Parameterizations of truncated food web models from the perspective of an end-to-end model approach

Abstract

Modeling of marine ecosystems is broadly divided into two branches: biogeochemical processes and fish production. The biogeochemical models see the fish only implicitly by mortality rates, while fish production models see the lower food web basically through prescribed food, e.g., copepod biomass. The skill assessment of ecological models, which are usually truncated biogeochemical models, also involves the question of how the effects of the missing higher food web are parameterized. This paper contributes to the goal of bridging biogeochemical models and fish-production models by employing a recently developed coupled NPZDF-model, Fennel [Fennel, W., 2007. Towards bridging biogeochemical and fish production models. Journal of Marine Systems, doi:10.1016/j.jmarsys.2007.06.008]. Here we study parameterizations of truncated NPZD-models from the viewpoint of a complete model. The effects of the higher food web on the cycling of the state variables in a truncated NPZD-model cannot be unambiguously imitated. For example, one can mimic effects of fishery by export fluxes of one of the state variables. It is shown that the mass fluxes between the lower and upper part of the full model food web are significantly smaller than the fluxes within the NPZD-model. However, over longer time scales, relatively small changes can accumulate and eventually become important.