Effects of a filter-feeding fish [silver carp, Hypophthalmichthys molitrix (Val.)] on phyto- and zooplankton in a mesotrophic reservoir: results from an enclosure experiment

SUMMARY 1. Silver carp, Hypophthalmichthys molitrix (Val.), feeds on both phyto- and zooplankton and has been used in lake biomanipulation studies to suppress algal biomass. Because reports on the effects of silver carp on lake food webs have been contradictory, we conducted an enclosure experiment to test how a moderate biomass of the fish (10 g wet weight m⁻³) affects phytoplankton and crustacean zooplankton in a mesotrophic temperate reservoir.
2. Phytoplankton biomass <30 μm and particulate organic carbon (POC) <30 μm were significantly higher in enclosures with silver carp than in enclosures without fish, whereas Secchi depth was lower. Total copepod biomass declined strongly in both treatments during the experiment, but it was significantly higher in fish-free enclosures. Daphnid biomass was also consistently higher in enclosures without fish, although this effect was not significant. However, the presence of fish led to a fast and significant decrease in the size at maturity of Daphnia galeata Sars. Thus, the moderate biomass of silver carp had a stronger negative effect on cladoceran zooplankton than on phytoplankton.
3. Based on these results and those of previous studies, we conclude that silver carp should be used for biomanipulation only if the primary aim is to reduce nuisance blooms of large phytoplankton species (e.g. cyanobacteria) that cannot be effectively controlled by large herbivorous zooplankton. Therefore, stocking of silver carp appears to be most appropriate in tropical lakes that are highly productive and naturally lack large cladoceran zooplankton.     

Ecological studies on crater lakes in West Cameroon The blood of endemic cichlids in Barombi Mbo in relation to stratification and their feeding habits

The West Cameroonian crater lake Barombi Mbo is stratified, with no detectable oxygen below 20 m. Of the endemic cichlid fishes in this lake Konia dikume is exceptional in having a mean haemoglobin concentration of 16.55 g/100 ml, while the other ten endemic cichlids range from 5.55 to 8.70 g/100 ml blood. The erythrocyte count is also higher in K. dikume than in the other species.
The phytoplankton in the lake is most abundant between 10 and 15 m from the surface.
The single species of cyclopoid copepod and the rotifers in the zooplankton remain in the top 20 m throughout the day and night, but the larvae of Chaoborus spend the daytime at depths below 20 m, mostly between 30 and 70 m. At night these larvae ascend into the upper layers, and feed on rotifers.
Konia dikume feeds on the larvae of Chaoborus , and the high concentration of haemoglobin in the blood of this fish, coupled with a large blood volume, appears to be an adaptation for the storage of oxygen. This would enable the fish to dive into the deoxygenated hypolimnion and extend the time available for feeding when the Chuoborus larvae are ascending at dusk and descending at dawn. Konia dikume appears to be a visual feeder, and in the short tropical twilight any extension of the feeding period could be of survival value.     

The impact of planktivorous flsh on the structure of a plankton community

SUMMARY. 1. The abundance of pianktivorous juvenile yellow perch, Perca flavescens , was manipulated in three 750 m³ enclosures in a eutrophic lake.
2. There was a significant negative relationship between fish and zoopiankton biomasses. At high fish densities the zooplankton community was dominated by small filter-feeding cladocera. primarily bosmi- nids. At low fish densities the zooplankton community was dominated by large filter-feeding cladocera, primarily daphnids.
3. There was no significant relationship between zooplankton and phytoplankton biomasses when considered over the whole experiment but there was a trend towards lower phytoplankton biomass in the enclosure dominated by daphnids during mid-summer.
4. We conclude that although planktivorous fish have a strong negative impact on zooplankton community biomass and size structure, the relationship at the next lower trophic level, zooplankton and phytoplankton, is much weaker. Therefore, the biomanipulation of planktivorous fish populations as a management technique to control phytoplankton abundance is largely ineffective.     

Some effects of enclosure on the zooplankton in a small lake

Observations on the effect of enclosure on zooplankton by introducing two experimental tubes, each holding some 18 000 m³ of water, into a small lake, showed that a limnetic community could be maintained within tubes of this capacity throughout a period of 22 months. Two species, the copepod Diaptomus gracilis and the cladoceran Daphnia hyalina were dominant both in the open lake and in each tube, but Diaptomus gracilis was relatively more abundant in the tubes than in the lake, whereas Daphnia hyalina, especially in spring and autumn, was more abundant in the lake than in the tubes. Several scarce species became relatively more abundant in the tubes than in the lake. These changes are more probably associated with reduced predation, especially by larvae of Chaoborus, than with artificially induced changes in the phytoplankton, changes only poorly correlated with changes in the zooplankton.     

The effect of small zooplankton on the microbial loop and edible algae during a cyanobacterial bloom

SUMMARY 1. We studied the effect of the small crustacean zooplankton on heterotrophic micro-organisms and edible phytoplankton in a eutrophic lake during a cyanobacterial bloom.
2. Small (15 L) enclosures were filled with natural or screened (100 μm) lake water and incubated for 5 days in the lake. Screening removed crustacean zooplankton but the initial density of rotifers and phytoplankton remained the same in control and removal treatments. Changes in the abundance and biomass of bacteria, autotrophic picoplankton (APP), heterotrophic nanoflagellates (HNF) and ciliates were measured daily.
3. The crustacean zooplankton, dominated by the small cladoceran Chydorus sphaericus , did not affect cyanobacteria, the main phytoplankton group during the experiment.
4. The removal of the crustacean zooplankton induced a higher abundance of ciliates and reduced that of the HNF, indicating the importance of ciliates in controlling HNF in this system.     

Changes in the plankton community following introduction of filter-feeding planktivorous fish

1. We conducted enclosure experiments in a shallow eutrophic lake, in which a biomass gradient of the filter-feeding planktivore, silver carp, Hypophthalmichthys molitrix Valenciennes, was created, and subsequent community changes in both zooplankton and phytoplankton were examined.
2. During a summer experiment, a bloom of Anabaena flos-aquae developed (≈ 8000 cells mL⁻¹) solely in an enclosure without silver carp. Concurrent with, or slightly preceding the Anabaena bloom, the number of rotifer species and their abundance increased from seven to twelve species (1700–14 400 organisms L⁻¹) after the bloom in this fish-free enclosure. Protozoans and bacteria were generally insensitive to the gradient of silver carp biomass.
3. During an autumn experiment, on the other hand, large herbivorous crustaceans were more efficient than silver carp in suppressing the algae, partly because the lower water temperature (≈ 24 °C) inhibited active feeding of this warm-water fish and also formation of algal colonies. Heterotrophic nanoflagellate and bacterial densities were also influenced negatively by the crustaceans.
4. Correspondence analysis (CA) was applied to the weekly community data of zooplankton and phytoplankton. A major effect detected in the zooplankton community was the presence/absence of silver carp rather than the biomass of silver carp, whereas that in the phytoplankton community was the fish biomass before the Anabaena bloom, but shifted to the presence/absence of the fish after the bloom.     

Effectiveness of phytoplankton control by large-bodied and small-bodied zooplankton

Employingin situ enclosures containing inocula of the lake zooplankton (mainlyDaphnia galeata, Daphnia cucullata andBosmina spp.) from a moderately eutrophic Lake Ros (Northern Poland) or large-bodiedDaphina magna, the following observations on succession of phytoplankton were made: 1) whereasD. magna could control the density of all the photoplankton size classes, the lake zooplankton could not suppress the large-sized phytoplankters or net phytoplankton; 2) the lake zooplankton was able to control the density of small algae (< 50μm), but its effect on large algae may be opposite: a promotion of net phytoplankton growth by removing small-sized algae which can out-compete net phytoplankton for limited PO₄-P resources (<5μg P l⁻¹). Since efficiency of phytoplankton density control byD. magna decreased with an increase in net phytoplankton abundance, biomanipulation could not be successful without introducing or maintaining a high population of large-bodied cladoceran species before high densities of large algae would make the control of phytoplankton inefficient.     

The role of light for fish–zooplankton–phytoplankton interactions during winter in shallow lakes – a climate change perspective

1. Variations in the light regime can affect the availability and quality of food for zooplankton grazers as well as their exposure to fish predation. In northern lakes light is particularly low in winter and, with increasing warming, the northern limit of some present-day plankton communities may move further north and the plankton will thus receive less winter light.
2. We followed the changes in the biomass and community structure of zooplankton and phytoplankton in a clear and a turbid shallow lake during winter (November–March) in enclosures both with and without fish and with four different light treatments (100%, 55%, 7% and <1% of incoming light).
3. In both lakes total zooplankton biomass and chlorophyll- a were influenced by light availability and the presence of fish. Presence of fish irrespective of the light level led to low crustacean biomass, high rotifer biomass and changes in the life history of copepods. The strength of the fish effect on zooplankton biomass diminished with declining light and the effect of light was strongest in the presence of fish.
4. When fish were present, reduced light led to a shift from rotifers to calanoid copepods in the clear lake and from rotifers to cyclopoid copepods in the turbid lake. Light affected the phytoplankton biomass and, to a lesser extent, the phytoplankton community composition and size. However, the fish effect on phytoplankton was overall weak.
5. Our results from typical Danish shallow eutrophic lakes suggest that major changes in winter light conditions are needed in order to have a significant effect on the plankton community. The change in light occurring when such plankton communities move northwards in response to global warming will mostly be of modest importance for this lake type, at least for the rest of this century in an IPCC A2 scenario, while stronger effects may be observed in deep lakes.     

Top-down control of phytoplankton: the role of time scale, lake depth and trophic state

SUMMARY 1. One of the most controversial issues in biomanipulation research relates to the conditions required for top-down control to cascade down from piscivorous fish to phytoplankton. Numerous experiments have demonstrated that Phytoplankton biomass Top-Down Control (PTDC) occurs under the following conditions: (i) in short-term experiments, (ii) shallow lakes with macrophytes, and (iii) deep lakes of slightly eutrophic or mesotrophic state. Other experiments indicate that PTDC is unlikely in (iv) eutrophic or hypertrophic deep lakes unless severe light limitation occurs, and (v) all lakes characterised by extreme nutrient limitation (oligo to ultraoligotrophic lakes).
2. Key factors responsible for PTDC under conditions (i) to (iii) are time scales preventing the development of slow-growing inedible phytoplankton (i), shallow depth allowing macrophytes to become dominant primary producers (ii), and biomanipulation-induced reduction of phosphorus (P) availability for phytoplankton (iii).
3. Under conditions (iv) and (v), biomanipulation-induced reduction of P-availability might also occur but is insufficient to alter the epilimnetic P-content enough to initiate effective bottom-up control (P-limitation) of phytoplankton. In these cases, P-loading is much too high (iv) or P-content in the lake much too low (v) to initiate or enhance P-limitation of phytoplankton by a biomanipulation-induced reduction of P-availability. However, PTDC may exceptionally result under condition (iv) if high mixing depth and/or light attenuation cause severe light limitation of phytoplankton.
4. Recognition of the five different conditions reconciles previous seemingly contradictory results from biomanipulation experiments and provides a sound basis for successful application of biomanipulation as a tool for water management.     

经典与非经典生物操纵理论及其应用

湖泊是我国的重要水资源之一,为人类提供了无法替代的生态及社会服务功能。但我国湖泊富营养化日趋严重。以食物网为基础的经典与非经典生物操纵成为湖泊富营养化修复的重要理论支撑。论文综述了经典与非经典生物操纵理论的原理、发展与应用,分析了鱼类(肉食性、滤食性)、浮游动物在控制藻类数量上发挥的功能,并讨论了两种理论的适用条件及实际应用中遇到的问题,以期为我国富营养化湖泊修复工作提供参考。经典与非经典生物操纵理论均是通过改变食物网结构控制藻类,分别利用浮游动物、滤食性鱼类控制藻类数量,但两者都未降低水中N、P含量。因此,实施有效的藻类水华生物操纵应与其它修复措施联合使用。     

Ecological studies on crater lakes in West Cameroon Zooplankton of Barombi Mbo, Mboandong, Lake Kotto and Lake Soden

The zooplankton of four crater lakes in North-west Cameroon has been studied by means of vertical hauls from the bottom to the surface. Two of the lakes are deep, with sparse phytoplankton, while the other two are shallow with dense blooms of blue-green algae. All the lakes are without planktonic Cladocera. In Lake Kotto, Mboandong and Barombi Mbo the dominant zooplankter is Mesocyclops (Thermocyclops) hyalinus , which is replaced in Lake Soden by Mesocyclops leuckarti. Larvae of Chaoborus are present in all the lakes. In the two deep lakes, Soden and Barombi Mbo, the dominant rotifer is Polyarthra dolichoptera , but in the two shallow lakes, Kotto and Mboandong, three species, Hexarthra mira, Brachionus fakatus and B. caudatus , make up the bulk of the rotifers.
In Barombi Mbo during the daytime most of the zooplankton was in the top 20 m, but most of the Chaoborus larvae were below 20m.
The standing crops of zooplankton in these lakes are low compared with those of lakes in the Kigezi District of Uganda.
The most remarkable feature of the zooplankton in these lakes is the occurrence of dwarfed forms. Both species of cyclopoid copepods and the eight identifiable species of Rotifera are smaller here than in most other localities. Factors which may be associated with this dwarfing are low latitude with constant high temperature, low phytoplankton availability, and the low ratio of drainage area to lake area.     

Implications of Chaoborus pupation and ecdysis in cold water

1. Cold water acted differently to delay and lengthen the pupation period for the larvae of two species of the zooplankton predator Chaoborus (Diptera: Chaoboridae). During Chaoborus pupation, the zooplankton community is released from predation, while the dark-coloured Chaoborus pupae are more susceptible to their own predators.
2. Fourth instar larvae of Chaoborus americanus and C. trivittatus , collected from an oligotrophic lake, were reared individually at 5 °C in the dark. Chaoborus americanus was also reared at 9 and 12 °C under spring photoperiod conditions (L : D, 16 : 8 h). Individuals were observed through pupation to emergence (ecdysis) or death.
3.  Chaoborus americanus pupated at 5, 9 and 12 °C with substantial emergence only at 12 °C. In comparison, C. trivittatus emerged at 5 °C. Light was not a necessary cue for pupation and ecdysis, contrary to previous reports. Cold water delayed the onset and lengthened and increased the variability of the duration of pupation.
4. In Shirley Lake, C. americanus pupated in late June–early July while C. trivittatus pupated first in April and again in June–July.
5.  Chaoborus americanus pupae needed a temperature cue to complete ecdysis. The ecdysis temperature threshold helps to explain the difference in pupation timing, and the geographical distribution, of C. americanus and its relatively inflexible life history contrasted with C. trivittatus . Delayed predator pupation in years with low spring temperature can affect the community dynamics of the prey.     

Long term study on the influence of eutrophication, restoration and biomanipulation on the structure and development of phytoplankton communities in Feldberger Haussee (Baltic Lake District, Germany)

Feldberger Haussee provides a classic example of eutrophication history of hardwater lakes in the Baltic Lake District (Germany) and of changes in their algal flora during the 20th century. The lake originally was regarded as slightly eutrophic. A process of drastic eutrophication from the 1950s until the end of the 1970s caused mass developments of blue-green and green algae. A restoration program was started in the 1980s to improve the water quality of the lake using both diversion of sewage outside the catchment area, and biomanipulation by altering the fish community. This restoration program led to positive changes in the lake ecosystem. Direct effects of biomanipulation resulted in an increase of herbivorous zooplankton, a decrease of phytoplankton biomass, and an increase of water transparency. The recovery of Feldberger Haussee also may have been indirectly enhanced by an increase in nutrient sedimentation as a consequence of intensified calcite precipitation, decrease in phosphorus remobilization due to a pH-decrease, increased NIP-ratio, and recolonization of the littoral zone by macrophytes. This paper concentrates on the long term development of the phytoplankton community as a response to changes in the food web structure as well as to alterations in the chemical environment of the algae. Both are reflected in four major stages passed by the algal assemblage between 1980 and 1994: (1) From 1980-summer 1985 dense green algal populations were found indicating similar conditions as in the 1970s during the period of maximum eutrophication. (2) A diverse phytoplankton community during summer 1985–1989 showed the first effects of a recovery. (3) From 1990–1992 the phytoplankton was characterized by ungrazeable filamentous blue-green algae first of all as a response to increased herbivory of zooplankton on edible species and to increasing N/P-ratios. (4) Finally, the algal species diversity increased in 1993 and 1994 whereas the phytoplankton biomass decreased showing the success of the combined restoration measures.     

Effects of planktivorous and benthivorous fish on organisms and water chemistry in eutrophic lakes

The effects of planktivorous and benthivorous fish on benthic fauna, zooplankton, phytoplankton and water chemistry were studied experimentally in two eutrophic Swedish lakes using cylindrical enclosures. In enclosures in both lakes, dense fish populations resulted in low numbers of benthic fauna and planktonic cladocerans, high concentration of chlorophyll, blooms of blue-green, algae, high pH and low transparency. In the soft-water Lake Trummen, total phosporus increased in the enclosure with fish, but in the hard-water Lake Bysjön total phosphorus decreased simultaneously with precipitation of calcium carbonate. Enclosures without fish had a higher abundance of benthic fauna and large planktonic cladocerans, lower phytoplankton biomass, lower pH and higher transparency.The changes in enclosures with fish can be described as eutrophication, and those in enclosures without fish as oligotrophication. The possibility of regulation of fish populations as a lake restoration method is discussed.This paper was presented at the XXth SIL Congress in Copenhagen in 1977.     

Effects of piscivore-mediated habitat use on growth, diet and zooplankton consumption of roach: an individual-based modelling approach

SUMMARY 1. We used an individual based modelling approach for roach to (i) simulate observed diel habitat shifts between the pelagic and littoral zone of a mesotrophic lake; (ii) analyse the relevance of these habitat shifts for the diet, activity costs and growth of roach; and (iii) quantify the effects of a hypothetical piscivore-mediated (presence of pikeperch) confinement of roach to the littoral zone on roach diet, activity costs and growth.
2. The model suggests that in the presence of pikeperch, roach shifts from zooplankton as the primary diet to increased consumption of less nutritious food items such as macrophytes, filamentous algae and detritus.
3. The growth of roach between May and October was predicted to be significantly higher in the absence of pikeperch, although the net activity costs were about 60% higher compared with the scenario where pikeperch were present.
4. These modelling results provide quantitative information for interpreting diel horizontal migrations of roach as a result from a trade-off between food availability and predation risk in different habitats of a lake.
5. Altering the habitat selection mode of planktivorous roach by piscivore stocking has the potential to reduce zooplankton consumption by fish substantially, and could therefore be used as a biomanipulation technique complementing the reduction of zooplanktivorous fish.     

Top-down control of natural phyto- and bacterioplankton prey communities by Daphnia magna and by the natural zooplankton community of the hypertrophic Lake Blankaart

Biomanipulation, through the reduction of fish abundance resulting in an increase of large filter feeders and a stronger top-down control on algae, is commonly used as a lake restoration tool in eutrophic lakes. However, cyanobacteria, often found in eutrophic ponds, can influence the grazing capacity of filter feeding zooplankton. We performed grazing experiments in hypertrophic Lake Blankaart during two consecutive summers (1998, with and 1999, without cyanobacteria) to elucidate the influence of cyanobacteria on the grazing pressure of zooplankton communities. We compared the grazing pressure of the natural macrozooplankton community (mainly small to medium-sized cladocerans and copepods) with that of large Daphnia magna on the natural bacterioplankton and phytoplankton prey communities. Our results showed that in the absence of cyanobacteria, Daphnia magna grazing pressure on bacteria was higher compared to the grazing pressure of the natural zooplankton community. However, Daphnia grazing rates on phytoplankton were not significantly different compared to the grazing rates of the natural zooplankton community. When cyanobacteria were abundant, grazing pressure of Daphnia magnaseemed to be inhibited, and the grazing pressure on bacteria and phytoplankton was similar to that of the natural macrozooplankton community. Our results suggest that biomanipulation may not always result in a more effective top-down control of the algal biomass.     

The seasonal dynamics and distribution of Chaoborus flavicans larvae in adjacent lake basins of different morphometry and degree of eutrophication

1.  Seasonal dynamics, spatial distribution and population size of the phantom midge Chaoborus flavicans in different parts of the eutrophic Lake Hiidenvesi (30.3 km²) were studied.
2.  Density of larvae was low in the shallow, most eutrophic parts of the lake, while the deep Kiihkelyksenselkä basin was inhabited by a dense population. In the deepest part of Kiihkelyksenselkä (33 m) density was 13 989 ± 3542 m^–2 in May, declined to 1102 ± 274 m^–2 in July and recovered to 7225 ± 1314 m^–2 by October. In spring and autumn the majority of larvae were benthic while, during high summer, few larvae were found in the sediment.
3.  Horizontal distribution fluctuated seasonally. On 3 June < 5% of the population inhabited areas shallower than 10 m. On 6 July the limnetic fraction was still restricted to regions deeper than 10 m, but 43% of benthic larvae were found between 6 and 10 m depths. In October both limnetic and benthic larvae were concentrated in areas deeper than 20 m.
4.  Within the lake, distribution was mainly regulated by stratification characteristics, degree of eutrophy being less important. The seasonal horizontal movements were probably induced by food shortage. Larvae could not meet their energetic demands in stratified areas and dispersed to shallower water, reducing predation risk by use of the benthic habitat.     

A multivariate analysis of phytoplankton and food web changes in a shallow biomanipulated lake

1. Phytoplankton dynamics, food chain changes and resilience in Lake Zwemlust, a shallow lake in The Netherlands, are described for the period 1986–94.
2. After biomanipulation in 1987, the lake moved through two alternative states, while the external nutrient loadings were maintained. A clear-water phase, mostly dominated by macrophytes, persisted from 1987 to 1991, and a rather turbid state, dominated by algae, occurred in the summers of 1992–94, after several consecutive and sustained perturbations affecting different parts of the food web in the lake. These two periods were characterized by different community structures.
3. The phytoplankton assemblage gradually changed in a pattern that reverted in later years towards that of the pre-biomanipulation stage, although the same species composition was not regained. This agrees with some mathematical models. During the clear-water phase, nutrient shortage, light climate and zooplankton feeding selected in favour of small, high surface : volume ratio and rapidly reproducing algae. However, in mid-summer of 1992–94, nutrient availability and cladoceran grazing on edible algae favoured cyanophytes.
4. Nutrients were transferred to higher trophic levels or lost from the system at relatively high rates when the lake was in a piscivore–macrophyte-dominated state, while they tended to accumulate in the algae in a planktivore-dominated chain without macrophytes. The role of weed beds was central for nutrient competition (mostly nitrogen) with algae, as well as a refuge and a base for alternative food sources to grazers. Weed beds seemed to have a strong effect in increasing connectedness, resilience and stability of the lake community.
5. The complete return of Zwemlust to a turbid state dominated by phytoplankton seems to have depended upon turnover of the limiting nutrient, which was retarded by macrophytes and stimulated by planktivorous fish and waterfowl.     

Biomanipulation development in Norway

Since 1974 several studies have been carried out in Norway to investigate the interactions between planktivorous fish, zooplankton, phytoplankton and water chemistry. Since 1978 a long-term national research program has been conducted by the Norwegian Council for Scientific and Industrial Research (NTNF). In this program several whole lake manipulations of the fish stocks have been performed to test hypotheses about trophic interactions. It was predicted that manipulations of planktivorous fish populations, might also improve water quality in lakes undergoing eutrophication. Two examples are given to illustrate the achieved results. I: Whole lake fertilization experiment (1974–1978) carried out by Langeland and Reinertsen. The results revealed the importance of top-down effects in the lake ecosystem. When cladocerans dominated, the zooplankton community was able to maintain a more or less constant phytoplankton biomass and a rather low phytoplankton production even when nutrient levels were increased. During years with rotifer dominance, algal biomass and productivity increased, despite the low amounts of added nutrients. II: Experiment performed by Reinertsen, Jensen, Koksvik, Langeland and Olsen in the eutrophic Lake Haugatjern, total elimination of the fish populations by rotenone in late 1980, resulted in a 4-fold decrease in the algal biomass. The species composition changed from the dominance of large-sizedAnabaena flos-aquae andStaurastrum luetkemuelleri to smaller, fastgrowing species and gelatinous green algae. The results are discussed in relation to management of inland waters by combined techniques of biomanipulation and reduced external nutrient supply which increase food-chain efficiency.     

A multivariate analysis of phytoplankton and food web changes in a shallow biomanipulated lake

1. Phytoplankton dynamics, food chain changes and resilience in Lake Zwemlust, a shallow lake in The Netherlands, are described for the period 1986–94.
2. After biomanipulation in 1987, the lake moved through two alternative states, while the external nutrient loadings were maintained. A clear-water phase, mostly dominated by macrophytes, persisted from 1987 to 1991, and a rather turbid state, dominated by algae, occurred in the summers of 1992–94, after several consecutive and sustained perturbations affecting different parts of the food web in the lake. These two periods were characterized by different community structures.
3. The phytoplankton assemblage gradually changed in a pattern that reverted in later years towards that of the pre-biomanipulation stage, although the same species composition was not regained. This agrees with some mathematical models. During the clear-water phase, nutrient shortage, light climate and zooplankton feeding selected in favour of small, high surface : volume ratio and rapidly reproducing algae. However, in mid-summer of 1992–94, nutrient availability and cladoceran grazing on edible algae favoured cyanophytes.
4. Nutrients were transferred to higher trophic levels or lost from the system at relatively high rates when the lake was in a piscivore–macrophyte-dominated state, while they tended to accumulate in the algae in a planktivore-dominated chain without macrophytes. The role of weed beds was central for nutrient competition (mostly nitrogen) with algae, as well as a refuge and a base for alternative food sources to grazers. Weed beds seemed to have a strong effect in increasing connectedness, resilience and stability of the lake community.
5. The complete return of Zwemlust to a turbid state dominated by phytoplankton seems to have depended upon turnover of the limiting nutrient, which was retarded by macrophytes and stimulated by planktivorous fish and waterfowl.