Behavioural strategies for species co‐existence during environmental extremes: competition between trout and salmon parr during dewatering

Competition between species of animals can be predicted to be intense under extreme environmental conditions during which niche overlap increases. Fluvial aquatic systems fluctuate naturally across a broad range of time‐scales. Land management activities impose further extreme fluctuations, particularly when water is abstracted for irrigation during summer. This study focused on the interaction between Atlantic salmon and brown trout parr during acute dewatering events. Brown trout are known to compete strongly for pool habitat whereas salmon can coexist by using riffle areas during normal flows. It is not known, however, how competition between the species affects their behavioural responses to extreme low flows when riffle areas decrease. Replicated groups of salmon were held in allopatry and sympatry with trout in sections of a large indoor stream, each of which was landscaped into riffles and pools. Space use and behaviours of the fish were recorded by direct visual observations and a network of Passive Integrated Transponder detectors. Here we report the response of the fish to dewatering and consider the system as a model for natural and forced responses of communities to environmental extremes.     

The evolution of cooperation and advanced social behaviour

Competition between species of animals can be predicted to be intense under extreme environmental conditions during which niche overlap increases. Fluvial aquatic systems fluctuate naturally across a broad range of time‐scales. Land management activities impose further extreme fluctuations, particularly when water is abstracted for irrigation during summer. This study focused on the interaction between Atlantic salmon and brown trout parr during acute dewatering events. Brown trout are known to compete strongly for pool habitat whereas salmon can coexist by using riffle areas during normal flows. It is not known, however, how competition between the species affects their behavioural responses to extreme low flows when riffle areas decrease. Replicated groups of salmon were held in allopatry and sympatry with trout in sections of a large indoor stream, each of which was landscaped into riffles and pools. Space use and behaviours of the fish were recorded by direct visual observations and a network of Passive Integrated Transponder detectors. Here we report the response of the fish to dewatering and consider the system as a model for natural and forced responses of communities to environmental extremes.     

Impact of Beaver Dams on Abundance and Distribution of Anadromous Salmonids in Two Lowland Streams in Lithuania

European beaver dams impeded movements of anadromous salmonids as it was established by fishing survey, fish tagging and redd counts in two lowland streams in Lithuania. Significant differences in abundancies of other litophilic fish species and evenness of representation by species in the community were detected upstream and downstream of the beaver dams. Sea trout parr marked with RFID tags passed through several successive beaver dams in upstream direction, but no tagged fish were detected above the uppermost dam. Increase in abundances of salmonid parr in the stream between the beaver dams and decrease below the dams were recorded in November, at the time of spawning of Atlantic salmon and sea trout, but no significant changes were detected in the sections upstream of the dams. After construction of several additional beaver dams in the downstream sections of the studied streams, abundance of Atlantic salmon parr downstream of the dams decreased considerably in comparison with that estimated before construction.     

Moon phase and nocturnal density of Atlantic salmon parr in the Sainte‐Marguerite River, Québec

Nocturnal underwater counts of Atlantic salmon Salmo salar parr were made on four consecutive occasions (two lunar cycles, samplings at both the full and new moon) at four 40 m long sites in the Sainte‐Marguerite River, Québec, Canada, between 30 June and 14 August 2003. Atlantic salmon parr counts did not differ significantly between moon phases. Cloud cover ranged from 0 to 100% during the study, and had no significant effect on parr counts. There were significantly more Atlantic salmon parr in the near shore than in the midstream areas. The findings of this study suggest that the sampling strategy of summer studies aimed at assessing population abundance or developing habitat quality models can be designed without taking moon phase or cloud cover into consideration, but it should account for the higher relative abundance of Atlantic salmon parr in the near shore areas as compared to areas closer to the middle of streams.     

Infectious salmon anaemia virus in wild fish from Scotland.

Following the outbreak of infectious salmon anaemia (ISA) at salmon farms in Scotland, UK, a survey was established to determine the extent of infection in wild fish. All fish tested were free from the clinical symptoms of ISA. Isolations of ISAV were made from 5 sea trout within areas where ISA affected salmon farms were located. Evidence for ISAV in other sea trout was provided by ISA RT-PCR diagnostic tests. Results from ISA RT-PCR tests reveal evidence for ISAV being present in salmon parr, adult salmon and juvenile brown trout in rivers distant from salmon farms and indicate that, at the time of the survey (1998-1999), ISAV may have been widely distributed. Nucleotide sequence analysis of segments 2 and 8 showed that for most sequences from wild fish there was 100% homology with ISAV isolated from clinically affected farmed fish although evidence is presented which indicates variability in ISAV sequences from wild fish. Modelling the RT-PCR findings indicates that ISAV among salmonid fish was spatially non-random. Brown trout, sea trout and salmon (adult and parr) show a pattern of occasionally large numbers of positive samples against a background of very low numbers.     

Is Atlantic salmon production limited by number of territories?

Thirty Atlantic salmon Salmo salar parr (mean 135 mm total length, L _T) in the River Alta, Norway, were radio‐tagged and tracked during a 11–13 day period. The parr stayed within defined home ranges with a 95% probability of localization within an average area of 1286 m²(241–3484 m²). On average, each parr had overlapping home ranges with 7·7 other parr, and the overlap between pairs of fish covered on average 24% of their home range areas. Mean length of the river stretch used was 90 m (22–383 m). On average, during 38%(16–77%) of the tracking surveys, the fish had moved >10 m since the previous survey. Mean total distance moved during the whole study was 402 m (208–862 m). The Atlantic salmon were most often recorded in riffles, but 27% of the parr alternated between riffles and pools. The extensive movements, flexible habitat utilization and relatively large home ranges, together with the fact that all the parr had home ranges partly overlapping with other radio‐tagged parr, indicate a flexible and variable behaviour and, thereby, a more complex social structure than a rigid defence of a fixed territory. Thus, Atlantic salmon production seems not to be limited by the highest potential number of territories for large parr in a given area.     

The effects of intra- and inter-specific competition on the distribution of stocked juvenile Atlantic salmon, Salmo salar L., in relation to depth and gradient in an upland trout, Salmo trutta L., stream

The relative effects of inter- and intra-specific competition on the distribution of stocked salmon in relation to depth and gradient were investigated in an upland stream during two summer sampling periods. The stream was divided into two areas by an impassable fish barrier, and trout were removed from the upstream section prior to 2 years of salmon stocking. A small amount of trout re-immigration to the cleared area occurred. Under sympatric conditions in the downstream section the fry of both species were significantly more abundant in shallow water, whereas the yearling and older fish tended to inhabit the deeper, slow-flowing areas. Under near allopatric conditions in the cleared section salmon fry changed their distribution to include deeper areas, with shallow, fast-flowing water becoming the least preferred habitat. This trend for a wide distribution of salmon fry over all the available habitat in the upstream section was maintained in the second year when parr were present but trout remained at very low densities. It was therefore concluded that high inter-specific competition from trout was responsible for restricting the distribution of salmon fry to shallow habitat in the control area. Intra-specific competition from older salmon apparently only affected the growth and survival of salmon fry. The regulating mechanisms involved in these inter- and intra-specific effects are discussed in terms of competition for stream resources and predation. Recommendations based on the findings are made for stocking and habitat management of salmonid waters.     

Effect of mink, Mustela vison Schreber, predation on cohorts of juvenile Atlantic salmon, Salmo salar L., and brown trout, S. trutta L., in three small streams

Two cohorts of Atlantic salmon parr and one of brown trout were studied in periods with and without the presence of mink, Mustela vison . In all localities a marked increase in mortality rate was observed during periods when mink were present. Mink were observed catching salmon parr, and approximately 10% of the parr had bite marks, especially on the tail fins. In the smallest stream with brown trout, the mortality rate was 0.80 during a few days with mink present; remnants of trout were found along the stream. The present study suggests that mink predation may be a major cause of mortality of salmonids in small streams. The results indicate that predation efficiency may vary with characteristics of the habitat, especially stream width and discharge, and fish density.     

Effects of sexual maturation, castration, and androgen implants on growth in one- and two-year-old parr in a Baltic Atlantic salmon (Salmo salar L.) stock

The effects of maturation, castration, and androgen implants on growth in 1- and 2-year-old Baltic salmon ( Salmo salar L.) were studied in a stock originating from the Umeälven (Ume river) in northern Sweden. 1 + male parr that later matured were larger than non-maturing parr in the spring. Later in the summer, during the period of gonadal growth, the growth rate of maturing parr was depressed compared to that of non-maturing males. Implantation of capsules with 11-ketoandrostenedione (OA) or testosterone (T) in sexually immature 1-year-old fish, accelerated the specific growth rate in spring and early summer, particularly considering growth in weight, resulting in an increase in the condition factor. In contrast, the growth rate was depressed from late June onwards, the period when the gonads normally develop in 1 + parr. T, but not OA, also promoted sexual maturation. Castration or OA treatment did not affect total summer growth in 2-year-old salmon. This study suggests a causal relationship between androgens and the growth pattern observed in maturing 1 + male parr.     

Production of juvenile salmonids in small Norwegian streams is affected by agricultural land use

1. We estimated the biomass and production of juvenile anadromous brown trout (Salmo trutta) and Atlantic salmon (Salmo salar) (parr) in 12 streams in the Skagerrak area of Norway to identify controlling environmental factors, such as land‐use and water chemistry. 2. Production estimates correlated positively with fish density in early summer, but not with the size of the catchment. The summer biomass of age‐0 brown trout and Atlantic salmon was smaller than that of age‐1 and constituted 27.4 and 25.7%, respectively, of the total biomass of the two groups. 3. Mean production of brown trout from July to September varied between streams, but in most cases it was below 2 g 100 m^?2 day^?1. Yearly cohort production from age‐0 in July to age‐1 in July was 10 g m^?2 or less, with mean annual production of 1.32 g 100 m^?2 day^?1, equivalent to 4.8 g m^?2 year^?1. The corresponding annual cohort production of Atlantic salmon was 0.38 g 100 m^?2 day^?1 or 1.4 g m^?2 year^?1. Annual production to biomass ratio (P/B) for brown trout of the same cohort in the various streams was between 1.47 and 4.37; the overall mean (±SD) for all streams was 2.25 ± 0.94. Mean turnover rate of Atlantic salmon was 2.73 ± 0.24. 4. Production of 0+ brown trout during the summer correlated significantly with the percentage of agricultural land and forest/bogs in the catchment, with maxima at 20 and 75%, respectively. Age‐0 brown trout production also correlated with concentration of nitrogen and calcium in the water, with maxima at 2.4 and 14 mg L^?1, respectively. 5. The results support the hypothesis that brown trout parr production reflects the quality of their habitat, as indicated by the dome‐shaped relationship between percentage of agricultural land and the concentration of nitrogen and calcium in the water.     

Infection of Atlantic salmon, Salmo salar L., by Gyrodactylus salaris, Malmberg 1957, in the River Lakselva, Misvær in northern Norway

Gyrodacrylus salaris was most probably introduced to the River Lakselva in 1975 through stocking of Atlantic salmon from an infected hatchery. The parasite population grew rapidly, and the parasite spread throughout the entire watercourse during the summer of 1976. This epidemic situation led to mortality among the young Atlantic salmon, and the density of salmon parr was heavily reduced from 1976 to 1977. The density of salmon parr has remained close to zero since then, while there are no apparent trends toward decrease or increase in the density of brown trout. In spite of the reduced density of young salmon, a new epidemic has developed each year among the few young 0+ and 1+ Atlantic salmon present in the river. Results from successive sampling during the summer of 1987, 1988 and 1989 indicate that most of the presmolt salmon are attacked during their first summer or autumn of life. The infection develops into an epidemic during the first autumn, winter or the next summer. The build-up of the parasite burden on the fish leads in turn to mortality. Norwegian Atlantic salmon probably have no resistance against G. salaris, since the parasite has recently been introduced to Norwegian rivers.     

Migration speeds and orientation of Atlantic salmon and sea trout post-smolts in a Norwegian fjord system

We recorded the observed and actual swimming speeds of Atlantic salmon and sea trout post-smolts in a Norwegian fjord system, and initiated studies on the orientation mechanisms of the post-smolts. We tracked Atlantic salmon and sea trout with acoustic transmitters for up to 14 h after release. The actual swimming speed and direction of a fish relative to the ground is the vector sum of the observed movements of the fish and the movements of the water. We determined actual swimming speeds and directions of the post-smolts, which reflect their real swimming capacities and orientation, by corrections for the speed and direction of the water current. The post-smolts were actively swimming. The observed direction of movement was dependent on the actual movement of the fish and not the water current. Water currents were not systematically used as an orientation cue either in Atlantic salmon or sea trout, as the actual movements were random compared to the direction of the water current. The actual movement of sea trout were in all compass directions, with no systematic pattern. The Atlantic salmon also moved in all compass directions, but with the lowest frequency of actual movement towards the fjord.     

Is there a threshold size regulating seaward migration of brown trout and Atlantic salmon?

We investigated the influence of variation in body size and growth rate on age of smolting in Atlantic salmon and brown trout in four different Norwegian rivers. In Atlantic salmon smolt ages varied between 2 and 6 years, and in brown trout between 2 and 7 years. Smolt age was negatively correlated with parr growth, and positively correlated with smolt size. Age at smolting was more variable in the two northern than the two southern rivers. Smolt sizes and ages were also more variable in brown trout than in Atlantic salmon. Based on the observed variation in smolt size and age, we reject the hypothesis that a threshold size alone regulates age at smolting. Within populations smolt age depends on growth rate so that fast-growing parr smolted younger and smaller than slow-growing parr. We hypothesize that smolt size and age is a trade-off between expected benefits and costs imposed by differences in individual growth rate.     

Movements and habitat use by PIT-tagged Atlantic salmon parr in early winter: the influence of anchor ice

1. Movements and habitat use by Atlantic salmon parr in Catamaran Brook, New Brunswick, were studied using Passive Integrated Transponder technology. The fish were tagged in the summer of 1999, and a portable reading system was used to collect data on individual positions within a riffle‐pool sequence in the early winter of 1999. Two major freezing events occurred on November 11–12 (Ice 1) and November 18–19 (Ice 2) that generated significant accumulations of anchor ice in the riffle. 2. Individually tagged parr (fork length 8.4–12.6 cm, n = 15) were tracked from 8 to 24 November 1999. Over this period, emigration (40%) was higher from the pool than from the riffle. Of the nine parr that were consistently located, seven parr moved <5 m up‐ or downstream, and two parr moved more than 10 m (maximum 23 m). Parr moved significantly more by night than by day, and diel habitat shifts were more pronounced in the pool with some of the fish moving closer to the bank at night. 3. During Ice 2, there was relatively little movement by most of the parr in the riffle beneath anchor ice up to 10 cm in thickness. Water temperature was 0.16 °C above the freezing point beneath anchor ice, suggesting the existence of suitable habitats where salmon parr can avoid supercooling conditions and where they can have access to low velocity shelters. To our knowledge, these are the first data on habitat use by Atlantic salmon parr under anchor ice.     

Susceptibility of brown trout to Gyrodactylus salaris (Monogenea) under experimental conditions

Brown trout ( Salmo trutta ) from anadromous River Lierelva, resident Lake Tunhovd, and resident Nordmarka stocks were exposed to Gyrodactylus salaris -infected salmon parr. The brown trout were fed pellets before the experiments, except for one group of the Nordmarka stock which was starved for 19 days before the experiments. The mean number of parasites declined directly and rapidly post infection for all groups of trout. There were no pronounced differences in resistance between the anadromous and the resident stocks. G. salaris infections tended to persist longer on starved than on fed trout of the Nordmarka stock. The maximum parasite persistence on trout was 50 days, and as parasite numbers increased on some fish parasite reproduction must have occurred on those trout. However, the limited susceptibility and marked innate resistance of trout to G. salaris establishment, development and reproduction, suggest parasite metapopulations will not survive on this species. Nevertheless, trout may still play a role in the dispersal of G. salaris within and between rivers.     

Production of juvenile Atlantic salmon, Salmo salar L., and brown trout, Salmo trutta L., within different sections of a small enriched Norwegian river

Growth, density and production of juvenile Atlantic salmon and brown trout were studied in three different sections of the Kvassheimsåna River in south-western Norway from 1979 to 1983. Section 1. in the upper part of the river, is located above a waterfall impassable for migratory salmonids and is surrounded by grazing land. Sections 2 and 3, in the middle and lower parts of the river, are influenced by agricultural activity. Total nitrogen concentration varied between 250 and 1000 μg l ^?1 in section 1 and 1500 and 2500 μg l^?1 in sections 2 and 3. Total phosphorus (Tot-P) concentrations also increased with decreasing altitude: 19–46 μg l^?1 in section I and 31–101 μg l ^?1 in sections 2 and 3. The number of 0 + salmon in sections 2 and 3 varied between 30.1 and 167.8 specimens 100 m ^?2, with means 90.2 and 95.2 specimens 100 m ^?2:, respectively; the density of 1 + salmon, with mean values of 16.3 and 51.0 specimens 100m^?2 was significantly correlated with the original fry density. The growth rate of 0+ salmon was not inversely related to cohort density, but was significantly so for 1 + salmon. Mean annual salmon production in section 2 was 1595 g 100 m^?2 year ¹, and in section 3 was 841 g 100m^?2 year ¹. A logarithmic function gave the best curve fit between salmon production and mean annual biomass. Thus, production levelled off for the highest values recorded in section 2, and perhaps approached the carrying capacity of the stream. A multiple regression analysis showed that yearly variation in 1 + salmon density was the single factor accounting for most of the total variability in production (60%). Variation in water temperature and nutrient content were not significantly related to variation in fish production. Densities of brown trout were low in all sections (<20 specimens 100m ^?2). Fry density was highest in section 3 and parr density in section 1. All age groups of sympatric brown trout grew significantly faster in sections 2 and 3 compared with allopatric brown trout in section 1.     

Competition between brown trout and Atlantic salmon parr over pool refuges during rapid dewatering

Variations in distributions and behaviours of Atlantic salmon Salmo salar in allopatry (homogeneous) and in sympatry with brown trout Salmo trutta (mixed) were observed before, during and after 2 day periods of dewatering in a large glass-sided indoor stream at densities typical of Scottish upland streams. Brown trout utilized pools more than Atlantic salmon at normal flows and in both species the majority of fishes moved into pools during dewatering. There was no significant effect of brown trout, which was the more dominant species, on the overall ability of Atlantic salmon to use pool habitat as a refuge during dewatering. Within mixed and homogeneous groups, average feeding levels decreased during dewatering. The highest ranking fish, which was always a brown trout in mixed groups, predominantly monopolized the pool and other individuals in pools adopted a more cryptic, stationary behaviour. Dewatering effectively increased local population density with the result that dominance status became much more important in maintaining food intake, and polarization between the top ranking fish and others increased. During the first day of dewatering, there was extreme behavioural polarization such that the dominant fish exhibited most aggression and least feeding within the group. Among dominant fish on the second day of dewatering, aggression had largely abated and feeding had returned to pretreatment levels despite the reduced average feeding within the group. The main difference between mixed and homogeneous groups was in the behaviour of the most dominant Atlantic salmon, which was near-despotic in allopatry and subordinate to brown trout in sympatry.     

Effects of food availability on temporal activity patterns and growth of Atlantic salmon

1. Patterns of sheltering and activity are of fundamental importance in the ecology of animals and in determining interactions among predators and prey. Balancing decreased mortality risk when sheltering with increased feeding rate when exposed is believed to be a key determinant of diel patterns of sheltering in many animals. 2. Despite lower foraging efficiency at night than during the day, Atlantic salmon Salmo salar parr are nocturnal during winter and at low summer temperatures. Nocturnal activity also occurs at warm water temperatures during summer, but little is known about the functional significance of this behaviour. 3. This study aimed to determine: (1) the preferred activity and shelter pattern of Atlantic salmon parr during warm summer months, and (2) their response to variations in food availability when balancing growth rate (G) and mortality risk (M), as expressed through time out of shelter. We differentiated among four potential responses to reduced food availability: (1) no response; (2) G decreases but M remains constant; (3) G remains constant but M increases; and (4) G decreases and M increases. 4. Time exposed from shelter was inversely related to food availability. Fish subject to high food availability were significantly less active during the day than those with restricted rations. However, food availability had no significant effect on the extent to which fish were active at night. There was no evidence of variation in growth rate with food availability. 5. Salmon were predominantly nocturnal at high ration levels, consistent with their previously reported behaviour during winter. Rather than switching to diurnal behaviour at high temperatures per se, as previously was supposed, it appears that the fish are diurnal only to the extent needed to sustain a growth rate, and this extent depends on food availability. 6. Atlantic salmon parr modulate the amount of time they are active rather than growth when responding to variations in food availability over an order of magnitude.     

Individual growth and phase differentiation of lacustrine masu salmon, Oncorhynchus masou, under artificial rearing conditions

 In lacustrine masu salmon, Oncorhynchus masou, originated from anadromous fish and inhabiting an artificial lake (Shumarinai Lake), we examined the relationship between individual growth during the juvenile stage and phase differentiation under artificial rearing conditions. In females, the mean fork length of potential (subsequent) 1+ smolts and potential 1+ parr (1+ parr show fish that neither smolted nor matured until 1+ autumn) always differed after their first summer (0+ summer), the former being larger. In males, the juveniles that grew faster during their first spring became 0+ mature male parr. After their first summer, potential 1+ smolts were always larger in body length than potential 1+ mature parr and potential 1+ parr. These results were similar to previous reports of anadromous masu salmon, suggesting that the lacustrine masu salmon studied seems to have maintained the phase differentiation as in the original fish, likely because only 60 years have passed since the formation of the lacustrine population.     

Effects of body size on sympatric shelter use in over‐wintering juvenile salmonids

The study tests two hypotheses: (1) the degree of shelter dominance in Atlantic salmon Salmo salar and brown trout Salmo trutta increases progressively with increasing size differential between heterospecific fish in a pair and (2) shelter dominance, standardized to size differential, correlates with aggression. The results support the first but not the second hypothesis, suggesting that the fitness consequences of high growth performance during the summer are likely to become evident during winter. At this time of year, when mortality is high among both Atlantic salmon and brown trout, shelter dominance may increase the chances of survival.