Habitat use and life history of brown trout (Salmo trutta) and Arctic charr (Salvelinus alpinus) in some low acidity lakes in central Norway

Habitat utilization and the life history of browntrout Salmo trutta and Arctic charr Salvelinus alpinus were investigated in fivesympatric populations and five allopatric brown troutpopulations in Høylandet catchment, a atmosphaericlow deposition area in Mid Norway. There was asignificant inverse correlation in abundance ofepibenthic Arctic charr and brown trout in theselakes, indicating that the latter species is dominant.The largest numbers of sympatric brown trout andArctic charr were caught in epibenthic habitat. In twolakes, brown trout to some extent also occurredpelagically, while pelagic individuals of Arctic charrwere found in all five lakes. The main food items forboth epibenthic and pelagic brown trout wereterrestrial surface insects and chironomid pupae.Zooplankton was the primary food item for Arctic charrin both habitats. Although the age distribution wasvery different in the populations, neither speciesseem to suffer from recruitment failure. There was nosignificant difference in survival rates betweensympatric populations of brown trout and Arctic charr.We found a significant inverse correlation betweenepibenthic catches of brown trout and the mean weightof 4+ fish, the most abundant age group. However, ifusing weight data for three-year-old fish, no suchrelationship was found for Arctic charr. Brown troutand Arctic charr reached asymptotic lengths of197–364 mm and 259–321 mm, respectively. Both speciestypically reached sexual maturity at age 2–3, and nomaturation-induced mortality was evident. We concludethat fish populations in Høylandet lakes areregulated throughout their lifes by inter- andintraspecific competition.     

August diet of age-0 Pacific halibut in nearshore waters of Kodiak Island,Alaska

Synopsis The objective of this study was to describe the diet of age-0 Pacific halibut,Hippoglossus stenolepis, for the inshore waters of Kodiak Island, Alaska during August 1991. Stomach contents were identified from 170 age-0 halibut captured inshore of the eastern and southern coasts of Kodiak Island, and were analyzed in relation to halibut size, location, depth and substrate of capture. One hundred sixty-eight of 170 fish had eaten crustaceans, of which the predominant prey taxa were Mysidacea (34.3%), Cumacea (33.1%), Gammaridea (26.6%) and Caridea (3.9%). In five of six capture locations, mysids and amphipods were predominant prey. In the remaining area, Sitkinak Strait, cumaceans were the primary food source. At depths less than 10 m, mysids were the predominant prey taxa. Gammarid amphipods were of primary importance at depths of 10–30 m. Halibut captured from 30–70 m fed mainly on cumaceans. Cumaceans and gammarid amphipods were consumed by halibut caught on gravel substrate. Fish caught on substrates of sand and mud fed mainly on mysids and amphipods. Cumaceans were also consumed on sandy substrates. Fish 45 mm fed on cumaceans. An ontogenetically related shift in diet occurred at 46–55 mm TL, at which size the halibut's primary prey began to shift from cumaceans to mysids. Fish of 46–75 mm consumed increasingly greater proportions of mysids, amphipods and shrimps. The diet of age-0 Pacific halibut along the Kodiak coast during August was related to predator size, and location, depth, and substrate type of capture.     

Temporal variations in the diet of brown trout (Salmo trutta L.) and rainbow trout (S. gairdneri R.) in Rutland Water

The diet of brown trout (Salmo trutta L.) and rainbow trout (S. gairdneri Richardson) in Rutland Water were compared during the first two fishing seasons (April–October 1977 and 1978).Fortnightly samples of approximately forty stomachs were obtained from boat and bank, rod-and-line caught trout giving a total of 1046 stomachs over the two seasons.During 1977 seasonal changes in the diet were divided into two phases; the first being a period of abundant drowned terrestrial food until June. This was followed by a period of more stable water level from July onwards when chironomid larvae and pupae were consistently the most important food items and the diversity of food also increased.In 1978 the proportion of chironomid pupae and larvae declined and they were replaced in the diet by Gammarus and Asellus.     

Use of space and food by resident and migrant brown trout,Salmo trutta

Synopsis Parr and resident forms of brown trout,Salmo trutta, from Vangsvatnet Lake, Norway live in freshwater, while migrant forms live in coastal waters during summer and in freshwater during winter. About 80% of parr and residents live at depths <5 m, smolts and migrants are more confined to near-surface water. Brown trout partly segregate by size, age and sex from spring through autumn. More than 90% of parr and residents in the tributaries are 0–2 years old, 2–14 cm in length, in the littoral zone 0–3 years old, 7–24 cm in length, and in the pelagic zone 2–6 years old, 18–32 cm in length. The mean body length of equal-aged fish increases from tributaries through littoral to pelagic zones in the lake. Smolts (2–7 years, 14–29 cm) leave the lake from April through August and return during September–October. Migrants (2–11 years, 23–67 cm) leave the take in April–May and return during August–September; sexually mature fish return earlier than immatures. Female brown trout are less stream-dwelling, but more migrant and pelagic than males. Most individuals in the lake population spend the winter in the littoral zone. In the tributaries, diet differs significantly between age-groups of parr; young fish feed on smaller food items than do older fish. In the lake, parr and residents living in the same habitats feed on the same food items. Littoral brown trout feed mainly on insect larvae and chironomid pupae, pelagic brown trout feed on zooplankton and surface insects. Migrants feed little while staying in freshwater, except for matures which feed on young salmonids and surface arthropods during the 2 first months after they had returned from coastal waters. The results are discussed in relation to growth possibilities and mortality risks of the different habitats.Reprint request to B. Jonsson.     

Feeding by Arizona trout (Salmo apache) and brown trout (Salmo trutta) at different light intensities

Synopsis Brown trout (Salmo trutta) were more efficient than Arizona trout (Salmo apache) in eating brine shrimp at starlight (10^–4 fL) light levels. Arizona trout required light levels moonlight (10^–3 fL) to feed. In bright light (50 fL), brown trout utilized cover to a much greater extent in both field and laboratory. Our study indicates that factors other than competition for food or habitat are probably causing the displacement of Arizona trout by brown trout when browns are stocked into the native habitat of Arizona trout.     

The Relevance of Individual Size to Management of Arctic Charr, Salvelinus Alpinus, Populations

Arctic charr, Salvelinus alpinus, tend to form stunted populations presumably due to competition for limited resources. In this context a long-term intensive fishery programme aiming at reducing charr density, and thereby increasing growth and harvestable sizes, was initiated in the early 1980s in a Norwegian alpine lake. Here we present long-term data on catch statistics and changes in mean weights of charr caught with gill nets on spawning grounds, as well as changes in mean weight of juvenile charr caught with funnel traps during the 1990s. Furthermore, we present results from shorter-term studies on growth, size- and age-distribution, and size-related habitat and resource utilisation of Arctic charr and brown trout, Salmo trutta, in this lake. Mean weight (± SD) of charr caught on the spawning grounds increased significantly from 129.8g (±11.9) in the years 1982–1990 to 213.1g (±37.8) in the years 1996–1999, whereas catch per unit of effort decreased significantly. Mean weight of juveniles caught in funnel traps increased significantly from 21.2g (±6.9) in 1993 to 41.9g (±14.8) in 1999. Apparently the increase in weight of spawning charr coincided with the onset of trap fishing for juvenile charr. Compared to generally shallow dwelling trout, charr grew rapidly and were generally found in deeper areas of the lake. Charr went through a distinct size-related niche shift from mainly consuming small zooplankton in the pelagic to consuming large benthic prey in shallower waters. Resource and habitat utilisation in different size-groups of charr and trout are discussed with respect to possible competitive and predatory intra- and interspecific interactions, and with regard to management of charr populations.     

On the age and growth of flounder Paralichthys orbignyanus (Jenyns, 1842) in Bahía Blanca Estuary, Argentina

Estimates of age and growth of flounder Paralichthys orbignyanus (Jenyns,1842) were made by analysing fish from commercial catches in Bahía Blanca estuary (39° LS). A total of 823 fish caught by fishing fleets operating in the estuary were collected between February 1997 and January 1998. Age was determined via scale reading and growth estimation parameters using von Bertalanffys equation. Annulus formation occurred in August. The age for the total population, which ranged from 7.7 to 87.5 cm TL was 0–7. Age ranged from 0 to 6 years (21.5–71.9 cm TL) in males and from 0 to 7 years (24.9–87.5 cm TL) in females. Length–weight relationships were W(g)=0.0093 L(cm) exp. 3.03 for the total population, W(g)=0.0147 L(cm) exp. 2.91 for females, and W(g)=0.0168 L(cm) exp. 2.87 for males, respectively. As from age three, females were longer and heavier than males. The growth parameters estimated for the total population, for females and for males were L=83.29, 79.66, and 46.11 cm, respectively; k values were 0.18, 0.23, and 0.92, respectively, and t₀ values were –1.87; –1.54, and –0.62, respectively. Similarly to other flounders,P. orbignyanus is a typical inhabitant of estuaries and coastal regions. Adults stay in the study area mainly during spring and summer, they go outside or offshore during the spawning period and then they return to the estuary for feeding and recovering.     

Use of three types of stream-margin habitat by age-0 brown trout late in the growing season

The use of stream-margin habitat by age-0 salmonids has been studied, but differences in use among various types of habitat along stream margins has not been addressed. We described the nighttime use of habitat features by age-0 brown trout (Salmo trutta) among three types of stream-margin habitat late in the growing season (August–September) and assessed the extent to which use of habitat features within each type differed over the sampling period. Differences in water depths, water velocities, distances from shore, and substrate at the locations of fish along the margins of pools, the margins of riffles, and in backwaters were studied. Variation in habitat use also was observed during the study period as fish increased in length. Our observations are important considerations when developing habitat suitability criteria for assessment of instream-flow needs of age-0 brown trout.     

Book reviews

The numbers of perch Perca fluviatilis in the Cikola backwater system of the river Danube were estimated by mark-recapture techniques using multiple fishings. Fish were caught by fyke nets and electrical fishing. The latter method was unselective for sex and the catch data could be used to correct the sex bias in the fyke net catches. A population estimate in April of 6059 ind ha^–1 (95% C.L. 5135–7386) for perch > 5 cm was obtained for a 0.6 ha area of backwater and 1665 ind ha^–1 (95% C.L. 1204–2692) in May for a 0.4 ha area of a small bay.     

Stomach contents of the Antarctic Silverfish Pleuragramma antarcticum from the southern and eastern Weddell Sea (Antarctica)

Summary The stomach contents of 142 Pleuragramma antarcticum from the southernmost part of the Weddell Sea (Gould Bay: 77°19S) and from the eastern coast (72°–74°S) were analyzed. The fish were collected in February 1983 and stored frozen. Size range of the investigated specimens was from 6.4 cm SL (7.3 cm TL) to 21.9 cm SL (24.1 cm TL). The fish were caught by bottom trawl (Gould Bay) and Agassiz Trawl and pelagic RMT-m net in the upper 300 m layer at the eastern coast, respectively. Thirty-six stomachs were empty or contained sand only. The most important prey in terms of biomass were euphausiids, which occurred in 49 stomachs at a mean number of 2. Their dry weight was estimated to be 15–50 times that of the next important food items, which were copepods, gastropods and gammarids. Krill (Euphausia superba) constituted up to 14% of the euphausiids in the diet of fish caught in the eastern Weddell Sea. The rest was usually made up by Euphausia crystallorophias. This species was taken with preference even when the abundance of E. superba in accompanying plankton catches was 16 times higher. In the Gould Bay, gastropods were frequently eaten despite high numbers of copepods in the plankton, whereas in the eastern Weddell Sea, copepods were abundant both in plankton and in the stomach content. The almost complete absence of the early stages of fish in the diet of Pleuragramma antarcticum in the southern and eastern Weddell Sea is due to a pronounced vertical segregation of fish of different sizes. This distribution pattern is thought to be an adaptive mechanism to avoid intraspecific predation, as 90% of the ichthyoplankton in that area is Pleuragramma antarcticum.     

Short-term cues used by foraging trout in a California stream

Stream salmonids choose foraging locations to maximize the energy benefit of foraging within the constraints of size-mediated dominance hierarchies and predation risk. But, because stream habitats are temporally variable, fish must use a search process to monitor changing habitat conditions as a means of locating potentially-better foraging locations. I explored the cues used by the cutthroat trout, Oncorhynchus clarki clarki, when searching for food at the pool scale by artificially increasing prey availability at different locations by using special feeders and by manipulating pool velocities. Behavior of individually marked fish was monitored from stream bank platforms under unmanipulated control conditions and under seven experimental sets of conditions involving different combinations of feeder location and velocity manipulation. Under natural conditions fish elected to forage in the deepest (>50 cm), fastest (0.10–0.25 m s⁻¹) locations and within 1 m of structure cover, but would readily move to shallower (<30 cm) water away from cover if velocities were manipulated to be highest there. Although fish did not locate feeders unless they were placed in high-velocity areas, when high velocity was provided fish would move into very shallow water (<20 cm) if prey were delivered there. Responses of individual trout to manipulations indicated that water velocity was the main physical cue used by fish to decide where to forage, and that fish could also learn about new food sources by observing conspecifics. Overall, results indicated fish were not “perfect searchers” that could quickly locate new food resources over short time scales, even when the new resources were within a few meters of the fish’s normal foraging location. When given the correct cues, however, fish could detect new food sources and defend them against subordinate fish. Movement of new fish into and out of the study pools during the ten-day observation period was common, consistent with the idea that trout used movement as a means of exploring and learning about habitat conditions at the reach scale.     

Comparison of the physical properties of real and artificial salmonid eggs and of their performance when drifting in an experimental stream channel

The paper compares three batches of brown trout (Salmo trutta Linnaeus) eggs, two batches of Atlantic salmon (Salmo salar Linnaeus) eggs and the artificial trout eggs described by Ottaway (1981), in terms of their measurable physical characteristics. Attention was given to the effects of temperature and of the stage of development of the real eggs. Comparisons were also made between brown trout eggs and artificial eggs of the rate of fall in a water column, of the value and speed of attainment of terminal velocity when falling through water and of the pattern of settlement within an experimental stream channel.The main measurable characteristics of brown trout eggs (after water hardening) and of artificial eggs (values in parentheses) were: diameter 0.52–0.54 mm (0.54), fresh weight 0.078–0.086 g (0.09), density 1.071–1.075 g ml^–1 (1.066), and volume 0.073–0.081 ml (0.085). There was no evidence of any major change in these values as egg development proceeded. Salmon eggs had a similar density to trout and artificial eggs but had 30 to 70% greater weight and volume. The water-hardening of eggs caused their dry matter content to fall from 37–39% to 31–34% and there were corresponding increases in volume and fresh weight.The rate of fall of trout eggs and artificial eggs through water when timed from rest at the water surface to a depth of 125 cm was similar for both types of egg at 10 °C. This rate rose with increasing temperature at the same low rate of c. 0.02 cm s^–1 °C^–1 for both types of egg. Both types of egg had a terminal velocity of c. 8.8 cm s^–1 and both achieved terminal velocity in less than 1.5 s after release.The patterns of settlement of the two types of eggs in an experimental channel were similar.     

Thermal stratification and benthic foraging patterns of white sucker

Abundances of white sucker, 100–500 mm FL, were not significantly different among the epilimnia, metalimnia and portions of the hypolimnia shallower than 20 m in each of two lakes. However, small suckers < 200 mm were captured most frequently in the epilimnion and no white suckers were captured in the deepest region, 20–38 m, of the two lakes. White suckers consumed prey from all three temperature zones in each lake. Prominent food items were Hyalella azteca and the chironomid larvae Heterotrissocladius, Djalmabatista and Procladius. Despite differences in relative densities of benthic invertebrates among thermal zones of the two lakes, suckers in neither lake foraged exclusively on prey of epilimnetic origin. Suckers captured in the metalimnia foraged on invertebrates that were common to all three thermal zones. And, only 0–4% of the suckers captured in the hypolimnia of the two lakes contained prey that were unique to the epilimnia. Suckers caught in the hypolimnia mainly consumed deep water invertebrates; 83% of the suckers foraged in the metalimnion and hypolimnion of Islets Lake and 45% foraged in the hypolimnion in Burnt Island Lake. Consequently there was little evidence of a massive inshore feeding migration followed by a post-feeding return to the hypolimnion. Northern pike and lake trout rarely fed on white suckers in these lakes and thus piscivory was an unlikely factor in the observed distribution of suckers.     

Body Size, Diet and Growth of Landlocked Brown Trout, Salmo trutta, in the Subarctic River Laxá, North-East Iceland

Studies on diet preferences of stream-dwelling salmonids have mostly been limited to a relatively small range in body size. This study examined the influence of salmonid body size on prey size and diet composition, and its consequences for growth, in landlocked stream-resident brown trout, Salmo trutta (2.5–61.3cm) in the Laxá River, N-E Iceland. The most common prey of the 1622 trout sampled, were blackflies, Simulium vittatum, chironomid midges and the freshwater snail, Lymnaea peregra, which represented 56.3%, 21.8%, 10.8% of the stomach content volume, respectively. In general, the Laxá trout showed a consistent, but moderate, shift towards larger prey with increased body size. The relatively stable growth and the large body size attained were probably due to the high production of small benthic invertebrates and only secondarily to the ontogenetic shift towards larger prey.     

Winter ecology of Arctic charr (<Emphasis Type="Italic">Salvelinus alpinus</Emphasis>) and brown trout (<Emphasis Type="Italic">Salmo trutta</Emphasis>) in a subarctic lake,Norway

We studied habitat choice, diet, food consumption and somatic growth of Arctic charr (Salvelinus alpinus) and brown trout (Salmo trutta) during the ice-covered winter period of a subarctic lake in northern Norway. Both Arctic charr and brown trout predominantly used the littoral zone during winter time. Despite very cold winter conditions (water temperature <1°C) and poor light conditions, both fish species fed continuously during the ice-covered period, although at a much lower rate than during the summer season. No somatic growth could be detected during the ice-covered winter period and the condition factor of both species significantly declined, suggesting that the winter feeding rates were similar to or below the maintenance requirements. Also, the species richness and diversity of ingested prey largely decreased from summer to winter for both fish species. The winter diet of Arctic charr <20 cm was dominated by benthic insect larvae, chironomids in particular, and Gammarus lacustris, but zooplankton was also important in December. G. lacustris was the dominant prey of charr >20 cm. The winter diet of brown trout <20 cm was dominated by insect larvae, whereas large-sized trout mainly was piscivorous, feeding on juvenile Arctic charr. Piscivorous feeding behaviour of trout was in contrast rarely seen during the summer months when their encounter with potential fish prey was rare as the small-sized charr mainly inhabited the profundal. The study demonstrated large differences in the ecology and interactions of Arctic charr and brown trout between the winter and summer seasons.     

The ecology of ruffe,Gymnocephalus cernuus (Pisces: Percidae) introduced to Mildevatn,western Norway

Synopsis Diet, habitat use, diel and seasonal activity and a number of population parameters were studied on ruffe,Gymnocephalus cernuus, introduced to Mildevatn, western Norway. This lake is sited outside the natural range of the ruffe and has a lower fish diversity and a different fish species composition than within its native range. From June through September the ruffe was planktivorous and mainly caught at 4 to 6 m depth in the benthic zone. At other times of year ruffe was feeding on zoobenthos and caught deeper in the benthic zone. Ruffe was mainly day active. Zooplankton feeding during summer is the clearest difference compared to ruffe populations living within its natural range. Presence of large zooplankton organisms available for ruffe is suggested as the main reason for the difference found in food choice. The availability of large zooplankton is probably due to community structure caused by a predator and lack of interspecific competition for zooplankton in the deeper parts of the lake. Piscivorous brown trout.Salmo trutta, restrict the habitat of threespined stickleback,Gasterosteus aculeatus, to the zone of littoral vegetation, allowing high densities of larger zooplankton species likeBythotrephes longimanus to be present in the lake. Brown trout is present only in the upper light and well oxygenated parts of the lake, leaving a refuge for the ruffe, where they can feed on the rich zooplankton community.     

Community structure and photosynthetic activity of epilithon from a highly acidic (pH≤2) mountain stream in Patagonia,Argentina

We explored a benthic community living on stones in an acidic (pH2) stream of active volcanic origin from Patagonia, Argentina, by combining in situ measurements (temperature, pH, conductivity, dissolved oxygen), photosynthesis of intact biofilms (measured with microsensors by the light–dark shift method), pure-culture experiments on isolated algae, and confocal laser scanning microscopy on the biofilms. The epilithon of the Agrio River was dominated (99% of total biomass) by one species: Gloeochrysis (Chrysophyceae). This species was observed as brown, mucilaginous, 200-m-thick films on stones, growing in clumps in a dense matrix of fungal hyphae, bacteria, and inorganic particles held together by extracellular polymeric substances. Gloeochrysis was isolated and cultivated. The photosynthetic rate measured at saturation irradiance was 120 mol oxygen (mg chlorophyll a)^–1h^–1 under laboratory conditions, and the saturation rate of photosynthesis by carbon dioxide was 90 mol oxygen (mg chlorophyll a)^–1 h^–1 for oxygen evolution. Photosynthetic activity of the biofilm was light-dependent and saturated above 200 mol photons m^–2 s^–1. In the dark, the stone surface became anoxic. Our data suggest that primary production in the Agrio River was not limited by light, carbon, or phosphorus but instead, nitrogen-limited.     

Spatial and temporal variation of benthic Cladocera (Crustacea) studied with activity traps in Lake Myvatn, Iceland

Benthic Cladocera were studied with a modified type of an activity trap in Lake Myvatn in 1990–1992. After feasibility experiments, the operation time and the distance of the traps from the benthic substrate were adjusted in order to minimize the effects of diurnal variation and the trapping of planktonic organisms. The trap catches of Eurycercus lamellatus were positively correlated with their abundance at the bottom as estimated by grab sampling (r = 0.910, P < 0.001). The usefulness of the activity trap was demonstrated by: (1) a lake-wide survey of the benthic Cladocera; (2) a study of the seasonal variation in the size distribution, abundance and sex ratio of E. lamellatus; and (3) a study of the seasonal succession of Chydorus sphaericus and Alona affinis. The variation of benthic Cladocera among 21 trap sites distributed on a 1–2-km scale across the lake exceeded the within-site variation. The sampling sites could be divided into five main groups based on cluster analysis. Eurycercus lamellatus was the most common species in the mat of filamentous green algae (Cladophorales). Alonella nana dominated the area of spring water inflow in the north basin and Macrothrix hirsuticornis the area of spring water inflow in the southeastern part of the lake. In other parts of the lake either Chydorus sphaericus or Alona quadrangularis tended to dominate. The size distribution and sex ratio of E. lamellatus was followed at two sites through one summer. In early summer most individuals were females less than 1.45 mm long. Around mid-summer they had grown to 0.69–3.1 mm. By the end of August the size distribution had become bimodal, with a large number of small males and a smaller number of much larger females. The seasonal succession in the abundance of E. lamellatus, A. affinis and C. sphaericus was followed at four sites over two seasons. With some exceptions the abundance of a species followed a similar seasonal trajectory on the different stations in any one year. There was, however, a marked difference between the two years (1991 and 1992), probably related to different temperatures.     

Habitat and shelter requirements of the stone crayfish,Austropotamobius torrentium Schrank

The attractiveness of shelters and the relationship between crayfish size and refuge size were investigated in the stone crayfish Austropotamobius torrentium. Shelter suitability was mainly determined by water velocity, stone surface area and stone width. Water depth and stone height had no influence on crayfish occupancy of stones. Austropotamobius torrentium avoided shelters in areas of water velocities exceeding 25 cm s^–1. They preferred large stones with a bearing surface greater than 300 cm². The probability of crayfish occupation rose quickly as stones' bearing surface increased up to 900 cm². Crayfish were exclusively caught underneath stones at least 3.19 times longer and 1.25 times wider than the carapace length of the refuge occupant. Large males occupied larger stones, where no relationship was detected between female size and stone refuge size.     

Habitat utilization by sympatric arctic charr Salvelinus alpinus L. and brown trout Salmo trutta L. in Lake Atnsjø, south-east Norway

SUMMARY. 1. Habitat utilization, as well as inter- and intraspecific relations of different size groups of arctic charr (Salvelinus alpinus (L.)) and brown trout (Salmo trutta L.) in Lake Atnsjø, south-east Norway, were investigated by analysing food and spatial niches from monthly benthic and pelagic gillnet catches during June-October 1985.
2. Small individuals (150–230 mm) of both arctic charr and brown trout occurred in shallow benthic habitats. However, they were spatially segregated as arctic charr dominated at depths of 5–15 m and brown trout at depths of 0–5 m.
3. Larger (>230 mm) arctic charr and brown trout coexisted in the pelagic zone. Both species occurred mainly in the uppermost 2-3 m of the pelagic, except in August, when arctic charr occurred at high densities throughout the 0–12 m depth interval. On this occasion, arctic charr were segregated in depth according to size, with significantly larger fish in the top 6 m. This was probably due to increased intraspecific competition for food.
4. The two species differed in food choice in both habitats, Arctic charr fed almost exclusively on zooplankton, whereas brown trout had a more variable diet, consisting of surface insects, zooplankton. aquatic insects and fish.
5. The data suggest that the uppermost pelagic was the more favourable habitat for both species. Large individuals having high social position occupied this habitat, whereas small individuals lived in benthic habitat where they were less vulnerable to agonistic behaviour from larger individuals and less exposed to predators. The more aggressive and dominant brown trout occupied the more rewarding part of the benthic habitat.