From biogeochemical to ecological models of marine microplankton

Models must simplify the complexity of real marine pelagic ecosystems. How much simplicity is needed? A series of increasing numbers of state variables is used to illuminate this issue and to illustrate biogeochemical (element-conserving) and ecological (semi-freely dynamically interacting) models of the marine microplankton, defined as all organisms less than 200 μm. The models are those of Riley [Riley, 1946. Factors controlling phytoplankton populations on Georges Bank, J. Marine Res., 6, 54-73.] for phytoplankton, a nitrogen-conserving version of Riley, the microplankton model (MP) of Tett [Tett, P., 1990. A three layer vertical and microbiological processes model for shelf seas. Proudman Oceanographic Laboratory, Report 14, 85 pp.] with a constant ratio η of microheterotrophs to total microplankton, a related model (AH) with Lotka-Volterra dynamics for interacting autotrophs and heterotrophs, and a simple microbial loop model (ML). It is argued that models must be (at least) biogeochemical. The biogeochemical model MP simulated changes during a microcosm experiment better than the simplest ecological-biogeochemical model AH, which is, as here parameterised, an unstable system. ML, with protozoans able to switch food sources, is more stable than AH, and satisfactorily simulates the seasonal cycle of chlorophyll in the North-East Atlantic. MP performed well when forced with a time-series of η taken from ML. Time-varying η is one way of providing the smallest amount of microbial 'ecology' needed in biogeochemical pelagic models without introducing instability. (C) 2000 Elsevier Science B.V.