Predicting indirect effects of fishing in Mediterranean rocky littoral communities using a dynamic simulation model

Modelling may significantly enhance our understanding of the potential impacts of fisheries at larger spatial scales and on groups that would otherwise be very difficult to study. An aggregated biomass-based simulation model of a Mediterranean infralittoral zone was developed and used to carry out fishing ‘experiments’ where fishing intensity and catch selection were varied. The model was constructed for the Bay of Calvi, Corsica, using the Ecopath with Ecosim software, and was composed of 27 compartments, including seabirds, 11 groups of fish, 12 groups of invertebrates, 2 primary producers, bacteria and detritus. l instances of indirect fishing effects (‘trophic cascades’ and ‘keystone predation’) have been proposed from anecdotal evidence in the western Mediterranean. Model outcomes provided little support for the widely accepted paradigm that fishing, by removing invertebrate-feeding fish, allows increases in the biomass of sea urchins and as a consequence the formation of overgrazed ‘barrens’ of bare substrate. Simulated harvesting of sea urchins by humans did, however, results in an increase of macroalgal biomass as reported previously. Intensified fishing pressure on ‘macrocarnivorous’ fish resulted in a ‘release’ of small fish species (e.g. blennies), and as a consequence a decline in the biomass of some small invertebrates on which they feed (e.g. amphipods). Increased fishing on large ‘piscivores’ resulted in increases in other small fish groups and consequential effects on other benthic invertebrate groups (e.g. polychaetes). Depletion of piscivorous fish resulted in a dramatic increase in the biomass of seabirds, which apparently compete with piscivores for small demersal and pelagic fish. An intensification of fishing pressure overall resulted in an increase in cephalopod biomass. ses of target species to increased fishing pressure were most marked within the first 5 years of the new fishing regime. Indirect responses exhibited varying degrees of inertia, and biomasses of many groups did not assume a new equilibrium within the first 20 years of the simulation. diterranean infralittoral rocky-bottom ecosystem was predicted to be relatively resilient to pulses of increased fishing and exhibited a high degree of detritus recycling. However, the speed and magnitude of ecosystem responses was shown to depend greatly on the extent of ‘top-down’ or ‘bottom-up’ control assumed for components within the system.