The ocean covers two thirds of the planet's surface and represents 96% of the volume of its biosphere. It plays a central role in climate, notably by sequestering large quantities of atmospheric carbon. In particular, marine plankton contribute strongly to the flux of organic carbon from surface waters to the ocean floor (acting as a biological "carbon pump").

Since 2010, and in connection with Tara Oceans, the Genoscope has undertaken an inventory of plankton diversity. Very early on, planktonic community composition emerged as being strongly linked to temperature.

DNA was sampled and sequenced from hundreds of samples taken around the world, yielding a quantity of information equivalent to 8,000 times the content of the human genome. This analysis highlighted the genomic partitioning of the ocean into regions or provinces, according to the terminology of oceanographers. These provinces correspond to the grouping of sampling sites that share the same genomic characteristics, for one of the six collected plankton fractions. Different sets of sieves allowed the selection of very distinct planktonic communities, from viruses at the smallest to zooplankton of 2 millimeters in size at the largest.

The multidisciplinary study conducted by the Genoscope and the LSCE demonstrates that these provinces correspond to ecological niches, by establishing a statistical link between the genomic composition of plankton and the environmental parameters.

To accomplish this, the researchers implemented supervised machine learning methods for each of the identified genomic provinces. By utilizing the climate models, they were able to spatially extrapolate "climato-genomic" provinces. These provinces are quite large, comparable to the large eddy structures (gyres) observed in the major oceans, and some include areas belonging to several oceans.

The scientists discovered that the geographical distribution of planktonic organisms is very different depending on their size.

• The distribution of large planktonic organisms (mainly zooplankton, 20-2000 µm in size) is very latitudinal (following the north-south temperature gradient) and can extend beyond the limits of an ocean basin.
• The distribution of small organisms (phytoplankton, bacteria and viruses, 0-200 µm in size) is more fragmented, while still being very latitudinal.

Finally, the climatologists introduced these biogeographical data into their model, which they ran under a business-as-usual scenario of significant warming (RCP 8.5) up to the year 2100.

Some provinces are expanding, while others are shrinking. Most are migrating toward the poles.

• The composition of plankton is changing: organisms with the ability to use atmospheric nitrogen (diazotrophs) are developing in equatorial zones, especially in the Pacific.
• The flux of organic carbon to the ocean floor is projected to decrease by 4%.

A decrease in the biological carbon pump can only reinforce global warming.
 
This work, launched thanks to the DRF-Impulsion transversal program, shows that it is possible to establish links between genomic data and major biogeochemical cycles. The task now is to build mechanistic and explanatory models.