Shelter availability affects behaviour, size-dependent and mean growth of juvenile Atlantic salmon

1. Anthropogenic disturbances of the physical habitat and corresponding effects on fish performance are key issues in stream conservation and restoration. Reduced habitat complexity because of increased sediment loadings and canalization is of particular importance, but it is not clear to what extent fish populations are influenced directly by changes in the physical environment, or indirectly through changes in the biotic environment affecting the food availability. 2. Here, we test for the direct effect of habitat complexity on the performance (growth) of juvenile Atlantic salmon by manipulating shelter availability (interstitial spaces in the substrate) across 20 semi‐natural stream channels without altering the substrate composition, and stocking each channel with a common density of fish. A simple method for measuring salmonid shelters using flexible PVC tubes was developed and tested. Daytime sheltering behaviour and growth rates were compared across the channels differing in shelter availability. 3. Measured shelter availability was strongly negatively correlated with observed number of fish not finding shelters and mass loss rates of the fish (growth performance) increased with decreasing number of measured shelters. Number and mean depth of interstitial spaces explained up to 68% and 24% of the among‐channel variation in sheltering behaviour and growth performance, respectively. Furthermore, negative effects of shelter reduction increased with fish body size. Thus, changes in habitat structure may even influence the size selection gradients. 4. Shelter availability is an easily measured variable, possibly affecting the population demographics and long‐term evolutionary processes, and is therefore a key habitat factor to be considered in stream restoration and habitat classification.     

Effects of life history variation on size and growth in stream-dwelling Atlantic salmon

A large size variation amongst life histories for stream-dwelling Atlantic salmon Salmo salar was found and the relative effect of life histories on size varied over time. As early as December (age 0+ years), fish that later smolted at age 2+ years were significantly larger than fish that did not smolt at age 2+ years. In contrast, there were no mass differences at age 0+ years between fish that would mature or not at age 1+ years (October). The mass differences between smolts and non-smolts persisted until smolting, and differences between mature and immature fish first appeared in May (age 1+ years). Following September (age 1+ years), there was also a significant interaction between smolting and maturity. Previously mature and immature age 2+ year smolts were not significantly different in size, but immature age 2+ year non-smolts were much lighter than mature age 2+ year non-smolts. Based on mass differences, the apparent 'decision' to smolt occurred c . 5 months before (winter, age 0+ years) the decision to mature (late spring, age 1+ years). In addition to strong seasonal growth variation, sizes of freshwater Atlantic salmon were largely structured by the complex interaction between smolt-age and maturity.     

Social status, access to food, and compensatory growth in juvenile Atlantic salmon

Juvenile Atlantic salmon Salmo salar subjected to three weeks of cooler temperatures were 8·5% smaller than controls at the end of the temperature manipulation, but had caught up in size 20 weeks later. The behavioural means is examined by which this catch-up or compensatory growth is achieved. While on average compensating fish did not spend more time feeding, dominant fish within each group gained more exclusive access to the feeding area during periods of catch-up growth. Therefore the extent to which compensatory growth could be achieved was dependent on both the social status of the individual and the dominants' ability to monopolize the food patch.     

Seasonal growth of wild Atlantic salmon juveniles and implications on age at smoltification

The seasonal growth trajectories of wild Atlantic salmon Salmo salar juveniles by age group within the Margaree River, Canada, are described. Circuli counts from scales were used to infer growth rates at different ages and these were used to predict the proportions of age 2‐ and 3‐year old smolts from different portions of the watershed. In the wild Atlantic salmon juveniles from the Margaree River, there was no bimodality in fork length frequencies and no 1 year old smolts were produced. Water temperature differences during the growing season were insufficient to explain the differences in growth rates and size at age among the sites sampled. There was a positive association between the growth rate in the first year and the subsequent age at smoltification. In the Margaree River, differences in tributary specific growth rates and size at age were expected to produce important differences in the relative ages at smoltification.     

Life-history patterns in a southern population of Atlantic salmon

Juvenile Atlantic salmon Salmo salar in the river Esva, Asturias, north-west Spain, developed a bimodal growth pattern during their first growing season. Segregation between the two modes was apparent by late autumn. All fish in the upper modal group (UMG) grew throughout the year and migrated downstream in the following spring (by April) at the age of 14 months. Some lower modal group (LMG) fish (56% of the main stem, 31% and 50% of the two tributaries) silvered like smolts and apparently migrated downstream the same spring, although c. 1·5 months after UMG fish. Larger LMG fish appeared more likely than smaller ones to migrate. LMG fish did not grow in winter, but they grew fast between March and May. Maturing male parr were detected first at 5 months old in July, and they occurred initially among the faster growers, >6·8 cm long. In spite of this, length of maturing males did not change from late summer and throughout the winter, whereas that of non-maturing UMG and LMG fishes continued to diverge from September onwards. These findings suggest that favourable conditions for growth (very early start of a long growing season) at the southern limit of the species' range may influence the life-history pattern of this population.     

Social status and growth rates as determinants of site attachment in juvenile Atlantic salmon

This paper describes an experimental study of the effects of food supply, growth rates and social interactions on homing by juvenile Atlantic salmon Salmo salar in response to displacement. Groups of five fish were housed in a section of an artificial stream and given either rations allowing maximum growth (the rich condition) or 0·1 of this amount (the poor condition); daily specific growth rates were significantly higher in the rich condition. After a 6-day settlement period, the fish were captured, displaced downstream and their movements recorded over the next 3 h. Prior to displacement, the fish showed a high degree of site fidelity and high levels of aggression. Dominant fish and those with stronger site attachment grew faster prior to displacement, these effects being independent. Following displacement, 24% of all fish returned to their previously favoured site and stayed there, 23% returned home initially, but subsequently moved on, 5% settled in a new site and 49% failed to move. The distribution of responses was identical for the rich and poor conditions, but fish that homed were dominant and had grown faster during the pre-displacement period.     

Size and age of maturity of Atlantic salmon correlate with the North Atlantic Oscillation Index (NAOI)

Sea‐age at maturity of Atlantic salmon Salmo salar decreased with increasing values of the seasonal NAOI from February to April. Body mass increment from smolts to adults of one‐sea‐winter Atlantic salmon increased with increasing NAOI in May at the time when the juveniles moved to sea.     

The effects of initial length and body curvature on shrinkage of juvenile Atlantic salmon during freezing

Fork length was measured in two groups of salmon parr (32–139 mm, frozen in a straight posture and frozen in a curved posture) before (L₁) and after (L₂) freezing and thawing. All the fish shrank. The decrease in length was significantly greater in the curved fish than the straight fish. The absolute reduction in length (L₁–L₂) was related directly to L₁, whereas the percentage reduction in length [(L₁–L₂)/L₁× 100] was related inversely to L₁.     

Compensatory sea growth of male Atlantic salmon, Salmo salar L., which previously mature as parr

The proportion of mature male parr in 11 families of Atlantic salmon, Sulrno sulur, reared under similar conditions in fresh water varied from 0–43%. The mature males were smaller than their siblings in December as 1 + and in late March. After individual tagging and transfer to a sea cage in early April. the previously mature males grew faster than previously immature salmon during the next 6 months. This compensatory growth resulted in almost equal size between the two groups. The results are discussed in relation to the different life strategies of salmon.     

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.     

Differences in growth between maturing and non-maturing male Atlantic salmon, Salmo salar L., parr

Sibling male Atlantic salmon parr that matured tended to be the larger fish in January, but their monthly specific growth rates between January and July did not differ from those of non-maturing fish. Maturing fish had lower condition factors in March, but greater increases in condition factor during April, exceeding those of non-maturing males by May. In maturing males, feeding rates between July and September, and specific growth rates in August and September, were lower than those of immature fish. Consequently, the mean size of immatures equalled or exceeded that of maturing males by October. Maturation rates were strongly correlated with increases in mean condition factor only during April.     

Marine post-smolt growth and age at maturity of Atlantic salmon

The annual variation in sea-age of maturation for a hatchery dependent stock of Atlantic salmon was compared to variation in post-smolt growth as evidenced by circuli spacing patterns. The proportion of returns of 1-seawinter (1 SW) and 2 SW salmon and the fraction of the smolt year class or cohort that maturated as 1 SW fish, were compared to seasonal growth indices determined from circuli spacing on the scales of smolt class survivors returning as 1 SW and 2 SW spawners. Using image processing techniques, we extracted inter-circuli distances from scales from 2244 recaptured fish. Spacing data for the first year at sea were collected and then expressed as seasonal growth indices for the spring period, when post-smolts first enter the ocean; the summer, when growth appears maximal; and winter, when growth appears to be at a minimum. In general, circuli spacings were wider for 1 SW than for the 2 SW returns of the same smolt cohort. The 1 SW fraction was significantly and positively correlated with late summer growth, suggesting that growth during this season is pivotal in determining the proportion of a smolt class that matures early.     

Patterns of growth and smolting in autumn migrants from a Scottish population of Atlantic salmon, Salmo salar L.

In late November 1990 salmon parr, Salmo salar L., from the Girnock Burn in northern Scotland were either caught on their feeding territories (n=25) or trapped during downstream migration (n= 18). They were then housed in a laboratory rearing tank and their food intake and growth rates were tracked, until their smolting status was ascertained in the following May. Female fish were predominant in both groups; although the range of ages was the same, the total age of migrants was 2+ while that of residents was 1+. In November, compared to resident fish of the same year class, migrants were larger, heavier and in better condition. Although growth rates dropped during the winter in both groups before increasing in spring, migrants ate more and consistently grew faster than residents. In seawater tolerance tests conducted in May, more residents than migrants failed to adapt. These results confirm the suggestion that autumn migrants smolt in the following spring and suggest that they represent the faster-growing component of their cohort.     

Diel activity patterns of juvenile Atlantic salmon in rivers with summer water temperature near the temperature‐dependent suppression of diurnal activity

Direct day and night underwater observations of juvenile Atlantic salmon Salmo salar during summer and autumn showed a duality in response to temperature between 7 and 11° C for young‐of‐the‐year (YOY) Atlantic salmon. They were predominantly diurnal in early summer and nocturnal in late summer although water temperatures were similar. Post‐YOY Atlantic salmon did not show a strong response to temperature as they were mostly nocturnal during the study period. It is suggested that the difference in activity patterns between YOY and post‐YOY Atlantic salmon can be explained by size‐dependent trade‐off between growth and predation risk.     

Food consumption and growth of Atlantic salmon Salmo salar parr in sub-Arctic rivers: empirical support for food limitation and competition

The quantitative food consumption and somatic growth of Atlantic salmon Salmo salar parr were compared between three sub-Arctic rivers in northern Norway and Finland, addressing the potential occurrence of resource limitation and interspecific competition. In one of the rivers, previous resource partitioning studies have suggested severe dietary competition between juvenile S. salar and a dense population of alpine bullheads Cottus poecilopus . It was hypothesized that S. salar parr in this river would have restricted food consumption and growth rates compared to the S. salar populations in the other two rivers where interspecific competition was less likely to occur. The feeding and growth performance differed significantly between the S. salar populations. The lowest food acquisition and growth rates were in the S. salar parr population living in sympatry with C. poecilopus , confirming a restricted food supply for the S. salar parr and providing empirical support for the presence of resource limitation and interspecific food competition in this river system. The study reveals that S. salar parr in sub-Arctic rivers may experience food limitations resulting in diminished growth rates.     

Changes in plasma protein patterns in smolting Atlantic salmon, Salmo salar L., are not dependent on changed growth rates

Juvenile Atlantic salmon, Salmo salar L., were raised under simulated natural photoperiod (SNP) and constant light (CL). SNP fish displayed typical characteristics of smolts in the spring including changes in colouration, elevated gill Na⁺/K⁺ ATPase activity and ability to pass a seawater challenge. CL fish grew at rates comparable to SNP fish, but did not demonstrate any of these molt characteristics.
Electrophoretic analyses of plasma proteins from both groups of fish revealed that SNP fish displayed reproducible qualitative changes in protein patterns from February to May with May patterns being the most complex. CL fish displayed patterns similar to SNP fish. The timing of the appearance of patterns was influenced by photoperiod as indicated by the fact that CL fish had a progression of pattern changes retarded by at least one month relative to SNP fish. The changes in plasma protein patterns in smolting Atlantic salmon were not dependent on changed growth rates.     

Counterintuitive migration patterns by Atlantic salmon Salmo salar smolts in a large lake

What little is known about the seaward migration of Salmo salar smolt migration through standing waters indicates that it is both slow and results in high mortality rates, compared with riverine migration. This may be partly because smolts in lakes need to swim more actively and require more complex directional cues than they do in rivers. In this telemetry study of smolt migration through Loch Lomond, S. salar smolts made repeated movements in directions away from the outflowing river, which considerably increased migration time.     

Growth rate correlations across life-stages in female Atlantic salmon

Within smolt ages there was a strong negative correlation between pre- and post-smolt growth in wild three-sea-winter female Atlantic salmon Salmo salar . This indicates that trait values providing, or resulting from, increased growth in one environment may cause growth to decrease in the other, potentially constraining growth rate evolution.     

Selection against late emergence and small offspring in Atlantic salmon (Salmo salar)

Timing of breeding and offspring size are maternal traits that may influence offspring competitive ability, dispersal, foraging, and vulnerability to predation and climatic conditions. To quantify the extent to which these maternal traits may ultimately affect an organism's fitness, we undertook laboratory and field experiments with Atlantic salmon (Salmo salar). To control for confounding effects caused by correlated traits, manipulations of the timing of fertilization combined with intraclutch comparisons were used. In the wild, a total of 1462 juveniles were marked at emergence from gravel nests. Recapture rates suggest that up to 83.5% mortality occurred during the first four months after emergence from the gravel nests, with the majority (67.5%) occurring during the initial period ending 17 days after median emergence. Moreover, the mortality was selective during this initial period, resulting in a significant phenotypic shift toward an earlier date of and an increased length at emergence. However, no significant selection differentials were detected thereafter, indicating that the critical episode of selection had occurred at emergence. Furthermore, standardized selection gradients indicated that selection was more intense on date of than on body size at emergence. Timing of emergence had additional consequences in terms of juvenile body size. Late-emerging juveniles were smaller than early-emerging ones at subsequent samplings, both in the wild and in parallel experiments conducted in seminatural stream channels, and this may affect success at subsequent size-selective episodes, such as winter mortality and reproduction. Finally, our findings also suggest that egg size had fitness consequences independent of the effects of emergence time that directly affected body size at emergence and, in turn, survival and size at later life stages. The causality of the maternal effects observed in the present study supports the hypothesis that selection on juvenile traits may play an important role in the evolution of maternal traits in natural populations.