Oceanic nutrient cycles are coupled, yet carbon-nitrogen-phosphorus (C:N:P) stoichiometry in marine ecosystems is variable through space and time, with no clear consensus on the controls on variability. Here, we analyze hydrographic, plankton genomic diversity, and particulate organic matter data from 1970 stations sampled during a global ocean observation program (Bio-GO-SHIP) to investigate the biogeography of surface ocean particulate organic matter stoichiometry. We find latitudinal variability in C:N:P stoichiometry, with surface temperature and macronutrient availability as strong predictors of stoichiometry at high latitudes. Genomic observations indicated community nutrient stress and suggested that nutrient supply rate and nitrogen-versus-phosphorus stress are predictive of hemispheric and regional variations in stoichiometry. Our data-derived statistical model suggests that C:P and N:P ratios will increase at high latitudes in the future, however, changes at low latitudes are uncertain. Our findings suggest systematic regulation of elemental stoichiometry among ocean ecosystems, but that future changes remain highly uncertain.

Concentrations and elemental stoichiometry of suspended particulate organic carbon, nitrogen, phosphorus, and oxygen demand for respiration (C:N:P:-O₂) play a vital role in characterizing and quantifying marine elemental cycles. Here, we present Version 2 of the Global Ocean Particulate Organic Phosphorus, Carbon, Oxygen for Respiration, and Nitrogen (GO-POPCORN) dataset. Version 1 is a previously published dataset of particulate organic matter from 70 different studies between 1971 and 2010, while Version 2 is comprised of data collected from recent cruises between 2011 and 2020. The combined GO-POPCORN dataset contains 2673 paired surface POC/N/P measurements from 70°S to 73°N across all major ocean basins at high spatial resolution. Version 2 also includes 965 measurements of oxygen demand for organic carbon respiration. This new dataset can help validate and calibrate the next generation of global ocean biogeochemical models with flexible elemental stoichiometry. We expect that incorporating variable C:N:P:-O₂ into models will help improve our estimates of key ocean biogeochemical fluxes such as carbon export, nitrogen fixation, and organic matter remineralization.