Inside marine snow's tiny megacities, bacteria are dissolving the ocean's carbon storage system

Proceedings of the National Academy of Sciences, March 2026

Oceanographers call them the megacities of the sea. Marine snow particles, tiny clumps of dead organisms and debris drifting down through the water column, are teeming with microbial life. And those microbes, it turns out, may be quietly undermining one of the planet's most important climate defenses.

A study led by Benedict Borer, assistant professor of marine and coastal sciences at Rutgers University, has demonstrated how bacteria living inside marine snow dissolve calcium carbonate, the mineral that gives these particles the weight they need to sink into the deep ocean. The findings, published in the Proceedings of the National Academy of Sciences, could reshape how climate scientists model the ocean's capacity to store carbon.

The dissolution mystery

The ocean's biological carbon pump works like this: phytoplankton at the surface absorb carbon dioxide from the atmosphere. When they die, their remains sink as marine snow, carrying carbon downward. Some phytoplankton, particularly coccolithophores, build shells of calcium carbonate, which acts as dense ballast that accelerates sinking. The deeper the carbon goes, the longer it stays sequestered.

In the cold, acidic deep ocean, calcium carbonate eventually dissolves. That is expected. What has puzzled oceanographers for years is evidence that calcium carbonate also dissolves in the upper ocean, where temperature and pH conditions should keep it intact. The chemistry simply does not favor it at those depths.

Recent work had suggested that acidic microenvironments, tiny pockets where conditions differ sharply from the surrounding water, might explain the discrepancy. Zooplankton guts, for example, create acidic conditions that can dissolve calcium carbonate. Borer's study demonstrates that the interiors of marine snow particles are another such environment.

A chip that mimics sinking

To test the hypothesis, Borer built a three-layer microfluidic chip at MIT and Woods Hole Oceanographic Institution. The middle layer held synthetic marine particles containing calcite, the crystalline form of calcium carbonate, along with marine bacteria. Artificial seawater flowed through narrow channels between the layers, simulating the conditions a particle would experience as it sinks.

By manipulating gas pressure, temperature, oxygen levels, and bacterial abundance, the team could precisely control the environment around the particles and measure how bacterial activity affected calcite. The results were clear: as bacteria respired, they increased local acidity, which accelerated calcite dissolution. Less calcite meant less ballast, which meant slower sinking.

The process creates a feedback loop. Slower sinking means more time in shallow, warmer water where bacterial activity is higher. More bacterial activity means more calcite dissolution. More dissolution means even slower sinking. The particle's fate spirals toward remaining near the surface, where its carbon is more likely to be released back into the atmosphere rather than locked away in the deep ocean.

Consequences for carbon accounting

The biological carbon pump moves billions of tons of carbon from the surface to the deep ocean each year. If microbial dissolution of ballast minerals is more significant than current models assume, the pump may be less efficient than we think. Carbon that models predict will reach the deep ocean may instead be recycled in shallower waters.

This has direct relevance for proposals to combat climate change by enhancing the ocean's natural carbon sequestration. Several geoengineering approaches aim to increase the biological pump's capacity. If bacteria are simultaneously working to slow that pump by dissolving ballast, the effectiveness of such interventions could be overestimated.

Borer acknowledged the uncertainty with candor: the biological carbon pump could become either more or less efficient in future climate scenarios. Whether warming ocean temperatures will accelerate bacterial dissolution of ballast faster than other factors might compensate remains unknown.

From lab to open ocean

The study's primary limitation is its laboratory setting. The microfluidic chip simulates conditions within a sinking particle, but the real ocean adds complexity: varying temperatures with depth, diverse bacterial communities, particles of different sizes and compositions, and the constant influence of currents and turbulence.

Field validation will be essential before the findings can be incorporated into global carbon cycle models. Borer described the results as a critical first step rather than a definitive answer, though the mechanism is clear enough to warrant attention from climate modelers now.