The 2008 meeting, three years on, gave the community an opportunity to gauge progress and define new key challenges. The conference was attended by around 160 scientists with significant representation from the Americas, Australasia, and Asia, as well as Europe and the UK. The entire meeting was held in plenary, a deliberate strategy to facilitate cross disciplinary communication and fertilisation. The conference comprised nine sessions over the four days covering climate related change, evaluation, operational models, complexity, processes, biodiversity, higher trophic levels, end to end ecosystems and environmental management.

The three year gap between meetings provided sufficient time for scientists to respond to challenges, as well as for the community as a whole to gauge progress. In particular, the marked increase in papers that seriously address the issue of model evaluation is encouraging (Lynch et al., 2009 and papers within; Los and Blaas, 2010-this issue), as is the development in methods of treating complexity and model interpretation. What, then, are the key challenges facing those developing marine ecosystem models today?

The issue of model complexity remains perhaps the key challenge facing the modelling community, which was brought into focus by Anderson (2005). The debate continues, focussing on the continuum between Nutrient–Phytoplankton–Zooplankton (NPZ) and functional group models, the veracity of their process descriptions and the accuracy of parameterisations. Recent studies have suggested that the inclusion of additional articulation, such as variable stoichiometry or extra trophic links, in models may lead to greater generality and portability, although only if the underlying mechanisms are accurately represented (Friedrichs et al, 2007 andWard et al, 2010-this issue). The goal of producing an ecosystemmodel for climate studies that includes key system feedbacks, has a single parameter set, and which is both accurate and globally robust (Fasham, 1993), has not yet been realised. In part, difficulties in achieving this aim arise from the imperfect physical descriptions of regional and global systems that underpin the descriptions of biogeochemistry, and the sensitivity of biological parameterisations to small variations between physical models (Sinha et al., in press; Allen et al., 2010-this issue). Physical processes impact upon many biological processes, including turbidity (see Le Fouest et al., 2010-this issue), phytoplankton competition (Perruche et al., 2010-this issue) and planktonic and larval distribution (Savina et al., 2010-this issue), underscoring the importance of the appropriate treatment of physical properties.... Full journal article

Link to Special Issue: Volume 81, Issues 1-2, Pages 1-206 (April 2010) .Contributions from Advances in Marine Ecosystem Modelling Research II 23-26 June 2008, Plymouth, UK.