Pelagic calcium carbonate production and shallow dissolution in the North Pacific Ocean

The marine calcium carbonate (CaCO3) cycle is central to the global carbon cycle and intimately linked to atmospheric CO2. Pelagic calcification by planktonic organisms regulates surface alkalinity, contributes to ballasting of organic matter, and exports alkalinity, thereby influencing atmospheric CO2. Despite its importance, large uncertainties persist in both the total amount of pelagic CaCO3 production (estimates span ~0.7–4.7 Pg C yr−1) and the relative contributions of major calcifying taxa (coccolithophores, foraminifera, pteropods). Traditional views based on sediment traps and sediments suggest coccolithophores and foraminifera each contribute roughly half of global production and sedimentation, while newer work points to a potentially large role for aragonitic pteropods. The North Pacific, spanning strong gradients from subtropical to subpolar waters and containing some of the most undersaturated waters with respect to calcite and aragonite, provides a critical natural laboratory to constrain living CaCO3 standing stocks, production, and export, and to resolve the apparent discrepancy between satellite/model-derived production and shallow sediment trap exports. This study quantifies the living CaCO3 standing stock and converts it to production by major taxa, comparing production to observed export fluxes, to determine contributors to pelagic calcification, the calcite versus aragonite partitioning, and the extent of shallow dissolution.

Previous assessments of pelagic CaCO3 production indicate wide uncertainty, with satellite- or model-based approaches typically yielding higher production than export-based estimates from sediment traps. Traditional budgets and sediment observations have suggested coccolithophores and foraminifera dominate pelagic CaCO3 production and burial (~50/50), while recent analyses proposed a major role for pteropods in shallow export. Taxon-specific vulnerabilities to ocean acidification and mineralogical differences (calcite vs aragonite; Mg content) influence solubility and PIC/POC ratios, with implications for rain ratio and carbon cycling. Observations of excess alkalinity above the saturation horizon in the North Pacific have fueled debate about where, and in what form, CaCO3 dissolves in the upper water column. Discrepancies with the Buitenhuis et al. (2019) modeling study, which emphasized aragonite production and pteropod dominance, may stem from parameter choices (low coccolithophore PIC/POC, similar turnover times for disparate taxa, and assumptions that all shallow dissolution above the calcite saturation horizon is aragonite).

Sampling was conducted along a Hawaii–Alaska transect in August 2017 during the CDisK-IV (KM1712) cruise aboard R/V Kilo Moana, with five primary stations spanning subtropical gyre, transition zone, and subpolar gyre waters, plus four intermediate plankton tow stations. For calcifying zooplankton (pteropods, heteropods, foraminifera), vertically integrated oblique net tows used a 0.5 m diameter, 90 µm mesh net from the surface to below the chlorophyll maximum (150–300 m depending on site). Tow volume was calculated using a flowmeter. Samples were preserved (4% buffered formalin), split, with large pteropods/heteropods (>1 mm) picked before splitting. Foraminifera were wet picked, counted, separated (>/<125 µm), and weighed; empty tests were <2%. Pteropods and heteropods were identified (families Cavoliniidae, Cymbuliidae, Limacinidae; Atlantidae, Carinidae) and shell size measured; CaCO3 biomass was estimated from size–weight relations, conversion of wet/dry weight to POC, and PIC/POC ratios, converting PIC to CaCO3 using a factor of 8.33. For heteropods, species-specific equations were developed from measured shell length and ash weight (550 °C, 5 h) as a proxy for CaCO3. Coccolithophores and biogeochemical parameters were measured from Niskin rosette casts (CTD-equipped). Seawater (2.1–6.0 L) was filtered on 0.45 µm membranes for coccolithophore analysis. Filters were dried and analyzed by polarized light microscopy; coccosphere concentrations were computed using filtration area and counted area. Taxa identification followed standard nannoplankton taxonomy. Coccolithophore CaCO3 was estimated by converting coccospheres to coccolith numbers and applying coccolith mass calibrations, including mass variation by Emiliania huxleyi calcification degree. Loose coccoliths were also quantified, but only intact coccospheres were used to estimate production. Integrated living coccolithophore calcite standing stocks were computed from the shallowest sampling depth to the 1% fluorescence depth (ranges: St.1 6–180 m; St.2 5–215 m; St.3 5–135 m; St.4 5–130 m; St.5 5–130 m). Production calculation: Living CaCO3 standing stocks (mg m−2) were divided by taxon-specific turnover times (days) to obtain daily production (mg m−2 d−1), then scaled to annual production, with uncertainty propagated via flat probability distributions for turnover time ranges. Turnover times used: coccolithophores 1.5–10 days (0.1–1.5 divisions d−1); pteropods 5–16 days; heteropods 5–16 days; foraminifera 14–28 days (faster, shallow dwellers dominating assemblages). A steady-state assumption for living standing stocks was applied, justified by the rapid disaggregation of coccospheres upon death and fast sinking of zooplankton shells. Seasonality/interannual correction: Coccolithophore production was corrected using satellite PIC (MODIS) ratios of August 2017 to mean annual PIC (2009–2019). Foraminiferal production was adjusted using satellite chlorophyll-a ratios (August 2017 to 2002–2019 mean), reflecting known coupling to primary production. Pteropod/heteropod production was seasonally adjusted using long-term zooplankton biomass seasonality at Stations ALOHA and PAPA, scaled by latitude (summer biomass ~2× at PAPA, ~1.2× at ALOHA relative to annual mean). An alternative estimate used the MAREDAT database (n=1793 upper-250 m observations in the North Pacific), converting carbon biomass to CaCO3 via PIC:POC and turnover times, with uncertainties handled via bootstrapping and Monte Carlo simulations, and truncated kernel densities to accommodate skewness and zeros. Export flux comparison: Floating PIT sediment traps (polycarbonate, 70 cm × 10 cm; 12 tubes per trap) were deployed at 100 m and 200 m for ~52–78 h at each station; samples were poisoned (HgCl2), swimmers removed, filtered, weighed, and analyzed (XRD for mineralogy; Picarro for PIC and total C). Production–export comparisons also included long-term shallow sediment traps at Station ALOHA (150 m) and Ocean Station PAPA (200 m). A simple linear regression of total production versus satellite PIC at stations was used to estimate a first-order global production from global PIC climatology, acknowledging key caveats.

• Standing stocks: Total living CaCO3 standing stock increased from subtropical gyre (~560–900 mg m−2) to subpolar gyre (~1700–4500 mg m−2). Coccolithophores dominated standing stocks at all stations, averaging ~79% (range 62–96%); pteropods ~14% (3–29%); foraminifera ~6% (0.1–22%); heteropods ~1% (0–2%). Mineralogically, calcite (coccolithophores + foraminifera) comprised ~86% (71–96%) of standing stock; aragonite ~14% (4–30%).
• Coccolithophores: Living CaCO3 concentrations ranged from 0.13 mg m−3 at 175 m (St.1) to 110 mg m−3 at 30 m (St.5); integrated stocks 753–3048 mg m−2. Loose coccoliths could contribute up to 44–64 mg m−3 during blooms.
• Pteropods: Concentrations 0.2–8.6 mg m−3; integrated stocks ~−64 to 111 mg m−2 (subtropical) and −215 to 1306 mg m−2 (subpolar). MAREDAT-based typical North Pacific values: 0.5 mg m−3 CaCO3 (0.2–1, 32–68% CI) in upper 250 m; integrated 122 mg m−2 (50–269); daily production 12 mg m−2 d−1 (5–27). Global MAREDAT dataset is highly skewed (skewness 13.3), indicating previously reported global mean pteropod CaCO3 biomass is not representative.
• Foraminifera: Integrated CaCO3 stocks 9–37 mg m−2 (subtropical) and 182–404 mg m−2 (subpolar); numbers up to 190,000–250,000 ind. m−2 in subpolar sites.
• Production by taxa: Coccolithophores accounted for ~86% (67–97%) of total annual CaCO3 production across sites; pteropods ~10% (2–17%); heteropods ~0.3% (0–1%); foraminifera ~2% (0.02–9%). Overall ~89% (70–97%) of production was calcite, remainder aragonite.
• Annual production: Seasonally corrected totals were 0.2–0.4 mol m−2 yr−1 (subtropical), and 0.9–1.0 mol m−2 yr−1 (transition/subpolar), consistent with independent estimates at OSP (1.2 mol m−2 yr−1; and 0.9 ± 0.1 mol m−2 yr−1 from seasonal DIC/alkalinity cycles).
• Production versus export: At Stations 1 and 5, annual production was 0.4 (0.2–2.1, 95% CI) and 0.9 (0.5–3.8) mol m−2 yr−1, respectively, far exceeding shallow trap exports at ALOHA (0.08 mol m−2 yr−1 at 150 m) and PAPA (0.16 mol m−2 yr−1 at 200 m). This implies ~80% of photic-zone CaCO3 production is remineralised before export; only ~20% is exported below the photic zone.
• Global extrapolation: A first-order global pelagic CaCO3 production estimate of ~3.1×10^14 mol yr−1 (3.7 Pg C yr−1) derived from the relationship between total production and satellite PIC, within the upper range of prior estimates (0.9–3.9×10^14 mol yr−1), and much higher than global shallow export estimates (~0.5–0.6×10^14 mol yr−1), reinforcing extensive shallow dissolution.
• Exporters: Despite contributing a small fraction to production (~1–2% at Station 5), foraminifera efficiently export CaCO3 to depth (e.g., 2–6 mg m−2 d−1 flux at 3800 m at OSP), consistent with their low organic content and rapid sinking.
• Discrepancy with prior modeling (Buitenhuis et al. 2019) likely arises from lower coccolithophore PIC/POC ratio assumptions, similar turnover times across disparate taxa, and the assumption that all dissolution above the calcite saturation horizon is aragonite.

This study directly quantifies living pelagic CaCO3 standing stocks and converts them to production rates by major calcifying taxa along a North Pacific transect. The results clarify that coccolithophores dominate both the standing stock and production of CaCO3, with pteropods and foraminifera playing secondary roles in production. By comparing production to shallow export fluxes, the work demonstrates a strong decoupling between surface production and export at 100–200 m, requiring extensive shallow dissolution (remineralisation) of CaCO3 within the photic and upper twilight zones. This finding reconciles the long-standing discrepancy between higher satellite/model-derived production estimates and lower shallow sediment trap exports. It also supports recent observations of excess alkalinity and in situ dissolution above the saturation horizon in the North Pacific, indicating that processes such as respiration-driven microenvironments, dissolution within grazer guts, and aggregate dynamics can drive significant dissolution of even calcite in undersaturated or near-saturated waters. The dominance of coccolithophore production but significant contribution of foraminifera to deep flux and sediments (owing to efficient export) explains the mismatch between production and sediment composition (often ~50/50 coccolith/foraminifera in sediments versus ~90/2 in production here). The study’s results emphasize that future projections of the CaCO3 cycle and its feedbacks on atmospheric CO2 must consider not only changes in calcification rates under acidification but also the poorly constrained processes governing photic-zone dissolution versus export, including community composition, PIC/POC ratios, grazing and aggregation dynamics, and environmental controls on particle microchemistry.

The study provides the first comprehensive, taxon-resolved quantification of living CaCO3 standing stocks and production across the North Pacific, showing that coccolithophores overwhelmingly dominate pelagic CaCO3 production (~86% of total), with aragonitic pteropods and foraminifera contributing less. Pelagic production notably exceeds shallow export at key sites (ALOHA, PAPA), implying that ~80% of CaCO3 produced in the photic zone dissolves before export, which reconciles satellite/model estimates with sediment trap data and underscores the importance of shallow dissolution in the marine carbonate budget. A first-order global production estimate of ~3.1×10^14 mol yr−1 lies within the higher end of published ranges and far exceeds shallow export, reinforcing the central role of remineralisation above the saturation horizons. Future research should prioritize resolving mechanisms and rates of photic-zone dissolution versus export (e.g., gut and respiration-driven dissolution, aggregate dynamics), quantifying how PIC/POC ratios and community composition respond to warming and acidification, and improving observational constraints (e.g., depth-resolved PIC, autonomous sensors) to reduce uncertainties in turnover times, seasonality, and global extrapolations.

Key limitations include: (1) temporal coverage restricted to August 2017, requiring seasonality corrections via proxies (satellite PIC for coccolithophores; chlorophyll-a for foraminifera; zooplankton seasonality for pteropods/heteropods) that cannot fully capture interannual variability for zooplankton; (2) assumptions of steady-state living standing stocks and flat probability distributions for turnover times, which may overestimate uncertainties and mask true variability; (3) taxon turnover time ranges derived from literature and may vary with environmental conditions and life stage; (4) satellite PIC only senses the upper optical depth, potentially biasing comparisons where production is deep; (5) global extrapolation from a regression driven largely by one high-PIC station, with the assumption that deep production bias scales similarly worldwide; (6) pteropod biomass datasets are highly skewed with zeros and outliers, influencing statistical descriptors; (7) short sediment trap deployments may miss time lags between production and export, though long-term trap data still indicate large production–export discrepancies.