Back

Cellular-level control on global ocean deoxygenation driven by phytoplankton ecophysiology

Sharoni, S.; Inomura, K.; Dutkiewicz, S.; Jahn, O.; Britten, G. L.; Follows, M. J.

2025-10-08 ecology
10.1101/2025.10.08.681239 bioRxiv
Show abstract

Phytoplankton elemental composition shapes the distribution of dissolved nutrient concentrations and thus plays a key role in ocean biogeochemistry. While the carbon-to-nitrogen-to-phosphorus ratio (C:N:P) in marine phytoplankton has been extensively studied, the hydrogen (H) and oxygen (O) content have received less attention despite their critical role in determining dissolved oxygen (O2) consumption rates in the ocean. Here, we estimated the elemental composition of marine phytoplankton, including the H and O content, from first principles, using a cellular allocation model embedded in a global ocean model. We estimated that an average phytoplankton cell has a chemical formula of C107H190N16O53P, with an O2 demand of 149 mol O2/mol P and respiration quotients of 1.40 mol O2/mol C, suggesting a lower H and O content, and higher O2 demand than commonly assumed. We found global variations in the O2 demand of organic matter respiration driven by population structure and cellular reorganisation under different environmental conditions. By testing how shifts in the macromolecular composition of phytoplankton cells affect the oceans O2 budget, we found that O2 consumption increases significantly when shifting cell composition from carbohydrate-rich to protein- or lipid-rich cells. As a result, low-O2 (hypoxic) zones in the ocean expanded by 75%. These findings demonstrate that cellular-level processes in marine phytoplankton shape the global O2 cycle and large-scale patterns of ocean biogeochemistry.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.