Pelagic bacteria, heterotrophic and autotrophic nanoflagellates, and ciliates were quantified in the marine shallow-water sound, Limfjorden, Denmark, from March to November 1983. During summer the populations showed pronounced oscillations with time scales of days or weeks. Concentrations of bacteria, heterotrophic nanoflagellates and ciliates ranged from 0.5 to 15.2 X 106 ml" (mean: 6.3), 0.2 to 15.2 X 103 ml-' (mean: 2.0), and 1.4 to 162.0 ml-' (mean: 17.1), respectively. Population sizes of bacteria and heterotrophic nanoflagellates were coupled, as were populations of total nanoplankton and ciliates; in contrast there was no direct coupling between bacteria and ciliates. It is estimated that heterotrophic nanoflagellates on average cleared 45 YO (range: 5 to 365 %) of the watercolumn for bactena per day during summer. Ciliates on average cleared 93 % (range: 4 to 352 %) of the water-column for nanoplankton per day.

During 9 to 25 June 1987, carbon budgets were established for estuarine enclosures manipulated by additions of nutrients and suspension-feeding bivalves An intensive sampling program and a detailed examination of autotrophic and heterotrophic nlicroorganlsms enabled construction of carbon budgets of the microbial food web and comparison of flow rates through a number of microbial components. Phytoplankton biomass and production covaried, and, as expected, lowest values were recorded in enclosures with added mussels, and highest values in enclosures with added nutrients. Bacteria and heterotrophic nanoflagellates peaked a few days after maxima in phytoplankton biomass and production. In enclosures with added mussels, biomasses were lower for bacteria and microzooplankton, and mesozooplankton, but slightly higher for heterotrophic nanoflagellates. Bacteria, flagellates, and microzooplankton, mostly ciliates, dominated heterotrophic processes, whereas larger mesozooplankton ingestion did not exceed 5 % of phytoplankton primary production. Microzooplankton and flagellate clearances were higher in enclosures with added nutrients, whereas no such changes were found in the macrozooplankton, probably because the duration of the experiments did not allow full development of the macrozooplankton. The added mussels dominated heterotrophic consumption and controlled organisms > 20 Fm. Exclusion of mussels induced a primary dominance of microzooplankton followed by a subsequent increase of mesozooplankton. Additions of nutrients and filtration by suspension-feeding bivalves caused qualitative and quantitative changes at all levels in the microbial food web. These changes were measured from a large number of microbial components and allowed balances of the carbon budgets to be made as well as identification of factors controlling the structure and function of the pelagic carbon cycle.