The Atlantic salmon in fresh water: spawning, rearing and production

Summary

Fluvial salmonids have evolved to use the diversity of habitats in natural streams for different life history stages and at different seasons. Required freshwater habitat of Atlantic salmon can be classified generally as that suitable (i) for spawning, (ii) for feeding during the major growing period, and (iii) for overwintering.

Spawning habitat of salmon is usually in rapid water at the tail of pools on the upstream edge of a gravel bar, ideally with depths about 25 cm, in mean water velocities of about 30–45 cm s^-1, with maximum velocities about 2 body lengths s^-1, and with a substrate of irregularly shaped stones of cobble, pebble, and gravel.

Underyearling salmon (<7 cm TL) are most common in shallow (<15 cm) pebbly riffles, whereas older and larger parr (>7 cm TL) are usually in riffles deeper than 20 cm with a coarse substrate. Depth preference increases with size. Multiple linear regression models quantifying parr habitat have identified substrate as an important variable, with a positive relationship to an index of coarseness. Negative relationships were found with mean stream width, range of discharge, and overhanging cover. Water chemistry, especially alkalinity, nitrates, and phosphates, are important regulators of production. Although similar variables had importance, coefficients among rivers differed. Interactions occur among variables. Further studies are required to quantify productive capacity of habitat for parr. Results suggest that useful models can be derived and if a river system is mapped, and stratified by habitat, then smolt yield could be predicted and the required egg deposition could be estimated.

In winter, young salmon shelter among coarse substrate or move to pools, but continue feeding, with larger parr being more active.

Feeding is in general opportunistic. Food consists mainly of insects, taken primarily in the water column, but also from the surface and at the bottom. Young salmon in flowing water are highly territorial but are less so in slow or still waters. In fast water, parr use their large pectoral fins to apply themselves to the substrate, allowing them to occupy this type of habitat with little expenditure of energy. Height above the substrate decreases with water velocity, but increases with temperature and social status. Although riffles are preferred habitat, and are relatively more productive, lentic waters can be occupied where there are few predators or severe competitors and may provide significant smolt yield in some systems. Selective segregation minimizes competition between salmon and brook charr or brown trout, but brook charr and brown trout may have negative effects on underyearling salmon, and on parr in pools, whereas salmon have negative effects on small brook charr and brown trout in riffles and flats. Competition by both interference and exploitation results in interactive segregation when the resource, mainly food, becomes limiting.

Limited downstream movement of underyearling salmon may occur during the summer. Older juveniles may make upstream movements, but generally migrate downstream, with most movements in the spring, and a lesser peak of activity in the autumn. Dispersal tends to be mainly downstream, indicating that for full distribution, spawning areas are best located upstream. High densities of yearling parr may have negative effects on growth and survival of underyearlings in some river systems, but apparently not in others, so that future research is required in this regard. Density-dependent growth is evident where food is limiting, and can provide an indicator of densities of cohorts so that if a quantitative relationship has been derived, mean size from a sample can give an estimate of the density at that station, with minimum size occurring at carrying capacity. Such regressions vary between habitats with differing productive capabilities, so that future research could provide useful models for assessing productive capacity of a habitat, and optimum densities. Life history strategies can change with changes in density-dependent growth rates. Present stock-recruitment functions do not take environmental variables into consideration, and have limited applicability. Further research is required to determine optimum spawning requirements for salmon in different types of river systems in different geographical areas.