A diverse community constitutes global coccolithophore calcium carbonate stocks

By employing a global, species-resolved machine-learning approach, this study reveals that the widely studied *Gephyrocapsa huxleyi* contributes only about 7% of global coccolithophore calcium carbonate stocks, demonstrating that a diverse assemblage of other species, rather than this single species, underpins marine CaCO3 cycling.

The Gist

Imagine the ocean is a giant, bustling city. For decades, scientists have been trying to understand how this city handles its "construction materials"—specifically, calcium carbonate, which is the stuff that makes up the tiny, chalky shells of microscopic plankton called coccolithophores.

Here is the problem: Scientists have been looking at this city through a very narrow keyhole. They have been staring almost exclusively at one specific resident: a tiny plankton named Gephyrocapsa huxleyi. It's like trying to understand the entire economy of New York City by only studying the people who work in one specific coffee shop on 5th Avenue.

This new study says, "Hold on a minute. We've been ignoring the rest of the neighborhood!"

Here is what the researchers found, broken down into simple concepts:

1. The "Star" is Overrated

G. huxleyi is famous. It's abundant, it's everywhere, and it's the "celebrity" of the plankton world. Because it's so famous, most computer models and satellite images are calibrated to look for it.

• The Old View: Scientists thought this one species was doing about 60% of the heavy lifting for the ocean's calcium carbonate production.
• The New Reality: The study used a super-smart computer program (machine learning) to look at the whole city, not just the coffee shop. They found that G. huxleyi actually only contributes about 7% of the total calcium carbonate stock. It's a big player, sure, but it's definitely not the boss.

2. The "Heavy Lifters" are the Quiet Giants

If G. huxleyi is the small, fast worker, the real heavy lifters are the "gentle giants" of the plankton world. The study found that three other species, which are larger and carry much heavier shells, do the real work.

• The Analogy: Imagine G. huxleyi is a delivery person on a bicycle. It moves fast and there are a lot of them, but they carry light packages.
• The Real Heroes: The new study found three other species (Coccolithus pelagicus, Calcidiscus leptoporus, and Florisphaera profunda) that are like delivery trucks. They are fewer in number, but each one carries a massive load of calcium carbonate.
• The Result: Just these three "truck" species, plus a few others, are responsible for half of all the calcium carbonate in the ocean.

3. The "Deep Divers" and the "Sunbathers"

The study also looked at where these plankton live, and it's not just at the surface where the sun shines.

• The Sunbathers: Some species love the surface waters (the "subtropical" zones).
• The Deep Divers: One of the heavy hitters, Florisphaera profunda, actually prefers to hang out in the "Sub-Euphotic Zone." This is the dimly lit twilight zone of the ocean, far below the surface where sunlight barely reaches.
• The Surprise: About one-third of the total calcium carbonate stock is hiding in this deep, dark twilight zone, not in the bright surface waters where we usually look.

4. Why This Matters for Climate Change

Why should you care about tiny plankton shells? Because they are the ocean's thermostat.

• The Carbon Cycle: These shells help move carbon from the air into the deep ocean, acting like a giant sponge that cools the planet.
• The Risk: If we only study the "bicyclist" (G. huxleyi), we might think we understand how the ocean reacts to climate change. But the "trucks" (the heavy species) and the "deep divers" react differently to warming water and acidification.
• The Metaphor: If you are trying to predict how a car will handle a storm, you can't just test the tires on a sunny day. You need to test the engine, the brakes, and the suspension. Similarly, to predict how the ocean will handle climate change, we need to understand the diverse community of plankton, not just the famous one.

The Bottom Line

This study is like a massive census of the ocean's microscopic population. It tells us that the ocean's calcium carbonate budget isn't run by a single superstar. Instead, it's a diverse community of different shapes, sizes, and depths working together.

The takeaway: To accurately predict our future climate, we need to stop looking at just one species and start appreciating the whole, diverse team of plankton that keeps our planet's chemistry in balance.

Technical Summary

1. Problem Statement

Coccolithophores are microscopic calcifying plankton responsible for the majority of marine calcium carbonate ($CaCO_3$) production, playing a critical role in the global carbon cycle and climate regulation. However, scientific understanding and climate modeling have historically been biased toward a single, highly abundant species: Gephyrocapsa huxleyi (formerly Emiliania huxleyi).

• The Gap: Most global estimates of coccolithophore distribution, stocks, and production rely on satellite algorithms and climate models parameterized almost exclusively on G. huxleyi.
• The Risk: This narrow focus risks misrepresenting global $CaCO_3$ cycling because other species differ significantly in cell size, calcification rates, motility, and environmental responses. Consequently, current models may fail to accurately predict how the broader coccolithophore community will respond to climate change (e.g., warming, stratification, and ocean acidification).
• Objective: To generate the first global, species-resolved assessment of coccolithophore biomass and inorganic carbon (CIC) stocks to determine the true contribution of G. huxleyi versus the diverse community of other species.

2. Methodology

The authors employed a data-driven, machine learning (ML) approach to generate global spatial-temporal species distribution models (SDMs).

• Data Sources:

  • Biological Data: The CASCADE dataset (de Vries et al., 2024), containing 6,166 water samples with species-specific abundance measurements from 1964–2019. The study focused on 58 species occurring in >5% of locations.
  • Environmental Data: Interpolated monthly global grids (1° resolution, 5m depth intervals up to 200m) for temperature, nutrients (nitrate, phosphate, silicate), oxygen, PAR (photosynthetically active radiation), DIC, and total alkalinity.
  • Conversion Factors: Species-specific cellular organic carbon (COC) and inorganic carbon (CIC) estimates derived from the CASCADE database.

• Machine Learning Pipeline (ABIL.PY):

  • Algorithm: A Python ensemble-based pipeline utilizing Random Forests (RF), Bagged K-Nearest Neighbors (b-KNN), and Extreme Gradient Boosting (XGBoost).
  • Two-Phase Modeling:
    1. Classifier: Predicts species presence/absence based on environmental conditions.
    2. Regressor: Predicts biomass concentration for voxels where presence is predicted.
  • Handling Absence Data: Since the training data lacked explicit zero observations (absence), the authors generated pseudo-absences based on "Area of Applicability" (AOA) estimates. A 1:1 ratio of presence to pseudo-absence was selected to optimize model performance.
  • Validation: Ten-fold cross-validation was used for hyperparameter tuning. Uncertainty was quantified using ensemble prediction distributions to generate 95% confidence intervals.

• Post-Processing:

  • Aggregation of stocks into global totals, ocean biomes (based on Fay and McKinley, 2014), and photic zones (Upper, Lower, and Sub-Euphotic).
  • Calculation of Shannon diversity indices and cumulative contribution curves (to determine how many species account for 80% of stocks).

3. Key Results

A. Global Stock Estimates

• Total Biomass: The global annual mean coccolithophore biomass is estimated at 15.3 Tg COC (top 200m) and 10.3 Tg CIC (top 200m).
• Comparison: This CIC stock is 50 times greater than planktonic foraminifera but ~30% lower than pteropod stocks. However, due to faster turnover rates (1 day$^{-1}$), coccolithophores contribute disproportionately to production and export fluxes compared to their standing stock.
• Satellite Discrepancy: The model estimates are lower than satellite-derived Particulate Inorganic Carbon (PIC) estimates. The authors attribute this to the exclusion of detached coccoliths in their model, suggesting detrital inorganic carbon is a major component of total ocean PIC.

B. Species Contributions: The G. huxleyi Myth

• Abundance vs. Stock: While G. huxleyi dominates total abundance (contributing 28.5% of total cells), it contributes only 7.2% of the total global CIC stock.
• Reason: G. huxleyi has a lower cellular calcite content and a lower CIC:COC ratio compared to other abundant species.
• Dominant Contributors: A diverse assemblage of large, heavily calcified species drives the majority of the stock:
  • Coccolithus pelagicus: ~24.7% of global CIC stock.
  • Florisphaera profunda: ~13.0% of global CIC stock.
  • Calcidiscus leptoporus: ~11.7% of global CIC stock.
  • Collective Impact: These three species alone account for nearly 50% of the global coccolithophore $CaCO_3$ stock.
• Diversity: To account for 80% of global CIC stocks, 13 species are required. Rare species (contributing <5% individually) collectively make up 43.3% of the global stock, particularly in subtropical and equatorial regions.

C. Spatial and Vertical Distribution

• Biomes: While subpolar regions have the highest concentrations (20.9 µmol C m$^{-3}$), subtropical biomes (Seasonally Stratified + Permanently Stratified) contribute the largest share of integrated stock (56.3%) due to their vast surface area.
• Vertical Distribution: Approximately 32% of the global CIC stock is found in the Sub-Euphotic Zone (<1% surface irradiance), driven largely by F. profunda, a deep-dwelling, potentially mixotrophic species.
• Decoupling: There is a spatial decoupling between organic (COC) and inorganic (CIC) carbon stocks. The CIC:COC ratio is not uniform; it is elevated (1.5–2.2) in deep subtropical and subpolar regions due to the dominance of heavily calcified species, whereas it is lower elsewhere.

4. Significance and Implications

• Paradigm Shift: The study challenges the prevailing view that G. huxleyi underpins global $CaCO_3$ cycling. Instead, it demonstrates that a morphologically and functionally diverse community is the primary driver.
• Climate Modeling: Current Earth System Models (ESMs) and CMIP models often treat coccolithophores as a single functional group parameterized by G. huxleyi. This study highlights that such models likely misrepresent:
  • The spatial heterogeneity of the CIC:COC ratio.
  • The response of the community to climate change (e.g., G. huxleyi is small and non-motile, while dominant stock contributors like F. profunda are large, motile, and potentially mixotrophic).
  • The vulnerability of heavily calcified species to ocean acidification.
• Future Directions: The authors call for a shift in research focus toward characterizing non-G. huxleyi species in laboratory and in situ studies. Future modeling efforts must incorporate multiple plankton functional types (PFTs) with distinct traits (size, motility, CIC:COC ratios) to accurately simulate the ocean carbon cycle under climate change.

Conclusion

This paper provides the first species-resolved global quantification of coccolithophore stocks, revealing that the global calcium carbonate budget is sustained by a diverse community rather than a single dominant species. The findings necessitate a revision of how coccolithophores are represented in global biogeochemical models to ensure accurate predictions of the ocean's role in climate regulation.