Ocean Acidification: Impact on Phytoplankton Chemistry and Climate

Oceans have absorbed over 30% of the carbon dioxide emitted by humans since the start of the industrial era, causing their pH to drop by about 0.1 units. This slight but systematic acidification disrupts the biochemical processes of the smallest but most crucial organisms on the planet: phytoplankton. An international team, published in Nature Geoscience, has just demonstrated that dissolved CO₂ modifies the internal chemical composition of these microalgae, with potential implications for oxygen production and marine ecosystem dynamics.

What the study reveals: Precise results and key figures

The researchers analyzed cultures of several phytoplankton species exposed to pH levels simulating conditions expected by the end of the century. They found a reorganization of metabolic pathways, including a reduction in the synthesis of complex carbon compounds and an increase in the production of volatile organic acids. This internal chemical modification translates into reduced photosynthesis efficiency, which could reduce these organisms' ability to absorb atmospheric CO₂.

Moreover, the study shows that the proportion of essential nutrients such as iron and nitrate is redistributed within cells, increasing their sensitivity to deficiencies. Although the authors did not provide an exact percentage, they emphasize that the changes observed are "statistically significant" and consistent with trends detected in Copernicus satellite data, which indicate a decrease in chlorophyll in some oceanic zones.

How: Research method and data used

To reach these conclusions, the team combined laboratory experiments with in situ observations. Cultures were tracked using Raman spectroscopy and gas chromatography, allowing for quantification of internal metabolites. Meanwhile, scientists exploited satellite data from ESA and NASA, integrated into the Copernicus system, to correlate the chemical variations observed in the lab with global trends in phytoplankton productivity.

The climate models used included the predictive ECMWF model as well as neural networks trained via machine learning to refine acidification projections. These models, similar to GraphCast or Pangu-Weather in the field of weather forecasting, allowed for estimation of prediction uncertainty related to CO₂ emission scenarios and evaluation of potential impact on ocean biogeochemical cycles.

What it changes: Implications