Microbial responses to ocean alkalinity enhancement in seasonally hypoxic coastal sediments

Abstract

Background:

Enhanced benthic weathering (EBW) is a promising marine carbon dioxide removal (mCDR) strategy, where alkaline minerals are added to the seafloor. Upon weathering, these artificially deposited minerals release alkalinity to the water column, which ultimately leads to CO₂ uptake from the atmosphere. However, feedbacks of the benthic bacterial community to this mCDR approach remain unknown.

Methods:

To investigate the effects of EBW, sediment core incubations with added alkalizing minerals were carried out and shifts in bacterial communities identified and linked to sedimentary properties. Surface sediments with supernatant water from a seasonally hypoxic basin in the Baltic Sea off the eastern coast of northern Germany were collected and either calcite (limestone) or silicate (dunite) added and kept under oxic (avg. 263 ± 34 μM O₂) or oxygen-limited (0 to maximum 162 μM O₂) conditions.

Results:

Bacterial surface sediment communities responded to the different oxygen availabilities in the overlying bottom waters independent of the mineral addition as reflected in cluster formation via Principal Coordinates Analysis (PCoA). Although, by contrast, mineral addition induced less pronounced changes, the relative abundances of several rare taxa changed significantly, such as chemolithoautotrophs associated with the sulfur and nitrogen cycles. Candidatus Electrothrix was selectively promoted by calcite addition under oxygen-limited conditions in two of three cores, whereas Nitrosomonas exhibited increased growth tendencies following dunite addition under both oxic and oxygen depleted bottom water conditions. Sulfur oxidizing bacteria (SOB) of Beggiatoaceae declined under oxygenated water conditions, independent of the type of mineral added. The dunite amendment was associated with an increase in other SOB while calcite amendment led to the enrichment of organotrophic taxa.

Conclusion:

Our findings provide important insights into the response of microbial communities to EBW, but also highlight the necessity of field-based studies to accurately assess the effects of EBW on coastal ecosystems.

1 Introduction

Atmospheric carbon dioxide (CO₂) levels have been increasing since the industrial revolution, resulting in warming of Earth’s climate (Hansen et al., 2010; Howe, 2015; MacFarling Meure et al., 2006; Keeling et al., 2001; Petit et al., 1999; Valipour et al., 2021). The oceans are fundamental in helping to slow down climate change by taking up heat and CO₂ from the atmosphere (DeVries et al., 2017; Gruber et al., 2023). This has caused rapid changes of the oceans including higher water temperatures, deoxygenation and acidification, in turn affecting oceanic life and biogeochemical cycles (Schmidtko et al., 2017; Brauko et al., 2020; Breitburg et al., 2018). To mitigate or even counteract anthropogenically induced changes to Earth’s climate, CO₂ must be actively removed from the atmosphere and stored over long time scales (IPCC, 2022). One of the currently tested marine CO₂ removal (mCDR) strategies is ocean alkalinity enhancement (OAE) (Gattuso et al., 2018; Kheshgi, 1995) via enhanced benthic weathering (EBW) (Dale et al., 2024; Fuhr et al., 2023; Fuhr et al., 2024; Fuhr et al., 2025). In EBW, minerals like calcite (calcium carbonate) or silicate (dunite rock) are added to marine surface sediments, where their dissolution generates alkalinity (A_T) and stimulates CO₂ uptake of the water body (Dale et al., 2024; Fuhr et al., 2023; Meysman and Montserrat, 2017; Montserrat et al., 2017; Fuhr et al., 2025; Fuhr et al., 2024).

The Baltic Sea features regions where bottom waters are exposed to seasonal or permanent oxygen depletion. These temporally or permanently calcite-undersaturated deep waters offer favorable conditions for achieving most efficient EBW rates (Dale et al., 2024; Fuhr et al., 2023; Fuhr et al., 2024). Economic evaluations indicate that near coastal regions are more viable for large-scale CDR strategies compared to the open ocean (Fuhr et al., 2025; Köhler et al., 2013; Renforth and Henderson, 2017). Bottom waters at the shallow coastal Boknis Eck site in Eckernförde Bay (SW Baltic Sea, Germany) are regularly exposed to seasonal hypoxia to anoxia in late summer (Alampalli et al., under review; Perner et al., 2022). Due to these recurring low-oxygen conditions, Boknis Eck is an ideal natural laboratory to evaluate the effectiveness of EBW as a CDR strategy and to investigate potential environmental implications in a realistic temperate costal setting (Fuhr et al., 2023, 2024). Boknis Eck sediments are rich in organic matter (~5 wt.%) (Dale et al., 2013; Wallmann et al., 2022) and contain up to 4 wt.% Fe-sulfide minerals, i.e., iron mono-sulfide (FeS) and pyrite (FeS₂) (Perner et al., 2022). Carbonate distribution of surface sediments in the region is spatially patchy (0.5 to 7 wt.%) (Wallmann et al., 2022; Dale et al., 2013). The undersaturation of calcite in the sediments of Boknis Eck reaches a depth of 0.4 cmbsf in summer and 1.3 cmbsf in winter (Dale et al., 2024). This is primarily driven by elevated rates of acidifying aerobic respiration in winter, when bottom waters are well-ventilated, and reduced aerobic rates in summer under oxygen-depleted conditions. These seasonal variations are accompanied by changes in bottom water carbonate chemistry. While the natural variation in A_T is rather small (1.8 mmol L⁻¹ – 2.1 mmol L⁻¹), the accumulation of metabolic CO₂ leads to undersaturation with respect to aragonite and calcite during times on hypoxia and anoxia (Melzner et al., 2013).

Possible feedback mechanisms of the benthic microbial community to the addition of alkalizing minerals for EBW under varying oxygen availabilities have so far not been investigated. In this study, sediment cores from Boknis Eck were incubated in the dark under oxic and low-oxygen conditions after depositing alkaline minerals (calcite and dunite) to the sediment surfaces. Shifts in the composition and function of benthic microbial communities and associated biogeochemical cycling were monitored over the course of the experiments. Microbial data were combined with EBW-related chemistry changes in the water body and sediment porewaters, to identify taxa showing significant correlation with the observed geochemical shifts. Our results show distinct changes in benthic microbial communities, providing first insights into ecosystem-scale implications of EBW.

2 Materials and methods

2.1 Description of the study site

Boknis Eck is a well-studied coastal site in the southwestern Baltic Sea (10°2,46′E, 54°31,8’N) that is subject to pronounced seasonal bottom-water oxygen fluctuations (Melzner et al., 2013). The site experiences regular transitions between oxic and anoxic conditions, making it an ideal natural laboratory to investigate microbial and biogeochemical dynamics under changing redox regimes (Alampalli et al., under review; Perner et al., 2022). Following the spring phytoplankton bloom, enhanced export of organic matter to the seafloor stimulates aerobic respiration by benthic microbes, thereby increasing oxygen demand in surface sediments (Hepach et al., 2024; Lennartz et al., 2014). Together with rising surface-water temperatures and the onset of water column stratification, this enhanced respiration drives progressive oxygen depletion in bottom waters (Hansen et al., 1999). Typically, anoxic conditions develop by late summer and persist until winter storms disrupt stratification and reintroduce oxygenated surface waters to deeper layers (Lennartz et al., 2014; Dietze and Löptien, 2021).

2.2 Experimental design and sampling description

The experimental setup of the oxic and oxygen-limited sediment core incubations has been described in detail elsewhere (Fuhr et al., 2024; Fuhr et al., 2023). Here, we provide a brief overview, with a focus on the aspects relevant to microbial community dynamics (Figure 1).

Figure 1

The benthic microbial communities at the sampling site are known to reflect seasonal hypoxic environmental endpoint conditions (Alampalli et al., under review; Perner et al., 2022). In order to mirror these seasonally adapted bacterial communities, the incubations aimed at simulating oxygenated bottom waters in late winter and oxygen-depleted deep waters in early fall. Eight sediment cores were retrieved for each of the incubation conditions (oxic and oxygen-limited) and kept in the dark at 12.5 °C, reflecting typical summer bottom-water temperatures at the natural site (Melzner et al., 2013). Cores for the oxic incubation were sampled during the well mixed winter period on 13 January 2022, when bottom waters were fully oxygenated (oxygen = 243 μM), following seasonal re-ventilation of the water column (Alampalli et al., under review; Fuhr et al., 2023). In contrast, cores for the oxygen-limited incubation were collected on 26 October 2022, approximately 4 weeks after bottom waters at Boknis Eck were anoxic (oxygen <1 μM) and coincided with the onset of re-ventilation of the deep waters reaching an oxygen concentration of 179 μM (Alampalli et al., under review; Fuhr et al., 2024).

Immediately after retrieval, cores were sealed with rubber stoppers, transported to the incubation facility and prepared for incubation on the same day. Sediment cores were subjected to three treatments: (i) unamended control (C1, C2), calcite addition (Cal1, Cal2, Cal3) and dunite addition (Dun1, Dun2, Dun3). Due to technical limitations of the experimental setup (limited pump capacity), the unamended control treatment was conducted in duplicate, whereas amended treatments were conducted in triplicate. The amount of added dunite and calcite was based on the rain rate of particulate organic carbon measured at Boknis Eck, equaling 0.5 mmol cm⁻² a⁻¹ (Dale et al., 2011) and then doubled to obtain a measurable weathering signal (Fuhr et al., 2024). Thus, in the mineral amended sediment incubations 7.86 g limestone calcite; Cal1, Cal2 and Cal3 or 4.48 g olivine AFS 80 dunite; Dun1, Dun2 and Dun3 were evenly distributed on the sediment surfaces. Mineral compositions for dunite were ~89% olivine, ~7% pyroxene, ~2% serpentinite and others <2% and for limestone ~98% calcite. Average grain size distribution of the washed calcite material was 87 μm and of the dunite 20–30 μm (Fuhr et al., 2023).

For the oxic incubation stirring heads with continuous oxygenated-water exchange were installed after minerals were added The cores were then left to equilibrate until the first subsampling event of the 0–2 cm sediment horizon, which took place 4 days after mineral addition (t0_oxic = day 4) to capture the initial community composition and chemical settings in all eight cores. A second subsampling was conducted 46 days after mineral addition (t2_oxic = day 46), when one core per treatment (C1, Cal2 and Dun2) was subsampled. The final sampling occurred 69 days after mineral addition (t3_oxic = day 69).

For the oxygen-limited incubation, sediment cores were kept sealed for 7 days until the thin oxidized surface layer became anoxic. First subsamples from all eight oxygen-limited incubations were then taken from the top 2 cm of sediment (t0_{oxygen-limited} = −6 days). Then stirring heads were installed for logging of pH and oxygen in waters. Sediment core waters were continuously exchanged with Boknis Eck water bubbled with dinitrogen (N₂) gas (flow rate 700 μL min⁻¹). Although the gas phase above the water column was kept as small as possible, oxygen occasionally invaded the waters in the different sediment core incubation experiments. The overall oxygen invasion was due to technical limitations, whereby two peristaltic pump channels had to be used for continuous exchange of deoxygenated water of each individual treatment (1: inlet, 2: outlet). If the peristaltic pump channels did not run 100% synchronously (due to, e.g., different deformation of the tubing, or minimal manufactural differences), a minimal under-pressure could develop that sucked in ambient air and caused oxygen invasion over time. The second set of subsamples was taken 6 days after stirring heads were installed, just 1 day before mineral addition (t2_{oxygen-limited} = −1 day) and the third subsampling was performed after 19 days (t3_{oxygen-limited} = 19 days). Subsamples were collected with a custom syringe designed to minimize sediment-surface disturbance and maintain experimental integrity (Fuhr et al., 2023). Portions of the subsamples were preserved and stored for microbiological analyses according to the procedures outlined in the methods sections below.

2.3 Quantitative analyses: cell numbers and CARD-FISH

For total cell counts via diamidino-2-phenylindole (DAPI) and catalyzed reporter deposition–fluorescence in situ hybridization (CARD-FISH) subsamples of the 0–1 cm and 1–2 cm sediment horizons from the eight oxic and eight oxygen-limited cores were fixed in 4% formaldehyde/phosphate buffered saline (PBS) overnight at 4 °C. After washing the samples twice with sterile PBS buffer, they were stored at −20 °C in PBS/EtOH [1:1] until further processing. Cells were then detached from the sediment by ultrasonication (2 times: 20%, 20 cycles, 20 s) and diluted with sterilized PBS [1:4000] before they were concentrated onto polycarbonate filters (0.2 μm pore size, Nucleopore, Whatman, UK) using vacuum filtration (−200 mbar). Cell enumeration was carried out using a Zeiss Axio Imager. M2 epifluorescence microscope, equipped with EGFP 214 and DAPI filter sets. For cell imaging, a Zeiss Axio Cam MRM. Rev2 in combination with the 215 ZEN 2.3 software package was used. At least 500 cells or 20 grid squares were counted per sample to ensure statistical robustness.

CARD-FISH analyses to further investigate cable bacteria morphology were conducted following the procedure described by Pernthaler et al. (2002) and Thiele et al. (2011) with minor modifications. Briefly, filters were embedded in 0.2% Metaphor agarose, and filter sections designated for bacterial cell permeabilization were incubated with lysozyme solution (10 mg ml⁻¹ lysozym, 0.5 M EDTA, 1 M Tris–HCl, pH 8) at 37 °C for 60 min, followed by treatment with achromopeptidase solution (180 U ml⁻¹ achromopeptidase, 0.01 M NaCl-Tris buffer, pH 8) at 37 °C for 15 min. Subsequent inactivation was carried out in 0.01 M HCl solution for 10 min at room temperature (22 °C). Hybridization was performed overnight at 35 °C in hybridization buffer (0.9 M NaCl, 0.02 M Tris–HCl pH 7.4, 10% blocking reagent, 2% w/v dextran sulfate, 0.01% SDS, 35% formamide) including the probe DSB706 diluted 1:1510. This probe targets most members of the Desulfobulbaceae, including cable bacteria as well as species (Loy et al., 2002). For signal amplification, filters were incubated in tyramide mix at 37 °C for 30 min. Filters were subsequently counterstained with DAPI for 10 min at room temperature. The NON-EUB probe (Wallner et al., 1993) labeled with FITC was used as a negative control to verify the absence of non-specific binding.

2.4 Generating and processing of 16S rRNA gene amplicons from RNA

Frozen sediments (−80 °C) from the top 2 cm of the unamended control (C1, C2), the calcite (Cal1, Cal2, Cal3) and the dunite (Dun1, Dun2, Dun3) treatments for the three subsampled timepoints in the oxic (t1-t3_oxic) and oxygen-limited (t1-t3_{limited-oxygen}) incubation were used for RNA isolation, reverse-transcribed into cDNA and used to generate 16S rRNA gene amplicons. RNA was isolated from approximately 500 mg sediment sample using the NucleoBond RNA Soil Mini Kit (Macherey-Nagel, Düren, Germany). Subsequently, DNase I digestion was performed applying the RapidOut DNA Removal kit (Thermo Fisher Scientific GmbH, Waltham, MA, USA). Extracted RNA was reverse transcribed into cDNA according to the Invitrogen SuperScript® VILO™ cDNA Synthesis Kit (Life Technologies™, Darmstadt, Germany) and immediately purified (Zymo Clean & Concentrator-5, Irvine, CA, USA). For amplicon library construction the hypervariable V3 and V4 regions of the reverse transcribed bacterial 16S rRNA gene were amplified using primer pairs S-D-Bac-0341-b-S-17 and S-D-Bac-0785-a-A-21 for bacterial 16S tags (Klindworth et al., 2013). Sequencing was done in collaboration with the Competence Centre for Genomic Analysis (CCGA, Kiel, Germany) in a 2 × 300 bp paired-end sequencing run on the Illumina MiSeq platform (Illumina, St. Diego, USA) as described before (Böhnke et al., 2019, Perner et al., 2022).

Processing of amplicon data was done in a Qiime2 environment (Bolyen et al., 2019). Filtering and merging of demultiplexed raw reads were conducted using the dada2 plugin with default settings and removal of the previous primer sequences (Callahan et al., 2016). The SILVA database release 138 (Quast et al., 2013) was pre-trained with the respective primer pairs for bacteria (Pedregosa et al., 2011; Bokulich et al., 2018) to obtain taxonomic assignments, which were computed using the feature-classifier plugin (classify sklearn) with default settings (Bokulich et al., 2018). The phylogeny was calculated using the align-to-tree-mafft-fasttree pipeline (Chistoserdova, 2010).

2.5 Data analysis and statistical methods

Data analysis, statistics and visualizations were performed using R software (R Foundation for Statistical Computing, Vienne, Austria. URL¹) with the “microeco” package (Liu et al., 2021). Prior to downstream analyses the dataset was filtered to retain bacterial sequences only, and remove non-target taxa including mitochondria, chloroplasts, Archaea, and Eukaryota. Samples were then rarefied to the lowest sequencing depth (=31,500 reads per sample) before calculating relative abundances and statistics.

Alpha diversity was assessed using the Shannon diversity index, and pairwise comparison among treatments were performed using two-sided Wilcoxon rank-sum tests with no p-value adjustment. Beta diversity was evaluated based on weighted and unweighted UniFrac distances and visualized by Principle Coordinate Analysis (PCoA). The ellipses were calculate based on the treatment, i.e., control, calcite and dunite in all plots. Spearman rank correlations were performed to evaluate associations between relative abundances of bacterial 16S rRNA gene amplicons and environmental parameters including A_T, O₂, B, Mn, Ca²⁺, Fe, Na, Mg, Sr., Si, Ba, Li, K, DIC, pH and H₂S. The corresponding geochemical data are provided in Supplementary Table S1 and have been published and described in detail in Fuhr et al. (2023, 2024). The correlation was done with p-adjust method = “none” and correlation methods = “spearman” (Benjamini and Hochberg, 1995) on the genus level. Differential abundance analysis (DAA) was performed using the linear discriminant analysis (LDA) effect size (LEfSe) method (Segata et al., 2011), with no p-value adjustment. Taxa were ranked according to their LDA scores, which reflect the magnitude of their contribution to the differentiation between groups. Higher LDA scores indicate a greater effect size and stronger discriminatory power.

2.6 Geochemical analysis

Bottom water samples (12 mL) were collected from the outflow of each core over several minutes, filtered through 0.2 μm cellulose membrane filters, and stored refrigerated until analysis. Sampling frequency decreased from daily during the first 2 weeks to every 2–4 days toward the end of each experiment. Total A^T was determined by titration of 1 mL sub-samples with 0.02 N HCl following Ivanenkov and Lyakhin (1978), with continuous nitrogen degassing to remove CO₂ and H₂S, and standardized against an IAPSO seawater standard. Anion concentrations (SO₄²⁻, Cl⁻, Br⁻) were measured by ion chromatography (IC; METROHM 761 Compact, conductivity mode). Acidified sub-samples (3 mL + 30 μL suprapure HNO₃) were analyzed for major and trace elements (Si, Na, K, Li, B, Mg, Ca, Sr., Mn, Ni, Fe) by inductively coupled plasma optical emission spectroscopy (ICP-OES; Varian 720-ES). In the oxic experiment (Fuhr et al., 2023), nutrients (NH₄⁺ by manual photometry; PO₄³⁻, NO₂⁻, NO₃⁻ by Seal-Analytical™ QuAAtro autoanalyser) were additionally measured from frozen 5 mL aliquots.

At the end of each experiment, cores were sliced and pore waters recovered by centrifugation (50 mL Falcon tubes, 3,000 rpm, 10 min), filtered (0.2 μm), and transferred to polyethylene vials under oxygen-free conditions. Dissolved Fe²⁺ was stabilized with ascorbic acid and quantified by Ferrozin complexation within 30 min; H₂S was fixed with zinc acetate gelatin solution following dilution with oxygen-free artificial seawater (Dale et al., 2014; Dale et al., 2016). Freeze-dried and ground solid phase samples were analyzed by flash combustion for total carbon (TC), total organic carbon (TOC), total nitrogen (TN), and total sulfur (TS) — using a Carlo Erba NA-1500 (oxic experiment) or a EuroEA 3,000 (EuroVector; anoxic/hypoxic experiment). Total inorganic carbon (TIC) was calculated as TC − TOC, with method accuracy verified using BBOT and an internal sediment standard.

Bottom water pH and O₂ were continuously recorded using Pyroscience™ optodes (PHROBSC-PK8 and OXROB3-SUB). In the anoxic/hypoxic experiment, a self-constructed 16-channel logging device (PICO-O2-SUB and PICO-PH-SUB, Pyroscience) was used. All pH optodes and micro-sensors were calibrated using three TRIS buffers for brackish water conditions (Pratt, 2014; Müller et al., 2018); pH is reported on the total scale (pH_T; Dickson, 1993). Sediment micro-profiling was carried out with a motorised Unisense™ micromanipulator using Clark-type micro-sensors for O₂ (Ox-50; 100 μm steps), pH (pH-100; 300 μm steps), and H₂S (SULPH-100; 300 μm steps; calibration at 109 μmol L⁻¹). Total dissolved sulfide concentrations were calculated from combined H₂S and pH profiles following Millero et al. (1988) and Jeroschewski et al. (1996). Carbonate system properties (DIC) were calculated from measured pH and A_T following Zeebe and Wolf-Gladrow (2001), using equilibrium constants appropriate for the respective salinity, temperature, and pressure.

3 Results

3.1 Chemical dynamics of oxic vs. oxygen-limited treatments

Detailed chemical characterizations of the incubations, i.e., pore- and bottom-water chemistry, solid-phase composition, and sediment-water fluxes, have been reported in Fuhr et al. (2023) for oxic and in Fuhr et al. (2024) for oxygen-limited treatments. Here, we provide a brief summary focusing on oxygen and A_T to contextualize the microbial responses in relation to oxic and oxygen-depleted water in the core incubation (Figure 2).

Figure 2

In the oxic incubations, oxygen concentration remained largely stable between 270 and 300 μM across all eight sediment cores, with two brief technically induced decreases: one shortly after mineral addition and another between days 70 and 80. Neither decrease in oxygen, however, resulted in hypoxic water conditions (Figure 2).

In contrast, under oxygen limited water conditions, the reproducibility of oxygen levels was more challenging. Of the oxygen-limited treatments, Cal1 and Dun3, oxygen concentrations remained persistently below 2 μM. Other sediment cores experienced episodic entrainment of oxygen, up to 148 μM in C1, 136 μM in Cal2, 142 μM in Cal3, 162 μM in Dun1, and 77 μM in Dun2, yet microoxic to anoxic water conditions dominated the overall incubation period (Figure 2). By the end of the experiment on day 19, oxygen concentrations of overlying waters had declined to microoxic levels (<3 μM) in all cores, except for a brief late-stage rise to 86 μM in C1.

Clear differences in sediment surface features were observed after exposure to oxic and oxygen-limited waters (Figure 2). Under oxic conditions, the sediment surfaces showed pronounced signs of bioturbation and meiofaunal activity, resulting in particle mixing and downward transport of the added minerals into deeper sediment layers, reaching down to ~1.5 cm sediment depth (Fuhr et al., 2023). In contrast, sediments from the oxygen-limited incubation showed no visible signs of bioturbation and the minerals persisted as a distinct, unmixed layer on the sediment surface throughout the experiment. The sediments in the oxygen-limited incubation appeared much darker, likely driven by Fe-sulfide formation which is characteristic for these anoxic sediments (Zhou et al., 2023; Perner et al., 2022). Notably, the Cal2 core in the oxygen-limited incubation, which experienced the longest phase of oxygen intrusion (day3 to day16), showed a distinct rust-brown coloration of the sediment surface (Figure 2) indicative of iron (oxyhydr)oxide formation (Emerson et al., 2010).

In the oxic incubations, alkaline mineral additions induced a rapid increase of total A_T in calcite (Cal1, Cal2, and Cal3) and two of three dunite (Dun1 and Dun2) treated cores. This initial mineral driven A_T rise leveled off within a few days, after which A_T continued to increase gradually, likely due to ongoing dissolution of added minerals along with natural alkalinity generating processes, such as pyrite burial and natural weathering. Towards the end of the experiments (day 69) A_T values peaked between 2.9 mM to 3.6 mM. In contrast, Dun3 exhibited an A_T increase prior to mineral addition, showing an early peak of 4 mM between days 14 and 17, followed by a decline to 3.0 mM by the end of the experiment. Untreated cores C1 and C2 displayed substantial A_T rise despite no mineral additions: C1 increased more gradually, reaching a peak of 3.7 mM at the end of the experiment, while C2 showed an early peak between days 14 and 17 of 5.4 mM, similar to Dun3 A_T pattern, and declined to 3.6 mM, which was in the upper range of values observed in mineral-treated cores. Under oxygen-limited conditions, A_T concentrations were generally low compared to the oxic treatments, reflecting the minimal activity of natural A_T-generating processes, as evident from the A_T patterns observed in the control cores C1 and C2. Unlike in the oxic incubation, calcite treatments, showed a significant increase over the course of the experiment relative to both controls (two-sided unpaired Student’s t-test, C1: p-values = 0.001, 0.003 and 0.006 for Cal1 - Cal3; C2: p-values = 1.4 · 10⁻⁵, 3.4 · 10⁻⁵ and 2.1 · 10⁻⁵ for Cal1-Cal3). In dunite treated cores A_T concentration also increased in overlaying water. However, the observed increase was only significant relative to C2 (<0.005) but not to C1. A clearer picture emerges when examining the A_T efflux data: calcite treated sediment cores incubated under oxygen-limited conditions exhibited the highest benthic A_T efflux (Supplementary Figure S1). The addition of calcite enhanced benthic A_T release by 2.94 μmol cm⁻² d⁻¹, while dunite addition increased fluxes by 1.12 μmol cm⁻² d⁻¹ relative to the unamended controls (Fuhr et al., 2024). In contrast, experiments conducted under oxic water conditions resulted in lower A_T fluxes, namely 0.91 μmol cm⁻² d⁻¹ for calcite and 0.22 μmol cm⁻² d⁻¹ for dunite, with calcite showing a more pronounced effect than dunite, similar to what was observed in the oxygen-limited incubations.

3.2 Cell abundance

Mean cell numbers were calculated across replicates, to identify overarching patterns (Figure 3). However, it should be noted that there is a natural heterogeneity among parallel sediment cores within the same treatment. Individual sediment core data for all replicates within one treatment are shown in Supplementary Figure S2. Overall, cell numbers consistently range between 10⁹ and 10¹⁰ cells per μm², regardless of treatment, which is in line with previous observations for Boknis Eck sediments (Perner et al., 2022). In the controls, total cell numbers for the two investigated sediment depth layers (0–1 cmbsf and 1–2 cmbsf) were 4.8·10⁹ ± 1.2·10⁹ and 8.4·10⁹ ± 2.3·10⁹ cells per μm² for the oxic incubation, and 6.6·10⁹ ± 2.2·10⁸ and 9.4·10⁹ ± 5.1·10⁸ cells per μm², for the oxygen-limited incubation, respectively are estimated. Despite seasonal differences in sampling, these values did not differ on a level of significance. Calcite addition under oxic incubation conditions did not result in significant changes in absolute cell abundances in either the 0–1 cmbsf layer or the 1–2 cmbsf layer. By contrast, under oxygen-limited conditions, the addition of calcite led to a significant increase in cell numbers in the 0–1cmbsf layer. Yet, in the sediment horizon 1–2 cmbsf calcite did not cause any significant change in cell numbers, although it is worth noting that there was pronounced variability among replicates (Cal1 = 4.6·10⁹ cells per μm², Cal2 = 6.2·10⁹ cells per μm², Cal3 = 1.2·10¹⁰ cells per μm², see Supplementary Figure S2), likely masking possible treatment effects. Dunite amendment under oxygen-limited conditions resulted in a significant increase in absolute cell abundances at both investigated sediment depths.

Figure 3

3.3 (Dis)similarity of bacterial community composition and diversity

Shannon diversity was assessed to evaluate within-sample bacterial diversity (Supplementary Figure S3). Across all sediment samples, Shannon diversity averaged 6.56 ± 0.20 (mean ± SD; n = 43). No significant differences in alpha diversity were observed between oxic and oxygen-limited incubations (Wilcoxon rank-sum test, n = 43, p = 0.34) (Supplementary Figure S3A). Likewise, no significant treatment effects were detected within the two incubation conditions, as Shannon diversity did not differ significantly between control and mineral-amended sediments (all Wilcoxon rank-sum tests, p > 0.05) (Supplementary Figures S3B,C).

Despite the comparable alpha diversity across treatments and incubation conditions, beta diversity analyses revealed a clear separation of benthic bacterial communities driven by seasonally controlled oxygen availability and temperature effects in the overlaying bottom water, distinguishing fully oxic, cold (7 °C at 28 m depth) February samples from oxygen-limited, warmer (13 °C at 28 m depth) October samples in both untreated and mineral-treated sediment cores (Figures 4A,B).

Figure 4

Although the degree of clustering differed between weighted and unweighted UniFrac distance algorithms (Figures 4A,B). While the unweighted PCoA resulted in a distinct grouping of samples from February and October sediment core incubations (Figure 4B), the weighted PCoA gave less pronounced clusters exhibiting some overlap (Figure 4A). These patterns suggest that the control and both mineral treatments share abundant bacterial taxa, as reflected by their proximity in the weighted PCoA (Figure 4A), while differences in low abundant taxa contribute to the greater separation observed in the unweighted PCoA (Figure 4B). Given that the sum of the percentages of the first two principle coordinates, A (9.5%) and B (5.1%), in the unweighted PCoA is less than 15% (Figure 4B), it can be concluded that the axes explain only a small proportion of the total variability in the unweighted data. Using the weighted algorithm, at the interface between the two clusters “oxic” and “oxygen-limited,” mostly communities from endpoint sampling are found, regardless of the oxygen contents of the bottom water or added minerals (Figure 4A).

PCoAs were also performed on subsets of oxic (Figures 4C,D) and oxygen-limited (Figures 4E,F) treatments using both weighted and unweighted algorithms. Although each experimental setup of sediment core incubations had similar bacterial starter communities, community compositions changed over time in all treatments. However, these changes appeared either random, as in the oxic calcite and dunite treatments with no consistent separation across parallel incubations (Figures 4C,D) or, as in the oxygen-limited incubations, communities developed distinctly over time, but untreated experiments and mineral treatments clustered together (Figures 4E,F).

3.4 Bacterial community composition

All surface sediments were largely dominated by cyanobacterial Aphanizomenon NIES81 (Nostocaceae), potential sulfate-reducing members of the Sva0081 sediment group (Desulfosarcinaceae) and members of the potential sulfide-oxidizing Thiogranum (Thiotrichaceae) and B2M28 (Gammaproteobacteria). Also, various genera within the organotrophic Pirellulaceae were detected with key players including Rhodopirellula, Blastopirellula, Pirellula, and Rubripirellula which, together with an unspecified Pirellulaceae genus, make Pirellulaceae the most abundant family across all treatments (Figure 5, Supplementary Figures S4, S5). Around 25% of the bacterial community could not be assigned to any known phyla and were classified as unidentified taxa (Supplementary Figure S5). Their potential feedbacks to mineral addition remain enigmatic as they were excluded from statistical analyses.

Figure 5

According to differential abundance analyses (DAA), unamended treatments under oxic or oxygen-limited water conditions did not give significant shifts of bacterial taxa during the experiments. In contrast, significant shifts in certain taxa were detected for mineral treatments kept under oxic or oxygen-limited bottom waters (Figures 6A–D). Under oxic incubation conditions relative abundance of members of the Desulfobacterium catecholicum group (Desulfobacterota), the Pla3 lineage (Planctomycetota) and the denitrifying genus Run-SP154 increased on a level of significance, independent of the type of mineral added, whereas Candidatus Parabeggiatoa and Candidatus Isobeggiatoa (both Thiotrichaceae) showed a marked decline (Figures 6A,B). After dunite addition with oxic overlying waters specific SOB became enriched including the genera Thiogranum (Ectothiorhodoceae), Sulfurovum (Sulfurovaceae) and Candidatus Thiobios (Gammaproteobacteria) (Figure 6B). Ammonia oxidizing bacteria (AOB) of the genus Nitrosomonas (Nitrosomonadaceae) were significantly enriched in dunite treatments, regardless of oxygen availability (Figures 6B,D).

Figure 6

Under oxygen-limited incubation conditions and amended calcite, the enrichment of Candidatus Electrothrix (cable bacteria performing long distance electron transport) after 21 days (t3) was most striking (Figure 6C). However, the associated standard deviation is high, reflecting natural variability among replicates, as cable bacteria enrichment was observed in only two out of the three parallel calcite-amended cores (0.02% in Cal1, 6.3% in Cal2 and 1.5% in Cal 3, Supplementary Figure S6). CARD-FISH analysis verified cable bacteria enrichment in the 0–2 cm sediment horizon of Cal2 under oxygen-limited bottom waters DSB706 signals were from cable bacteria-like filamentous cells, but also from non-cable bacteria rod-shaped single cells or cell aggregates (Supplementary Figures S7A–D). Cable bacteria-like filaments were thin, elongated cells (0.6 μm width and 5.0 μm length) and thick cells (1.9 μm wide and 3.5 μm long) indicating two cable bacteria morphotypes, positing at least one further cable bacterium type other than Candidatus Electrothrix in the sample. A possible unrecognized cable bacteria in Boknis Eck sediments has previously been suggested, since sediment pH profiles gave the typical cable bacteria pH excursions, but less than 1% of bacterial 16S rRNA genes could be assigned to Candidatus Electrothrix (Fuhr et al., 2023). The NON-EUB probe indicated no unspecific binding (Supplementary Figures S7E–H). Under dunite amendment and oxygen-limited waters representatives of the Desulfobacterium catecholicum group increased only in the dunite-amended core incubations but not when calcite was added, similar to the organotrophic Blastopirellula (Pirellulacea) (Figure 6D).

3.5 Correlation of community composition with sediment properties

Overall, from five taxa that were shown to be dominant across all treatments, three exhibited no or comparatively few significant correlations with measured environmental parameters (Figure 6), suggesting a high degree of ecological robustness and limited sensitivity to short-term environmental variation. This includes members of the Sva0081 sediment group, Thiogranum and B2M28. The Sva0081 sediment group showed positive correlations on a level of significance with A_T, O₂, bottom water DIC and bottom water pH, and a negative correlation with Mn (Figures 7, 8). Thiogranum also displayed a significant negative correlation with dissolved Mn concentrations (p = 1.93 × 10⁻⁰⁵), Si, Ba and the depth at which H₂S was detectable with microelectrodes (H₂S front), while showing a positive correlation with Mg and K. The B2M28 show no statistically significant relationship to any of the tested environmental parameters.

Figure 7

Figure 8

Several genera that exhibit significant correlations with nearly all tested environmental parameters (except Mn and Ca) were observed, suggesting that these taxa are highly responsive to the prevailing geochemical gradients. Based on the correlation patterns two groups can be distinguished (Figure 7). The first group shows positive correlations with A_T in bottom water (AT_BW), dissolved inorganic carbon (calculated for bottom water, DIC_cal_BW), oxygen (from microprofile measurements, O₂_MP), pH (calculated for bottom water, pH_cal_BW), the oxygen penetration depth (O₂ depth_MP), silicate (Si_BW), the depth of the hydrogen sulfide front (H₂S front_MP), and barium (Ba_BW), alongside negative correlations with hydrogen sulfide (from microprofile measurements, H₂S_MP), boron (B_BW), magnesium (Mg_BW), potassium (K_BW), sodium (Na_BW), strontium (Sr_BW), the ratio between pH and DIC (based on microprofile measurements, ratio_pH: DIC_MP), pH (from microprofile measurements, pH_MP), lithium (Li_MP), dissolved inorganic carbone (calculated based on microprofile measurements, DIC_cal_MP), and total dissolved iron (Fe_BW). This group includes the genera OM190, Candidatus Electrothrix, Rhodopirellula, Nitrosomonas and Candidatus Thiobios. Most striking here, however, is Candidatus Electrothrix, exhibiting the strongest negative correlations with H₂S, K, Sr., and Fe, and pronounced positive correlations with O₂ and the onset of the anoxic zone (O₂ depth_MP) (Figures 7, 8), underlining its role in oxygen-dependent sulfide and iron oxidation.

Compared to the first group, the second group shows an opposite correlation pattern. This group includes the genera Sulfurimonas, Desulfatiglans, Aphanziomenon NIES81, Desulfoconvexum, Cyanobium PCC-6307, R76-B128, and Sulfurovum. The latter stands out due to its strong positive correlations with H₂S, Mg, K, Na, and Sr., as well as strong negative correlations with O₂, the oxygen penetration depth (O₂ depth_MP), and the position of the H₂S front (Figures 7, 8), consistent with an adaptation to more reduced, sulfide-rich microenvironments.

4 Discussion

The concept of EBW relies on the idea that alkalizing minerals such as calcite or dunite are added to the seafloor, where they release alkalinity as they dissolve and thereby increase atmospheric CO₂ uptake by the ocean. To evaluate the environmental implications of this approach, sediment core incubations with and without mineral addition were performed under contrasting oxygen conditions. The geochemical responses of these incubations have been comprehensively assessed in Fuhr et al. (2023, 2024), demonstrating that calcite amendments under oxygen-limited bottom waters produced the highest benthic A_T fluxes and thus the strongest apparent mCDR potential (Supplementary Figure S1). Building on this geochemical data assessment, the present study focusses on the microbial responses to EBW. Specifically, we have examined how mineral-driven A_T enhancement via dunite and calcite dissolution affected sediment microbial communities and how microbial taxa dynamics correlated with key geochemical parameters. 16S rRNA gene amplicons were generated from reverse-transcribed environmental RNA (cDNA) to specifically target the metabolically active bacterial community in the top 2 cm sediment horizons across all treatments. Overall, the bacterial community composition of the natural sediments closely resembled previous reports from Boknis Eck (Alampalli et al., under review; Perner et al., 2022), implying that the communities in the experiments are representative of this study site. Alpha diversity was evaluated to determine whether bacterial community diversity differed among incubation conditions and mineral treatments. The observed Shannon diversity was remarkably consistent across all samples, averaging 6.56 ± 0.2 (n = 43), and fell within the range previously reported for Boknis Eck sediments (Perner et al., 2022) and other organic-rich Baltic Sea sediments (Seidel et al., 2026). The absence of significant differences in Shannon diversity between incubation conditions and mineral treatments suggests that microbial richness and evenness remained largely stable throughout the experiment, despite the geochemical changes induced by mineral addition. Patterns of beta diversity were analyzed to resolve treatment- and incubation-specific differences in bacterial community structure. Principal coordinate analysis (PCoA) revealed a clear separation of samples primarily driven by oxygen availability in the overlying waters, indicating that redox regime and, potentially, seasonal differences at the time of sampling, represented the dominant structuring factors of the bacterial communities (Figures 4A,B). Mineral-specific effects were, by contrast, less pronounced, as evidenced by the lack of treatment specific clustering in the PCoA (Figures 4C–F). However, differential abundance analysis (DAA) identified specific taxa that contributed to community differentiation caused by alkaline mineral addition on a level of significance (Figure 6). Notably, no significant changes in taxa abundance were observed in the control incubations of oxic and oxygen-limited incubation over time, respectively. While this suggests limited temporal and redox-driven variability under the two experimental conditions, the results should be interpreted with caution given the limited replication in the control treatment (n = 2 due to technical constrains).

Mineral addition under oxic conditions was associated with a significant decline in relative abundances of members of Beggiatoaceae (sub-family within the Thiotrichacea), including Candidatus Parabeggiatoa and Candidatus Isobeggiatoa. In the calcite treatment, the relative abundance of Candidatus Parabeggiatoa decreased from 5.3% at t0 to 0.01% at t6 (99.8% decline), while Candidatus Isobeggiatoa decreased from 1.3% at t0 to 0.3% t6 (76.9% decline). Similar trends were observed in the dunite treatments, where relative abundances declined over the course of the experiment from 2.7 to 0.04% (98.5% decline of Candidatus Parabeggiatoa) and from 1.1 to 0.3% (72.7% decline Candidatus Isobeggiatoa), respectively. Beggiatoaceae are chemolithotrophic bacteria that oxidize sulfide to elemental sulfur using oxygen or nitrate (NO₃⁻) as a terminal electron acceptor (Teske and Salman, 2013), making them characteristic inhabitants of sulfidic sediments such as those at Boknis Eck (Perner et al., 2022; Dale et al., 2011). Here, they are of particular ecological relevance as they contribute to the regulation of sulfide (H₂S) concentrations (Preisler et al., 2007). By oxidizing H₂S, they prevent release of H₂S from the sediment into the overlying water column, thereby reducing the toxic effects of sulfide for higher organisms (Teske and Salman, 2013; Schulz and Jørgensen, 2001). Although less pronounced and not statistically significant compared to the calcite treatments, downward trends in the relative abundance of Candidatus Parabeggiatoa and Candidatus Isobeggiatoa were also observed in the control and dunite treatments (Figure 5), suggesting that the experimental conditions independent of mineral addition, may have imposed a general growth disadvantage for Beggiatoaceae spp. While Candidatus Parabeggiatoa showed a significant correlation with Mn and Ca concentrations in the overlying water (Spearman’s rank correlation, r_s = 0.42, p = 5.18 · 10⁻³ and r_s = −0.35, p = 0.021, respectively), Candidatus Isobeggiatoa significantly correlated with almost all tested environmental parameters (p < 0.05), with the exception of Mn (r_s = 0.23, p = 0.14) (Figure 7). This indicates that Candidatus Isobeggiatoa is more sensitive to environmental changes induced by redox regime and mineral addition compared to Candidatus Parabeggiatoa. The similar orientation of both genera along the same RDA gradient defined by the H₂S front and bottom water Mn (Figure 8), however, likely reflects their response to a shared redox-related environmental gradient as a main driver of their abundance. The decrease of Beggiatoaceae may have opened an ecological niche for other SOB. Consistent with this interpretation, various SOB taxa increased in relative abundance, including Thiogranum (Ectothiorhodoceae), Sulfurovum (Sulfurovaceae), Candidatus Thiobios (Gammaproteobacteria) and HOC36 (LEfSe; n = 3, all p < 0.05) (Figure 6B). This pattern indicates a potential restructuring within the sulfide-oxidizing microbial community, in which different SOB taxa may partially replace the functional role previously dominated by Beggiatoaceae. This is further supported by stable H₂S concentrations in the top 2 cm horizon (Supplementary Figure S7 in Supplementary material to Fuhr et al. (2023). However, when calcite was added under oxic bottom water conditions, the decline of Beggiatoacaea was not compensated by an increase in other SOB taxa. Instead, an enrichment of organotrophic taxa was observed including Latescibacterota, Pla3_lineage and Lentimicrobiaceae (Figure 6A). Calcite addition under oxic bottom waters may thus directly or indirectly limit the growth of other SOB taxa while providing a relative advantage to organotrophic microorganisms. However, the decline in Beggiatoaceae spp. and the lack of a compensatory increase in other SOB taxa are not consistently reflected in H₂S profiles, as only slight increases in H₂S concentrations in the top 2 cm horizon were observed in Cal1 and Cal2, whereas H₂S concentrations in Cal3 decreased (Supplementary Figure S7 in Supplementary material to Fuhr et al. (2023).

Candidatus Electrothrix (Desulfobulbaceae) appears particularly responsive to calcite amendments under oxygen-limited bottom water conditions (LEfSe; n = 3, p = 0.03, LDA score = 4.2) (Figure 6C, Supplementary Figure S6). Members of the genus Candidatus Electrothrix are widespread in the Baltic Sea and can greatly modify porewater chemistry, even at low abundances (Marzocchi et al., 2020; Hermans et al., 2019). These filamentous multicellular bacteria belong to the so-called group of cable bacteria and as such they allow for vertical long-distance electron transport over centimeter distance, coupling sulfide oxidation in the anoxic zone with oxygen reduction at the sediment surface (Marzocchi et al., 2014; Nielsen et al., 2010; Pfeffer et al., 2012). Subsurface sulfide oxidation by anodic cable bacteria cells releases 10 moles of protons per mole of sulfide (H₂S) oxidized (Burdorf et al., 2024). This porewater acidification leads to dissolution of iron sulfides (FeS) and Ca/Mn carbonates, mobilizing Fe, Ca and Mn, which diffuse upwards, where higher pH and oxygen availability facilitate Fe/Mn-oxide and Mg-calcite precipitation (Geerlings et al., 2019 and references therein). However, porewater acidification induced by cable bacteria may be counteracted by the level of the sediment pH buffering capacity, i.e., primarily carbonate dissolution (Rao et al., 2016; Burdorf et al., 2024). This may affect the availability of ferrous Fe (Fe²⁺) and thus Fe-oxide formation. The extent of Fe-oxide precipitation in low-oxygen systems is an important consideration because Fe-oxides efficiently bind free sulfide and thus act as a barrier to sulfide escaping from sediments during seasonal anoxia (Seitaj et al., 2015). By contrast, calcite addition to carbonate-rich sediments would increase sediment buffer capacity and thus counteract cable bacteria acidification to a certain degree, thus essentially making less iron available for near surface Fe-oxide precipitation. This could weaken the potential sedimentary barrier for sulfide effluxes during seasonal anoxia (Seitaj et al., 2015; Burdorf et al., 2024). The response of Candidatus Electrothrix was inconsistent across the three parallel calcite treated cores (0.02% in Cal1, 6.3% in Cal2 and 1.5% in Cal 3, Supplementary Figure S6) incubated under oxygen-limited water conditions (Supplementary Figure S6), which likely reflects temporary oxygen intrusion into Cal2 and Cal3, but not in Cal1 (Figure 2, right panel). Since no cable bacteria growth was observed in the control (C2) and the dunite treatments (Dun1 and Dun2) despite temporary exposure to oxygen during incubation, it can be concluded that the addition of calcite stimulates the growth of cable bacteria under periodically low-oxic water conditions. Cable bacteria are capable of autotrophic carbon fixation via the Wood-Ljungdale pathway and encode periplasmatic carbonic anhydrases (CAs) on their genomes (Teske, 2019; Kjeldsen et al., 2019). CAs (EC 4.2.1.1)) catalyze the reversible interconversion between HCO₃⁻ and CO₂ and can enhance CO₂ fixation in autotrophic microbes (Kupriyanova et al., 2017). In the context of EBW, where mineral alkalinity is added to the seafloor to convert CO₂ to HCO₃⁻ (Dale et al., 2024), stimulated microbial CA activity might affect the effectiveness of OAE. Under the given experimental conditions, the genus Candidatus Electrothrix was highly responsive to chemical dynamics, showing significant correlations with all measured environmental parameters except Mn and Ca (Figure 7). The strongest positive correlation was observed with pH (Spearman’s rank correlation, r_s = 0.71, p = 9.83 · 10⁻⁸), followed by O₂ (r_s = 0.67, p = 1.06 · 10⁻⁶), O₂ penetration depth (r_s = 0.67, p = 9.13 · 10⁻⁷), A_T (r_s = 0.62, p = 7.23 · 10⁻⁶), H₂S front (r_s = 0.60, p = 1.80 · 10⁻⁵) and DIC (r_s = 0.56, p = 9.28 · 10⁻⁵). In the RDA plot, Candidatus Electrothrix showed an orientation similar to Nitrosomonas and Rhodopirellula (Figure 8), reflecting comparable responses to geochemical gradients linked to alkalizing mineral addition, but also revealed links between microbial dynamics and geochemical changes in the incubation.

Members of the genus Nitrosomonas (family Nitrosomonadacaea) were enriched in both dunite treatments, exhibiting a 1.9-fold increase in relative abundance under oxic conditions (0.7 to 1.3%) and a 5.5-fold increase under oxygen-limited conditions (0.2 to 1.1%) (Figures 6B,D). At the species level, this included unassigned Nitrosomonas species as well as relatives of the species Nitrosomonas cyrotolerance. Although Nitrosomonas accounted for less than 2% of the total microbial community in the present sediment core incubations, previous studies have shown that even low-abundance Nitrosomonas populations can exhibit high transcriptional and metabolic activity (Cholet et al., 2024). Cultivated representatives of Nitrosomonas are obligate chemolithoautotrophs that oxidize ammonia (NH₃) to nitrite (NO₂⁻), which can be further oxidized to nitrate (NO₃⁻) by nitrite oxidizing bacteria, completing the nitrification (Prosser et al., 2014). Enhanced nitrogen availability in coastal environments commonly promotes eutrophication, which can ultimately result in oxygen depletion and the development of water column hypoxia or anoxia (Gruber and Galloway, 2008). Nitrogen removal processes such as microbial denitrification and anerobic ammonium oxidation (anammox) counteract this accumulation by converting reactive nitrogen species (NO₂⁻; NO₃⁻) into dinitrogen gas (N₂) (Damashek et al., 2014). In this context, nitrifying organisms including Nitrosomonas, play a major role in the nitrogen cycle as ammonia oxidation regulates the availability of both NO₂⁻ and NO₃⁻ that serve as substrates for downstream nitrogen loss pathways (Seitzinger et al., 2006; Cholet et al., 2024; Wang et al., 2025). The observed increase of Nitrosomonas following dunite addition could therefore represent a beneficial side effect, contributing to mitigate eutrophication of the Baltic Sea (Kuss et al., 2026; Murray et al., 2019). Previous studies have shown that ammonia-oxidizing microbes can adapt to acidification by increasing their substrate affinity as a compensatory response (Tong et al., 2026). However, in the sediment cores collected in October 2022 for the oxygen-limited incubation experiment, Nitrosomonas abundances tended to increase not only in the dunite-treatment, but to some extent across all sediment cores (Figure 5). Thus, conditions in the oxygen-limited incubation appear to promote the growth of Nitrosomonas, while the dunite addition seems to provide an additional growth boost. Nitrosomonas spp. are classically considered aerobic chemolithoautotrophic ammonia oxidizers, although several species have been shown to perform nitrifier denitrification under reduced oxygen concentrations, using NO₂⁻ as terminal electron acceptor to form gaseous products like dinitrogen (N²), nitrous oxide (N₂O) and nitric oxide (NO), which are ultimately released from the ecosystem (Koops et al., 1991; Herbert, 1999; Schmidt et al., 2004). This metabolic strategy likely explains their persistence in the present sediment cores incubated under oxygen-limited bottom water conditions. Across the measured environmental gradients, Nitrosomonas abundance correlated with almost all tested parameters, with the strongest correlation observed for A_T (Spearman’s rank correlation, r_s = 0.67, p = 1.10 · 10⁻⁶), dissolved inorganic carbon (r_s = 0.66, p = 1.57 · 10⁻⁶), and bottom-water oxygen (r_s = 0.65, p = 2.57 · 10⁻⁶). The positive correlation with oxygen is consistent with a possible change in the metabolism from ammonia oxidation to nitrifier denitrification due to oxygen availability, while the correlations with a A_T and DIC may reflect the altered carbonate chemistry induced by mineral weathering and the autotrophic lifestyle of Nitrosomonas, which relies on the fixation of inorganic carbon.

5 Conclusion

To our knowledge, this study provides the first assessment of the effects of enhanced benthic weathering (EBW) as a marine carbon dioxide removal (mCDR) strategy on benthic microbial communities. Laboratory-based sediment core incubations demonstrated that the addition of calcite and dunite induced significant shifts in the relative abundances of specific bacterial taxa, while the overall community structure and diversity remained largely stable. These findings suggest that microbial responses to EBW are primarily expressed through changes in selected community members rather than through a large-scale restructuring of the bacterial community.

The ecological and biogeochemical consequences of these taxonomic shifts, however, remain to be resolved. The hypothesized feedbacks derived from the chemistry of overlying bottom waters and sediment porewaters should therefore be validated through complementary analyses, including transcriptomics and microbial metabolic rate measurements, including, e.g., sulfate reduction, oxygen consumption, and carbon assimilation.

While sediment core incubations provide a controlled framework for assessing microbial responses to mineral addition, they cannot fully capture the spatial heterogeneity, biological interaction, and seasonal variability of natural coastal sediments. Consequently, the long-term ecosystem effects of EBW remain uncertain. Nevertheless, since EBW has proven to be an effective and cost-efficient strategy for mCDR (Fuhr et al., 2024; Dale et al., 2024), our results provide an important basis for future field-scale studies aimed at evaluating the ecological feasibility and environmental consequences of alkaline mineral deployment on the seafloor.

References

• 1

  AlampalliS. V.Böhnke-BrandtS.Adam-BeyerN.Laufer-MeiserK.WallmannK.BangeH. W.et al. (under review). Seasonal hypoxia at a coastal study site in the Baltic Sea serves as a model for predicting future climate driven deoxygenation and sulfide expansion in bottom waters. Nat. Commun.in revision

  • Google Scholar

• 2

  BenjaminiY.HochbergY. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Statist. Soc. Ser. B57, 289–300.

  • Google Scholar

• 3

  BöhnkeS.SassK.GonnellaG.DiehlA.KleintC.BachW.et al. (2019). Parameters governing the community structure and element turnover in Kermadec volcanic ash and hydrothermal fluids as monitored by inorganic electron donor consumption, autotrophic CO₂ fixation and 16S tags of the transcriptome in incubation experiments. Front. Microbiol.10:2296. doi: 10.3389/fmicb.2019.02296,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  BokulichN.KaehlerB.RideoutJ. R.DillonM.BolyenE.KnightR.et al. (2018). Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2’s q2-feature-classifier plugin. Microbiome6:90. doi: 10.1186/s40168-018-0470-z,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  BolyenE.RideoutJ. R.DillonM. R.BokulichN. A.AbnetC. C.Al-GhalithG. A.et al. (2019). Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat. Biotechnol.37, 852–857. doi: 10.1038/s41587-019-0209-9,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6

  BraukoK. M.CabralA.CostaN. V.HaydenJ.DiasC. E. P.LeiteE. S.et al. (2020). Marine heatwaves, sewage and eutrophication combine to trigger deoxygenation and biodiversity loss: a SW Atlantic case study. Front. Mar. Sci.7:2020. doi: 10.3389/fmars.2020.590258

  • CrossRef
  • Google Scholar

• 7

  BreitburgD.LevinL. A.OschliesA.GrégoireM.ChavezF. P.ConleyD. J.et al. (2018). Declining oxygen in the global ocean and coastal waters. Science359:eaam7240. doi: 10.1126/science.aam7240,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8

  BurdorfL. D. W.CookP. L. M.RobertsonE. K.TramperA.Hidalgo-MartinezS.Vasquez-CardenasD.et al. (2024). Electrogenic sulfur oxidation by cable bacteria in two seasonally hypoxic coastal systems. Estuar. Coast. Shelf Sci.297:108615. doi: 10.1016/j.ecss.2024.108615

  • CrossRef
  • Google Scholar

• 9

  CallahanB. J.McMurdieP. J.RosenM. J.HanA. W.JohnsonA. J. A.HolmesS. P. (2016). DADA2: high-resolution sample inference from Illumina amplicon data. Nat. Methods13:581–+. doi: 10.1038/nmeth.3869,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  ChistoserdovaL. (2010). Functional metagenomics: recent advances and future challenges. Biotechnol. Genet. Eng. Rev.26, 335–351. doi: 10.5661/bger-26-335

  • CrossRef
  • Google Scholar

• 11

  CholetF.AgoguéH.IjazU. Z.LachausséeN.PineauP.SmithC. J. (2024). Low-abundant but highly transcriptionally active uncharacterised Nitrosomonas drive ammonia-oxidation in the Brouage mudflat France. Sci. Total Environ.946:174312. doi: 10.1016/j.scitotenv.2024.174312,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12

  DaleA. W.BerticsV. J.TreudeT.SommerS.WallmannK. (2013). Modeling benthic–pelagic nutrient exchange processes and porewater distributions in a seasonally hypoxic sediment: evidence for massive phosphate release by Beggiatoa?Biogeosciences10, 629–651. doi: 10.5194/bg-10-629-2013

  • CrossRef
  • Google Scholar

• 13

  DaleA. W.GeilertS.DiercksI.FuhrM.PernerM.ScholzF.et al. (2024). Seafloor alkalinity enhancement as a carbon dioxide removal strategy in the Baltic Sea. Commun. Earth Environ.5:452. doi: 10.1038/s43247-024-01569-3

  • CrossRef
  • Google Scholar

• 14

  DaleA. W.SommerS.BohlenL.TreudeT.BerticsV. J.BangeH. W.et al. (2011). Rates and regulation of nitrogen cycling in seasonally hypoxic sediments during winter (Boknis Eck, SW Baltic Sea): sensitivity to environmental variables. Estuar. Coast. Shelf Sci.95, 14–28. doi: 10.1016/j.ecss.2011.05.016

  • CrossRef
  • Google Scholar

• 15

  DaleA. W.SommerS.LomnitzU.BourbonnaisA.WallmannK. (2016). Biological nitrate transport in sediments on the Peruvian margin mitigates benthic sulfide emissions and drives pelagic N loss during stagnation events. Deep-Sea Res. I Oceanogr. Res. Pap.112, 123–136. doi: 10.1016/j.dsr.2016.02.013

  • CrossRef
  • Google Scholar

• 16

  DaleA. W.SommerS.RyabenkoE.NoffkeA.BohlenL.WallmannK.et al. (2014). Benthic nitrogen fluxes and fractionation of nitrate in the Mauritanian oxygen minimum zone (eastern tropical North Atlantic). Geochim. Cosmochim. Acta134, 234–256. doi: 10.1016/j.gca.2014.02.026

  • CrossRef
  • Google Scholar

• 17

  DamashekJ.SmithJ. M.MosierA. C.FrancisC. A. (2014). Benthic ammonia oxidizers differ in community structure and biogeochemical potential across a riverine delta. Front. Microbiol.5:743. doi: 10.3389/fmicb.2014.00743

  • CrossRef
  • Google Scholar

• 18

  DeVriesT.HolzerM.PrimeauF. (2017). Recent increase in oceanic carbon uptake driven by weaker upper-ocean overturning. Nature542, 215–218. doi: 10.1038/nature21068

  • CrossRef
  • Google Scholar

• 19

  DicksonA. G. (1993). The measurement of sea water pH. Mar. Chem.44, 131–142. doi: 10.1016/0304-4203(93)90198-W

  • CrossRef
  • Google Scholar

• 20

  DietzeH.LöptienU. (2021). Retracing hypoxia in Eckernförde bight (Baltic Sea). Biogeosciences18, 4243–4264. doi: 10.5194/bg-18-4243-2021

  • CrossRef
  • Google Scholar

• 21

  EmersonD.FlemingE. J.McBethJ. M. (2010). Iron-oxidizing bacteria: an environmental and genomic perspective. Ann. Rev. Microbiol.64, 561–583. doi: 10.1146/annurev.micro.112408.134208,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22

  FuhrM.DaleA. W.WallmannK.BährleR.KalapurakkalH. T.SommerS.et al. (2025). Calcite is an efficient and low-cost material to enhance benthic weathering in shelf sediments of the Baltic Sea. Commun. Earth Environ.6:106. doi: 10.1038/s43247-025-02079-6

  • CrossRef
  • Google Scholar

• 23

  FuhrM.WallmannK.DaleA. W.DiercksI.KalapurakkalH. T.SchmidtM.et al. (2023). Disentangling artificial and natural benthic weathering in organic rich Baltic Sea sediments. Front. Climate5:1245580. doi: 10.3389/fclim.2023.1245580

  • CrossRef
  • Google Scholar

• 24

  FuhrM.WallmannK.DaleA. W.KalapurakkalH. T.SchmidtM.SommerS.et al. (2024). Alkaline mineral addition to anoxic to hypoxic Baltic Sea sediments as a potentially efficient CO₂-removal technique. Front. Climate6:2024. doi: 10.3389/fclim.2024.1338556

  • CrossRef
  • Google Scholar

• 25

  GattusoJ.-P.MagnanA. K.BoppL.CheungW. W. L.DuarteC. M.HinkelJ.et al. (2018). Ocean solutions to address climate change and its effects on marine ecosystems. Front. Mar. Sci.5:2018. doi: 10.3389/fmars.2018.00337

  • CrossRef
  • Google Scholar

• 26

  GeerlingsN. M. J.ZetscheE. M.Hidalgo-MartinezS.MiddelburgJ. J.MeysmanF. J. R. (2019). Mineral formation induced by cable bacteria performing long-distance electron transport in marine sediments. Biogeosciences16, 811–829. doi: 10.5194/bg-16-811-2019

  • CrossRef
  • Google Scholar

• 27

  GruberN.BakkerD. C. E.DeVriesT.GregorL.HauckJ.LandschützerP.et al. (2023). Trends and variability in the ocean carbon sink. Nat. Rev. Earth Environ.4, 119–134. doi: 10.1038/s43017-022-00381-x

  • CrossRef
  • Google Scholar

• 28

  GruberN.GallowayJ. N. (2008). An earth-system perspective of the global nitrogen cycle. Nature451, 293–296. doi: 10.1038/nature06592,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  HansenH. P.GiesenhagenH. C.BehrendsG. (1999). Seasonal and long-term control of the bottom water oxygen deficiency in a stratified shallow water coastal system. ICES J. Mar. Sci.56, 1054–3139.

  • Google Scholar

• 30

  HansenJ.RuedyR.SatoM.LoK. (2010). Global surface temperature change. Rev. Geophys.48, RG4004. doi: 10.1029/2010RG000345

  • CrossRef
  • Google Scholar

• 31

  HepachH.PiontekJ.BangeH. W.BarthelmeßT.von JackowskiA.EngelA. (2024). Enhanced warming and bacterial biomass production as key factors for coastal hypoxia in the southwestern Baltic Sea. Sci. Rep.14:29442. doi: 10.1038/s41598-024-80451-w,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  HerbertR. A. (1999). Nitrogen cycling in coastal marine ecosystems. FEMS Microbiol. Rev.23, 563–590. doi: 10.1111/j.1574-6976.1999.tb00414.x,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  HermansM.LenstraW. K.Hidalgo-MartinezS.van HelmondN.WitbaardR.MeysmanF. J. R.et al. (2019). Abundance and biogeochemical impact of cable bacteria in Baltic Sea sediments. Environ. Sci. Technol.53, 7494–7503. doi: 10.1021/acs.est.9b01665,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  HoweJ. P. (2015). This is nature; this is un-nature: reading the keeling curve. Environ. Hist.20, 286–293. doi: 10.1093/envhis/emv005

  • CrossRef
  • Google Scholar

• 35

  IPCC (2022). “Summary for policymakers,” in Global Warming of 1.5°C: IPCC Special Seport on Impacts of Global warming of 1.5°C above pre-Industrial Levels in context of Strengthening Response to Climate Change, Sustainable Development, and Efforts to Eradicate Poverty, ed. Intergovernmental Panel On Climate C (Cambridge: Cambridge University Press).

  • Google Scholar

• 36

  IvanenkovV. N.LyakhinY. I. (1978). Determination of the alkalinity of seawater. Methods Hydrochem. Stud. Ocean110:115.

  • Google Scholar

• 37

  JeroschewskiP.SteuckartC.KühlM. (1996). An Amperometric microsensor for the determination of H₂S in aquatic environments. Anal. Chem.68, 4351–4357. doi: 10.1021/ac960091b

  • CrossRef
  • Google Scholar

• 38

  KeelingC. D.PiperS. C.BacastowR. B.WahlenM.WhorfT. P.HeimannM.et al. (2001). Exchanges of atmospheric CO₂ and ¹³CO₂ with the terrestrial biosphere and oceans from 1978 to 2000. I. Global Aspects, SIO Reference Series, No. 01–06, 88.

  • Google Scholar

• 39

  KheshgiH. S. (1995). Sequestering atmospheric carbon dioxide by increasing ocean alkalinity. Energy20, 915–922. doi: 10.1016/0360-5442(95)00035-F

  • CrossRef
  • Google Scholar

• 40

  KjeldsenK. U.SchreiberL.ThorupC. A.BoesenT.BjergJ. T.YangT.et al. (2019). On the evolution and physiology of cable bacteria. Proc. Natl. Acad. Sci. USA116, 19116–19125. doi: 10.1073/pnas.1903514116,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 41

  KlindworthA.PruesseE.SchweerT.PepliesJ.QuastC.HornM.et al. (2013). Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res.41:e1. doi: 10.1093/nar/gks808,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42

  KöhlerP.AbramsJ.VoelkerC.HauckJ.Wolf-GladrowD. (2013). Geoengineering impact of open ocean dissolution of olivine on atmospheric CO2, surface ocean pH and marine biology. Environ. Res. Lett.8, 14009–14009. doi: 10.1088/1748-9326/8/1/014009

  • CrossRef
  • Google Scholar

• 43

  KoopsH. P.BöttcherB.MöllerU. C.Pommerening-RöserA.StehrG. (1991). Classification of eight new species of ammonia-oxidizing bacteria: Nitrosomonas communis sp. nov., Nitrosomonas ureae sp. nov., Nitrosomonas aestuarii sp. nov., Nitrosomonas marina sp. nov., Nitrosomonas nitrosa sp. nov., Nitrosomonas eutropha sp. nov., Nitrosomonas oligotropha sp. nov. and Nitrosomonas halophila sp. nov. Microbiology137, 1689–1699. doi: 10.1099/00221287-137-7-1689

  • CrossRef
  • Google Scholar

• 44

  KupriyanovaE.ProninaN.LosD. (2017). Carbonic anhydrase — a universal enzyme of the carbon-based life. Photosynthetica55, 3–19. doi: 10.1007/s11099-017-0685-4

  • CrossRef
  • Google Scholar

• 45

  KussJ.HoltermannP.UmlaufL.DellwigO.PrienR. D.WaniekJ. J. (2026). The changing Baltic Sea: between nutrient load reduction and a warming climate. Annu. Rev. Mar. Sci.18, 219–244. doi: 10.1146/annurev-marine-040324-020707

  • CrossRef
  • Google Scholar

• 46

  LennartzS. T.LehmannA.HerrfordJ.MalienF.HansenH. P.BiesterH.et al. (2014). Long-term trends at the Boknis Eck time series station (Baltic Sea), 1957–2013: does climate change counteract the decline in eutrophication?Biogeosciences11, 6323–6339. doi: 10.5194/bg-11-6323-2014

  • CrossRef
  • Google Scholar

• 47

  LiuC.CuiY.LiX.YaoM. (2021). Microeco: an R package for data mining in microbial community ecology. FEMS Microbiol. Ecol.97:fiaa255. doi: 10.1093/femsec/fiaa255,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48

  LoyA.LehnerA.LeeN.AdamczykJ.MeierH.ErnstJ.et al. (2002). Oligonucleotide microarray for 16S rRNA gene-based detection of all recognized lineages of sulfate-reducing prokaryotes in the environment. Appl. Environ. Microbiol.68, 5064–5081. doi: 10.1128/AEM.68.10.5064-5081.2002,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49

  MacFarling MeureC.EtheridgeD.TrudingerC.SteeleP.LangenfeldsR.van OmmenT.et al. (2006). Law dome CO₂, CH₄ and N₂O ice core records extended to 2000 years BP. Geophys. Res. Lett.33, L14810. doi: 10.1029/2006GL026152

  • CrossRef
  • Google Scholar

• 50

  MarzocchiU.PalmaE.RossettiS.AulentaF.ScomaA. (2020). Parallel artificial and biological electric circuits power petroleum decontamination: the case of snorkel and cable bacteria. Water Res.173:115520. doi: 10.1016/j.watres.2020.115520,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51

  MarzocchiU.TrojanD.LarsenS.MeyerR. L.RevsbechN. P.SchrammA.et al. (2014). Electric coupling between distant nitrate reduction and sulfide oxidation in marine sediment. ISME J.8, 1682–1690. doi: 10.1038/ismej.2014.19,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 52

  MelznerF.ThomsenJ.KoeveW.OschliesA.GutowskaM.BangeH.et al. (2013). Future Ocean acidification will be amplified by hypoxia in coastal habitats. Mar. Biol.160, 1875–1888. doi: 10.1007/s00227-012-1954-1

  • CrossRef
  • Google Scholar

• 53

  MeysmanF. J. R.MontserratF. (2017). Negative CO₂ emissions via enhanced silicate weathering in coastal environments. Biol. Lett.13, 20160905. doi: 10.1098/rsbl.2016.0905,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54

  MilleroF. J.PleseT.FernandezM. (1988). The dissociation of hydrogen sulfide in seawater. Limnol. Oceanogr.33, 269–274. doi: 10.4319/lo.1988.33.2.0269

  • CrossRef
  • Google Scholar

• 55

  MontserratF.RenforthP.HartmannJ.LeermakersM.KnopsP.MeysmanF. J. R. (2017). Olivine dissolution in seawater: implications for CO₂ sequestration through enhanced weathering in coastal environments. Environ. Sci. Technol.51, 3960–3972. doi: 10.1021/acs.est.6b05942,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56

  MüllerJ. D.BastkowskiF.SanderB.SeitzS.TurnerD. R.DicksonA. G.et al. (2018). Metrology for pH measurements in brackish waters—part 1: extending electrochemical pHT measurements of TRIS buffers to salinities. Front. Mar. Sci.5:2018. doi: 10.3389/fmars.2018.00176

  • CrossRef
  • Google Scholar

• 57

  MurrayC. J.Müller-KarulisB.CarstensenJ.ConleyD. J.GustafssonB. G.AndersenJ. H. (2019). Past, present and future eutrophication status of the Baltic Sea. Front. Mar. Sci.6:2019. doi: 10.3389/fmars.2019.00002

  • CrossRef
  • Google Scholar

• 58

  NielsenL. P.Risgaard-PetersenN.FossingH.ChristensenP. B.SayamaM. (2010). Electric currents couple spatially separated biogeochemical processes in marine sediment. Nature463, 1071–1074. doi: 10.1038/nature08790,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 59

  PedregosaF.VaroquauxG.GramfortA.MichelV.ThirionB.GriselO.et al. (2011). Scikit-learn: machine learning in Python. J. Mach. Learn. Res.12, 2825–2830.

  • Google Scholar

• 60

  PernerM.WallmannK.Adam-BeyerN.HepachH.Laufer-MeiserK.BohnkeS.et al. (2022). Environmental changes affect the microbial release of hydrogen sulfide and methane from sediments at Boknis Eck (SW Baltic Sea). Front. Microbiol.13:1096062. doi: 10.3389/fmicb.2022.1096062,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 61

  PernthalerA.PernthalerJ.AmannR. (2002). Fluorescence in situ hybridization and catalyzed reporter deposition for the identification of marine bacteria. Appl. Environ. Microbiol.68, 3094–3101. doi: 10.1128/AEM.68.6.3094-3101.2002,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 62

  PetitJ. R.JouzelJ.RaynaudD.BarkovN. I.BarnolaJ. M.BasileI.et al. (1999). Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature399, 429–436. doi: 10.1038/20859

  • CrossRef
  • Google Scholar

• 63

  PfefferC.LarsenS.SongJ.DongM.BesenbacherF.MeyerR. L.et al. (2012). Filamentous bacteria transport electrons over centimetre distances. Nature491, 218–221. doi: 10.1038/nature11586

  • CrossRef
  • Google Scholar

• 64

  PrattK. W. (2014). Measurement of pHT values of Tris buffers in artificial seawater at varying mole ratios of Tris:Tris·HCl. Mar. Chem.162, 89–95. doi: 10.1016/j.marchem.2014.03.003

  • CrossRef
  • Google Scholar

• 65

  PreislerA.de BeerD.LichtschlagA.LavikG.BoetiusA.JørgensenB. B. (2007). Biological and chemical sulfide oxidation in a Beggiatoa inhabited marine sediment. The ISME Journal1, 341–353. doi: 10.1038/ismej.2007.50,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66

  ProsserJ. I.HeadI. M.SteinL. Y. (2014). “The family Nitrosomonadaceae,” in The Prokaryotes: Alphaproteobacteria and Betaproteobacteria, eds. RosenbergE.DelongE. F.LoryS.StackebrandtE.ThompsonF. (Berlin, Heidelberg: Springer Berlin Heidelberg).

  • Google Scholar

• 67

  QuastC.PruesseE.YilmazP.GerkenJ.SchweerT.YarzaP.et al. (2013). The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res.41, D590–D596. doi: 10.1093/nar/gks1219,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68

  RaoA.MalkinS.Hidalgo-MartinezS.MeysmanF. (2016). The impact of electrogenic sulfide oxidation on elemental cycling and solute fluxes in coastal sediment. Geochim. Cosmochim. Acta172, 265–286. doi: 10.1016/j.gca.2015.09.014

  • CrossRef
  • Google Scholar

• 69

  RenforthP.HendersonG. (2017). Assessing Ocean alkalinity for carbon sequestration. Rev. Geophys.55, 636–674. doi: 10.1002/2016RG000533

  • CrossRef
  • Google Scholar

• 70

  SchmidtI.van SpanningR. J. M.JettenM. S. M. (2004). Denitrification and ammonia oxidation by Nitrosomonas europaea wild-type, and NirK- and NorB-deficient mutants. Microbiology150, 4107–4114. doi: 10.1099/mic.0.27382-0,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 71

  SchmidtkoS.StrammaL.VisbeckM. (2017). Decline in global oceanic oxygen content during the past five decades. Nature542, 335–339. doi: 10.1038/nature21399,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 72

  SchulzH. N.JørgensenB. B. (2001). Big bacteria. Ann. Rev. Microbiol.55, 105–137. doi: 10.1146/annurev.micro.55.1.105,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 73

  SegataN.IzardJ.WaldronL.GeversD.MiropolskyL.GarrettW. S.et al. (2011). Metagenomic biomarker discovery and explanation. Genome Biol.12:R60. doi: 10.1186/gb-2011-12-6-r60,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 74

  SeidelL.LiS.Hanna-EliasS.RulaI.AhlbergL.ForsmanA.et al. (2026). Warming causes a decline in Baltic Sea coastal sediment microbial abundance. Environ. Microbiol.28:e70256. doi: 10.1111/1462-2920.70256,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 75

  SeitajD.SchauerR.Sulu-GambariF.Hidalgo-MartinezS.MalkinS. Y.BurdorfL. D. W.et al. (2015). Cable bacteria generate a firewall against euxinia in seasonally hypoxic basins. Proc. Natl. Acad. Sci. USA112, 13278–13283. doi: 10.1073/pnas.1510152112,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 76

  SeitzingerS.HarrisonJ. A.BöhlkeJ. K.BouwmanA. F.LowranceR.PetersonB.et al. (2006). Dentitrification across landscapes and waterscapes: a synthesis. Ecol. Appl.16, 2064–2090. doi: 10.1890/1051-0761(2006)016[2064:DALAWA]2.0.CO;2,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 77

  TeskeA. (2019). Cable bacteria, living electrical conduits in the microbial world. Proc. Natl. Acad. Sci.116:201913413. doi: 10.1073/pnas.1913413116

  • CrossRef
  • Google Scholar

• 78

  TeskeA.SalmanV. (2013). The Family Beggiatoaceae. The Prokaryotes: Gammaproteobacteria, (Rosenberg, E., DeLong, E.F., Lory, S., Stackebrandt, E., and Thompson, F., eds.), Berlin, Heidelberg: Springer Berlin Heidelber. pp. 93–134.

  • Google Scholar

• 79

  ThieleS.FuchsB. M.AmannR. I. (2011). “Identification of microorganisms using the ribosomal RNA approach and fluorescence in situ hybridization,” in Treatise on Water Science, ed. WildererP. (Oxford: Elsevier).

  • Google Scholar

• 80

  TongS.ShenH.HanL.-L.TangJ.-M.WanX. S.ZhengZ.-Z.et al. (2026). Ammonia oxidizers offset acidification stress via adaptive substrate affinity in aquatic ecosystems. Nat. Commun.17:2083. doi: 10.1038/s41467-026-68747-z,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 81

  ValipourM.BateniS. M.JunC. (2021). Global surface temperature: a new insight. Climate9, 81. doi: 10.3390/cli9050081

  • CrossRef
  • Google Scholar

• 82

  WallmannK.DiesingM.ScholzF.RehderG.DaleA. W.FuhrM.et al. (2022). Erosion of carbonate-bearing sedimentary rocks may close the alkalinity budget of the Baltic Sea and support atmospheric CO₂ uptake in coastal seas. Front. Mar. Sci.9:968069. doi: 10.3389/fmars.2022.968069

  • CrossRef
  • Google Scholar

• 83

  WallnerG.AmannR.BeiskerW. (1993). Optimizing fluorescent in situ hybridization with rRNA-targeted oligonucleotide probes for flow cytometric identification of microorganisms. Cytometry14, 136–143. doi: 10.1002/cyto.990140205,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 84

  WangY.ZhaoW.BaiM.QinY. (2025). Research progress on ammonia oxidizing microorganisms: the discovery, nitrogen metabolic pathways, influencing factors and contribution rate. J. Environ. Chem. Eng.13:117477. doi: 10.1016/j.jece.2025.117477,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 85

  ZeebeR.Wolf-GladrowD. (2001). CO₂ in Seawater: Equilibrium, Kinetics, Isotopes. Amsterdam: Elsevier Oceanography Book Series. 65, 346.

  • Google Scholar

• 86

  ZhouZ.HenkelS.KastenS.HoltappelsM. (2023). The iron “redox battery” in sandy sediments: its impact on organic matter remineralization and phosphorus cycling. Sci. Total Environ.865:161168. doi: 10.1016/j.scitotenv.2022.161168,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar