We hypothesise that Prymnesiales cells and especially the cells of P. polylepis, even if not always toxin producers, are food items of low quality for filtering blue mussels. Support for this hypothesis comes from studies of blue mussels during the P. polylepis bloom in 1988 in the Kattegatt and Skagerack region. For example, Loo (1989; pers comm) found that blue mussels reduced their filtering activity and postponed spawning until cell concentrations of P. polylepis declined. Other occasions when algal blooms, either for toxic reasons or reasons related to cell features, have led to reduced filtration rate or reduced soft body mass in blue mussels have also been described earlier ( Gainey and Shumway, 1988, Pieters et al., 1980 and Tracey, 1988). Between March and May 2008, Prymnesiales cells including cells of P. polylepis constituted at some Baltic monitoring sites of up to 80% or more of the total phytoplankton biovolume. We therefore suggest that the P. polylepis bloom in early spring 2008 reduced or delayed the soft body mass increase and gonad build-up of blue mussels, and thereby reduced the availability of high quality food for eiders at foraging sites close to their nesting sites during the pre-laying period. Intensive blooms of P. polylepis have not been recorded before in the Baltic Sea. Such blooms therefore cannot explain the steady decline of the Baltic eider population since the mid 1990s. However, the spatial and temporal match between the P. polylepis bloom and the extensive non-breeding of eiders observed in this study does suggest that breeding eiders can be significantly affected by species interactions at low trophic levels as well as by the quality of the food obtained close to the breeding sites during the pre-laying period.