A three-dimensional hydrodynamic model (ELCOM) coupled with a biological model (CAEDYM) was calibrated with field data from Lake Simcoe (2008) and used to examine the expected impact of dreissenid mussels on the distribution of phytoplankton and nutrients, mussel energetic, and the interactions with hydrodynamic conditions. In accordance with the near-shore shunt hypothesis and the actual distribution of mussels, the model predicted a considerable impact of mussels on phytoplankton in nearshore areas (<15 m depth; −38 % phytoplankton). Horizontal advection could account for the lower (−15 % phytoplankton) offshore impact despite the absence of mussels within this region of the lake. These predictions were similar to observed long-term changes in phytoplankton biomass after the dreissenid mussel invasion. Simulated decreases in the initial mussel biomass by 50 % led to increased phytoplankton nearshore, increased mussel growth and decreased phosphorus excretion relative to ingestion. In contrast, increases in initial mussel biomass by 50 % resulted in minor changes in phytoplankton, since near-bottom depletion of phytoplankton caused growth and energy-limitation of mussels. The model, which was the first to capture metabolically-driven variations of mussel-mediated nutrient cycling, predicted a partial uncoupling of carbon and phosphorus cycling by mussels, driven by varying controls on mussel growth. A 60 % increase in the intensity of three storms during the summer had little net-effect on phytoplankton biomass, but reduced energy-limitation of mussels and increased mussel growth due to increased supply of phytoplankton. Our model demonstrates that near-bottom depletion of phytoplankton is an important process limiting the impact of mussels, and indicates that it affects the mussels’ cycling of energy and nutrients. It further suggests that stronger storm events could lead to further increase of flux of organic matter from the pelagic zone to the benthos, further altering energy pathways in the lake.