Ecosystem approach to fisheries and marine ecosystem modelling: review of current approaches

Martin Stecken, Pierre Failler

    Research output: Contribution to journalArticlepeer-review

    218 Downloads (Pure)


    This paper presents a review of existing approaches on ecosystem modelling and assesses them in the light of their capacity to integrate main elements of the ecosystem and their pertinence for fishery management. It completes existing EAF reviews by adding substantial analysis of the way to incorporate ecosystem complexities in models. The two most used approaches use nearly the same structure of boxes with flows of biomass for EwE or flows of elements for BM2. The main difference which gives a more efficient modelling is the method of estimates coming from the steady state model ECOPATH. This method is one of the strength of the EwE approach giving it a good predictive power. However, it is in the same time one of its main weakness in assuming an equilibrium situation that is far to be always the case. BM2 and biogeochemical models in general will have a better predictive power for low trophic levels. The problem in raising trophic levels is the higher number of parameters which raises in the same time the uncertainty of such models. The last model, ECGEM, is a recent model with an original approach. It presents, however, several important weaknesses as its ecological pertinence remains to be proved. Despite this, the model requires less parameters and supply methods needed by standard models and thus allows to process more complex situations.
    Original languageEnglish
    Article number199
    JournalJournal of Fisheries & Livestock Production
    Issue number4
    Early online date30 Aug 2016
    Publication statusPublished - Dec 2016


    • ecosystem
    • biodiversity
    • biomass
    • biogeochemical
    • planktivorous


    Dive into the research topics of 'Ecosystem approach to fisheries and marine ecosystem modelling: review of current approaches'. Together they form a unique fingerprint.

    Cite this