Emergence of negative trophic level-size relationships from a size-based, individual-based multispecies fish model

Modeling the mechanisms underlying trophic interactions between individuals allows the food web structure to emerge from local interactions, which constitutes a prerequisite for assessing how marine ecosystems respond to various anthropogenic pressures. Using a multispecies spatially explicit individual-based model, the emergence of trophic patterns was explored in the eastern English Channel ecosystem, where pelagic-benthic trophic coupling was recently studied empirically. The OSMOSE model was applied to this ecosystem by explicitly representing the life cycle of 13 fish species and one squid group, forced by pelagic and benthic prey fields that are variable over time and space. A matrix defining possible accessibilities between life stages was added to the model to link benthic and pelagic communities through overlap of vertical distribution. After optimizing some parameters of the model to represent the average state of the fish community during the 2000-2009 period, the simulated trophic structure was explored and compared to empirical data. The simulated and stable-isotope-derived trophic levels of fish were in relatively good agreement. Intraspecific variability of the trophic level is high in the five stable-isotope datasets but is well encompassed by the model. Despite the hypothesis of opportunistic size-based predation, the simulation showed a decreasing trend of trophic level with size for four benthic species, a pattern observed empirically for a different set of species in the ecosystem. Model exploration showed that this emerging pattern varies spatially and is both explained by the spatial variability of prey availability and by the independence of trophic and size structures of benthic invertebrates. The combination of individual-based models, stomach contents and intrinsic tracers, such as stable isotopes, appears to be a promising tool to better understand the causes of observed trophic patterns.