Fish–duck interactions in boreal lakes in Finland as reflected by abundance correlations

We studied the hypothesis that fish play an important role in lake use by ducks (pairs and broods) in boreal lakes. The study was based on densities of different duck and fish species in 28 boreal lakes in southern Finland. We focused on the three most common duck species (mallard Anas platyrhynchos, green-winged teal A. crecca and common goldeneye Bucephala clangula) and on the three most common fish species (perch Perca fluviatilis, roach Rutilus rutilus and pike Esox lucius) in the region. We considered both competitive and predatory interactions between ducks and fish, the perch and roach being potential competitors with ducks and the pike a potential predator of ducks. We found a negative association between green-winged teal brood density and total fish density, the other duck species having only a weak association with total fish density. When the three fish species were considered separately, a negative association, suggesting food competition, was found between perch, green-winged teal and goldeneye, whereas the role of roach as a food competitor seemed to be of minor importance. We did not find any clear signs of predatory effects of pike on ducks. Our results suggest that food competition is a more important factor than pike predation in affecting lake use by ducks in oligotrophic boreal environments in southern Finland.     

Trophic dynamics in a whole lake experiment: size-structured interactions and recruitment variation

Horizontal and vertical heterogeneity as a result of size‐structured processes are important factors influencing indirect effects in food webs. In a whole‐lake experiment covering 5 years, we added the intermediate consumer roach (Rutilus rutilus) to two out of four lakes previously inhabited by the omnivorous top predator perch (Perca fluviatilis). We focused our study on the direct consumption effect of roach presence on zooplankton (and indirectly phytoplankton) versus the indirect effect of roach on zooplankton (and phytoplankton) mediated via effects on perch reproductive performance. The patterns in zooplankton and phytoplankton abundances were examined in relation to population density of roach and perch including young‐of‐the‐year (YOY) perch in the light of non‐equilibrium dynamics. The presence of roach resulted in changed seasonal dynamics of zooplankton with generally lower biomasses in May–June and higher biomasses in July–August in roach lakes compared to control lakes. Roach presence affected perch recruitment negatively and densities of YOY perch were on average higher in control lakes than in treatment lakes. In years when perch recruitment did not differ between lakes as a result of experimental addition of perch eggs, total zooplankton biomass was lower in treatment lakes than in control lakes. Phytoplankton biomass showed a tendency to increase in roach lakes compared to control lakes. Within treatment variation in response variables was related to differences in lake morphometry in treatment lakes. Analyses of the trophic dynamics of each lake separately showed strong cascading effects of both roach and YOY perch abundance on zooplankton and phytoplankton dynamics. Consideration of the long transients in the dynamics of top predators (fish) in aquatic systems that are related to their long life span involving ontogenetic niche shifts is essential for making relevant interpretations of experimental perturbations. This conclusion is further reinforced by the circumstance that the intrinsic dynamics of fish populations may in many cases involve high amplitude dynamics with long time lags.     

Development of fish communities in lakes after biomanipulation

Biomanipulation measures in the Netherlands are usually a combination of a drastic fish stock reduction and an introduction of pike fingerlings. In three small shallow lakes (Noorddiep, Bleiswijkse Zoom and Zwemlust) these measures resulted in a clear water state and the development of macrophytes. After the measures the fish community developed differently because of the new physical and biological conditions. Results of lake Noorddiep and lake Bleiswijkse Zoom showed that the fish community became more divers. Bream and carp became less dominant and were partly replaced by roach and perch. The importance of the main predator pike-perch was strongly reduced and replaced by pike and perch. The share of piscivorous fish in the total fish stock increased at all sites. The recruitment of young-of-the-year was similar or even higher in the clear overgrown areas than in the turbid water before the measures, but the recruitment of young-of-the-year to older fish 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 from bream to roach, because of selective predation on bream in the open water, while roach was hiding in the vegetation. The macrophytes provide new refugia and feeding conditions that favour roach and perch, but offer relatively poor survival conditions for bream and carp.     

Pike predators induce morphological changes in young perch and roach

The morphological response of two 1 year-old prey species, perch Perca fluviatilis and roach Rutilus rutilus to the presence of predatory pike Esox lucius , was tested in a 6 week aquarium experiment. The growth of perch was higher than that of roach but there was no difference in growth of prey between predator and control treatments. The presence of pike caused morphological changes in both perch and roach. Perch showed an increase in body depth during the experiment whereas the morphological response of roach to pike was more related to a displacement of the dorsal and pelvic fins and width of the anal fin. Differences in morphological responses of perch and roach to pike can be interpreted by their overall differences in antipredator strategies.     

Substitution of top predators: effects of pike invasion in a subarctic lake

1. Invasions of top predators may have strong cascading effects in ecosystems affecting both prey species abundance and lower trophic levels. A recently discussed factor that may enhance species invasion is climate change and in this context, we studied the effects of an invasion of northern pike into a subarctic lake ecosystem formerly inhabited by the native top predator Arctic char and its prey fish, ninespined stickleback. 2. Our study demonstrated a strong change in fish community composition from a system with Arctic char as top predator and high densities of sticklebacks to a system with northern pike as top predator and very low densities of sticklebacks. A combination of both predation and competition from pike is the likely cause of the extinction of char. 3. The change in top predator species also cascaded down to primary consumers as both zooplankton and predator‐sensitive macroinvertebrates increased in abundance. 4. Although the pike invasion coincided with increasing summer temperatures in the study area we have no conclusive evidence that the temperature increase is the causal mechanism behind the pike invasion. But still, our study provides possible effects of future pike invasions in mountain lakes related to climate change. We suggest that future pike invasions will have strong effects in lake ecosystems, both by replacing native top consumers and through cascading effects on lower trophic levels.     

Effects of brown and turbid water on piscivore–prey fish interactions along a visibility gradient

1. Environmental changes such as eutrophication and increasing inputs of humic matter (brownification) may have strong effects on predator–prey interactions in lakes through a reduction in the visual conditions affecting foraging behaviour of visually oriented predators. 2. In this experiment, we studied the effects of visual range (25–200 cm) in combination with optically deteriorating treatments (algae, clay or brown humic water) on predator–prey interactions between pike (Esox lucius) and roach (Rutilus rutilus). We measured effects on reaction distance and strike distance for pike and escape distance for roach, when pike individuals were exposed to free‐swimming roach as well as to roach held in a glass cylinder. 3. We found that reaction distance decreased with decreasing visual range caused by increasing levels of algae, clay or humic matter. The effect of reaction distance was stronger in turbid water (clay, algae) than in the brown water treatment. 4. Strike distance was neither affected by visual range nor by optical treatment, but we found shorter strike distances when pike attacked roach using visual cues only (roach held in a cylinder) compared to when pike could use multiple senses (free‐swimming roach). Escape distance for roach was longer in turbid than in brown water treatments. 5. Changes in environmental drivers, such as eutrophication and brownification, affecting the optical climate should thus have consequences for the strength of predator–prey interactions through changes in piscivore foraging efficiency and prey escape behaviour. This in turn may affect lake ecosystems through higher‐order interactions.     

Changes in the fish community and water quality during seven years of stocking piscivorous fish in a shallow lake

SUMMARY 1. Piscivores (annual stocking of 1000 individuals ha^?1 of 0+ pike and a single stocking of 30 kg ha^?1 of large 20–30 cm perch) were stocked in seven consecutive years in a shallow eutrophic lake in Denmark. The stocking programme aimed at changing food‐web structure by reducing zooplanktivorous and benthivorous fish, with resultant effects on lower trophic levels and ultimately water quality. 2. The fish community and water quality parameters (Secchi depth, concentrations of total phosphorus, chlorophyll a and suspended solids) were monitored between 1996 and 2000 and relationships were evaluated between predatory fish and potential prey and between zooplanktivorous or benthivorous fish and water quality parameters. In addition, potential consumption of piscivorous fishes was calculated. 3. The density of fish feeding on larger zooplankton or benthos (roach >15 cm, crucian carp >15 cm) declined distinctly during the study period. This effect was attributed to predation by large (>50 cm) pike. Based on scale readings, we cautiously suggest that the stocking of 0+ pike boosted the adult pike population to produce an unexpected impact in later years. Conversely, no direct impact of stocked 0+ pike was detected on 0+ roach. 4. A major decline in the recruitment strength of 0+ roach was observed in 2000. A combination of (i) the indirect effect of large pike preying on adult roach, with negative effect on roach reproduction and (ii) the direct predation effect of 0+ pike and/or 1+ and 2+ perch recruited in the lake, provides the most likely explanation of this phenomenon. 5. A marked increase in Secchi depth in 2000 and declining trends in suspended solids, chlorophyll‐a and total phosphorus concentrations were observed. These changes may also be attributable to changes in the fish community, although the relationships were not straightforward. 6. This 7‐year study indicates that piscivorous fish may be a significant structuring force in shallow eutrophic lakes, suggesting that stocking piscivores can increase predation pressure on cyprinids. However, the general lack of impact of 0+ pike points to the need of refining current stocking practices in several countries across Europe.     

Chernobyl-derived radiocaesium in fish as dependent on water quality and lake morphometry

¹³⁷Cs concentrations in perch Perca fluviatilis , pike Esox lucius and roach Rutilus rutilus obtained from lakes of different size and water quality in an area which received about 10–67 kBq m⁻²¹³⁷Cs, were compared with environmental data. Radiocaesium concentrations were highest in pike, and were about two to three times higher in the pike and perch than in the roach. The largest perch had about four times more ¹³⁷Cs than the smallest ones, but the activities in the pike and roach were independent of fish size. All of the water quality parameters examined correlated with ¹³⁷Cs concentrations in the fish, but the concentrations of ¹³⁷Cs in the bottom sediment did not. ^l37Cs fallout and water retention time in lakes showed a positive correlation with fish ¹³⁷Cs, while the size of the catchment area, phosphorus content of the water, water colour, pH and electrical conductivity showed a negative one. Thus there was more ¹³⁷Cs in the fish from the oligotrophic lakes than from the eutrophic ones, from the less humic lakes than from the more humic ones, from the more acid than from the less acid ones and from the lakes containing less potassium than from those containing more potassium. The depth, area and volume of the lake did not affect radiocaesium in the fish. Regression equations are presented for predicting ¹³⁷Cs in fish on the basis of environmental data. Fallout, total phosphorus and pH together had the highest influence in multiple regression models but were less significant in the roach than in the perch and pike.     

The distribution and growth of coarse fish in gravel-pit lakes in south-east England

SUMMARY. Age and growth rates of the coarse fish of thirty-nine gravel-pit lakes in south-east England are compared. Species composition was very variable, the commonest species being roach ( Rutilus rutilus ), bream ( Abramis brama ), tench ( Tinca tinca ), perch ( Perca fluviatilis ) and pike ( Esox lucius ). Two main kinds of fish community were identifiable from the survey: (1) those consisting of a mixture of cyprinids, usually roach and bream; and (2) those in which cyprinids were rare or absent and the fish community was composed of perch and pike. The majority of the lakes fell into the first category. Comparison of growth curves derived by back-calculation showed that, in all species, variation occurred from lake to lake and from year to year. The majority of gravel-pit lake fishes had 'medium' growth rates when compared with published data by other workers on different waters. Seasonal growth of roach infected with plerocercoids of Ligula intestinalis was observed at monthly intervals. Annulus formation on the scales of these fish occurred at the onset of growth in May. The heavy Ligula burden in 89% of these fish had not caused the formation of any additional or false rings.     

The influence of the invasive black bullhead Ameiurus melas on the predatory efficiency of pike Esox lucius L.

The influence of the invasive black bullhead Ameiurus melas on the predatory efficiency of the pike Esox lucius was investigated using an additive experimental design. Pike predatory success on 0+ years roach Rutilus rutilus was significantly reduced in the presence of black bullhead. Among the different hypotheses that may explain such a pattern, the hypothesis of direct competition between pike and black bullhead was not verified, as black bullhead hardly fed on roach. Similarly, pike predatory efficiency did not decrease with turbidity, rejecting therefore the hypothesis of an indirect effect through black bullhead-generated turbidity. Therefore, the reduced predatory efficiency of pike was probably related to behavioural interference between pike and black bullhead. These laboratory results confirm the potential negative impact of black bullhead on native European fauna, with a particular emphasis on pike, which is a top predator considered as vulnerable in some European regions.     

Environmental correlates of population variables of perch (<Emphasis Type="Italic">Perca fluviatilis</Emphasis>) in boreal lakes

We examined relationships among perch population variable parameters in two types of lakes, lakes with perch (P-lakes, n?=?15) and lakes with perch and roach (PR-lakes, n?=?10) using redundancy analysis (RDA) to relate population variables to environmental factors. Effects from environmental factors were tested for significance by means of permutation tests (Monte Carlo). Three factors, pH, altitude and fraction of roach (by number) in the gill net catches, explaining 47.9% of the variation, had significant effects on perch population variables. The significance of pH was improved by partialing out the effect of conductivity and roach. Similarly, the significance of altitude was improved by partialing out the effect of pH and roach, and the significance of roach was improved by partialing out the effect of pH and altitude. When the fraction of pike in the catch was included in the analysis, the effect of roach was not significant and vice-versa, as roach and pike fractions were correlated with each other. The effect of pike was significant when roach was not included, but the effect was not as strong as the effect of roach. A biplot was constructed by plotting population variables on the first and second RDA axis, with arrows showing five selected environmental factors. Growth of 3+ to 5+ perch was positively related to pH and altitude, perch catch per unit effort was negatively related to pH and altitude, and age specific perch weight was negatively related to fraction of roach. The relationship between growth of 2+ perch and pH was not as strong as the relationship between pH and the growth of older perch. Moreover, the growth of 2+ perch was negatively related to the fraction of roach, probably due to competition between young zooplankton feeding perch and roach.     

Piscivore efficiency and refuging prey: the importance of predator search mode

In predator-prey interactions, the efficiency of the predator is dependent on characteristics of both the predator and the prey, as well as the structure of the environment. In a field enclosure experiment, we tested the effects of a prey refuge on predator search mode, predator efficiency and prey behaviour. Replicated enclosures containing young of the year (0+) and 1-year-old (1+) perch were stocked with 3 differentially sized individuals of either of 2 piscivorous species, perch (Perca fluviatilis), pike (Esox lucius) or no piscivorous predators. Each enclosure contained an open predator area with three small vegetation patches, and a vegetated absolute refuge for the prey. We quantified the behaviour of the predators and the prey simultaneously, and at the end of the experiment the growth of the predators and the mortality and habitat use of the prey were estimated. The activity mode of both predator species was stationary. Perch stayed in pairs in the vegetation patches whereas pike remained solitary and occupied the corners of the enclosure. The largest pike individuals stayed closest to the prey refuge whereas the smallest individuals stayed farthest away from the prey refuge, indicating size-dependent interference among pike. Both size classes of prey showed stronger behavioural responses to pike than to perch with respect to refuge use, distance from refuge and distance to the nearest predator. Prey mortality was higher in the presence of pike than in the presence of perch. Predators decreased in body mass in all treatments, and perch showed a relatively stronger decrease in body mass than pike during the experiment. Growth differences of perch and pike, and mortality differences of prey caused by predation, can be explained by predator morphology, predator attack efficiency and social versus interference behaviour of the predators. These considerations suggest that pike are more efficient piscivores around prey refuges such as the littoral zones of lakes, whereas perch have previously been observed to be more efficient in open areas, such as in the pelagic zones of lakes.     

Crayfish in lakes and streams: individual and population responses to predation, productivity and substratum availability

1. In a correlative study, we investigated the relative importance of fish predation, refuge availability and resource supply in determining the abundance and size distributions of the introduced and omnivorous signal crayfish (Pacifastacus leniusculus) in lakes and streams. Moreover, the biomass and food selection of predatory fish was estimated in each habitat type and stable isotopes of carbon and nitrogen were measured in perch (Perca fluviatilis), the dominant predator in the lakes, and in its potential food sources (crayfish, juvenile roach and isopods). 2. In lakes, crayfish were the most frequent prey in large perch (46%), followed by other macroinvertebrates (26%, including the isopod Asellus aquaticus) and small fish (25%). Crayfish and fish dominated the gut contents of large perch with respect to biomass. Nitrogen signatures showed that perch were one trophic level above crayfish (approx. 3.4‰) and a two‐source mixing model using nitrogen isotope values indicated that crayfish (81%) contributed significantly more to perch isotope values than did juvenile roach (19%). A positive correlation was found between the abundance of crayfish and the biomass of large perch. Crayfish abundance in lakes was also positively correlated with the proportion of cobbles in the littoral zone. Lake productivity (chlorophyll a) was positively correlated with crayfish size, but not with crayfish abundance. 3. In streams, brown trout (Salmo trutta) were the most abundant predatory fish. Gut contents of large trout in a forested stream showed that terrestrial insects were the most frequently found prey (60%), followed by small crayfish (27%) and isopods (27%). In contrast to lakes, the relative abundance of crayfish was negatively correlated with the total biomass of predatory fish and with total biomass of trout. However, abundance of crayfish at sites with a low biomass of predatory fish varied considerably and was related to substratum grain size, with fewer crayfish being caught when the substratum was sandy or dominated by large boulders. The mean size of crayfish was greater at stream sites with a high standing stock of periphyton, but neither predator biomass nor substratum grain size was correlated with crayfish size. 4. Our results suggest that bottom‐up processes influence crayfish size in lakes and streams independent of predator biomass and substratum availability. However, bottom‐up processes do not influence crayfish abundance. Instead, substratum availability (lakes) and interactions between predation and substratum grain size (streams) need to be considered in order to predict crayfish abundance.     

Prey vulnerability to a gape-size limited predator: behavioural and morphological impacts on northern pike piscivory

The northern pike ( Esox lucius ) is a selective and important predator in lake ecosystems. Prey size in pike is limited by pike gape size, which is a linear function of pike body length. Here we show that the absolute gape-size limit in pike is greater than previously considered, and that maximum ingestible prey size is limited by prey body depth. Further, we experimentally show that pike prefer shallow-bodied roach before deeper-bodied common bream, and small prey sizes within each prey species. Handling time in pike increases with prey body depth, and since common bream are deeper-bodied than roach, handling time is longer for bream than for roach of the same length, but equal considering body depth. Prey handling time is suggested to be a major cost to the pike, since it increases the risk of losing the prey, as well as exposure to predation, kleptoparasitism and cannibalism. Consequently, prey vulnerability is determined by risk of predation and intraspecific interactions, and behavioural preferences in the pike, and not by pike gape-size limits. The consequences for natural populations is evaluated by analysing size structures of predator and prey fish populations in a eutrophic lake.     

The interactions of abiotic and biotic factors influencing perch Perca fluviatilis and roach Rutilus rutilus populations in small acidified boreal lakes

Four small, acidified boreal lakes, all sustaining populations of perch Perca fluviatilis, roach Rutilus rutilus and pike Esox lucius, were studied in four successive years. Three lakes were moderately acidified (mean pH of 5·61-5·83), while the fourth was more acidic (mean pH of 5·16) and had a sparse population of R. rutilus. Perca fluviatilis density was higher in this lake (1004 ha(-1)) than in the other three (355-717 ha(-1)), where R. rutilus dominated in terms of numbers (981-2185 ha(-1)). Large, potentially predatory, P. fluviatilis were most abundant in the lake with clearest water, and these seemed to have a negative effect on P. fluviatilis density. Perca fluviatilis mean mass was negatively correlated with R. rutilus biomass and was highest in the most acidic lake with the sparse R. rutilus and the highest P. fluviatilis density. Perca fluviatilis mass correlated positively with pH in two lakes (with the highest fish biomass), suggesting that low pH affected P. fluviatilis mass negatively. Perca fluviatilis growth correlated positively with summer (July to August) air temperature in the lake with sparse R. rutilus, thus differing from P. fluviatilis and R. rutilus growth in the other three lakes. The mean age of P. fluviatilis was generally lower than that of R. rutilus and was lowest in the two lakes with the highest fish biomass, indicating that adult mortality was affected by density-induced factors.     

Diet-dependent body morphology and ontogenetic reaction norms in Eurasian perch

Individual morphology and performance are directly or indirectly under the influence of variation in resource levels. To study the effects of different resource conditions and their effects on morphology and ontogenetic reaction norms in young‐of‐the‐year (YOY) perch (Perca fluviatilis), we used three different approaches. First, we examined the morphological trajectories over early ontogeny in relation to lake‐specific resources in a field study. Second, one lake that lacked perch recruitment was stocked with perch eggs from a control lake in a whole‐lake experiment to study ontogenetic reaction norms. Third, we compared the development of YOY perch in the three lakes that mainly inhabited the littoral zone with YOY perch experimentally confined to enclosures in the pelagic zone of the lakes.
Overall body morphology of the YOY perch changed both as a function of size and as a function of diet. As perch increased in size they developed a deeper body morphology corresponding to an increased proportion of benthic macroinvertebrates in their diet. In pelagic enclosures where perch were constrained to feed mainly on zooplankton they had a more fusiform body morphology than perch in the lakes that fed on a mixture of zooplankton and macroinvertebrates. Similarly, the ontogenetic reaction norm of perch was related to the diet and lake‐specific zooplankton levels in the whole‐lake experiment.
In the pelagic enclosures, perch with high growth rates had a more fusiform body morphology than slow‐growing perch, whereas the opposite was found in the lakes, where perch included more macroinvertebrates in their diets. Perch in lakes with a higher proportion of macroinvertebrates in the diet also had deeper body morphology. The opposite morphology – growth rate relationship found between perch in the pelagic versus those using the whole lake suggest a morphological trade‐off between foraging on zooplankton and foraging on macroinvertebrates. Our results suggest that YOY perch show different ontogenetic reaction norms as a function of lake‐specific resource levels, which may allow YOY recruitment to later stages. Our results further suggest that diet‐related changes in morphology are a rapid process.     

Piscivory of 0+ pike (Esox lucius L.) in a small eutrophic lake and its implication for biomanipulation

Stocking 0+ pike in shallow eutrophic lakes as an indirect way to facilitate clear water by increased phytoplankton grazing by zooplankton, has been used in several studies with varying success. Lack of piscivory could be one reason for poor impact of the pike stockings, on reducing the 0+ cyprinid fish population. In relation to this, we present data on food and feeding habits of 936 0+ pike (Esox lucius L.) (2–18 cm) sampled during five years in a small eutrophic lake undergoing biomanipulation. Overall, the food of 0+ pike followed a sequence of diet shifts including microcrustacea, insects, macrocrustacea and finally vertebrates (fish) with increase in pike length. Despite a high abundance of potential prey fish in the lake, in general less than 50% of the pike were piscivorous until a length of 10.1–13.7 cm. As piscivory increased, the principal prey fish (0+ roach) apparently displayed antipredatorial behaviour which decreased their availability for 0+ pike. We conclude that lack of piscivory caused by e.g. anti predator tactics by prey fish, may reduce the impact of 0+ pike stockings in lake restoration projects.

     

Piscivore-Prey Fish Interactions: Mechanisms behind Diurnal Patterns in Prey Selectivity in Brown and Clear Water

Environmental change may affect predator-prey interactions in lakes through deterioration of visual conditions affecting foraging success of visually oriented predators. Environmental change in lakes includes an increase in humic matter causing browner water and reduced visibility, affecting the behavioural performance of both piscivores and prey. We studied diurnal patterns of prey selection in piscivorous pikeperch (Sander lucioperca) in both field and laboratory investigations. In the field we estimated prey selectivity and prey availability during day and night in a clear and a brown water lake. Further, prey selectivity during day and night conditions was studied in the laboratory where we manipulated optical conditions (humic matter content) of the water. Here, we also studied the behaviours of piscivores and prey, focusing on foraging-cycle stages such as number of interests and attacks by the pikeperch as well as the escape distance of the prey fish species. Analyses of gut contents from the field study showed that pikeperch selected perch (Perca fluviatilis) over roach (Rutilus rutilus) prey in both lakes during the day, but changed selectivity towards roach in both lakes at night. These results were corroborated in the selectivity experiments along a brown-water gradient in day and night light conditions. However, a change in selectivity from perch to roach was observed when the optical condition was heavily degraded, from either brown-stained water or light intensity. At longer visual ranges, roach initiated escape at distances greater than pikeperch attack distances, whereas perch stayed inactive making pikeperch approach and attack at the closest range possible. Roach anti-predatory behaviour decreased in deteriorated visual conditions, altering selectivity patterns. Our results highlight the importance of investigating both predator and prey responses to visibility conditions in order to understand the effects of degrading optical conditions on piscivore-prey interaction strength and thereby ecosystem responses to brownification of waters.     

Growth in length and in body depth in young‐of‐the‐year perch with different predation risk

Body shape of young‐of‐the‐year (YOY) perch Perca fluviatilis , and number and size of potential predators (perch and pike Esox lucius ) were compared across five lakes in northern Sweden. Body depth and dorsal fin ray length of YOY perch differed between lakes, with high relative body depth and long dorsal fin rays found in the lakes where the number of large piscivores was the highest. The most slender fish were found in the lake where the perch population had the smallest individuals and pike occurred in very low numbers. The average body mass in YOY perch from the two lakes with highest and lowest body depth, respectively, were the same, which indicates a difference between lakes in the relation between growth in length and in depth. Both body depth and fin ray length were correlated with predation risk by pike. Fin ray length was also correlated with number of piscivorous perch. Selection for different body shapes can be caused by different biotic and abiotic factors, singly or in combination, and the results from this study indicates that predation risk is one of these factors affecting body depth and fin ray length in perch.