Does planktonic community structure determine downward particulate organic carbon flux in different oceanic provinces?

A previous study based on data from a NE Atlantic site provided evidence, using an existing foodweb/vertical-flux modelling approach, that the size-structure of the phytoplankton community - rather than primary production - can be the dominant control on downward particulate organic carbon (POC) flux. In order to assess whether taking community structure into account can also provide more reliable estimates of downward POC flux in other oceanic provinces, epipelagic observations (mainly size-fractionated primary production, biomass, community structure data and heterotrophic bacterial production) and POC flux data from deep-moored sediment traps were collated from the six different ocean regions for which suitable data are available. At each sire the epipelagic data were used in conjunction with two standard versions of the foodweb/vertical-flux model (one permits direct sinking of large ungrazed algae out of surface waters, the other does not) and published empirical depth/POC-flux algorithms to predict the POC flux to the deep ocean. Predictions were also made using published primary-production/POC-flux algorithms, and the two sets of predictions were compared to the deep-ocean POC flux measurements. While the version of the foodweb/vertical-flux model permitting the direct sinking of ungrazed algae provided the most reliable predictor of POC flux for five of the six sites, no conventional algorithm provided comparable predictions for more than two sites. The reliability of these predictions is discussed in the context of recent modelling studies that explore the extent to which Row fields in the water column overlying deep-moored sediment traps confound attempts to relate particle flux measurements to observations of surface-water processes. The present study suggests that the sinking of ungrazed large cells, probably diatoms, may be the key determinant of the magnitude of the downward POC flux in a variety of ocean regions. Planned ocean-observing programmes may provide sufficient epipelagic data to allow this approach to be used to improve the accuracy of basin-scale estimates of downward POC flux and hence reduce the uncertainty of the magnitude of this flux within the oceanic carbon budget. (C) 1999 Elsevier Science Ltd. All rights reserved.