This proposal addresses the cause of Heterosigma blooms and their variable toxicity to fish.  Heterosigma blooms in 1989-1990 killed more than 2 million pen-reared salmon worth $8 million dollars in Western Washington; fish killing blooms in many other parts of the world have also been reported. In this project, a new mechanism for causing the dense surface blooms was described.  In vertical columns, when distilled water was added to the surface, Heterosigma cells became concentrated in surface regions of lesser salinity water (Hershberger, 1995).   Although this behavior was demonstrated in laboratory cultures, the conditions seem consistent with those observed during field blooms. A fish kill involving wild, free swimming salmon was associated with a Heterosigma bloom at Case Inlet, Allyn, WA in late September 1994; data were collected on vertical distribution of temperature, salinity, pH, nutrients, and Secchi depth. The general hypothesis we are testing is that nutrient deficiencies or imbalances induce (otherwise non-toxic) Heterosigma to become toxic, or encourage the competitive dominance of toxic strains.  We succeeded in obtaining phosphorous limited cells, but no toxicity was demonstrated. Nitrate-limited cultures have not been obtained; cells have died when introduced to medium with less than 500 uM nitrate, whereas Puget Sound rarely exceeds 40 uM nitrate.  Cells rinsed free of nitrate did not demonstrate toxicity. We have also studied the effects of iron limitation, salinity, and ultra-violet light on growth and toxicity. Because nitrate levels were unusually low (0.22 uM) during the fish kill, we are continuing to investigate N deficiency, as well as other stresses associated with surface bloom formation.