Carbon turnover in sinking particles in the marine environment

Abstract

Fecal pellets and marine snow aggregates drive the biological carbon pump via sedimentation of organic matter from the surface ocean to the deep ocean where carbon can be sequestered for hundreds to thousands of years. However, the controlling mechanisms for carbon export from the surface ocean are still unclear. Studies of fecal pellet fluxes and retention have revealed that most pellets are degraded within the surface waters. However, the key degraders are poorly known. We observed copepods (Danish coastal waters) rejecting pellets immediately after capture, in some cases with pellet fragmentation as outcome. Therefore, coprorhexy (pellet fragmentation) seems the main impact by copepods, and they are not the key degraders. From size fractionation of a plankton community (Öresund, Denmark), we identified large heterotrophic dinoflagellates as major pellet degraders in the upper ocean. In deeper waters bacterial degradation dominates, only limited by pellet residence time, which is hypothesized to be dependent on ballast minerals. Bacterial degradation (laboratory) was independent of ballast material whereas sinking velocity increased with denser diet, leading to <10-fold higher pellet carbon flux with coccolithophorid and diatom diets than with flagellate diet. We calculated vertical carbon fluxes from in situ profiles of marine snow and aggregate size-specific abundance measured during a ship cruise off Cape Blanc (Mauritania), using estimated sinking speeds and aggregate masses. Calculated fluxes agreed well with sediment trap data for the same area and period. We estimated carbon consumption from the fluxes and used consumption changes to identify degradation processes at different depths. Calculated copepod abundances needed to remove the majority of carbon in the upper 75 meters were consistent with previous observations. In deeper waters carbon consumption fit measured bacterial degradation rates. Highest degradation rates were in the upper 200 meters and limited by aggregate residence time (sinking speed), which we argue depends on ballast material. We observed high sinking speeds for the aggregates formed on material from Cape Blanc due to presence of liths from coccolithophorids and high input of dust. In laboratory experiments bacterial carbon specific degradation was independent of ballast material whereas sinking speed increased with denser aggregate constituents. Thus, aggregated carbon flux should be greater for pure coccolithophorid than pure diatom aggregates.