Food web manipulation or biomanipulation in lakes aims at suppressing the biomass of planktivorous fish, at maintaining a high biomass of Daphnia, and thereby at reducing algal biomass. This tool of water management has been applied as a whole-lake experiment in Bautzen reservoir (Germany) since 1981 and was extented by an in-lake phosphate precipitation in 1996. However, large algae that are inedible for daphnids have been dominating the phytoplankton community regularly. At the same time, bacteria should be able to use the exudates of these large phytoplankton and, thus, should have the potential to be an important alternative food source for daphnids, which can feed on bacteria directly. The aim of this study was to answer three questions: (1) How are primary production (PP), bacterial production (BP), and the coupling between BP and PP influenced by the phosphate precipitation? (2) Is BP top-down controlled by strong grazing pressure of daphnids? (3) Do daphnids ingest bacteria and algae with similar efficiency, and how important are both as food sources for Daphnia?

PP and BP were determined in various size fractions before (1995) and during the phosphate precipitation (1996, 1997). An enclosure experiment comparing the development of the plankton community with and without Daphnia was performed to study the top-down effects on BP. Furthermore, clearance rates and ingestion rates of Daphnia galeata on bacteria and algae were measured.

Under hypertrophic conditions before the phosphate precipitation, both the biomass of Daphnia and the PP were high, with the bluegreen Microcystis spp. contributing a large fraction of the total PP. BP was relatively low during most periods but the coupling between the dynamics of BP and PP was very close. Phosphate precipitation improved the water quality and led to eutrophic conditions with decreased PP and low abundance of Microcystis spp. In contrast to PP, under eutrophic conditions and at simultaneously low biomass of Daphnia, BP was mostly higher than in 1995, and the dynamics of BP and PP were coupled only weakly. The most important top-down impact on BP in the lake was exerted by daphnids effectively suppressing protozoan grazers of the bacteria. In the enclosure experiment, after an initial stimulation of BP due to sloppy feeding, BP was reduced directly by grazing as well as indirectly due to decreased PP. Clearance rates of daphnids on bacteria and on algae in the reservoir often were equally high. As long as phytoplankton biomass was higher than bacteria biomass, algae were the more important food source for daphnids. However, during winter and during the spring clear water phase, ingestion of bacteria contributed up to 87 % and 42 %, respectively, of the total carbon ingested by daphnids. Thus, during certain periods, bacteria were an important food source for daphnids and supported the maintenance of a high biomass of Daphnia at a low biomass of algae. Bacteria might bridge over periods of food limitation of daphnids and thereby stabilize biomanipulation in lakes.