Interactions between fishes and the structure of fish communities in Dutch shallow, eutrophic lakes

E. Lammens

    Research output: Thesisexternal PhD, WU


    This thesis describes the structure of fish communities in Tjeukemeer (21 km 2) and some other surrounding very eutrophic lakes and emphasizes the interactions of the fishes with each other and their food organisms (predation and (exploitative) competition). It is a compilation of seven articles which appeared in international journals in the field of limnology and ecology and it was performed at the Limnological Institute of the Royal Dutch Academy of Sciences. Most data are collected in the period 1979-1985 and some less detailed data from the period 1971- 1978 are used too. Most emphasis is put on the cyprinids, of which bream Abramis brama, roach Rutilus rutilus and white bream Blicca björkna are the most important representatives and of which bream is most successful. In the first years of the study most emphasis was put on bream and gradually the other species were involved in the study in an attempt to understand the success of bream.

    Since these lakes are very turbid (seechi-disc 30-40 cm) and shallow (1.0-2.0 m), trawling is a very successfull fishing method. Moreover since floating and submerged plants are not present the lake is very homogenous in structure. As a consequence sampling zooplankton and chironomids can more easily be done than in a structured environment where zooplankton and macrofauna is strongly clustered. Therefore food organismen could be monitored with relatively little effort simultaneously with fishing.

    Bream is known as a typical benthivorous fish dependent on organisms living in the sediments and only occasionally feeding on zooplankton during very lean periods. Therefore it was a great surprise to find out that in those periods when condition, growth and gonad development were good the diet of bream was largely composed of zooplankton (Lammens 1982). In years when chironomids were the main food item growth stopped, condition deteriorated and gonad development was poor. The choice for zooplankton or chironomids was clearly related with the availability of the first, which can develop huge populations in these shallow hypertrophic lakes. If only dependent on the chironomids the bream population would be much smaller, because of the low chironomid production.

    The availability of zooplankton for bream was related to the size composition and density of particularly Daphnia hyalina. Former studies of Wim van Densen and Koos Vijverberg demonstrated the great impact of smelt Osmerus eperlanus predation on the composition of the Daphnia hyalina population. Because this fish continuously selects the largest individuals within the zooplankton and can retain very small specimens it can affect the composition considerably. However, since the size of the smelt population can vary considerably, the composition of the zooplankton population and its availability for the bream population varies too and competition between smelt and bream is not permanent. The large variation in the smelt population is caused by the great predation pressure on this fish, which allows only one to two yearclasses and by the varying yearly recruitment which is largely dependent on the hydrological input from lake IJsselmeer. The bream population is rather stable as it is composed of about 15 yearclasses, which can buffer the yearly fluctuations in recruitment.

    The varying success in smelt recruitment caused niche shifts in the other fish populations (Lammens et al. 1985). When the smelt population was abundant, the zooplankton was hardly available for the bream population. Smaller bream largely switched to benthic cladocerans and the larger ones to chironomids. The condition of particularly larger bream deteriorated and the chironomid population was overexploited. Consequently eel < 30 cm had to switch to fish and its condition deteriorated too, as the loss of chironomids could not be compensated for by the prey-fish smelt. The much more efficient piscivorous larger eel and pike-perch showed a good condition on a diet of mainly smelt. When the smelt recruitment was poor, zooplankton, was available for bream, and chironomids for small eel, whereas the larger eel switched largely to molluscs and pike-perch to small cyprinids and percids.

    Bream in the length range 30-40 cm with a mouth opening of 20-30 mm consumes rather efficiently zooplankton in the length range 1-2 mm. It seemed not very probable that these particles are consumed one by one, but some sort of filter-feeding was hypothesized. Relating the gut contents of different size classes of bream to the size composition and density of Daphnia hyalina and Bosmina coregoni it could be inferred that this mechanism is very probable (Lammens 1985). Three conditons had to be fulfilled: 1. since the openings between the gill-rakers increase with increasing fish length, the average length of the food organisms has to increase, and their standard deviation has to decrease in relation to the length of the fish, 2. the average length and standard deviation of the food organisms in the gut are positively related with those in the lake, 3. the average length and standard deviation of consumed zooplankton is not related with the density of food organisms in the lake. The mechanism of filter-feeding could only be rejected for the small bream feeding on Daphnia hyalina. Length classes smaller than 15-20 cm probably combine filter and particulate feeding or only act as particulate feeders. About 70% of the total variation in average length and standard deviation could be explained by these variables. The rest variation may probably be explained by some regulation of the distance between the gill-arches during filter feeding.

    Bream, roach and white bream showed a clear difference in diet and efficiency of feeding, which could be related to the functional morphology of the head (Lammens 1984, Lammmens et al. 1986). Comparing similar length classes the efficiency for zooplankton is highest for bream>10 cm and almost the same for roach and white bream>10 cm. This was clearly related to the structure of the gill-raker basket, which shows the smallest openings in bream. However, this fine structure of the basket in bream obstructs in the small specimens the easy passage of bottom substrate particles when feeding on chironomid larvae in the sediments. Benthivorous feeding is therefore more efficiently done by white bream comparing similar length classes. However, bream is more efficient on the population level since it reaches a much larger size than white bream and can dig much deeper in the sediments. Roach has the lowest efficiency for benthivorous feeding, which is not related to its gill-raker basket but to the very limited protrusibility of the upper jaw in contrast to bream and white bream. The latter two can protrude the upper jaw downward and forage more easily than roach. The latter is more specialized in hard to crush food such as snails, mussels and vegetation, which can easily be handled by the strong pharyngeal teeth. Only the largest bream can handle some mussels, as can white bream at a somewhat smaller length, but only roach can handle snails.

    However, apart from the feeding efficiency of each of these species, the predation risks of each species is also important in structuring the fish community. These three fish species are on their turn prey-fish for predatory-fish and have to evaluate their predation-risks when foraging. The most important predators are pike-perch and eel>30 cm, whereas perch and pike have a much too low density to have any impact on the prey-populations. Pike-perch is a typical light-avoiding open water fish in contrast to pike which is a typical representative of clear, shallow and vegetation rich waterbodies. Prey-fishes are in the first place fishes < 20 cm living in the open water area, that is, smelt and small pike-perch, whereas ruffe, perch and all small cyprinids are born in the littoral region and migrate gradually to the open water area. Comparative study of situations with and without a pike-perch gill-net fishery showed a difference in composition of the prey populations and of zooplankton. (Lammens et al. in prep.). The size of the smelt, perch, ruffe and roach population was inversely related to the pike-perch population, whereas that of bream, in particular the larger specimens, was positively correlated to the pike-perch population and white bream was intermediate showing hardly any change. So predation seemed to stimulate the (for predation invulnerable) bream population as the competion pressure with the small prey-fish was largely reduced. Therefore a very eutrophic condition is profitable for bream from both the competitive and predation point of view.

    In eutrophic conditions with relatively large amounts of zooplankton and chironomids bream seems best equipped to cope the food conditions. If only limited amounts of snails, molluscs and vegetation are present, the populations of roach and white bream will only be composed of small specimens. In less eutrophic systems with clear water, much vegetation, little zooplankton and relatively much macrofauna the role of bream will be more limited. Here alert visual feeders such as roach and perch are much more important, moreover the vegetation offers sufficient refugia for these small fishes.

    Original languageDutch
    QualificationDoctor of Philosophy
    Awarding Institution
    • Osse, J.W.M., Promotor
    • van Densen, W.L.T., Promotor, External person
    Award date31 Oct 1986
    Place of PublicationWageningen
    Publication statusPublished - 1986


    • fishes
    • biocoenosis
    • lakes
    • netherlands
    • animal communities

    Cite this