We verified an active uptake of kleptoplastids in the toxic and bloom-forming dinoflagellates of the genus Dinophysis from its preferred prey, the ciliate Myrionecta rubra, using a quantitative real-time PCR technique. During a 65 d starvation/feeding experiment with Dinophysis caudata, changes in plastid 16S rRNA, plastid autofluorescence and plastid/nuclear DNA ratio through the cell cycle were followed with quantitative real-time PCR and flow cytometry. During starvation, the cultures initially showed a rapid growth and a 3.5-fold increase of number of cells ml -1 , while at the same time, plastid DNA cell -1 showed a 3.5-fold decrease, and a 3.6-fold decrease in phycoerythrin fluorescence cell -1 . The decrease in plastid DNA cell -1 d -1 closely followed culture growth rate (Pearson correlation, r = 0.91), indicating that existing plastids were diluted within the growing population and that no new plastids were synthesised by the cells. When starved cells were re-fed by the ciliate M. rubra on Days 43 to 51 of the experiment, plastid DNA cell -1 increased 7-fold up to 14 000 16S DNA copies per cell, thereby directly revealing the kleptoplastic behaviour. The implication is that not only availability of the prey M. rubra itself, but also the supply of suitable kleptoplastids might be an important controlling factor for Dinophysis spp. bloom formation and decline.

The marine sand-dwelling dinoflagellate Polykrikos lebourae possesses obvious gold-brown pigmented plastids as well as taeniocyst-nematocyst complex structures. Despite of the presence of the visible plastids, previous attempts to establish this species in culture all failed and thus the unavailability of cultures of this species has posed a major obstacle to further detailed exploration of ecophysiology of the dinoflagellate. Here, we isolated P. lebourae from sandy sediment of an intertidal flat on Korean western coast, successfully established it in culture, and have been maintaining the stock culture over the past 3 years. Using this stock culture, we explored phagotrophy and potential prey resources of P. lebourae, growth and grazing responses of P. lebourae to different prey organisms, the effect of prey concentration on growth and grazing rates and gross growth efficiency (GGE) of P. lebourae when fed three different prey organisms, and the growth kinetics of P. lebourae under different light regimes. P. lebourae captured prey cells using a tow filament and then phagocytized them through the posterior end. The dinoflagellate was capable of ingesting a broad range of prey species varying in size, but not all prey species tested in this study supported its sustained growth. GGE of P. lebourae was extremely high at low prey concentration and moderate or low at high prey concentrations, indicating that P. lebourae grows heterotrophically at high prey concentrations but its growth seems to be more dependent on a certain growth factor or photosynthesis of plastids derived from the prey. In the presence of prey in excess, P. lebourae grew well at moderate light intensity of 40 µmol photons m -2 s -1, but did not grow at dim and high (10 or 120 µmol photons m -2 s -1 ) light intensities. Our results suggest that the benthic dinoflagellate P. lebourae is an obligate mixotroph, requiring both prey and light for sustained growth and survival.