Reduction of nutrient loading and biomanipulation as tools in water quality management:: Longterm observations on Bautzen Reservoir and Feldberger Haussee (Germany)

Long-term (1976-1999) biomanipulation in Bautzen Reservoir (BR) revealed that a combination of piscivore stocking and catch restrictions for piscivores led to the desired effects of low planktivorous fish biomass and enhanced biomass of large filter feeders (Daphnia galeata). Despite the hypertrophic status of BR, fisheries management shifted the planktivore-dominated fish community into a piscivore-dominated community. High winter (Jan-Mar) Daphnia biomass was a sensitive indicator of reduced planktivory. Although edible phytoplankton was suppressed by elevated Daphnia biomass, mean seasonal (May-Oct) total phytoplankton biomass remained unchanged due to growth of large inedible algae and cyanobacteria. Inedible and total phytoplankton biomass was primarily controlled by phosphorus availability. However, during clear water periods a reduction of total phytoplankton was achieved with drastically increased Secchi readings. In Feldberger Haussee (FH), despite intensive long-term manual removal of cyprinids (1985-2002) and stocking of piscivorous fish (1988-2002), biomanipulation only had restricted, delayed, or transient effects on the ecosystem. Mean proportion of piscivores within total yield increased but was below 20% in most years. Planktivore cyprinid yields dropped until 1990 and remained constant thereafter. Daphnia biomass slightly increased after biomanipulation became effective (1987-1989) but declined to pre-biomanipulation levels later in the experiment (1996-2005). Beginning in 1997, both edible and inedible phytoplankton biomass started to decrease. Finally, when biomanipulation had been implemented after a delay of 8 years water clarity increased significantly. Although external loading reductions and biomanipulation in both lakes resulted in moderate phytoplankton biomass reduction and Secchi depth enhancement, the reasons for the observed changes were different. Even though strong cascading effects were detected at the top of the food web in BR, the trophic cascade was largely decoupled between phyto- and zooplankton. External plus internal phosphorus loading still exceeded a critical threshold below which a top-down-induced indirect effect of phosphorus sedimentation and finally limitation could have reduced phytoplankton biomass. In constrast to BR, the critical phosphorus loading threshold in FH has probably been approached. Nonetheless, cascading effects were weak due to insufficient reduction of planktivorous cyprinids. Improved water quality was primarily a result of resource-related effects. Thus, the 2 long-term experiments reveal that (1) biomanipulation cannot be applied successfully without reducing nutrient loading below a critical threshold (BR), and (2) for successful biomanipulation, 30-40% piscivores within total fish standing stocks are required (BR and FH). Without optimum piscivory, manual removal of planktivores will hardly produce sustained cascading effects (FH).