The spatial and temporal variability of food abundance is likely to be a major factor determining the productivity of mussel (Perna canaliculus) farms. Phytoplankton are a major food source for mussels and depletion of phytoplankton could therefore be a significant factor in farm productivity. Measurements of phytoplankton biomass (as chlorophyll a) were made inside and outside mussel farm sites, over a 13-month period. High ambient chlorophyll a concentrations in the surface waters occurred during winter, when farms had a significant reducing impact on phytoplankton biomass. In five of the seven months when farms were surveyed, the highest chlorophyll concentrations were in deeper water (15-25 m), associated with a pycnocline. This led to one management option that could increase mussel productivity: the deployment of mussels to deeper waters to take advantage of the deep-water chlorophyll maxima. The influence of water current on phytoplankton depletion was then investigated. Two farms were compared that were predicted to have different ambient flow regimes. Current data showed an ambient mean speed of 8.3 cm s⁻¹ at Farm A, and 9.1 cm s⁻¹ at Farm C. It was calculated that this difference in mean speed would result in Farm C receiving 9.6% more chlorophyll a per unit time than Farm A. Surveys inside and outside the farms showed that the longline structure reduced water speed by an average of 20% at Farm A, and 17% at Farm C. Land-based chamber experiments showed that mussel effects on chlorophyll a decreased with increasing water speed. Mussels at an assumed maximum in situ density were able to reduce chlorophyll a in water speeds up to 15 cm s⁻¹, but beyond this speed they had little impact. The effects of mussel farms on phytoplankton biomass, cell counts, biovolume and taxonomic richness were investigated by sampling inside and outside a single farm every two months over an annual cycle. Small flagellates were numerically dominant, and the diatom Ditylum brightwellii was the most abundant species by biovolume. Mussels were shown to reduce phytoplankton abundance and biovolume. And while there was some evidence of mussels being able to select certain phytoplankton species, the analysis across the entire study period showed that mussels did not have a significant influence on taxonomic richness, and so removed cells in an unselective manner. A management option that arose in the first part of the study, deployment of mussel dropper ropes to deeper water to take advantage of the chlorophyll maxima, was then tested. The selected site showed thermal stratification of the water column, causing a high concentration of phytoplankton at 15 to 25 m. Mussels were grown simultaneously at 5 and 17 m, for up to 96 days. Mussel growth and condition index data showed no significant differences between the two depth treatments, indicating that there are unlikely to be substantial mussel productivity benefits from lowering mussel farms to the deep-water chlorophyll maximum layer. Possible reasons for the lack of a higher growth rate at depth include higher phytoplankton turnover rates in shallow waters, different mussel feeding behaviour between treatments, and greater water current speeds at depth. The relationships between mussels, nutrients, seasonal conditions, and phytoplankton biomass were then investigated experimentally using in situ enclosures. Four enclosure experiments were done, two in summer when ambient nitrogen was low and two in winter when it was high. In summer there was a highly significant increase of chlorophyll a in response to added nitrogen, showing that phytoplankton were nitrogen limited. At this time mussels caused an increase in phytoplankton biomass by converting particulate nitrogen to ammonium and thereby making the nitrogen available for phytoplankton growth. During winter, the highest ambient chlorophyll a concentrations coincided with high ambient nitrogen. At this time mussel grazing mediated a significant decrease in phytoplankton concentration, further indicating that within-farm depletion of phytoplankton is most likely to occur in winter. The key concepts arising from this study were that phytoplankton depletion does occur in mussel farms in the Marlborough Sounds, and that the magnitude of depletion depends on time of year, ambient nitrogen concentration, and water current speed.