In six deep, soft-water reservoirs, ranging from oligotrophic to eutrophic, fishery management has been guided by the use of biomanipulation to improve water quality and opportunities for recreational fishing. As evidenced by the establishment of larger-bodied daphnids, a low level of zooplanktivory could be maintained in the newly filled Grosse Dhü nn and refilled Brucher and Lingese Reservoirs only by regular stocking of piscivores (Sander lucioperca, Esox lucius, Salmo trutta lacustris, Onchorhynchus mykiss) in combination with size and bag limitations for the recreational fisheries. However, in the mesotrophic Pre-Reservoir Grosse Dhü nn, Bever Reservoir and the newly filled slightly eutrophic Wupper Reservoir it took between 8 and 10 years before the predatorresistant zooplankton community responded to management with a switch to larger daphnids. Except for oligotrophic Grosse Dhü nn Reservoir where angling is prohibited, the expected enhancement of piscivore biomass through stocking not least was prevented by anglers. However, growth of perch (Perca fluviatilis) benefited from the changed fishery management relying upon stocking piscivores only allowing them to reach the size of piscivory. The appearance of larger daphnids in Pre-Reservoir Grosse Dhü nn and Wupper Reservoir resulted in the biomass of the Daphnia spring peak to increase and occur earlier causing prolonged clear water conditions. Although the larger daphnids increased transparency, total summer mean chlorophyll concentrations in the euphotic zone only decreased in the slightly eutrophic reservoirs due to reduced phosphorus availability following unchanged external phosphorus loadings. Reduced phosphorus availability in these reservoirs caused a real oligotrophication. Although the edible seston fraction was controlled ''top down'' the results did not support the trophic cascade hypotheses because total phytoplankton remained controlled ''bottom up'', admittedly triggered by ''top down'' forces. In general, the results support the importance of indirect (non-lethal) effects as the driving forces for the successful biomanipulations, particularly in slightly eutrophic reservoirs. Not least, the delayed zooplankton response provides an indication that the underlying change of internal feedbacks was not driven by external forces to stabilize the reservoirs trophic state in these deep stratifying reservoirs.

Today's fish fauna of Wupper Reservoir is the result of natural development in combination with management. Manageability of the fish stock, a prerequisite for continuing biomanipulation, was limited. Despite protection and stocking, the abundance of the stocked predatory fishes (pike, pikeperch) did never exceed 10% of the total fish biomass since filling in 1988. Contrary to predictions, the ''juvenile'' cyclical perch population that became dominant after flooding, with the disappearance of gigantic perch in 1997, was not replaced by cyprinids (460% of total fish biomass). Instead, a strong, non-cyclical piscivorous perch population, plus cyprinids (o40% of total fish biomass) became established, giving rise to low planktivory and high water quality since 1999. There is compelling evidence that the introduction of a self-reproducing pikeperch population was a key-factor in the successful management of this slightly eutrophic reservoir. With the introduction of this new type of predator (pelagic, efficient at low light conditions) there are predators (pikeperch, perch, pike and eel) present in all habitats of the reservoir. Thus, antipredator behaviour of planktivorous perch and roach was enhanced, resulting in substantial habitat segregation. Consequently, the perch were released from competition and became large enough for piscivory. Hence, predator biomass was substantially enhanced, reaching at least 25% of the total fish biomass which was estimated to be in the range of 90 kg ha À1 in August 2003. Ultimately, a high level of piscivory driving the whole lake trophic cascade, and thus a clear-water regime, which seems to be driven and stabilized by internal feed-backs, was established in Wupper Reservoir.

Refilling of the formerly oligo-mesotrophic, softwater Brucher Reservoir commenced in April 1993 and took 11 months to completely fill. A severe 'trophic upsurge' in the sense of nutrient enrichment (phosphorus, dissolved organic material) as a result of the decomposition of the inundated vegetation occurred. However, algal crop and phosphorus utilization efficiency, expressed as chlorophyll concentrations per unit of total phosphorus, remained very low. In the absence of any fish stock, a single species, Daphnia galeata, monopolized the resources. Sustained by the detritus food chain, daphnids exerted a severe‚ 'top-down' control upon phytoplankton, thereby preventing any net algal growth. In 1994, artificial mixing prevented the occurrence of anoxic water conditions and internal nutrient enrichment of the lake. Although the decay of the inundated vegetation was still of importance, phosphorus concentrations in the water column approached equilibrium with the external input while dissolved organic material concentrations clearly declined. That year, the reservoir became stocked with minnows, sun bleak (Leucaspius delineatus) and trout. As food limitation, as a result of reduced heterotrophic production, became more severe in the face of an increasing predation pressure, the daphnid population density declined, resulting in a decreasing but still adequate community filtering rate providing pronounced 'clear-water phases' of up to 10 m that were features of the period 1995-1997. Although D. galeata defended its key position in the food web, its life-history traits (e.g. body size) changed. Submerged macrophytes, which since 1995 gradually colonized suitable areas of the reservoir, provided a favourable refuge for minnows from trout predation that resulted in reduced predation pressure upon pelagic daphnids. However, in 1998, ungrazeable algae became prominent, adversely affecting transparency. That year, the significant inverse relationship between chlorophyll : total phosphorus ratios and daphnids became uncoupled during the summer (July-August) by indigestible chlorococcalean algae. Nevertheless, the fishery management that was implemented was successful in sustaining not only the lowest yield of algae at the given nutrient concentration but also the most favourable species composition with respect to water quality.

Only a combination of nutrient load abatement and food-web management proved efficient for the management of water quality in the deep stratifying Wupper Reservoir. Reduction of nutrient loading, was completed in winter 1992/1993, but resulted only in reduced winter/spring mixing of phosphorus concentrations. Since the capacity of the diatom spring bloom to remove nutrients from the trophogenic layer of this slightly eutrophic waterbody was never exhausted, the surplus of total phosphorus available to support summer algal growth remained unchanged. Thus, nutrient reduction alone did not improve the water quality, as expected. Subsequent replacement of the smaller Daphnia cucullata by the larger Daphnia galeata-hyalina complex that was attributable to successful foodweb management did, however, result in a shift from a turbid to a clear water regime in 1999. Clearly, the zooplankton community, and therefore food-web structure, played an integral role in nutrient recycling and in the repartitioning of the phosphorus pool. As diatom settling and grazing became much more tightly linked with the appearance of the larger-bodied Daphnia galeata-hyalina complex, which exploits lower-level food resources as early as May, daphnids increasingly acted as a sink for phosphorus. This increased export fluxes out of the pelagic zone and leaves a smaller surplus of total phosphorus to support the accumulation of summer algae. Consequently, water transparency and total chlorophyll concentrations in summer improved with food-web restructuring, indicating real oligotrophication of Wupper Reservoir driven by internal feedbacks.

No abstract