Giant sponge grounds of Central Arctic seamounts are associated with extinct seep life

Dense sponge grounds act as ecosystem engineers and are typically linked to areas with enhanced particle concentrations and hydrodynamic features. However, the Central Arctic’s Langseth Ridge summits are permanently ice-covered, with very low primary production (<25 g C m−2 yr−1) and minimal export fluxes (<1 g C m−2 yr−1), conditions that should limit benthic biomass. This study investigates how a surprisingly dense community of bacteriosponges persists on these extinct volcanic seamount peaks. The research tests the hypothesis that sponges meet their carbon and nitrogen needs by exploiting refractory organic matter trapped within a thick spicule–tube mat, including remnants of an extinct seep community, and by leveraging autotrophic capabilities of their symbionts, rather than relying on contemporary particulate organic matter supply.

Previous work documents dense sponge grounds from shelf to abyssal depths, often at shelf breaks and regions with hydrodynamic concentration of particulate organic carbon. Arctic–Boreal sponge grounds are frequently dominated by high-microbial-abundance demosponges (e.g., Geodia spp.) co-occurring with Stelletta and Thenea, and can reach high biomass. Sponges can utilize both particulate and dissolved organic matter and significantly impact local biogeochemical cycles. Geodia spp. host diverse microbiomes (e.g., Chloroflexi, Poribacteria) that contribute to host nutrition, antibiotic production, and waste processing. Chemosynthetic communities occur at active vents and seeps on seamounts where reduced compounds fuel microbial primary production. In the Central Arctic, primary production and export are generally too low to support high benthic biomass, raising questions about alternative carbon sources. Prior studies report microbial lineages (e.g., SAR202) capable of degrading refractory DOM and suggest potential autotrophic pathways in sponge symbionts.

Study area: Langseth Ridge in the Central Arctic (three hydrothermally inactive summits: Northern Mount, Central Mount, Karasik Seamount; depths ~585–721 m on peaks). Surveys and sampling conducted during RV Polarstern expedition PS101 (Sep–Oct 2016). Seafloor mapping and habitat imaging: Ocean Floor Observation and Bathymetry System (OFOBS) towed camera/sonar collected ~3.2 km of transects and 696 images; image analysis quantified sponge abundance, size, and habitat categories (a–d) and computed terrain variables. Sponge biomass estimated using size–volume–weight conversions from literature. Spicule–tube mat thickness and organic matter content assessed from core photographs and ash-free dry weight. Hydrography: CTD and ADCP characterized productivity proxies and currents; no evidence for topography-enhanced productivity or active venting; bottom currents <0.1 m s−1. Sample collection: Sponges, spicule–tube mat, sediments, and macrofauna collected via TV-MUC, USNEL box corer, chain bag dredge, and Nereid Under Ice ROV; seawater via CTD-Rosette; zooplankton and Oikopleura fecal pellets via side-net. Samples preserved for isotope, lipid, and omics analyses. Stable isotope analyses: Bulk δ13C and δ15N measured for sponges (Geodia parva, G. hentscheli, Stelletta rhaphidiophora; plus a calcareous sponge), putative food sources (suspended POM at 10 m, sea-ice matter, fecal pellets), sediments (0–16 cm for stable isotopes), and associated macrofauna (bryozoans/hydrozoans, crustaceans, asteroids, siboglinid and serpulid tubes). Radiocarbon (Δ14C): Measured on sponge tissue (one large and one small G. parva), siboglinid tubes, surface sediment (0–1 cm), and isolated PLFAs; inorganic Δ14C from calcareous serpulid tubes, bryozoan/hydrozoan skeletons, and a bivalve shell. Reference Δ14C for DIC, DOC, and POM from nearby Arctic studies. Lipids and CSIA: Phospholipid-derived fatty acids (PLFAs) extracted (modified Bligh-Dyer), fractionated chromatographically, and derivatized; organism-specific biomarkers identified (bacterial i/ai-branched, MUFA, MBFA; algal PUFA; sponge-specific long-chain FA). Compound-specific δ13C measured on PLFAs. Microbiome and transcriptomics: For G. parva, 16S rRNA V3–V4 amplicon sequencing (DADA2; classification with SILVA); LEfSe to identify phyla enriched vs. seawater. Metagenomics and metatranscriptomics: RNA/DNA extraction, rRNA depletion, HiSeq sequencing; assemblies (MEGAHIT, Trinity), PFAM annotation (Trinotate), expression quantification (RSEM, TMM). Pathway completeness (MEBS) and expression profiling for enzymes in carbon (including refractory DOM utilization), nitrogen, and sulfur metabolism; searched for autotrophic pathway markers; assessed presence/absence of methanotrophy/thiotrophy genes. Carbon demand and pumping: Community water processing and O2 consumption estimated from allometric relationships for Geodia barretti; converted to carbon demand using respiratory quotient assumptions. Isotope mixing: SIAR mixing model explored source contributions using δ13C and δ15N with fractionation factors reflecting microbial processing and, alternatively, classical predator values.

- Discovery of the densest known Arctic sponge ground across >15 km² on Langseth Ridge summits (585–721 m). Densest aggregations on flat peak centers had 7–11 individuals m−2; mean densities by habitat category: c = 2.8 ± 1.1 ind. m−2 (n=211), d = 5.9 ± 1.7 ind. m−2 (n=54). - Thick underlying mat up to ~15 cm of sponge spicules interwoven with blackened, empty siboglinid and serpulid tubes and bivalve shells (spicule–tube mat). Average OM content of the mat: 8 ± 1% DW. - Dominant sponges: Geodia parva, G. hentscheli, and Stelletta rhaphidiophora (Demospongiae). Median sponge diameter 17 cm (n=10,839). - Biomass: 21.9 ± 12.5 kg WW m−2 (4.1 ± 2.3 kg DW m−2) across sponge-rich areas; highest at Northern Mount ~66 kg WW m−2. Organic carbon content of sponge tissue ~30 ± 5% DW; Corg standing stock: 1213 ± 690 g C m−2 in densest areas (categories c/d) and 456 ± 190 g C m−2 over entire surveyed area. - Reproduction and juveniles: substantial budding observed; juvenile abundance 3.4 ± 1.5 m−2 (max 29.3 m−2 at NM). Radiocarbon age of juveniles (1–5 cm diameter): 133 ± 7 years. - Carbon demand and pumping: Estimated community pumps ~1640 L water m−2 day−1 at the densest site (KM); average metabolic carbon demand ~110 g C m−2 yr−1 (range 82–182), far exceeding local particulate export (<0.6 g C m−2 yr−1) and regional primary production/export budgets. - Isotopes (bulk): Demosponges δ13C ~ −19.3 to −18.2‰ and δ15N ~ 6.6 to 8.4‰; juveniles had significantly lower δ15N than adults. Calcareous sponge showed more depleted δ13C (~ −25.58‰), closer to suspended POM. Sponge isotope signatures differed from most putative food sources except siboglinid tubes and sponge-associated macrofauna. - Radiocarbon (Δ14C): Sponge bulk Δ14C similar to bacterial FA (iC15:0) and within the range of DIC at similar depths, indicating significant incorporation of carbon derived from DIC via autotrophic symbionts. Δ14C of siboglinid tubes and sediment OM lower (older) than sponge tissue; bryozoans on sponges also similar to DIC. Ages: siboglinid/serpulid tubes and bryozoans ~2392 ± 449 years BP; sediments beneath mat older than exposed sediments by >1400 years; bivalve shell ~7162 ± 29 years BP. - PLFAs and CSIA: Bacterial PLFAs dominated (>60% of total), with mid-methyl-branched fatty acids (C16–C18 MBFA) contributing 22–32%; bacterial MUFAs (C16:1ω7, C18:1ω7) ~16–18%; sponge-specific long-chain PLFAs 10–23%; algal markers <2%. δ13C of MBFAs relatively high (≈ −19.5‰), consistent with non-methanotrophic/thiotrophic autotrophy; other bacterial markers −22 to −27‰. - Microbiome: G. parva dominated by Chloroflexi (SAR202 class ~85% of Chloroflexi) with substantial Acidobacteriota, Proteobacteria; metagenomics revealed Poribacteria (~5%) and archaeal Nitrosopumilus (~8%). - Metatranscriptomics: High expression of genes for refractory DOM utilization, proteases, carbohydrate-active enzymes; signatures for ammonia and nitrate assimilation and multiple sulfur metabolism pathways (e.g., sulfoacetaldehyde degradation, thiosulfate disproportionation). Absence of key methane/thiotrophic autotrophy genes (e.g., RuBisCO typical for thiotrophs, pmo). Presence of multiple autotrophic CO2 fixation pathways (e.g., reductive citrate cycle, dicarboxylate–hydroxybutyrate, 3-hydroxypropionate cycles, reductive acetyl-CoA), indicating substantial autotrophic potential. - Mixing model (SIAR): Using fractionation factors reflecting microbial processing (Δ13C 0.5 ± 0.5‰; Δ15N 1.5 ± 0.5‰) suggested a sizeable contribution (~40%) from siboglinid tube material to sponge nutrition; this fraction decreases under higher trophic enrichment assumptions. - No hydrographic or biogeochemical evidence of active venting/degassing during surveys; no methane/sulfur-oxidizing symbionts detected in sponges; yet abundant fossil siboglinid tube detritus indicates a formerly active seep/vent community.

Findings show that exceptionally dense bacteriosponge grounds can persist in the ultra-oligotrophic, permanently ice-covered Central Arctic by relying on alternative carbon and nutrient sources. Bulk and compound-specific isotopes, radiocarbon, lipid biomarkers, and expressed microbial genes collectively indicate a mixed nutrition strategy: (1) exploitation of locally accumulated, refractory organic matter trapped in a thick spicule–tube mat, prominently including sulfurized, nitrogen-rich chitin/protein from an extinct siboglinid seep community; (2) assimilation of refractory DOM via the sponge holobiont, mediated by a microbiome dominated by Chloroflexi (SAR202) and Poribacteria with enzymatic capacity for complex polymer and sulfurized OM degradation; and (3) significant autotrophic carbon fixation by symbionts, as evidenced by Δ14C similarities to DIC and expression of multiple CO2 fixation pathways. The mismatch between high community carbon demand and minimal contemporary particulate flux underscores the importance of these non-classical sources. Spatial patterns (largest sponges over thickest tube deposits) and older ages of underlying detritus support colonization upon extinct seep remains. Sponge locomotion and spicule deposition likely enhance detritus trapping and foraging. This study thus links a modern Arctic sponge hotspot to legacy chemosynthetic production and symbiont-mediated metabolism, expanding understanding of benthic resilience and carbon cycling in low-productivity oceans.

This work documents the most northerly and densest Geodia-dominated sponge ground known, occupying Central Arctic seamount summits despite extremely low surface productivity and export. Multiple lines of evidence connect this hotspot to fossil seep detritus and reveal a holobiont strategy combining refractory OM utilization and symbiont autotrophy to meet metabolic demands. The discovery highlights the role of legacy chemosynthetic production and microbial symbioses in sustaining rich benthic communities under marine desert conditions. Future research should quantify in situ DOM and detrital fluxes to the mat, constrain autotrophic vs. heterotrophic carbon incorporation over sponge lifespans, resolve spatial–temporal dynamics of mat accumulation and sponge locomotion/foraging, and assess vulnerability of such ecosystems to changing Arctic sea-ice and circulation regimes.

- Limited sample sizes for suspended POM and some fauna constrained isotope-based source characterization (e.g., POM n=4). - Δ14C integrates mixed-age carbon pools and cannot unambiguously partition ancient vs. recent sources without compound-specific radiocarbon across more biomarkers. - Isotope mixing model results depend on assumed fractionation factors and may be influenced by substantial internal nitrogen recycling in sponges. - Carbon demand and pumping rates were inferred from allometric relationships from related species (G. barretti) and extrapolated to community scale. - Observations represent a snapshot (single expedition); temporal variability in fluxes, microbial activity, and hydrodynamics remains unresolved. - No direct measurements of in situ degradation rates of siboglinid material or DOM uptake at the site; lack of direct evidence for historical venting timing beyond radiocarbon constraints.