Ecological shifts from diatoms to other phytoplankton species have been related to decreasing Si:N and Si:P nutrient ratios. The Bay of Brest is such a perturbated ecosystem where Si has become limiting but where diatoms continue to dominate the phytoplankton throughout the productive period. Several hypothesis have been invoked to account for this dilemma. The most recent combines suspension feeder activity and Si recycling. Suspension feeder activity, stimulated by the proliferation of the invasive species Crepidula fornicata, would be the driving force of a biologically active silicate pump which would (1) retain Si within the Bay during spring and (2) provide the Si necessary for diatoms during summer. During the year 2000, this hypothesis was successfully tested. Direct evidence of silicic acid limitation has been provided, and during summer, benthic fluxes measured at a site with a high density of C. fornicata are one order of magnitude higher than those measured at the site with no C. fornicata. Seasonal budgets of Si inputs and diatom demand demonstrated that diatom production during summer depends strongly on Si recycling at the sediment-water interface. Thus, if C. fornicata decreases benthic biodiversity and perturbates the development of the native Great Scallop (Pecten maximus), it also helps the Bay cope with elevated N inputs. The proposed removal of C. fornicata might be economically desirable from a fisheries economic viewpoint, but it clearly would be associated, in the present context of excessive N inputs, to a potential risk of harmful algal blooms during summer.