During April to September, plankton community structure was monitored in large enclosures (60 m3) in the presence and absence of planktivorous fish. Weekly sampling included inorganic nutrients, oxygen, pH, and transparency as well as biomass of phytoplankton, bacteria, heterotrophic flagellates, rotifers, and macrozooplankton.In fishless enclosures, concentrations of inorganic nutrients and transparency were higher than in enclosures with fish, whereas oxygen, pH, and biomasses of phytoplankton, picoplankton and rotifers were lower. As an average for the entire period, Chl a was 14 pg liter-' in the enclosures without fish and 46 pg liter-l in enclosures with fish.

Plankton community structure and major pools and fluxes of carbon were observed before and after culmination of a bloom of cyanobacteria in eutrophic Frederiksborg Slotssø, Denmark. Biomass changes of heterotrophic nanoflagellates, ciliates, microzooplankton (50 to 140 μm), and macrozooplankton (larger than 140 μm) were compared to phytoplankton and bacterial production as well as micro- and macrozooplankton ingestion rates of phytoplankton and bacteria. The carbon budget was used as a means to examine causal relationships in the plankton community. Phytoplankton biomass decreased and algae smaller than 20 μm replacedAphanizomenon after the culmination of cyanobacteria. Bacterial net production peaked shortly after the culmination of the bloom (510 μg C liter(-1) d(-1) and decreased thereafter to a level of approximately 124 μg C liter(-1) d(-1). Phytoplankton extracellular release of organic carbon accounted for only 4-9% of bacterial carbon demand. Cyclopoid copepods and small-sized cladocerans started to grow after the culmination, but food limitation probably controlled the biomass after the collapse of the bloom. Grazing of micro- and macrozooplankton were estimated from in situ experiments using labeled bacteria and algae. Macrozooplankton grazed 22% of bacterial net production during the bloom and 86% after the bloom, while microzooplankton (nauplii, rotifers and ciliates larger than 50 μm) ingested low amounts of bacteria and removed 10-16% of bacterial carbon. Both macro-and microzooplankton grazed algae smaller than 20 μm, although they did not control algal biomass. From calculated clearance rates it was found that heterotrophic nanoflagellates (40-440 ml(-1)) grazed 3-4% of the bacterial production, while ciliates smaller than 50 μm removed 19-39% of bacterial production, supporting the idea that ciliates are an important link between bacteria and higher trophic levels. During and after the bloom ofAphanizomenon, major fluxes of carbon between bacteria, ciliates and crustaceans were observed, and heterotrophic nanoflagellates played a minor role in the pelagic food web.

The impact of various zooplankton groups on the cycling and retention of phosphorus in a mesotrophic lake in late summer was estimated in a 1 O-d mesocosm experiment with addition of 33P. Specific assimilation rates for various dissolved and particulate fractions and individual zooplankton species were estimated by application of tracer models. By means of mathematical inverse methods, these estimates were used to construct a balanced P budget that included the different zooplankton groups.Altogether, the consumers (including the mixotrophic Dinobryon) were responsible for >60% of the estimated P sedimentation (-1% d-l). The microconsumers had the large$t specific P assimilation (0.28 d-l) and regenerated 45% of assimilated P; Daphnia had an intermediate P assimilation (0.17 d-l), but regenerated <20%, whereas Cyclops assimilated P quite slowly (0.07 d-l) and regenerated 230% of its assimilation of algal P ~20 pm, due to its feeding on microconsumers.The microconsumers, and the daphnids in particular, thus act mainly as sinks for P, whereas the cyclopoids act mainly as sources of P by assimilating and regenerating P otherwise trapped in the microconsumers. The mechanisms responsible for these different roles seem related to growth rate and P content of the consumers relative to that of their diet.Phosphorus controls algal growth in many lakes (Einsele 194 1;Dillon and Rigler 1974;Schindler 1977). Zooplankton and other phagotrophic consumers modify the P supply to phytoplankton in two ways: they act as P sinks by incorporating P and as P sources by regenerating Acknowledgments

The fate of primary production in the pelagic food web in mesotrophic Svinsjoen, southeastern Norway, was studied in a 10-d mesocosm experiment with 14C addition. Specific assimilation rates for all the main compartments in the system were estimated by applying tracer models and further combined to construct a balanced C budget by means of mathematical inverse methods. The specific carbon assimilation rates for the phytoplankton and bacteria were 0.24 and 0.49 d-l, respectively, and 70% of the bacterial production was fueled by phytoplankton exudates. The C budget shows that roughly half of the gross primary production was lost as exudates and 10% was respired. Only 20% of phytoplankton net production was assimilated by consumers. The rest of the net primary production was lost as dissolved and particulate detritus or by sedimentation. Of the net bacterial production, a third accumulated as bacterial biomass, a third was assimilated by bacterivores, and a third was lost as detritus. The coupling between the microbial loop and the classical grazer food web was insignificant, probably due to low Daphnia biomass. The microbial food web thus acted as a sink rather than as a source of C to the higher consumers. Altogether, <2% of gross primary production was incorporated into consumer biomass.