Elsevier

Fisheries Research

Volume 62, Issue 2, May 2003, Pages 143-170
Fisheries Research

Habitat requirements of Atlantic salmon and brown trout in rivers and streams

https://doi.org/10.1016/S0165-7836(02)00160-1Get rights and content

Abstract

The distributions and abundances of trout and salmon are strongly influenced by their habitat. The habitat includes both abiotic and biotic factors, which interact in complex webs. Habitat probably has strongest effects during population bottlenecks, when the standing stock approaches the carrying capacity of the environment. Various approaches to modelling interactions between habitat and population density and mean weight have been explored, but further work is needed in this area of investigation. The importance of depth, current, substrate, cover, and to a lesser extent, temperature and oxygen availability to the various stages of the life cycles of salmon and trout are briefly reviewed. By drawing on published data, it is possible to define broad ranges of acceptable conditions for the life stages of each species. However, it is not possible to partition this variation into between-population differences, within-population preferences, within-population tolerances, and effects of interactions between habitat variables. To pursue this important issue further, a structured approach using experimentation both in the field and in suitable laboratory systems is recommended. There is abundant evidence that habitat requirements of salmon and trout overlap. Trout tend to out-compete salmon except often in areas of particularly fast flows and, perhaps, remote from the river bank. The habitat requirements of year classes of salmon and trout overlap and therefore, there is scope for interactions between them depending on the spatial arrangement of habitats and the occurrence of bottlenecks. It is particularly important to understand where the bottlenecks to production lie and to focus on these in the first instance. Otherwise, there is a risk of manipulating habitat that is already in excess, or increasing numbers of a population that will subsequently be constrained, e.g., by over-wintering habitat. For this reason, it is prudent to accept that although manipulations of habitat may appear to be beneficial when considered locally, they should be measured and assessed where possible in terms of the production of returning adults and/or high quality smolts. Because of the complexity of interactions between salmon, trout, and the animals that eat them, it is at present difficult, or impossible, to derive good predictive models of the effects of manipulating habitats under many circumstances.

Introduction

This paper is part of a series dealing with the application of scientific knowledge to the management of Atlantic salmon, Salmo salar L., and brown trout, Salmo trutta L., in UK. The aim is to provide an overview of information available, predominantly in the primary literature, regarding the habitat requirements of salmon and trout, with an emphasis on its value to fisheries managers.

Habitat is usually understood to be the range of physical and chemical factors that affect an animal. Often these factors are considered to be those acting in the immediate vicinity of the animal. However, in reality factors may result from processes that impinge across a broad range of scales (Armstrong et al., 1999). In the spatial domain, scales may range from fractions of the home range, through reach, river, oceanic to global in extent. In the temporal domain, scales range from fractions of a second through to geological time. A review of the importance of scale to the application of science to the management of Atlantic salmon (Armstrong et al., 1999) concluded that it is not possible to generalise on how habitat limits fisheries, but that each population should be considered independently to identify the factor(s) that limit production.

Managers of salmonid fishes are usually concerned with manipulating, or protecting, habitat with the aim of maximising fish production (which we define here as the number of fish available to a fishery). In this context the habitat is best considered in terms of the range of physical, chemical and also biological factors that impinge on the growth and chances of survival of the species of interest. The important point here is that changes in the physical or chemical habitat may not have a direct effect on salmon or trout, however, they may have an indirect effect if they influence animals which then interact with the fish as predators, competitors or prey. Recent work has demonstrated experimentally that simple linear models do not provide adequate descriptions of the behaviour of freshwater communities, instead more complex networks are appropriate (Hulot et al., 2000). Johnson and Law (1995) aptly described the ecology of rivers as ‘intricate webs of interdependence’. It is within this context that habitat for salmonids should be considered. To give an idea of this complexity, examples of the biotic and abiotic factors that may affect salmon are shown in Fig. 1, Fig. 2. These figures are not necessarily complete accounts or unique classifications of the processes, nor do they apply to all river systems. Nevertheless, they show the challenge of complexity that faces managers searching for opportunities to manipulate populations of salmonid fishes. The focus of this review is on abiotic habitat, but some reference is made to biotic factors, particularly in relation to competition between trout and salmon and within species.

At any point in time during the development of a cohort of salmon or trout, habitat may limit growth and survival only if the densities of fish are sufficiently high relative to the size of the fish. There is an inverse relationship between size and density at carrying capacity (Bohlin et al., 1994). Which of these combinations of size and density maximises production of returning adults depends on the interaction between size and survival.

Salmon and trout may occupy their habitat fully only at some times of the year (Milner et al., 2003). For example, in a stream in which spawning was abundant, survival of brown trout was density-dependent during a critical period in the first several months after hatching, but not thereafter (Elliott, 1989). In this case, habitat may have been limiting only during the critical period. In some cases critical periods may occur later in life, e.g., during winter (Bjornn, 1971, Mason, 1976). To enhance habitat effectively, it is necessary to identify critical periods and focus in the first instance on these stages.

In principle, a dependence of stream salmonid abundance on habitat implies that it should be possible to derive predictive relationships between abundance and stream habitat features. If this is the case then such models may be useful tools for management by providing indices of habitat quality, predicting fish abundance (for comparison with observed abundance) and predicting the consequences of habitat manipulation (Milner et al., 1985, Milner et al., 1998). There have been many attempts to develop such empirical models, but comparatively few have been applied in management. Fausch et al. (1988) reviewed North American models and a more recent review of these and models from UK and other European work has been reported by Barnard and Wyatt (1995). These reviews showed that habitat models work, in that they can account for significant proportions of the variance measured in salmonid abundance. However, models developed for one localised stream type rarely work in other stream types or in other regions. The better models (in terms of explaining variance in fish abundance or predicting abundance in new data sets) combine both local site features (e.g. width, depth, substrate, cover, flow type, bankside vegetation) and catchment-scale variables (e.g. altitude, flow, stream order, geology, primary productivity indices).

One of the most successful modelling systems has been HABSCORE, which was developed originally for Welsh rivers (Milner et al., 1993) and was then extended to UK and Wales (Barnard et al., 1995, Wyatt et al., 1995). This suite of models, developed for stream-dwelling trout and Atlantic salmon, was derived from 600+ sites using a combination of site (typically 25–50 m long) variables measured by transect surveys and map-based variables, reflecting each site’s location in the catchment and the upstream subcatchment characteristics. The fish populations at each site were sampled by electrofishing. Models were derived using stepwise multiple regression for different size and age categories.

It is essential to evaluate the performance of habitat models in the context of the overall variability seen in salmonid populations, resulting from random stochastic factors as well as stock and recruitment processes (Milner et al., 2003). This variance, when measured over time and many different sites, includes spatial and temporal variance. Spatial variance is determined by factors associated with the location and physical features of each site and is thus the variance component that most habitat models attempt to explain. Thus variance has two important consequences for habitat models. First, if spatial variance is a low proportion of the total variance then no habitat (=spatial) models can have good explanatory power. Secondly, a direct estimate of model performance is the proportion of the existing spatial variance that it manages to explain. This varies depending on the ability of selected variables to represent the functional links between habitat and standing stock of fish. The national (UK and Wales) HABSCORE models for trout and salmon, when applied to independent test data sets, accounted for between 29 and 46% of overall variance (i.e. the maximum that any models could hope to explain) and between 49 and 70% of spatial variance (Wyatt et al., 1995). Ibbotson (1993) showed that a habitat quality index based on two factors (permanent instream cover and depth greater than 20 cm) explained 73% of the variation in trout population density. Many habitat models developed on data sets restricted to 1 year, explain up to 75% of population variance (Fausch et al., 1988), but this is misleading because they ignore temporal variation which can cause spatial variance to be comparatively low.

An analysis of HABSCORE models, based on smaller data sets where time series were available, showed that within 28 sites on the River Conwy, N. Wales, spatial variance was between 46 and 62% of the total for young-of-the-year (YOY) and post-young-of-the-year (PYOY) salmon and trout. In contrast, temporal variance ranged between 4 and 12% for the same groups (Milner et al., 1995). However, when variance was estimated at within-tributary level (five small streams <8 m wide), mean spatial variance ranged between 22 and 42% compared with a mean temporal variance range 24–39%. This means that, at tributary level, factors other than local site habitat (as measured by the scheme) were having an equally strong and synchronous influence on abundance. These results emphasise the importance of scale and put the importance of habitat into the context of overall variation in salmonid abundance.

HABSCORE models worked best for YOY and PYOY trout and YOY salmon, but were less effective for PYOY salmon. The reason for this finding is not clear, but may be because the method failed to measure effectively some habitat feature important to that group. An interesting point is that catchment features that are cheaply measured using GIS procedures, on their own explained significant proportions of the overall variance. For YOY trout, e.g., omission of the (expensive) site-specific variables still produced a model explaining 41% of spatial variance compared with 63% for the full model (Wyatt et al., 1995). The reason for this finding is the interdependence of stream channel structure with the fluvial dynamics and geology of a sub-catchment; catchment features are variously both independent factors in their own right and act as surrogates for the site-specific variables such as substrate size and flow types. The practical benefit of this finding is that simple models based on cheaper data may be useful in describing habitat quality on large scales and developing such methods is an area of current research (Milner et al., 1998).

A note of caution. Most models developed so far are empirical, statistical representations of the association between habitat features and abundance. The variables selected for such models are expected, from the literature and common understanding, to be those which have some functional determination of abundance. However, most habitat features are strongly correlated with each other (width, depth, velocity, substrate size, gradient, etc.) and the final suite of variables in the models may not necessarily include those which most influence a species/age group. This is because of the statistical procedures usually used (e.g. stepwise selection of variables), that cannot distinguish between factors directly controlling fish abundance from those which are merely surrogates, by virtue of their strong statistical association, with the proximate factor/s. Moreover, some important variables may be omitted from models because of dominance by correlated variables. Habitat–fish relationships are extremely complex and dynamic and it is naı̈ve to suppose that models are anything more than convenient simplifications of complex ecological processes. They are still very useful for predicting abundance, classifying river habitat and impact assessment, but caution should be applied when using them to predict the effects of altering specific habitat components to guide habitat improvement schemes.

Habitat models have been useful for highlighting habitat features likely to be of key importance, but have rarely, if ever, considered whether the populations measured were limited by habitat, numbers of spawning fish, or other bottlenecks outwith the monitoring period. By including density of fish early in summer as a potential explanatory variable for variation in density late in summer it was possible to increase significantly the explanatory power of a multivariate habitat model (Armstrong and Gardiner, 1995). This study was preliminary, but suggested that variation in spawning deposition may be dealt with. It has been suggested that a hierarchical approach to modelling the influence of habitat on production of salmon and trout may be a useful way forward to identify key limiting factors in any given system (Armstrong et al., 1999).

In populations limited by space and/or food, growth of fish is by necessity associated with mortality: a process termed self-thinning. Where self-thinning relationships (the regressions of log weight on log density and vice versa) apply it may be possible to measure habitat quality not by density or weight alone, but by the elevation of the relationship. It is not yet clear how widely this concept may be applied because evidence for self-thinning in natural populations is ambiguous (Armstrong, 1997, Milner et al., 2003). Self-thinning appears not to be compatible with the concept of a critical period in the first year of growth (Elliott, 1989, Egglishaw and Shackley, 1977), except during the critical period (Armstrong, 1997). If populations do self-thin during an early critical period then a negative relationship between weight and density of cohorts (as reported by Bohlin et al., 1994) would be expected to persist when the period has ended. The consequence would be that it may be possible to identify features of the habitat that limit the population during the critical period even if density and length of older fish are used as response variables or to generate thinning relationships. However, in the absence of knowledge of the dynamics of the population, correlations between number of parr and habitat characteristics may be misleading because they would have been driven by the relationships between habitat and fry densities.

A method that has been applied widely in N. America, particularly for regulating water flow, has been physical habitat simulation (PHABSIM) (Bovee, 1986) and similar models (Jorde, 1996; Capra et al., 1995; Heggenes et al., 1996a, Heggenes et al., 1996b; Peviani et al., 1996; Bartsch et al., 1996; Boudreau et al., 1996). These models include biological and hydraulic components. The biological component comprises habitat preference curves, which are plots of the number of organisms found in relation to the habitat variable (e.g. flow rate) biased by the availability of the habitat variable. So, e.g., if most salmon were found in areas of low flows and, moreover, low flows were rare, then the preference index for low flow would be very high. The hydraulic component describes the availability of habitat components, typically across a range of river discharges. In applying the models, it is usual to generate habitat preference curves at one discharge and to use these to predict how suitable the overall habitat will be as discharge changes. Recent work has shown that the method can be misleading when applied to Atlantic salmon and brown trout because preference curves are sensitive to density (Greenberg, 1994, Bult et al., 1999) and vary with discharge (Holm et al., 2002). New methods are needed to model effects of discharge on fish. The current focus is to develop bioenergetic models. It is not yet clear whether these models will be sufficiently generalisable to have widespread application in river management or whether their structures are sufficiently sound to avoid problems encountered with earlier models.

Despite the difficulties of developing models to link habitat variables to production of salmonid fishes, there is good evidence of which factors affect populations. The problem is quantifying the optimum level for each factor. In general, the population response to habitat variables may be described by one of two curves (Fig. 3). In the first case (Fig. 3a), the population increases as the variable increases until a plateau is reached. This may be the response to changes in over-wintering refuges. The proportion of the population safe from predation during the day increases until shelters are in excess, at which point the population may be limited by the number of fish that survive through the summer, so that further availability of refuges has no effect (Armstrong and Griffiths, 2001). In the second case, the population increases initially as the habitat variable increases. However, further increases in the habitat variable have a negative effect. This may represent the responses of populations to bankside vegetation. Initially, gains may be accrued. However, with a further increase in growth of vegetation the stream may become over-shaded with a concomitant reduction in production (O’Grady, 1993). There are likely to be interactions between habitat variables, which would be manifest as changes in the shapes of the curves depicted in Fig. 3. Moreover, the shapes of these curves are likely to vary between populations, being dependent on the genetic constitution of stocks (Youngson et al., 2003). Sometimes habitat variables may act together to limit production of fish. However, in other cases there may be one limiting factor. For example, an increase in boulders may increase production of fish, but only if food is not a limiting factor.

We have seen that quantifying the relationships between habitat variables and production of salmonids is a challenge that remains and will play an important part in developing sound river management plans for the future. However, it is also incumbent upon managers to use current knowledge to improve and conserve habitat quality and quantity (Kennedy, 1984a). A useful way of using our incomplete knowledge is to define in broad terms the ecological niches occupied by salmon and trout throughout their life cycles (Heggenes et al., 1999). The niche does not usually provide information on optimal habitat or the relative qualities of different habitats, but describes what types of habitat have been used successfully by the fish species. Such information provides a good starting point for fisheries managers.

Several abiotic factors are thought to be of particular importance in affecting production of trout and salmon, such as temperature, rate of water flow, fluctuations in discharge and availability of cover (see Binns and Eiserman, 1979). More recently, the consensus of opinion in the literature is that the habitat features most important to the distribution and abundance of salmonids are depth, current, substrate and cover (Heggenes, 1990). Here it is the aim to review these variables and additionally consider briefly the influence of oxygen concentrations and temperature on the habitat choice of Atlantic salmon and brown trout. Clearly, other factors may be of overriding importance. For example, chronic pollution and obstructions to upstream migration may devastate populations. These issues are dealt with by Hendry et al. (2003).

Our focus is on the lives of the salmon and trout in fresh water because here managers might have most influence on the fish. Habitat is considered within the categories: spawning, nursery, summer rearing, in-stream over-wintering and up-river migration stages of the life cycle. The distinction between nursery and rearing habitat follows Heggenes et al. (1999). For comparison and contrast, Atlantic salmon and brown trout are considered together within each of these categories. Information on how velocity, depth and substrate influence spawning, nursery and rearing habitat is summarised in Table 1, Table 2, Table 3. However, reference to the main text is strongly recommended to put this information in context.

Section snippets

Spawning habitats

Salmon and trout spawn during the autumn and winter months. In upland and northern areas of the range, spawning occurs mainly between October and December and may be confined to a period of only 2 or 3 weeks. However, in areas with less severe conditions the spawning period may be later (November–March) and extend over 4–7 months depending on temperature (Crisp, 1993, Bardonnet and Bagliniere, 2000). Compared to cyprinids, salmonids release a small quantity of large eggs: 1600–1800 per kg

Nursery habitat

The period during emergence and the establishment of feeding territories appears to be of critical importance in the dynamics of salmonid populations. This is a time when mortality can be very high and during which the strength of the cohort may be established and may be termed a critical period. In a study of trout the critical period extended for several months after emergence (Elliott, 1989), whereas in a study of salmon it extended until the first autumn after emergence (Egglishaw and

Rearing habitat

Parr may move from nursery areas during autumn and winter or may continue to grow near where they were spawned (Rimmer et al., 1983, Rimmer et al., 1984). However, they use different local habitats as they grow.

Over-wintering habitat

It has been recognised for many years that activity of juvenile salmon decreases as temperature declines. For example, Allen (1940) stated that salmon remained inactive in deep pools when water temperatures were below 7 °C. Underwater observations of Atlantic salmon fry in streams of the River Tay (Scotland) system suggested that activity levels during the day increased steadily as temperature increased above 5 °C (Gardiner and Geddes, 1980). Below 5 °C, most individuals were apparently hiding in

Habitat used during up-river migration

Cover, in the form of boulders, overhung banks and deep pools, is required by adult salmonids migrating upstream, both as protection from predators and to enable them to avoid bright sunlight (Crisp, 1996). Salmon migrating to spawning grounds may stay in pools and large water bodies for extensive periods (Hawkins and Smith, 1986; Gowans et al., 1999a, Gowans et al., 1999b). Temperature influences migratory activity of adult fish. The rate of upstream migration appears to be impaired by low

Discussion

This review illustrates wide differences between the results of studies reporting the habitats used by many of the life stages of salmon and trout. This variation may be explained by at least four factors. First, there may be genetic adaptation to local habitats. This adaptation may include changes in the morphology of the fish to suit local habitats. For example, the pectoral fins of salmon may be larger, relative to body length, in fast-flowing streams (Riddell et al., 1981). Because of this

Acknowledgements

We thank colleagues at the Salmonid 21C conference for general comments on this review and Ross Gardiner and Dick Shelton for comments on an earlier version of the manuscript.

References (145)

  • S.M. Einarsson et al.

    Utilisation of fluvial and lacustrine habitat by anadromous Atlantic salmon, Salmo salar L., in an Icelandic watershed

    Fish. Res.

    (1990)
  • K. Hendry et al.

    Management of habitat for rehabilitation and enhancement of salmonid stocks

    Fish. Res.

    (2003)
  • R.M. Acornley et al.

    Sediment transport and siltation of brown trout (Salmo trutta L.) spawning gravels in chalk streams

    Hydrol. Process.

    (1999)
  • J.S. Alabaster

    The temperature requirements of adult Atlantic salmon, Salmo salar L., during their upstream migration in the river Dee

    J. Fish. Biol.

    (1990)
  • K.R. Allen

    Studies on the biology of the early stages of the salmon (Salmo salar). 1. Growth in the river Eden

    J. Anim. Ecol.

    (1940)
  • K.R. Allen

    Studies on the biology of the early stages of salmon (Salmo salar). 2. Feeding habits

    J. Anim. Ecol.

    (1941)
  • J.D. Armstrong

    Self-thinning in juvenile sea trout and other salmonid fishes revisited

    J. Anim. Ecol.

    (1997)
  • Armstrong, J.D., Gardiner, R., 1995. Juvenile trout populations in the Loch Leven catchment. Fisheries Research...
  • J.D. Armstrong et al.

    Density-dependent refuge use among overwintering Atlantic salmon juveniles

    J. Fish Biol.

    (2001)
  • J.D. Armstrong et al.

    Redistribution of juvenile salmonid fishes after localised catastrophic depletion

    J. Fish Biol.

    (1994)
  • J.D. Armstrong et al.

    The application of science to the management of Atlantic salmon: integration across scales

    Can. J. Fish. Aquat. Sci.

    (1999)
  • G.P. Arnold et al.

    The role of the pectoral fins in station-holding of Atlantic salmon parr (Salmo salar L.)

    J. Exp. Biol.

    (1991)
  • R. Bachman

    Foraging behaviour of free ranging wild and hatchery brown trout in a stream

    Trans. Am. Fish. Soc.

    (1984)
  • R.J. Baldes et al.

    Physical parameters of microhabitats occupied by brown trout in an experimental flume

    Trans. Am. Fish. Soc.

    (1969)
  • A. Bardonnet et al.

    Freshwater habitat of Atlantic salmon (Salmo salar)

    Can. J. Fish. Aquat. Sci.

    (2000)
  • A. Bardonnet et al.

    The influence of potential predators on the habitat preferenda of emerging brown trout

    J. Fish Biol.

    (1994)
  • Barnard, S., Wyatt, R.J., 1995. An analysis of predictive models for stream salmonid populations. Bull. Fr.Pech Piscic....
  • Barnard, S., Wyatt, R.J., Milner, N.J., 1995. The development of habitat models for stream salmonids and their...
  • Bartsch, N., Chad, P.G., Hardy, T.B., 1996. Determining habitat criteria for the endangered fountain darter through...
  • K.F. Beland et al.

    Water depth and velocity preferences of spawning Atlantic salmon in Maine rivers

    N. Am. J. Fish. Man.

    (1982)
  • N.A. Binns et al.

    Quantification of fluvial trout habitat in Wyoming

    Trans. Am. Fish. Soc.

    (1979)
  • Bird, D.J., Lightfoot, G.W., Strevens, A.P., 1995. Microhabitat use by young salmon and trout in a southern UK chalk...
  • T.C. Bjornn

    Trout and salmon movements in two Idaho streams as related to temperature, food, stream flow, cover, and population density

    Trans. Am. Fish. Soc.

    (1971)
  • T. Bohlin

    Habitat selection and intercohort competition of juvenile sea-trout Salmo trutta

    Oikos

    (1977)
  • T. Bohlin et al.

    The energetic equivalence hypothesis and the relation between population density and body size in stream-living salmonids

    Am. Nat.

    (1994)
  • Boudreau, P., Bourgeois, P., Leclerc, M., Boudreault, A., Belzile, L., 1996. Two-dimensional habitat model validation...
  • Bovee, K.D., 1986. Development and evaluation of habitat suitability criteria for use in the instream flow incremental...
  • E.E. Bridcut

    A study of terrestrial and aerial macroinvertebrates on river banks and their contribution to drifting fauna and salmonid diets in Scottish catchments

    Hydrobiologia

    (2000)
  • T.P. Bult et al.

    Density-dependent habitat selection by juvenile Atlantic salmon (Salmo salar) in experimental riverine habitats

    Can. J. Fish. Aquat. Sci.

    (1999)
  • H. Capra et al.

    A new tool to interpret magnitude and duration of fish habitat variations

    Regul. Rivers

    (1995)
  • D.W. Chapman

    Food and space as regulators of salmonid populations in streams

    Am. Nat.

    (1966)
  • D.W. Chapman

    Critical review of variables used to define effects of fines in redds of large salmonids

    Trans. Am. Fish. Soc.

    (1988)
  • Clark, E.D., 1998. Microthermal habitats in British rivers. Ph.D. Thesis. University of Exeter, 294...
  • P. Collen

    The influence of riparian tree selection on stream invertebrate production

    Scott. For.

    (1994)
  • D.T. Crisp

    The environmental requirements of salmon and trout in fresh water

    Freshw. Forum

    (1993)
  • D.T. Crisp

    Dispersal and growth rate of 0-group salmon (Salmo salar L.) from point-stocking together with some information from scatter stocking

    Ecol. Freshw. Fish

    (1995)
  • D.T. Crisp

    Environmental requirements of common riverine European salmonid fish species in fresh water with particular reference to physical and chemical aspects

    Hydrobiologia

    (1996)
  • Crisp, D.T., 2000. Trout and Salmon. Ecology, Conservation and Rehabilitation. Blackwell Scientific Publications,...
  • D.T. Crisp et al.

    Observations on siting, dimensions and structure of salmonid redds

    J. Fish Biol.

    (1989)
  • R.A. Cunjak

    Behaviour and microhabitat of young Atlantic salmon (Salmo salar) during winter

    Can. J. Fish. Aquat. Sci.

    (1988)
  • R.A. Cunjak et al.

    Winter habitat utilization by stream resident brook trout (Salvelinus fontinalis) and brown trout (Salmo trutta)

    Can. J. Fish. Aquat. Sci.

    (1986)
  • R.A. Cunjak et al.

    Atlantic salmon (Salmo salar) in winter: the season of parr discontent

    Can. J. Fish. Aquat. Sci.

    (1998)
  • D.A. DeGraaf et al.

    Habitat use by and preference of juvenile Atlantic salmon in two Newfoundland rivers

    Trans. Am. Fish. Soc.

    (1986)
  • P.W. DeVore et al.

    Daytime responses of brown trout (Salmo trutta) to cover stimuli in stream channels

    Trans. Am. Fish. Soc.

    (1978)
  • H.J. Egglishaw et al.

    An experiment on faster growth of salmon Salmo salar L. in a Scottish stream

    J. Fish Biol.

    (1973)
  • H.J. Egglishaw et al.

    Growth, survival and production of juvenile salmon and trout in a Scottish stream, 1966–1975

    J. Fish Biol.

    (1977)
  • H.J. Egglishaw et al.

    Survival and growth of salmon, Salmo salar, planted in a Scottish stream

    J. Fish Biol.

    (1980)
  • H.J. Egglishaw et al.

    Influence of water depth on dispersion of juvenile salmonids, Salmo salar L. and S. trutta L., in a Scottish stream

    J. Fish Biol.

    (1982)
  • A. Eklov et al.

    Influence of water quality, habitat and species richness on brown trout populations

    J. Fish Biol.

    (1999)
  • S.R. Elliot et al.

    Spatial variation in environmental characteristics of Atlantic salmon (Salmo salar) rivers

    Can. J. Fish. Aquat. Sci.

    (1998)
  • Cited by (0)

    View full text