G.J.A. Kennedy
Fisheries Research Laboratory
Coleraine, Londonderry, Northern Ireland
ABSTRACT
Atlantic salmon were stocked at the same density in successive years in upland brown trout streams as eyed ova, green eggs and swim-up fry. Their survival was evaluated by electric fishing during the late summer, and the results were compared with the survival rates in the naturally spawning trout population. Eyed ova showed the best survival rate and swim-up fry were marginally less successful, although the latter was in a year when competition was more intense due to higher trout fry densities. Green eggs gave the lowest survival rate, and this was attributed to the scouring effect of floods. Similarly, the effect of scouring was related to the lower survival of eyed ova in a high gradient stream compared to one of low gradient.
The presence of trout and older age classes of salmon was also found to affect the survival and distribution of planted salmon. When a stream contained trout and salmon parr the survival of salmon stocked as eyed ova was less than half that found when other fish were absent. A reduction in the survival of stocked salmon due to the competitive influence of trout fry was also noted below a stream barrier where much trout spawning had taken place. The survival of stocked salmon was also reduced considerably by the presence of yearling salmon parr in the stream. Changes in the trout population were similarly recorded following salmon stocking, and these suggested that while salmon fry do not affect trout fry survival, the presence of salmon parr does cause a reduction in trout fry stocks.
These competitive effects were related to the habitat preferences of salmon and trout. When the two species were living sympatrically about 70 percent of the fry of both were captured in sites of less than 20 cm mean depth, whereas only about 30 percent of the salmon fry population were found in this depth range when no trout or older salmon were present. Yearling fish were found in all the depth ranges sampled, but trout were limited in their distribution to areas of lower flow while salmon were not. Older trout were found mainly in the deeper areas.
Finally, an electrofishing survey of a range of depth related habitat types below a stocked area indicated that dispersal of stocked salmon fry was very limited. This result and the other findings are discussed in relation to their implications for salmon stocking exercises in trout streams.
RESUME
On a repeuplé des torrents à truites avec des saumons atlantiques en utilisant plusieurs années de suite la même densité d'oeufs embryonnés, d'oeufs verts et de fretin non nourri. Les taux de survie ont été évalués par pêche électrique en fin d'été; on les a comparés aux taux de survie des populations de truites frayant naturellement. C'est avec les oeufs embryonnés que l'on a obtenu le taux de survie le plus élevé. Les alevins mis à l'eau non nourris venaient juste après (et encore s'agissait-il d'une année où le frai de truite était très dense, d'où une concurrence particulièrement forte). En revanche, les résultats obtenus avec les oeufs verts étaient nettement moins bons, ce qui serait dû à l'effet d'appouillement des crues. C'est ce même phénomène qui expliquerait que le taux de survie des oeufs embryonnés soit plus faible dans les cours d'eau à forte dénivellation.
On a également constaté que la présence de truites et de saumons plus âgés avait un effet sur la survie et la répartition du matériel de repeuplement. Ainsi, le taux de survie des oeufs embryonnés diminuait de plus de moitié en présence de truites et de saumoneaux. On a également noté une réduction du taux de survie là où le frai de truites était très dense, notamment en aval d'un obstacle où beaucoup de truites avait frayé. La réduction était très nette en présence de saumons d'un an. On a observé des modifications des populations de truites à la suite de ces opérations de repeuplement. A ce qu'il semble, le frai de saumon n'a pas d'influence sur la survie du frai de truite dont les stocks en revanche, diminuent en présence de saumoneaux d'un an.
Ces effets de concurrence étaient liés aux préférences du saumon et de la truite en matière d'habitat. Lorsque les deux espèces étaient sympatriques, on capturait environ 70 pour cent de leur frai dans des eaux de moins de 20 cm de profondeur en moyenne alors que l'on n'y capturait que 30 pour cent environ du frai de saumon lorsqu'il n'y avait pas de truites ou de saumons plus âgés. On a trouvé des juvéniles d'un an à toutes les profondeurs où l'on a prélevé des échantillons mais, à la différence des saumons, la répartition des truites se limitait aux zones à faible débit. Les truites âgées hautaient surtout les profondeurs.
Une étude par pêche électrique de différents types d'habitat, plus ou moins profonds, dans une zone repeuplée a montré que la dispersion du frai de saumon était très limitée. L'auteur tire de tous ces résultats un certain nombre de conclusions concernant le repeuplement en saumons du torrents à truites.
Harris (1978) defined the basic constraint on any fishery management programme for increasing the numerical abundance of salmon (Salmo salar L.) in a river as the productivity of the water and its ability to produce and sustain fish stocks. If a fishery manager feels that the full potential of a river for the production of salmon smolts is not being realized due to inadequate recruitment, inaccessibility or pollution kills, artificial stocking is frequently carried out using a variety of techniques. These efforts have met with varying degrees of success in different rivers (Harris, 1978; Symons, 1979) but an evaluation of the physical and competitive factors regulating the survival of stocked salmon, and the impact on resident fish populations have not yet been fully quantified.
The work presently being undertaken in Northern Ireland covers various aspects affecting the survival of stocked salmon including comparative stockings of eyed ova, green eggs and swim-up fry in the same stream environment, comparative stockings of eyed ova in different streams, and the influence of resident brown trout (Salmo trutta L.) and older age classes of salmon on the survival and distribution of stocked salmon. The reciprocal effect of the impact of stocked salmon on the growth and survival of the resident trout population has also been monitored.
The study areas are the two feeder streams to Altnahinch Reservoir in the upper catchment of the River Bush (Fig. 1). The dam at the reservoir acts as a barrier to upstream migration from the River Bush, and brown trout were found to be the only resident fish species in the streams prior to the experiments.
The streams rise at an altitude of about 360 m on heather moorland which is subject to reafforestation. The larger of the two streams (stream 1) has impossible falls approximately 2.4 km upstream from the reservoir, and this investigation was limited to the area below this natural barrier. The overall gradient on stream 1 is about 1:35, and the mean width 3.7 m. The corresponding values for stream 2 are 1:22 and 1.8 m. The total catchment for both streams is about 12 km2. The pH is about 7.2 in low flow conditions, but this can drop to pH 6.3 during floods due to run off from the peaty moorland. The bed of both streams consists mainly of coarse gravel and moss covered stones, although bedrock and peaty silt were found in some of the pools.
Fish barriers were erected on both streams during 1977 to prevent the immigration of trout to the areas above (see Fig. 1). On stream 1 this barrier was situated about 800 m upstream from the reservoir, providing a downstream control area which was accessible to trout and an upstream experimental area which was inaccessible. On stream 2 the barrier was situated at the bottom of the stream, thus isolating the whole area from trout immigration.
All the salmon stocked in the experimental streams were obtained from River Bush adults taken in traps on the lower reaches of the river. All eggs and fry for stocking were retained in a hatchery situated beside the trapping installations prior to the experiments. All ova were placed in artificial redds using the pipe method described by Sedgwick (1960). The redds were separated by distances of less than 150 m. Salmon fry at the “swim-up” stage, i.e., with some yolk sac left, were stocked out evenly over all the available habitat using buckets. Green eggs were stocked out during December, eyed ova during March and swim-up fry during May. In all the experiments salmon were stocked out at an overall stream density of 6.2/m2.
Population estimates were carried out during March and late August to early September by electrofishing using a two anode technique to obtain high efficiencies (Kennedy and Strange, 1981). The sites sampled during March were chosen so that each covered a range of habitat types, and varied in area from 170 m2 to 297 m2. The summer sites varied in area from 11.3 m2 to 182 m2, and stop nets were sited to enclose, as far as possible, one depth related habitat type at a time, i.e., deep pool, shallow pool, riffle, etc. Water depths were measured on a grid system, and gradients using a manometer method.
The fork length of all fish was measured to the millimetre below, and weighings were carried out both on anaesthetized fish in the field and on electrofishing casualties. In this work, 1 April was taken as the start of the salmonid year.
The mean summer densities and percentage survival of salmon fry which had been stocked in stream 1 by the three methods described are given in Table 1 for the years 1977–81. Eyed ova consistently produced the best survival rates with no significant difference between the three years when this stocking method was employed, the summer densities falling within the range 108.3 to 120.2 per 100 m2. The survival rate from swim-up fry was marginally significantly lower than the best from eyed ova (p < 0.1), with a summer density of 103.3 per 100 m2. However, this was in a year when trout fry densities were very high due to intense trout spawning in this area, and the resultant increased competition apparently influenced salmon fry survival (see section 4.3). Green eggs gave a highly significantly lower survival rate than eyed ova or fry stocking (p <.001), with a mean summer fry density of only 25 per 100 m2. Kennedy and Strange (1981a) confirmed this result by further sampling during March 1978 and March 1979, when annual survival rates from eyed ova and green eggs were found to be 10.3 percent and 1.4 percent, respectively. They attributed the lower survival rate from green eggs to the scouring effect of winter floods which destroyed several redds completely. It seems probable that man is not so efficient at building redds as adult salmon, although Harris (1978) recorded that up to 30 percent of natural redds can also be washed away before the eggs hatch out.
Plantings of eyed ova are not so subject to losses due to scouring since the redds constructed for them only have to last a relatively short period of time and do not have to face the brunt of winter floods. However, even with these the effect of scouring was noted at Altnahinch.
Eyed salmon ova were stocked at the same densities in the experimental areas above the barriers in streams 1 and 2 during March 1981. No other age classes of salmon and trout were present in these areas at this time. The artificial redds were similarly constructed in each stream. During the week following planting severe flooding occurred in the area. Observation subsequently indicated that some of the redds in the high gradient stream 2 had been scoured out and others damaged, whereas redds appeared to be intact in stream 1 which has an overall lower gradient. Although the gradient of stream 1 could not be considered low at 1:35, long stretches are much less steep than this, whereas stream 2 is more uniformly steep.
Electrofishing in the summer confirmed that there was very poor survival of the salmon stocked in stream 2 compared to the experimental area of stream 1 (Table 2). The mean density of salmon fry in the experimental area of stream l was over five times greater than that in stream 2, where the distribution of fry was patchy, i.e., some redds had survived and some had not. Thus although eyed ova stocking has consistently produced good survival rates in an area of lower gradient, this method of stocking may not be so effective in torrential high gradient areas if flooding occurs before the alevins leave the redds.
Stocking of a 490 m stretch of the control area below the barrier on stream l was carried out using swim-up salmon fry during May 1980. Care was taken to distribute the fry evenly over the whole area. However, when sites were electro-fished over the length of the stocked area during August and September 1980, it was found that there was a highly significant (p <.001) downstream trend for increasing abundance of salmon fry relative to trout fry abundance (Fig. 2). This trend was non-linear, and a quadratic regression gave a significantly better fit to the data (p <.01). Neither the trout fry themselves, nor any of the older fish present showed any significant trends over the length of the stocked area. However, as pointed out by Kennedy (1980), many adult trout spawned in the upstream end of the stocked area after their spawning migration was blocked by the impassable barrier. Many more trout fry would, therefore, have hatched out directly downstream from the barrier than could possibly acquire territories and survive. Kalleberg (1958) has pointed out that in interspecies territorial disputes trout tended to be more successful than salmon. It is likely, therefore, that during the first few critical weeks of establishing territories the salmon in the upstream end of the stocked area were under much greater pressure from the large numbers of emerging trout fry in that area than were the salmon stocked further downstream. Downstream dispersal in this environment was found to be very limited (see section 4.7) and competition for space from large numbers of trout fry directly below the barrier has, therefore, had a very localized influence on the survival of stocked salmon fry.
In earlier work described by Kennedy and Strange (1980) the survival of stocked salmon was also found to be influenced by the presence of older age-classes of salmon parr in the stream. The summer densities of salmon fry derived from ova stocking when streams contained trout and salmon parr was only about half that recorded when trout alone made up the resident stock (Fig. 3). Similarly, after removal of all the fish from the experimental area in stream 1 the summer salmon fry density from eyed ova stocked in 1981 was more than twice that found in the downstream control area which contained trout and salmon parr (Table 2). The salmon fry were also larger in the experimental area with a mean size of about 6.2 cm compared to only 5.2 cm in the control area. Although the work has not yet resolved the relative importance of inter and intra species competition, the biomass calculations of Kennedy and Strange (1980) suggest that there is also a certain amount of niche segregation between the two species. Both the streams at Altnahinch were found to be carrying only about two thirds of their standing crop capacity prior to the introduction of salmon (Fig. 4).
Egglishaw and Shackley (1973) suggested that the introduction of eyed salmon ova from a lowland hatchery gives the salmon an artificial advantage over the naturally spawned trout due to earlier hatching of the salmon eggs - advanced because of higher hatchery temperatures. Kennedy and Strange (1980) did not find any unequivocal influence of introduced salmon fry on trout fry survival at Altnahinch, although some increased competition may have been evident in a significant reduction in the growth rate of trout fry in stream 1 following salmon stocking. However, substantial declines in the densities of trout fry were recorded in both streams following the second introduction of salmon (Fig. 3). The inference is that although salmon fry do not affect trout survival, as also recorded by Le Cren (1965), the presence of salmon parr does cause a reduction in trout fry stocks. This is contrary to the results of Egglishaw and Shackley (1980), who found no evidence that salmon stocking had a detrimental effect on any age class of trout in a stream. Certainly the work at Altnahinch did not demonstrate any consistent change in yearling trout densities or growth rates which could be attributed to the introduced salmon.
Various authors have suggested that variation in fry survival may be related to the amount of predation by older age classes of both salmon and trout (McCrimmon, 1954; Mills, 1964). However, Kennedy and Strange (1980) noted only occasional predation by trout on fry, and no predation at all by salmon parr on fry. This led them to suggest that competition for space is a more important regulating influence on fry survival. Other work being carried out in the Altnahinch streams is designed to determine the extent of this competition for space by quantifying the distribution of each age-class of each species in relation to water depth and flow (Kennedy and Strange, 1982).
The results indicated consistent significant correlations of density with water depth over three years of sampling. The findings for one year (1977) are summarized in Fig. 5 as the density of each age class in each depth range expressed as a percentage of the overall abundance of that age class. Fry were highly significantly more abundant in shallow water sites than deeper sites (p < .001). During the three years 71.8 percent of the salmon fry and 64.5 percent of the trout fry were captured in sites of mean depth less than 20 cm. The older trout were highly significantly more abundant in deeper water (p < .001), with only 6.3 percent of these being captured in sites shallower than 20 cm mean depth. The yearling fish showed an intermediate relationship, being found in all the depth ranges sampled, but with a tendency for higher numbers in mid-range depths, significant in the case of trout (p < .05). There were similar correlations in the abundance of each age class with the actual areas of shallow, mid-range and deep water habitat available within sites, whatever their mean depth.
The distribution of salmonids in relation to water flow was assessed from site gradients (Table 3). Salmon fry were significantly more abundant in high gradient riffle areas in all years, but trout fry only showed this trend significantly in 1978. The older trout were significantly less abundant in areas of high gradient confirming that they are found in the slow-moving pools. Yearling trout were also highly significantly negatively correlated to stream gradient, but the yearling salmon did not show any significant correlation to this parameter in any year, i.e., yearling salmon are apparently able to exploit a whole range of flow regimes while trout are limited to slower flowing areas. Jones (1975) has previously pointed out that salmon are better adapted than trout for living in fast flowing water due to their having larger pectoral fins, which can be used for holding the fish to the bottom by deflection of the current. This ability of salmon to occupy areas to which trout are not so well adapted concurs with the results described in section 4.3, where trout on their own apparently did not occupy all the available habitat (Kennedy and Strange, 1980). These results also confirm that yearling salmon can occupy the same habitat as, and by inference compete with, fry of both species.
A change in the habitat occupied by a species resulting from the addition or removal of a closely related species is regarded as the strongest and most direct evidence of interspecific competition for a resource (Diamond, 1978). This criterion was used as a test of the interspecific influence of trout on salmon by carrying out the 1981 stocking of the experimental section of stream 1 after removal of all the resident fish. The distribution of salmon fry in this area and in the control area which contained all age classes of salmon and trout was investigated in relation to habitat type as before.
The results show that the distribution of salmon fry changes considerably in the near absence of competition from other fish (Fig. 6). In the control area the distribution of salmon fry is significantly correlated to that found in previous years (p .01) with 84.7 percent of them being captured in sites with mean depths less than 20 cm. However, in the experimental area there is no significant correlation of the salmon fry distribution with either those in the control area or in previous years. Only 30.4 percent of the salmon fry were captured in sites of less than 20 cm mean depth, and high densities of fry were found in all the depth ranges sampled. In fact, in the absence of other fish it appears that the shallowest areas are least preferred by salmon fry with the mode of abundance occurring at mean depths of 15 to 25 cm. This is a similar distribution to that previously recorded for 1+ salmon and may be related to the more rapid growth of fry in the absence of competition from other fish. A study of the distribution of fry in the presence of salmon parr but in the absence of trout has yet to be completed.
An assessment of the extent of downstream dispersal of salmon planted as swim-up fry was made at Altnahinch in 1980. The fry were planted out in May over a 490 m stretch below the barrier in stream 1. In addition to the electrofishing carried out to assess survival within the planted area (see section 4.3) a number of sites were sampled during late August for a distance of 330 m below the most downstream stocking point (Fig. 7). Salmon fry were almost totally absent from the deepwater sites in this unstocked area, and in the five sites which were less than 25 cm deep, the mean density of salmon fry was only 23.7 per 100 m2 (the mean density of these at comparable sites in the stocked area was 119.7 per 100 m2). Over 70 percent of the salmon fry captured below the stocked area were taken within 100 m of the lowest stocking point, and 200 m below the stocked area the density had fallen to 8.9 per 100 m2.
These findings are comparable to those of Egglishaw and Shackley (1973) in the Fender Burn, where the downstream density of salmon fry below a site planted with eyed ova tailed off following an exponential decay pattern until no fry were captured at sites further than 400 m below the stocking point. These authors also noted that very few fry moved upstream, and the density 120 m above a stocked site was only 10 percent of that recorded at the stocking site.
Both these studies, therefore, confirm limited dispersal of stocked salmon and are contrary to the earlier work of Elson (1962) on the Pollett River, where he reported that planting of salmon fry at sites 1.5 miles apart resulted in “… relatively uniform dispersal and good utilization of the stream throughout the entire experimental area”. However, Elson's plantings were carried out at the end of August, and fry planted at this stage must show a great tendency for dispersal than fry newly emerged from redds. In fact, the Altnahinch findings suggest that salmon stocked out as swim-up fry do not even disperse as far as fry emerging from a redd. This may be due to the limited yolk sac available to the fish at this stage, when they must quickly acquire a territory or die. Certainly these results emphasize the need for an even distribution of fry during planting out operations.
Percentage survival rates are frequently quoted as a measure of the success of stocking exercises, but as pointed out by Mills (1964) and Egglishaw and Shackley (1980) these are related to initial stocking densities, with mortality rates increasing as densities increase. A more definitive measure is the holding capacity of a stream for fry. Le Cren (1965) assessed this at about 7 to 10 fish per m2 during September for trout fry in small streams in the north of England in the absence of other fish. The present studies indicate tht the overall summer holding capacity of Altnahinch stream 1 for salmon fry in the presence of salmon parr and trout is in the range of 103.3 to 120.2 per 100 m2. This was increased to 229.8 salmon fry per 100 m2 in the absence of other fish. These figures were all derived from initial stocking densities of about 620 salmon per 100 m2. However, the minimum initial stocking density required to achieve these summer densities has not yet been investigated. As reviewed by Symons (1979), this type of assessment is vital if unnecessary mortality of stocked fry is to be avoided.
Elson (1975) considered that an egg deposition of 168 per 100 m2 (140 per 100 yards2) was the optimum required for the Miramichi River in Canada. However, the assessment of juvenile survival from which this was derived (Paloheimo and Elson, 1974) indicated that underyearling densities in this river system are much lower than those in the Altnahinch streams i.e., the holding capacity of the Miramichi for juvenile salmon appears to be much lower than at Altnahinch. It is not clear if this is due to less extreme climatic conditions in Northern Ireland, a greater degree of enrichment producing a more abundant food fauna in Northern Ireland streams, differences in the densities of other competing fish species, or is simply due to the fact that the assessment in the Altnahinch streams was carried out in a more suitable nursery habitat for juvenile salmon than the depth and flow regimes encountered in the Miramichi. A combination of these factors is the most likely, but the implications are clear - it is most unwise to extrapolate optimum stocking densities from one river system to another.
At Altnahinch it was found that the effect of scouring by floods can influence the survival of planted ova considerably. Mills (1973) suggested that under natural conditions gradient is one of the main factors limiting salmon spawning in certain streams, with gradients of greater than about 3 percent causing inaccessibility to adult salmon. However, the results described by Kennedy and Strange (1980) indicated that if scouring of planted eggs does not occur, the high overall gradient of stream 2 at Altnahinch did not in itself produce lower holding capacities in terms of fry densities. It was also confirmed that in the presence of salmon parr and trout, salmon fry can inhabit areas of localized high gradient (Kennedy and Strange, 1982). However, the results also indicated that the overwinter mortality or emigration of salmon fry from the higher gradient stream 2 was greater than in stream 1, and the standing crop in terms of biomass was consistently lower. The implication is that calculations of optimum stocking density should also be based on the holding capacity of a stream for salmon parr as well as salmon fry, otherwise the unnecessary mortality effect on stocked fish may simply be delayed from the first year of life to the second. Certainly, the stocking of high gradient streams with ova at any stage can be a hit and miss affair depending on the post stocking weather conditions. Fry stocking in such areas would probably give more consistent success, although this has not so far been tested. What has been confirmed is that optimum utilization of the available habitat in any stream will only occur if care is taken to distribute ova or fry evenly over the river. Stocking only adjacent to convenient bridges or access points will give unsatisfactory results.
While it appears that overall stream gradient can regulate salmon parr survival, the results at Altnahinch also indicated that these will exploit all habitat types in a lower gradient stream. For this reason, Kennedy and Strange (1982) consider that intraspecific competition for space between salmon parr and fry is a critical regulating factor affecting fry survival. Certainly Kennedy and Strange (1980) recorded only half the survival rate of salmon fry stocked in the presence of salmon parr compared to when salmon parr were absent. Stocking operations should, therefore, also be regulated by the density of salmon parr already present in a stream. There may even be a case for alternating years of heavy and light stocking in streams, but the efficacy of this has not yet been assessed.
Similarly, the evaluation of the effect of resident trout on the survival of planted salmon has not yet been completed. Initial results at Altnahinch have confirmed the influence of trout fry on salmon fry. Also, much higher summer densities, improved growth rates and a full utilization of all the available habitat by salmon fry have been recorded in the absence of salmon parr and trout. Only after observations on the survival and distribution of salmon fry in the presence of salmon parr without trout can the relative effects of inter- and intra-specific competition be properly assessed. This work is scheduled for the Altnahinch streams in 1982.
The results of the studies at Altnahinch have certainly confirmed that the survival of stocked salmon is dependent on a number of inter-related factors, and indicate the potential for production of in Northern Ireland rivers. However, optimum utilization of a salmon resource available for stocking requires a careful assessment of local conditions and resident fish populations. The holding capacity and recommended stocking densities for one area cannot be considered as definitive and extrapolated ubiquitously throughout the range for the species.
I wish to thank my colleagues Mr. C.D. Strange and Mr. B.T.M. Hart for their assistance and interest in all aspects of the work involved in this study.
Diamond, J.M., 1978 Niche shifts and the rediscovery of interspecific competition. Am.Sci., 66:322–31
Egglishaw, H.J. and P.E. Shackley, 1973 An experiment on faster growth of salmon Salmo salar (L.) in a Scottish stream. J.Fish Biol., 5:197–204
Egglishaw, H.J., 1980 Survival and growth of salmon, Salmo salar (L.) planted in a Scottish stream. J.Fish Biol., 16:565–84
Elson, P.F., 1962 Predator-prey relationships between fish-eating birds and Atlantic salmon. Bull.Fish.Res.Board Can., (133):87 p.
Elson, P.F., 1975 Atlantic salmon rivers, smolt production and optimal spawning; an overview of natural production. Int.Atl.Salmon Found.Spec.Publ.Ser. (6):96–119
Harris, G.S. (ed.), 1978 Salmon propagation in England and Wales. A report by the Association of River Authorities/National Water Council Working Party, 62 p.
Jones, A.N., 1975 A preliminary study of fish segregation in salmon spawning streams. J.Fish Biol., 7:95–104
Kalleberg, H., 1958 Observations in a stream tank of territoriality and competition in juvenile salmon and trout (Salmo salar L. and S. trutta L.). Rep.Inst.Freshwat.Res., Drottningholm, (39):55–98
Kennedy, G.J.A., 1980 A study of the factors influencing the production of salmon and trout in the River Bush. 1. Salmon/trout relationships at Altnahinch. Annu.Rep.Res.Tech.Work.Dep. Agric.N.Ire., (1980):142–4
Kennedy, G.J.A. and C.D. Strange, 1980 Population changes after two years of salmon (Salmo salar L.) stocking in upland trout (Salmo trutta L.) streams. J.Fish Biol., 17:577–86
Kennedy, G.J.A., 1981 Efficiency of electric fishing for salmonids in relation to river width. Fish.Manage., 12(2):55–60
Kennedy, G.J.A., 1981a Comparative survival from salmon (Salmo salar L.) stocking with eyed and green ova in an upland stream. Fish.Manage., 12(2):43–8
Kennedy, G.J.A., 1982 The distribution of salmonids in upland streams in relation to depth and gradient. J.Fish Biol., 20:579–91
Le Cren, E.D., 1965 Some factors regulating the size of populations of freshwater fish. Mitt.Int.Ver.Theor.Angew.Limnol., 13:88–105
McCrimmon, H.R., 1954 Stream studies on planted Atlantic salmon. J.Fish.Res.Board Can., 11(4):362–403
Mills, D.H., 1964 The ecology of the young stages of the Atlantic salmon in the River Bran, Ross-shire. Freshwat.Salm.Fish.Res., 32:1–58
Mills, D.H., 1973 Preliminary assessment of the characteristics of spawning tributaries of the River Tweed with a view to management. Int.Atl.Salmon Found.Spec.Publ.Ser., 4(1):145–55
Paloheimo, J.E. and P.F. Elson, 1974 Reduction of Atlantic salmon (Salmo salar) catches in Canada attributed to the Greenland fishery. J.Fish.Res.Board Can., 31(9):1467–80
Sedgwick, D.S., 1960 Planting salmon. Salmon Trout Mag., 160:204–10
Symons, P.E.K., 1979 Estimated escapement of Atlantic salmon (Salmo salar) for maximum smolt production in rivers of different productivity. J.Fish.Res.Board Can., 36(2):132–40
Table 1 Details of the population densities of trout fry and stocked salmon in stream 1 at Altnahinch as assessed by electrofishing during late August and early September in the years 1977–81. The percentage survival of salmon stocked each year at a standard density of 6.2/m2 as eyed ova, green eggs or swim-up fry is also included
| Year | Salmon stocking method | Mean summer fry density (N 100/m2) | ||
| Trout (from natural spawning) | Salmon (from artificial stocking) | Percentage survival of stocked salmon | ||
| 1977 | Eyed ova | 36.6 | 120.2 | 19.4 |
| 1978 | Green eggs | 34.9 | 25.0 | 4.0 |
| 1979 | Eyed ova | 24.9 | 118.8 | 19.2 |
| 1980 | Swim-up fry | 91.3 | 103.3 | 16.7 |
| 1981 | Eyed ova | 31.4 | 108.3 | 17.5 |
Table 2 Mean summer densities and percentage survival of salmon stocked as eyed ova at a standard density of 6.2/m2 in two streams of differing gradient at Altnahinch in 1981. All fish had been removed from the experimental areas prior to salmon stocking, whereas the control area contained salmon parr and all age classes of trout
| Stream | Overall stream gradient | Area sampled | Details of stocked salmon survivors | ||
| Mean Density (N 100/m2) | Percentage survival | ||||
| 1 | 1 : 35 | { | Control area | 108.3 | 17.5 |
| Experimental area | 229.8 | 37.1 | |||
| 2 | 1 : 22 | Experimental area | 41.9 | 6.8 | |
Table 3 Correlation coefficients of the summer densities of each age class and salmon and trout in relation to the gradient at each site during 1977, 1978 and 1979 in the Altnahinch streams
| Species | Age class | Year | ||
| 1977 | 1978 | 1979 | ||
| Trout | 0+ | 0.1147 | 0.3540* | 0.3702 MS |
| 1+ | -0.6636*** | -0.5265*** | 0.6818*** | |
| 2++ | -0.5588** | -0.6518*** | -0.5442** | |
| Salmon | 0+ | 0.4610** | 0.3877* | 0.7585*** |
| 1++ | -0.1844 | -0.2166 | -0.0472 | |
MS p <0.1 marginally significant
Fig. 1 Map of the experimental Altnahinch streams in Northern Ireland showing the position of fish barriers
and the electrofishing sites samples during August/September (
) and March (▼)
Fig. 2 Densities of stocked salmon fry expressed as percentages of trout fry densities at sites in a 490 m stretch below an impassable stream barrier during August and September 1980
Fig. 3 Mean population densities of trout and stocked salmon (numbers 100 m-2)
in the two Altnahinch streams during March 1976, 1977 and 1978
(0+□ ; 1+
; 2+ and older ■)
Fig. 4 Mean biomass of trout and stocked salmon (100 g m-2) in the two Altnahinch
streams during March 1976, 1977 and 1978 (trout □; salmon )
Fig. 5 Population densities of each age-class of salmon and trout captured in
various depth ranges, expressed as a percentage of their overall abundance,
in the two Altnahinch streams during August and September 1977
(0+□; 1+; 2+ and older ■)
Fig. 6 Population densities of salmon fry (N 100 m-2) captured in various depth ranges (a) in the presence of salmon parr and all age-classes of trout in the control area of stream 1 (■), and (b) in the absence of other fish in the experimental area of stream 1 (□) at Altnahinch during August and September 1981
Fig. 7 Relative densities of salmon fry during August and September 1980 expressed as
a percentage of the mean density in the stocked area (103.3 per 100 m2) at
varying distances downstream from the lowest stocked point in Altnahinch stream
1 (sites < 25 cm mean depth 
; sites > 25 cm mean depth 
). The curve has
been fitted by eye to the results from sites < 25 cm deep. Stocking of the
upstream area had been carried out with swim-up fry at a density of 6.2 m-2.
