Up to the mid 1950's Dutch lakes were characterised by clear water and luxurious macrophyte growth. In the 1960's eutrophication changed this situation and the lakes suffered from cyanobacterial blooms, turbid water and poor submerged vegetation. Restoration programmes were aimed at reducing the rates of the external phosphorus loading. Due to these measures, the nutrient- and chlorophyll-a concentrations decreased. The transparency of the water, however, increased only slightly. In the past one decade it has been recognised that eutrophication had given the turbid state various stabilising mechanisms through which it is very resistant to restoration measures centred on nutrient reduction.A drastic reduction of the fish stock, known as biomanipulation, was suggested as an alternative approach. This thesis examines and evaluates the potential of biomanipulation as an additional restoration method and tries to understand and explain the mechanisms involved.Alternative stable states (Chapter 2)The turbidity of lakes is generally considered to be a smooth function of their nutrient status. However, models and observations in lakes suggest that over a range of nutrient concentrations, shallow lakes can have two alternative equilibria: a clear state dominated by aquatic vegetation, and a turbid state characterised by high algal biomass. This bi-stability has important implications for the possibilities of restoring eutrophied lakes. The turbid state is stabilised by planktivorous fish consuming large zooplankton, the grazer of algae and by benthivorous fish resuspending the bottom sediment in search for food. Nutrient reduction alone may have only a little impact on water clarity, but an ecosystem disturbance like fish stock reduction may contribute to factors to bring the lake back to a clear state. Once the clear water state has led to colonisation of the lakes with macrophytes, the macrophytes will stabilise the clear water state by various mechanisms. They may increase sedimentation and reduce resuspension, they may provide refuges for zooplankton against fish predation, in addition the macrophytes will compete with algae for nutrients, especially nitrogen and some macrophyte species may excrete allelopathic substances, that may inhibit algal growth.The first Biomanipulation experiments (Chapter 3 and 4)Biomanipulation in the Netherlands started with an experiment in small ponds (Chapter 3). Ten ponds of 0,1 ha were divided into two compartments; one half of each pond was stocked with 0+ cyprinids, the other served as a reference. This replicated, controlled experiment demonstrated significant differences in transparency, algae and zooplankton between compartments with fish and without fish. Despite the repeated addition of nutrients in the ponds no eutrophic conditions could be realised, probably due to increased sedimentation in the relatively small mesocosms. Although the aim of the experiment was to demonstrate the impact of planktivorous 0+ cyprinids on zooplankton and algae, the 0+ carp appeared to resuspend the bottom sediment as well, as the turbidity could not be explained by algal biomass only.Lake Bleiswijkse Zoom was biomanipulated by dividing the small lake into two parts, and removing more than 75% of the original fish biomass from one part (Chapter 4). In the treated part the removal of bottom-stirring activity by benthivorous fish and the increase in zooplankton grazing causing low chlorophyll-a concentrations led to an increase of Secchi depth transparency from 20 cm to the bottom of the lake. Within two months the Characeae became abundant. The two parts differed significantly in inorganic suspended solids, algal biomass, transparency and total nutrient concentrations. Compared with the control compartment the number of fish species in the treated part increased.Impact of benthivorous fish on turbidity (Chapter 5)In two biomanipulated small lakes (Bleiswijkse Zoom and Noorddiep) the increase in transparency was not only caused by a decrease in algal biomass but also by a decrease in sediment resuspension by benthivorous fish. The biomass of benthivorous fish and the concentration of inorganic suspended solids were positively related. A model was used to calculate the impact of resuspended inorganic suspended solids on the turbidity. By combining these relations a direct effect of benthivorous fish on the Secchi depth was calculated. In addition, it is argued that the algal biomass was also indirectly influenced by removal of benthivorous fish, as the benthivorous fish prevented macrophytes from settling, whereas macrophytes were essential to keep the water clear.Comparison of results with a Danish lake (Chapter 6)In Chapter 6 the five year results of three biomanipulated Dutch lakes (Lake Zwemlust, Noorddiep and Bleiswijkse Zoom) are compared with one Danish biomanipulated lake (lake Vaeng). The fish stock reduction led in general to a low fish stock, low chlorophyll-a, increase in Secchi disk transparency depth and a high abundance of macrophytes. Large Daphnia became first abundant, but their density soon decreased due to food-limitation and predation by fish. The total nitrogen concentrations decreased due to N-uptake by macrophytes and enhanced denitrification. In lake Bleiswijkse Zoom the water transparency decreased and the clear water state was unstable. The fish stock increased and the production of young fish in summer was high. Clear water occurred only in spring . Large Daphnia decreased markedly in summer and the macrophytes dissappeared. Except in Bleiswijkse Zoom the water remained clear in all lakes during the first five years. In summer of the sixth year transparency decreased in the hypertrophic lake Zwemlust. Also in the less eutrophic lake Vaeng, a short-term turbid state (6 weeks) occurred in summer 1992 after a sudden collapse of the macrophytes. Deterioration of the water quality seems to start in summer and is apparently related to a collapse in macrophytes. At a low planktivorous fish stock the duration of the turbid state is shorter than in presence of a high planktivorous fish biomass which became apparent comparing Lake Vaeng with lake Zwemlust.Development of fish communities after biomanipulation (Chapter 7)The development of fish communities after biomanipulation has been studied in detail in three small lakes. In lake Zwemlust, after the biomanipulation, the introduced pike could not control the explosive growth of the introduced rudd. In lake Noorddiep and Bleiswijkse Zoom the fish population became more diverse. Bream and carp became less dominant and were partly replaced by roach and perch. The main predator pike-perch was strongly reduced and replaced by pike and perch. The share of piscivorous fish increased at all sites. The recruitment of young-of-the-year fish was similar or even higher in the clear overgrown areas than in the turbid water before the measures, but the recruitement of young -of-the-year to older species differed between the species. Predation by pike and perch could not control the young-of-the-year cyprinids, but their predation may have contributed to the shift form bream to roach, because of selective predation on bream in the open water, while roach was hiding in the vegetation. The macrophytes provided new refugia and feeding conditions that favour roach and perch, but offer relatively poor survival conditions for bream and carp.Biomanipulation on a large scale (Chapter 8)The drastic fish removal in Lake Wolderwijd (2670 ha) showed that also in a very large lake biomanipulation cause water clarity to increase. In this large lake in total 425 tons of fish were removed, i.e. ca. 75% of the original fish stock. In addition 625.000 specimen of 0+ pike were added to the lake. The success of biomanipulation in the lake was transient. In spring 1991 the transparency of the water increased as a result of grazing by Daphnia galeata . The clear-water phase lasted for only six weeks. Macrophytes did not respond as expected and most of the introduced young pike died. However, from 1991 to 1993 the submerged vegetation gradually changed. Chara began to spread over the lake (28 ha in 1991 to 438 ha in 1993). The water over the Chara meadows was clear, probably as a result of increased net sedimentation in these areas, due to reduced resuspension.Evaluation of Biomanipulation Projects (Chapter 9)Evaluation of eighteen biomanipulation projects showed that biomanipulation can be a very effective method to increase the transparency of the water in lakes. In almost half of all projects a return to the clear water state was obtained and in only ten per cent of the projects the biomanipulation failed to cause an increase in the Secchi depth. In all other cases, the water transparency increased significantly. The increase in Secchi depth was significantly stronger than the general trend observed in Dutch lakes where no measures have been taken. The improvement in Secchi depth and chlorophyll-a was also more marked in lakes where only the phosphorus load had been reduced.The critical factor for obtaining clear water was the extent of the fish reduction in winter. Significant effects were observed only after > 75% fish reduction. Success seems to require substantial fish reduction. In such strongly biomanipulated lakes, wind resuspension of the sediment did not influence water clarity. No conclusions can be drawn with respect to the possible negative impact of cyanobacteria or Neomysis on grazing by Daphnia and consequently on water clarity. In all lakes, with an additional phosphorus loading reduction, the fish stock had been reduced less drastically. In these lakes, the effect on transparency was less pronounced than in the lakes with > 75% fish removal.Daphnia grazing seems responsible for spring clear water in all but one clear lakes. The factors that determine the clear water in summer are less obvious and more diverse, but a high macrophyte cover seems to play an important role in all the lakes.The decrease of the Secchi depth over the years in almost all manipulated eutrophic lakes supports the idea that the clear, vegetation dominated state is not stable at high nutrient conditions. However, the return time to the turbid water state is long ( > 8 years). As we have not applied biomanipulation in mesotrophic lakes, we could not investigate if under those conditions the clear water state would be stable on the long term. For the stability of the clear water state it would be good to have a large piscivorous population. Under the present condition in The Netherlands, with "constant" waterlevels, scarce emergent vegetation and relatively high nutrient concentrations, no dense population of piscivores are likely to develop.Development of ideas (Chapter 10)How our ideas have evolved on biomanipulation on biomanipulation in the past fifteen years is subject of the last chapter. Biomanipulation seems a more effective tool than it was thought when it was started, especially for eutrophic lakes. Expected negative aspects like increase in inedible algae have hardly been found. Also the ideas on the working of biomanipulation have evolved. In accordance with earlier ideas, the increase in water clarity mostly occurs in spring and Daphnia grazing seems crucial for this spring clear-water phase. Reduced resuspension after removal of benthivores has been shown to be important in three out of the eighteen lake studies. However, in most lakes the fish removal caused a reduction in nutrient concentrations, which may have been caused by a reduction in the benthivores or by an increase of bottomalgae.According to early views, colonisation of macrophytes seems crucial for maintaining the clear water in summer. In lakes without macrophyte growth, the water frequently became turbid again in summer, if predation of 0+ fish on Daphnia was high. At least a coverage of 50->70% of the lake with macrophytes is required to keep the water clear in the whole lake. At lower coverage only locally clear water above the macrophytes may be achieved. Macrophytes induce many changes in the ecosystem, which help stabilising the clear water state. The expected increase of piscivore control of 0+ fish has not been found in Dutch lakes. Rather, the macrophytes seem to stimulate the production of 0+ fish in eutrophic lakes. Macrophytes can provide refuge for zooplankton, but this effect may be small if high densities of 0+ fish or Neomysis are present within the vegetation. In large, shallow lakes reduced resuspension and increased sedimentation between the macrophytes may significantly contribute to the water clarity. Furthermore, nitrogen limitation of the algal growth has proven to be important in some lakes, but in others a strong reduction of the chlorophyll-a nutrient ratio's suggest that other factors than nutrient limitation are responsable for the low algal biomasses in presence of plants. Those can be related to zooplankton grazing, allelopathic effects or increase of filtering zebramussels.An increase in benthic algae may play a role in this too, but there is not yet much knowledge on this aspect.Implication for the water quality managerThis thesis shows that biomanipulation can be a very effective restoration method provided the fish biomass reduction is substantial. Indeed, biomanipulation seems to be more effective in increasing water transparency than reducing the phosphorus load. Biomanipulation may even lead to a substantial reduction of the nutrient concentrations in the water column. Provided that a substantial fish reduction is carried out, the system may shift to a clear water state even at high nutrient concentrations.The costs of biomanipulation are low compared to the costs of phosphorus reduction measures (Boers et al, 1997). Although this study shows that in eutrophic lakes no long-term success can be obtained, the long return time to the turbid state implies that even in highly eutrophic lakes, where the clear state is not stable, biomanipulation may be an cost-effective management strategy, as in some lakes it may suffice to reduce the fish stock drastically once every five year.However, biomanipulation may not work in all conditions. In open systems with a a lot of boats, it may be difficult to remove a substantial part of the fish stock. In those lakes repeated fish reduction may also have effect (as shown in Finland and Sweden), but in The Netherlands there is not yet much experience with this method. Furthermore in lakes with a very high density of inedible cyanobacteria, and a high resuspension of sediment the potential of the method is still to be tested.With future applications attention should be paid to removal of more than 75% of the fish stock, removal of small fish in adjacent ditches, removal or disturbance of spawning fish to reduce the production of 0+ fish and to reservation of money for additional fisheries. Furthermore the potential for macrophyte growth should be investigated beforehand, and if necessary macrophyte stimulating measures should be taken, like reducing the water depth or addition of oospores or seeds of plants.The challenge for the future is to create a shift to the clear water state by means of biomanipulation in more open systems, in lakes with a high density of inedible cyanobacteria or floating layers of cyanobacteria and in lakes with a high resuspension of loose sediment.