. The pump was programmed to allow 1–2 hours after the initial descent for sediment to settle and then draw in water at 20–30 L min⁻¹ until the battery expired. At Senke, the pump ran for 18.25 hours and filtered a total of 31,586 L. At S3, the pump ran for 20 hours and filtered a total of 34,593 L. Filters were retrieved immediately following lander recovery, and specimens were sorted live in chilled seawater in a lab at sea. Specimens were individually photographed using a camera (Flir Flea3) in conjunction with a dissecting microscope (Olympus SZ61) and preserved in 95% ethanol.

Following the expedition, specimens were processed for DNA barcoding at Woods Hole Oceanographic Institution. DNA was extracted by incubating specimens with 200 μL Insta-Gene matrix (BioRad) and 20 μL proteinase K (Qiagen) at 56°C for 30 minutes and 100°C for 10 minutes. The mixture was then centrifuged for 3 minutes at 12,000 rpm, and DNA was recovered from the supernatant. The 18S locus has been used for the identification of phaeodarian species and as the basis for a phaeodarian phylogeny (Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015). Thus, we amplified this locus using PCR. Each reaction consisted of 14.1 μL nuclease-free water (Ambion), 5 μL Phusion high-fidelity Taq buffer, 200 μM each dNTP, 0.75 μL dimethyl sulfoxide, 0.3 U Taq DNA polymerase (New England Biolabs), 0.5 μM forward and reverse primer, and 2 μL template DNA for a total reaction volume of 25 μL. We used two sets of primers to cover the 18S locus: NS1 (5′-TAGTCATATGCTTGTCTC-3′) and NS4 (5′-CTTCCGTCAATTCCTTTAAG-3′) for the first half and NS5 (5’-ACTTAAAGGAATTGACGGAAG-3′) and NS8 (5′-TCCGCAGGTTCACCTACGGA-3′) for the latter half (White et al. Reference White, Bruns, Lee, Taylor, Innis, Gelfand, Sninsky and White1990). Thermal cycles included 4 min at 94°C for initial denaturation, followed by 30 cycles of (30 sec at 94°C for denaturation, 30 sec at 55°C for annealing, 1 min at 72°C for extension), followed by a final extension step of 4 min at 72°C. PCR products were visualized on an electrophoresis gel and sent to Sequegen (Worcester, MA) for Sanger sequencing.

Sequences were trimmed of primers, aligned (NS1-NS4 and NS5-NS8) in Geneious Prime 2024.0.2 (https://www.geneious.com), then concatenated in the forward direction. Complete consensus sequences for the 18S locus were compared to public sequences in GenBank using the megablast algorithm (Camacho et al. Reference Camacho, Coulouris, Avagyan, Ma, Papadopoulos, Bealer and Madden2009). Phylogenetic analysis was conducted by building a multiple sequence alignment (MSA) with GenBank-queried sequences and their closest identity sequences in GenBank, further expanded with near-branching clades from other phylogenetic studies (Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015). Sequences were aligned with MEGA (Koichiro et al. Reference Koichiro, Stecher and Kumar2021), and a maximum likelihood tree was generated with 100 pseudo-replications within the MEGA software. Specifically, hyper-variable and non-alignable regions were manually eliminated from the MSA. The MSA was further end-trimmed to ensure all sequences were of uniform length prior to phylogenetic testing.

Specimens from station Senke were similar to holotype and paratype sequences for Aulographis japonica (100% query cover, >99% identity, E-value 0.0) (Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji and Suzuki2013) and to Auloceros arborescens (100% query cover, >99% identity, E-value 0.0) (Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015). However, our specimens clustered together in a sister group to A. japonica and A. arborescens (Figure 2). This clustering indicated that our specimens were more closely related to each other than either A. japonica or A. arborescens, and that our sequences were as different from A. japonica and A. arborescens as these two species and genera were from each other. This clustering had high bootstrap support (>85%), meaning that the same tree topology was reproduced in >85% of 100 random data permutations. A. japonica and A. arborescens have been collected in the Sea of Japan and are not known to occur in the Fram Strait (Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji and Suzuki2013, Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015). Good agreement between our 18S tree and the phylogenetic tree of Nakamura et al (Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015) indicated that our specimens from station Senke belong with A. japonica and A. arborescens in Clade K, family Aulacanthidae. Haeckel (Reference Haeckel1887) described 6 species in the Aulacanthidae with distributions in the North Atlantic. It is possible that our specimens correspond to one of these 6 species, none of which have sequences in public repositories. Unfortunately, our specimens included central capsules only, most likely because the high shear stress of the pump disrupted the sclerocomas. Thus, we were unable to observe sclerocoma or lattice shell morphology (Figure 3). Nakamura et al (Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015) described Phaeodaria in Clade K having a sclerocoma called a ‘spherical veil’. Without morphological examination of these key structures, it is not possible to match our specimens to a taxonomically described species.

Figure 2.

Phylogenetic tree based on 18S sequences of our specimens compared to phaeodarian sequences available through GenBank.

Figure 3.

Representative specimens photographed with incident light under a dissecting microscope. (a) Specimen HG-2021-295; (b) specimen HG-2021-291; (c) specimen HG-2021-20; (d), scleracoma of specimen HG-2021-22, n = 3 individuals. Top scale bar is for panels (a–c).

Specimens collected from station S3 were similar to the holotype and paratype sequences of Auloscena pleuroclada (99% query cover, >98% identity, E-value 0.0) (Nakamura et al. Reference Nakamura, Imai, Tuji and Suzuki2016). However, 18S sequences from our specimens formed a sister clade to A. pleuroclada sequences, indicating they were more closely related to each other than to this species (Figure 2). This clustering had high bootstrap support (>85%), meaning that the same tree topology for sequences was achieved in >85% of 100 random data permutations. A. pleuroclada was described from the Sea of Japan and is not known to occur in the Fram Strait (Nakamura et al. Reference Nakamura, Imai, Tuji and Suzuki2016). Comparison of our 18S tree to the results of Nakamura et al. (Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015) suggested that our specimens from S3 belong in Clade E, family Aulosphaeridae, genus Auloscena. Lattice shells observed around our specimens provided further support for their inclusion in Clade E (Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015). Specifically, pyramidal and radial extensions of the sclerocoma strongly resembled Auloscena spp. specimens depicted by Haeckel (Reference Haeckel1887) (Figure 3). While Haeckel (Reference Haeckel1887) did not collect any specimens in the North Atlantic-Arctic region, Haecker (Reference Haecker and Chun1908) described Auloscena verticillus hamata as a bipolar, cold-water subspecies found in the Labrador Current and Irminger Sea. It is possible that our specimens belong to this subspecies or a new species of Auloscena.

In this study, the two phaeodarian taxa were collected in different environments, with Aulacanthidae occurring at a steep rocky reef station (Senke) and Auloscena sp. occurring at a station with flat bathymetry, soft sediments, and few stones (S3). Bottom water at both stations is Arctic water; however, predominant near-bottom flow is to the northwest near station Senke and to the southeast near S3 (von Appen et al. Reference von Appen, Schauer, Somavilla, Bauerfeind and Beszczynska-Möller2016). The presence of rocky substrata such as the rocky reef and dropstones also influences near-bottom flow patterns (Hasemann et al. Reference Hasemann, Bergmann, Kanzog, Lochthofen, Sauter, Schewe and Soltwedel2013, Meyer et al. Reference Meyer, Soltwedel and Bergmann2014). However, if phaeodarians passively sink from shallow waters, they may not be much affected by near-bottom flow. Whether and to what extent near-bottom water mass and flow direction influence phaeodarian distributions should be investigated in further studies.

Phaeodarians are classically considered pelagic organisms, so it was surprising that our specimens were collected ∼1 m above the seafloor (Nakamura and Suzuki Reference Nakamura, Suzuki, Ohtsuka, Suzaki, Horiguchi, Suzuki and Not2015). Nevertheless, most pelagic sampling devices are not brought so close to the seafloor, leaving near-bottom waters undersampled. High-throughput sequencing of Arctic sediments has revealed high abundances of cercozoan sequences (Pawlowski et al. Reference Pawlowski, Christen, Lecroq, Bachar, Shahbazkia, Amaral-Zettler and Guillou2011), raising questions about their vertical distribution in the Arctic. Some acantharians – another group of pelagic protists – sink into deep waters as part of their life cycle to release flagellated swarmers (Decelle et al. Reference Decelle, Martin, Paborstava, Pond, Tarling, Mahé, de Vargas, Lampitt and Not2013). It is possible that the phaeodarians we observed may similarly travel to deeper waters for reproduction, although further studies are needed to confirm this.

Baseline distributions of phaeodarian species are still poorly understood, even though rising temperatures have led to changing distributions of Phaeodaria in the eastern Arctic (Ikenoue et al. Reference Ikenoue, Bjørklund, Krabberød, Nishino and Wassmann2023). Our specimens likely correspond to species that have been taxonomically described, but phaeodarian taxonomy is complex and includes many sub-species, potentially cryptic species, and unresolved relationships (Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji, Not and Suzuki2015). Definitely identifying specimens such as ours will require integrative taxonomy, including morphological examination of well-preserved specimens and DNA sequencing (e.g., Nakamura et al. Reference Nakamura, Imai, Yamaguchi, Tuji and Suzuki2013).

Further research will undoubtedly reveal novel biodiversity in the deep Arctic Ocean. Long-term observatories such as HAUSGARTEN provide a powerful platform for sustained biodiversity research. Most importantly, our observations establish a baseline of sequences from the Arctic that can be referenced by researchers in the future.