Dataset: f-ratio MATLAB routine for an adaptive food web model to explain carbon cycling in the ocean, 2004 (U.S. JGOFS Synthesis & Modeling Phase project results)

The proposed research will be the development of a food web model that can be incorporated into a global circulation model (GCM) to predict export production and to characterize the nature of the carbon exported to the interior of the ocean. The model will be similar to the food web model developed conceptually at the 1999 Synthesis and Modeling (SMP) food web workshop in Keystone, Colorado. In the model phytoplankton are envisioned as consisting of five functional groups, small phytoplankton such as Synechococcus and Prochlorococcus, diatoms, coccolithophores, Phaeocystis, and nitrogen fixers. Division of the phytoplankton in this manner is hypothesized to be necessary to explain the dependence of export ratios on temperature and primary production, to account for the allocation of exported carbon between calcium carbonate, particulate organic carbon, and dissolved organic carbon, and to take into account various methods of ballasting (fecal pellets, calcium carbonate, and silica) that influence the sinking and remineralisation rates of particulate carbon. A distinguishing characteristic of the model is the assumption that open ocean biological communities adapt to environmental conditions in a way that tends to maximize the stability of the steady state condition toward which the communities evolve. This same hypothesis has previously been tested with a simpler food web model in which the phytoplankton are envisioned as consisting of only two functional groups, small and large phytoplankton. The success of that previous model, which was developed with funding from the first phase of the SMP, has provided the motivation for extending this same approach to the more complex model with five functional phytoplankton groups. It is hypothesized that a stable coupled physical-biological model of the ocean will require that the biological component be adaptive. With respect to export production, specific questions to be addressed with the model will include the following: 

1. How much of the organic carbon is exported as dissolved organic carbon (DOC) and how much as particulate carbon (PC) 
2. Of the PC export, how much is exported as particulate organic carbon (POC) and how much as carbonate carbon (CC) 
3. To what extent is the exported POC ballasted by silica and/or carbonate or sequestered by incorporation into encapsulated fecal pellets 

Once the adaptive parameters in the model have been determined through steady state analysis, a time-dependent version of the model will be run to explore model behavior when the forcing functions are time dependent. Questions to be addressed with the time-dependent version will include the role of DOC export and the ability of the model?s output to facilitate description of nutrient and oxygen profiles in the interior of the ocean. An additional issue to be explored with the time-dependent version will be the need to allow for non-Redfield stoichiometry in processes such as nutrient uptake and remineralisation. The food web model developed at the Keystone meeting is sufficiently complex as to make its incorporation into more elaborate GCMs problematic. Consequently, the final phase of the proposed research will involve examination of the impact on model performance of simplifications of the food web structure proposed at the Keystone meeting. In other words, the final phase of the project will attempt to identify the simplest possible food web model that can simulate the aspects of system behavior necessary to explain the pattern of export production in the ocean and the nature of exported carbon.