Erythrocyte lysis and fish mortality assays, in combination with high performance liquid chromatography-electrospray ionization-mass spectrometry (LC-ESI-MS) analysis, were investigated for bioassay-guided fractionation of cultured golden alga (Prymnesium parvum). Intracellular constituents from isolated cell pellets and extracellular supernatant growth medium were fractionated by a variety of common separation modes, including reversed phase and normal phase solid phase extraction step fractionation procedures. For reversed phase fractionation of extracellular growth medium, one fraction was obtained that displayed hemolytic activity and adversely affected fish survival. Effective dose concentrations for this sample were similar in both assays and the LC-ESI-MS analysis of the fraction showed a number of mass spectral signals which were distinct to this fraction. Fractions obtained from separation of an ethanol extract of the lyophilized cell pellet provided one sample that was highly hemolytic, but not toxic to fish. Discrepancies such as this, along with notable fish behavioral responses from other nonhemolytic cell pellet fractions, problems with the use of unbonded silica gel for fractionation, and misleading mass spectral signatures are interesting in the context of our current understanding of P. parvum toxicity and remain to be investigated further. This work provides an account of ongoing research aimed toward comprehensive elucidation of toxic constituents produced by golden alga for the purpose of providing a better understanding and means to potentially remediate the ecological impact of this harmful bloom organism.(KEY TERMS: erythrocyte lysis; fish mortality; liquid chromatography; electrospray ionization; mass spectrometry; prymnesin.)

The haptophyte flagellate Prymnesium parvum forms blooms in brackish waters and produces toxins that harm aquatic organisms. Batch cultures of P. parvum were grown in phosphorus-limited artificial seawater medium with 3 treatments: no aeration or buffering, continuous aeration, and buffering to a high, basic pH. Over a period of 32 d, frequent samples were taken to determine: cell abundance; cellular composition of carbon (C), nitrogen (N), and phosphorus (P); culture pH; and hemolytic activity. Only pH differed significantly among media treatments: it was basic in all treatments after 10 d of culture, but consistently highest in the buffered medium treatment. In all treatments, exponential population growth was observed during the first 10 d of culture, at rates of about 0.4 to 0.6 d -1. The cell quota for P declined rapidly over the first 8 d of culture and more slowly thereafter. A transition from exponential growth to stationary phase occurred over 10 to 21 d of culture. Population growth rate was related to cell quota for P according to Droop's equation, with an estimated quota for zero growth of about 5 fmol cell -1 . In all cultures, high hemolytic activity was seen on Days 8 and 12. All but one culture displayed oscillations of hemolytic activity thereafter. At times of high hemolytic activity, the cell quota for P was <100 fmol cell -1 and the cellular C:P ratio was at or above the Redfield ratio of 106. KEY WORDS: Harmful algae · Phytoplankton · Blooms · Toxins · Cell quota · Nutrient limitation · Phosphorus Resale or republication not permitted without written consent of the publisherAquat Microb Ecol 61: [141][142][143][144][145][146][147][148] 2010 represented 2 of the most sensitive assays for assessing the toxicity associated with P. parvum. Choice of a toxicity assay depends on the objectives of a study, and because the hemolytic assay provides relatively rapid information for a toxicological response, it was used here to examine variation in hemolytic toxicity over time in laboratory cultures of P. parvum (modification of Eschbach et al. 2001).Factors associated with the toxicity of Prymnesium parvum were recently reviewed by Granéli & Salomon (2010). They proposed that the abundance and stage of population growth were important, with older and denser populations being more toxic than young, sparse populations. As a population grows in a nutrientlimited medium, cellular nutrient content falls, growth slows, and toxicity rises (Dafni et al. 1972, Johansson & Granéli 1999, Granéli & Johansson 2003. Concurrently, pH can increase as CO 2 is removed from growth media (Grover et al. 2007). Some studies have found that high pH increases the toxic activity of P. parvum (Shilo & Aschner 1953, while others have found the opposite (Padilla 1970, Igarashi et al. 1998. Aeration can also increase the toxicity of P. parvum (Igarashi et al. 1995), prompting suggestions that strong wind-mixing might enhance the toxicity of blooms (Granéli & Salomon 2010).In the present study, Prymnesium parvum wa...