RECONCILING AGGREGATION-THEORY WITH OBSERVED VERTICAL FLUXES FOLLOWING PHYTOPLANKTON BLOOMS

Sediment trap data show that rapidly sinking pulses of phytodetritus form after phytoplankton blooms, even when bloom intensity is low. A numerical model of physical aggregation and sedimentation in the surface ocean was used to gauge whether predicted aggregation rates were high enough to generate postbloom sediment pulses. Initial models behaved inaccurately without a full range of particle sizes, abundant nonphytoplankton particles, and explicit hydrodynamic retardation of particle contact. Provision for background particles while tracking phytoplankton required implementation of a novel bookkeeping scheme. To address the degree of retardation for contact between particles, an expression for contact efficiency for collision by turbulent shear was developed. The most realistic way to produce model results that mimicked field data was to include background particles, to invoke particle stickiness in the range 0.1-1.0, and to make modest upward adjustments to contact efficiencies calculated for impermeable spheres. For all but the highest background particle concentrations, the magnitude of a postbloom sediment pulse scaled nonlinearly with vertically integrated cell number in the surface layer of our two layer model. The existence of a nonlinear relationship between pulse size and bloom intensity makes initial cell number per unit of area in the surface layer, and not productivity, the proximate determinant of carbon export flux. This result emphasizes the need for caution when applying established scalings between export flux and productivity. Further, it provides a mechanistic explanation both for tight pelagic-benthic coupling under waters prone to intense blooms and for interannual variability in export flux in polar regions.