Mixotrophy by the photosynthetic dinoflagellate Heterocapsa triquetra was investigated using fluorescently labelled algae (FLA) (size 1, 3, 6 mu m). Experiments were conducted in nitrogen-and phosphorus-replete/depleted medium under light and dark conditions. Incubations ranged from several hours to several days. The dinoflagellate was capable of phagotrophy when exposed to light and dark periods in nutrient-depleted medium. H. triquetra showed similar ingestion rates in the light (range: 0.1 to 0.4 FLA dino(-1) d(-1)) and in the dark (range: 0.06 to 0.4 FLA dino(-1) d(-1)). The dinoflagellate was able to selectively ingest the different FLA. The cyanobacterium Synechococcus sp, was not ingested, whereas a small round flagellate and the diatom Thalassiosira pseudonana were observed inside the cells. The flagellate was ingested at higher rates than the diatom in both the light and the dark. About 40% of the labelled flagellate was removed from the suspension by H. triquetra in the light and 22% in the dark. The diatom was removed from the suspension at the same rate (27 to 30%) in both light and dark treatments. After 5 d incubation in nutrient-depleted medium and no addition of FLA, the proportion of small H. triquetra cells (<1000 mu m(3)) increased from 50 to over 75%. In the presence of FLA, the distribution of H. triquetra cell volumes showed that the proportion of larger cells (>2000 mu m(3)) increased from 6 to >20% during long incubations in the light and the dark. Since the frequency of observed cells with ingested FLA varied from 3 to 20% for the same period, the proportion of larger cells may be the phagotrophic proportion of the population. However, only 8 (dark) to 12% (light) of the observed H. triquetra cell volume increase can be explained in terms of carbon from the ingestion of fluorescently labelled phytoplankton. We conclude that phagotrophy in H. triquetra may be important in maintaining the population in environments of low nutrient concentration and low light intensity.