Ironing out carbon export to the deep ocean.

In PNAS, Lopes et al. (1) present a novel approach to quantifying the efficiency of the biological carbon pump as Earth’s climate warmed from the ice bound glacial episode (26,000–18,000 y before present) to the equable climate of the present interglacial (10,000 y BP until present). The biological carbon pump begins with atmospheric CO2 transformation into chemical energy by marine photosynthesizers. The particulate organic carbon (POC) from these organisms is then transported into the deep ocean to be buried in sediments. An efficient biological pump increases carbon burial at the ocean floor and this can impact the Earth’s climate system. Past intervals of high marine carbon burial have been associated with atmospheric greenhouse gas removal leading to planetary cooling. Marine photosynthesizing organisms (phytoplankton) convert CO2 to POC in sunlit surface waters when sufficient nutrients are available to fuel primary productivity (PP). As nutrients are stripped out of surface waters by the phytoplankton, high PP can continue when either subsurface nutrientrich waters are returned to the surface through upwelling processes (Fig. 1A) or when terrestrial-derived nutrients are added via river discharge, dust, or glacial processes. Lopes et al. (1) estimated past PP in Northeast Pacific surface waters using the relative abundance of diatoms—a phytoplankton with a box-like silica skeleton or frustule— that are associated with nutrient-rich environments. However, most POC is recycled in the upper 200 m of the ∼4-km-deep ocean, as only efficiently packaged POC can sink fast enough to avoid decomposition (2). A number of mechanisms for the efficient transfer of POC to the seafloor have been suggested including the “ballast hypothesis,” where the density of sinking particles is increased by the addition of minerals (2, 3). Lopes et al. (1) produce a diatom size index to estimate the efficiency of POC transfer to the seafloor based on the assumption that large diatoms sink more rapidly, similar to the ballast hypothesis. However, the deglacial interval (18,000–12,000 y BP) containing the greatest proportion of large diatoms is decoupled from the Last Glacial interval (30,000–17,000 y BP) of high PP that Lopes et al. (1) associate with increased nutrients from high river discharge. The efficiency of the biological pump can be decoupled from the strength of nutrientdriven PP through changes in phytoplankton community composition. Ocean regions such as the Southern Ocean and North Pacific contain waters rich in nitrate yet support only low concentrations of phytoplankton because of micronutrient limitation (e.g., iron) (4). In these regions, diatoms consistently outcompete other phytoplankton in Fe-enrichment experiments, suggesting their productivity is limited by bioavailable Fe (5, 6). As a result, supplementing Fe to ocean waters can shift phytoplankton communities away from haptophyteto diatom-dominated assemblages (5). In addition, Fe-rich environments promote the production of heavily silicified and chain-forming diatoms relative to small diatoms (5, 6). Lopes et al. (1) posit that Northeast Pacific surface waters were Fe rich during deglaciation, increasing the relative abundance of large diatoms such as Chaetoceros and enhancing POC transfer efficiency to sediments even as PP declined. Similarly, in the Southern Ocean, high POC in Fe-rich regions is associated with blooms of Chaetoceros. These are major POC exporters as they convert their vegetative cells into grazerresistant resting spores when nutrients are exhausted, clumping into aggregates with other detritus and sinking en masse (7). How then can Fe concentrations vary through time to force changes in the composition of phytoplankton communities? The return of subsurface nutrient and Fe-rich waters is possible through upwelling processes (Fig. 1A), however, external to the ocean’s biogeochemical cycle; Fe is primarily added to surface marine waters from terrestrial sources via river discharge, aeolian dust, or glacial processes (Fig. 1B). In coastal upwelling regions similar to the Lopes et al. (1) Fig. 1. The availability of iron may decouple primary productivity from the biological carbon pump changing the ocean’s ability to remove CO2 from the atmosphere. (A) In highly productive nitrate-rich environments limited by Fe, small, lightly silicified diatoms dominate the phytoplankton community and are easily recycled in the upper water limiting the efficiency of the biological pump. (B) In an Fe-rich marine environment, however, large chain forming diatoms clump into rapidly sinking aggregates that efficiently transfer carbon to the sea floor (efficient biological carbon pump). The excess Fe may come from continental shelves, rivers, glacial sediment, or atmospheric dust. The relative flux of Fe from these different sources may vary as climate changes. Author contributions: I.L.H. wrote the paper.