Ranching of Atlantic Salmon (Salmo Salar L.) to the Rod from a Native and Non-Native System

CM 2002/T:05

Ranching of Atlantic salmon (Salmo salar L.) to the rod from a native and non-native system.

G. Rogan, R. Poole, N. O' Maoileidigh and K. Whelan.

Marine Institute, Salmon Management Service Division, Newport, Co. Mayo, Ireland.

ABSTRACT The behaviour of the Burrishoole salmon ranch strain, released from a non-native system, was compared with behaviour in its native system. The study showed that the juvenile Burrishoole salmon required a period of acclimation at the remote site prior to release due to the elevated stress levels associated with the transfer. Underwater observations showed that some of the reared smolts also required a period of acclimation after release. Results from the Irish coded wire tagging programme show the exploitation of Burrishoole fish, released at the remote site, in the Irish coastal drift net fishery occurred over a wide area along the Irish coast. Burrishoole adults returned to the location of release rather than their native system. The number of Burrishoole adults recaptured in freshwater from other systems was low and was dependent on local environmental conditions. A greater rod catch of Burrishoole fish was recorded in the non- native system.

INTRODUCTION

Went (1955) referred to the salmon, as the most valuable fish in the Irish economy. In recent years there has been increased awareness of the potential value of the recreational fishery in Ireland. Several studies have been carried out to calculate a value to the Irish economy in terms of income and employment from salmon angling (O’Connor, 1984; Whelan and Whelan, 1986; Whelan and Marsh, 1988). The sea trout decline on the Irish West Coast (Whelan, 1991) has shown how vulnerable recreational fisheries can be. If a similar decline were to happen with salmon could donor salmon stocks be ranched to give sufficient returns to a rod fishery?

The ability to predict the behavioural patterns of introduced stocks is a useful tool in relation to management decisions for ranching to the rod. This study was carried out to determine if the Burrishoole salmon ranched strain showed similar patterns of behaviour, at the smolt and adult stages, in a non-native system compared to it’s native system. This could provide valuable information for management decisions in relation to ranching of Atlantic salmon from non-native systems in Ireland. The Burrishoole ranch strain was derived from wild Burrishoole fish in 1964. The progeny were artificially reared at Burrishoole and used to establish a self sustaining ranch stock. Full trapping facilities and a rod fishery at Burrishoole, together with data from the Irish coded wire tag recovery programme, provides a unique set of data for the Burrishoole ranch strain. this includs marine survival, coastal exploitation, return rate to freshwater and angling exploitation rates.

Juvenile Burrishoole salmon were transferred in two successive years from Burrishoole to the

1 remote site. The transfer took place three months prior to release and growth rates monitored against a control group at Burrishoole. Following release post smolt behaviour was observed by scuba diving at both sites. Rod catch, rod effort and location of capture were compared between fisheries. In addition environmental conditions, inshore netting and straying were also examined.

Materials and Methods

Study Sites Burrishoole System

The Burrishoole system (Fig. 1.) flows into the north east corner of Clew Bay on the west coast of Ireland near Newport (53 57 N; 9 35 W). It is an oligotrophic system with a catchment of 109 square kilometres and consists of two main lakes, Lough Feeagh 410 ha and Lough Furnace a brackish lake of 172 ha which is connected to the sea by the Burrishoole River (3.5km). Lough Feeagh is a freshwater lake and is discharged through two channels, the Salmon Leap and the Mill Race into Lough Furnace. Permanent fish traps are situated on both discharge channels and trap all fish moving upstream or downstream between the two lakes. Rearing facilities for the ranched salmon stock is situated between the two Loughs on the Mill Race discharge channel and water supply is gravity fed from Lough Feeagh. Angling at Burrishoole is restricted to fly-fishing from boats on both Lough Feeagh and Lough Furnace.

Remote site (Delphi) The remote system (Fig. 1.) is situated 60 kilometers south of Burrishoole between the Mweelrea Mountains to the west and the Sheffry Hills to the east. The system consists of three lakes, Glencullin 54 ha, at the top of the system, Doo Lough 250 ha, and Finlough 20 ha. The system drains into Killary Harbour through the Bundorragha River, which is 2.5km long and drains a total catchment of 52 square kilometres. Killary Harbour is a fjord like inlet, which is 13km long, and 700m wide (Keegan and Mercer, 1986). Along with Delphi the other main river which drains into Killary is the Erriff which has a catchment of 166.3 square km2 (O'Farrell et al., 1989). The rearing facilities are situated between Doo Lough and Finlough. Angling is carried out on the three lakes and on the river. Boat fishing only is carried out on the lakes with a maximum of 2 rods per boat. A maximum of 3 boats are permitted on Doo Lough, 2 on Finlough and 1 on Glencullin. A maximum of 4 rods is permitted on the Bundorragha river. The majority of angling is by artificial fly but a limited amount of trolling for salmon is permitted on Doolough up to July.

Fig. 1. Location of Burrishoole and Delphi systems

The salmon stock used in this study were line bred Burrishoole ranched grilse. In both years a control group of 10,000 juvenile salmon were reared at the Burrishoole rearing unit until release as smolts. A second group of 10,000 juvenile salmon were transferred 60km by truck in an oxygenated transporter tank to a rearing unit at the remote location. These fish were on-reared for three months prior to release as smolts. Both smolt groups were coded wire tagged according to the methods of Browne (1982). Water temperatures were taken at both rearing sites. Length and weight measurements of smolts were taken at Burrishoole prior to the transfer and then monthly at both sites prior to release. In both years the Burrishoole fish were transferred to the remote site during February and released in the following April. At Burrishoole the control group was released into Lough Furnace and at the remote site smolts were released into Finlough. Underwater observations by SCUBA diving were also carried out at both sites for several days after the releases. Details of the experimental groups in Irish coastal waters were gathered as part of the Irish national coded wire tag recovery programme. Adult return rates to freshwater at Burrishoole were calculated from returns to the trapping facilities and rod fishery. Adult reared salmon captured in both rod fisheries, identified by an adipose finclip, were sampled and cored to remove the microtag and the microtags read by Department of Marine staff. Capture date, time, location weight (kg), length (cm) and sex were recorded. Fishing effort at both sites was measured in rod days based on an eight-hour day. A proportion of the fin clipped salmon catch from a draft net fishery adjacent to the remote site was purchased and sampled using the same procedure used in the rod fisheries.

Results

In year 1 the group of juvenile salmon transferred to the remote site had an average weight of 35.7g at the time of transfer in February and were released at an average weight of 42.8g. The control group at Burrishoole had an average weight of 37.3g in February and 49.3g at the time of release. In year 2 the transferred group had an average weight of 35.7g at the time of transfer in February and were released at an average weight of 41.1g. The control group at Burrishoole had an average weight of 34.7g in February and 45.1g at the time of release. There was no significant difference in mean weight between the two groups of pre-smolts at Burrishoole prior to the transfer of one of the groups to the remote site in either year, Year 1 (p = 0.98), Year 2 (p=0.27). At the time of release the control group was significantly heavier in both years, (p>0.0001). The specific growth rates were higher for the Burrishoole smolts reared at Burrishoole than those transferred to remote site during the three-month period in both years. A negative specific growth rate was recorded in February at the remote site in Year 1 (-0.04) and March in Year 2 (- 0.1). A negative specific growth rate was not recorded in either year at Burrishoole. Rearing densities at the remote site ranged from 12.9-14.7kgm-3 in both years and 12.4 - 27.4kgm-3 in year 2. Rearing densities at Burrishoole ranged from 12.4-27.4 kgm-3 in year 1 and 18.3-40.5 kgm-3 in year 2. The water temperatures in year 1 at the time of transfer were 3.9 0C at Burrishoole and 4.00C at the remote site. The water temperatures at the time of release were 8.5 0C at Burrishoole and 8.00C at the remote site. In year 2 water temperatures at the time of transfer were 6.0 0C at Burrishoole and 5.50C at the remote site. The water temperatures at the time of release were 7.5 0C at Burrishoole and 5.00C at the remote site. Underwater observations during the smolt releases at both sites showed that when smolts were released away from a water current they formed a shoal and moved away from the release site slower than when released close to a current. No smolts were observed to move upstream of the release point. Smolts were observed 10 days after release but not after 16 days at both sites. The return rate to the trapping facilities at Burrishoole was 2.3% in year 1 and 2% in year 2.

In year 1 of adult returns it was estimated that 3,000 salmon were caught in the draft nets. A sample of 944 fish examined 313 (33.2%) were reared fish of which: 86 (47.8%) were Burrishoole fish released at the remote site. The following year the Killary draft catch was estimated between 800 - 1000 salmon 26.5% were identified as reared fish of which 44 (40.7%) identified as Burrishoole fish released from the remote site in 1992, and 2 (1.9%), as Burrishoole fish released from the remote site in 1991. In both years Burrishoole fish released from Burrishoole were identified in the catch.

The rod catch of the control group at Burrishoole was 48 fish in year 1 and 25 fish in year 2. The rod catch of Burrishoole fish at the remote site was 37 fish in year 1 and 69 fish in year 2. The CPUE of the control group at Burrishoole was 0.05 in year 1 and 0.03 in year 2. The CPUE of Burrishoole fish at the remote site was 0.06 in year 1 and 0.07 in year 2.

Fig. 2. Location of capture of rod caught Burrishoole adults in the native and non-native systems

LOCATION OF CAPTURE OF CONTROL GROUP IN LOCATION OF CAPTURE OF BURRISHOOLE ADULTS IN NON- BURRISHOOLE ROD FISHERY NATIVE ROD FISHERY

40 80 30 60

above release point above r el ease poi nt ROD CATCH 20 ROD CATCH 40 below release point bel ow r el ease poi nt 10 20 0 0 Year 1 Year 2 Year 1 Year 2

The location of capture in the rod fisheries (Fig. 2.) shows that in both years the majority of rod catch Burrishoole fish were caught below the smolt release location.

A total of 10 Burrishoole fish were recorded on the Erriff system during the first year of adult returns from a total rod catch of 615 fish, 9 were caught on rod and line and one in the upstream trap. All of the Burrishoole fish were from the group released from the remote site. The following year 1 Burrishoole grilse released as a smolt from the remote site the previous year was reported in the Erriff fishery from a total rod catch of 845 salmon.

DISCUSSION

Ranching Atlantic salmon to the rod in a non-native system is dependent on their ability to imprint and accurately home to the location of release. The mechanisms for imprinting and homing have been reviewed by Hasler and Scholz, (1983) and Quinn and Dittman, (1992) and show that memory of the home stream is not inherited and juvenile salmon can imprint to the stream of release even if it is not its natal system. During the imprinting period juvenile salmon learn olfactory clues associated with their home water which will be used as a guide to return them as adults to these waters. It has been suggested that Atlantic salmon are particularly sensitive to olfactory learning during the parr smolt transformation, Morin et al. (1992). Therefore when ranching from a non-native system it is essential that the juvenile salmon are in the release location during this critical period. Juvenile salmon were transferred three months prior to release, firstly to ensure they were at the release location during the parr smolt transformation and secondly to provide a period of time to acclimate from the transfer. Schreck et al. 1989 reported that yearling coho salmon released to migrate to the ocean before they had time to recover from stress yielded significantly fewer adults. Handling and transportation have been shown to increase stress resulting in increased plasma cortisol levels Schreck et al., (1989).

5 Loading of fish onto transport vehicles being particularly stressful to salmonids, Barton et al., 1980; Specker and Schreck, 1980; Barton and Peter 1982; Maule et al., (1988). This study also demonstrated that transportation of smolts to a new location could result in increased stress levels. Although plasma cortisol levels in the pre smolts could not be measured at either site during the study the differences in growth rates between the two sites suggest that the transfer did result in increased stress levels. A negative specific growth rate was recorded only at the non-native site. This occurred in both years following transfer. Although rearing densities have been shown to influence growth whereby high densities can depress growth rates, Refstie and Kittlesen, (1976), in both years the highest rearing densities were recorded at Burrishoole. Similarly water temperature also effects growth, but the temperature profile was similar at both sites. The specific growth rates recorded at both sites during April, prior to smolt release, would suggest a reduction in stress levels in the fish at the remote site. It would appear therefore that in relation to ranching from a non native site, juvenile salmon do require a reasonable period of acclimation prior to release due to the potential for elevated stress levels, particularly if it effects the imprinting process.

The timing of wild salmon smolt migration has been selected through evolution to provide optimum survival Hvidsten, et al. (1995). In the case of ranched smolts the timing of release is generally pre-determined. As it was difficult to determine the readiness in relation to smolt migration of individual ranched Burrishoole smolts prior to release the wild Burrishoole smolt run was used as an indictator of release time. As the Burrishoole ranch strain used in this study originated from the wild Burrishoole strain, the peak downstream migration of wild smolts was used as an indicator of release time for the ranched groups. Hvidsten and Johnson (1993) showed a higher adult recapture rate for hatchery reared smolts released during the peak wild salmon smolt run. The historical data at Burrishoole shows that the main peaks of the wild salmon smolt run occur after periods of high water discharge. In relation to salmon smolt releases, particularly in systems with lakes, releasing smolts following increased water levels may provide directional clues for smolts to find their way out of the system. At both Burrishoole and the remote site all of the smolts were released on a single day into lakes. The underwater observations carried out showed that some of the reared smolts did require a period of acclimation as some smolts were observed in fresh water 10 days after release. Skilbrei et al. suggests that the chronology of smoltification may be complex and vary between individuals, (1994). This is evident on the Burrishoole system where the mean duration for the completion of 95% of the wild salmon smolt run since 1970 was 53 days (Byrne et al. in press). Therefore releasing smolts into a lake provides a period of acclimation for smolts which may not be fully ready for migration on the release date. At both sites when smolts were released near an outflow they orientated in the current and moved slowly away from the release point. However when smolts were released away from the current at the remote site they formed a large shoal close to the release site. This shoaling behaviour may have been a result of immediate disorientation when released away from the main current velocity. Shoaling has been suggested to be an effective means of mitigating threat from predators Neill and Cullen, (1974). Shoaling may therefore have been carried out as a natural protective mechanism when released into a new environment with reduced directional clues.

Results from the Irish coded wire tagging programme show the exploitation of Burrishoole fish

6 released at the remote in the Irish coastal fishery occurred over a wide area along the Irish coast. This is similar to the location of capture of Burrishoole adults released as smolts from Burrishoole since 1981 in the Irish coastal fishery. This programme has consistently shown that exploitation of these fish occur both to the north and south of the Burrishoole system O' Maoileidigh et al., (1994). Thus adult Burrishoole fish returning from the feeding grounds do not all arrive off the Irish coast at the same location or necessarily close to the river they are returning to. This is similar to the findings of Went (1970), who from tagging and releasing adult salmon off the Irish coast did not establish any definite pattern of migration in inshore waters. He concluded that inshore migration is determined by whether the salmon reached the coast at a point above or below its release site. In relation to the homing migration of maturing Atlantic salmon Hansen et al. (1993) suggest it consists of two phases; a crude navigation from the feeding grounds to home waters and a more precise navigation in the coastal and estuarine waters. They also suggest that salmon learn to locate their home river during their smolt migration rather than using a genetically fixed memory. The results from this study would agree with these findings, as Burrishoole fish released from the remote system were not attracted to the Burrishoole system as they moved along the Irish coast. Nor were they attracted by population specific pheromones back to the Burrishoole system as proposed by Nordeng (1977). The fact that some Burrishoole adult fish returned to the remote system after two sea winters suggests that the odour memory at the smolt stage is long lasting.

Along with an ability to home, salmon also have a natural tendency to stray into non-native systems whereby new habitats are colonised Labelle, (1992). Straying has been defined as the migration of mature individuals to spawn in a stream other than the one where they originated Quinn, (1993). The straying rates for Atlantic salmon are generally considered to be low, Altukhov and Salmenkova (1994) in reviewing straying estimated a mean of 3.9%. Although Burrishoole adults released as smolts from both Burrishoole and the remote site were recorded in the Killary draft nets in both years, only those imprinted to the remote site were recorded in the nearby Erriff system. The Erriff catchment is the nearest freshwater system to the remote site and has a much bigger catchment (166.3 square km compared to the remote site at 51.8 square km); it is also more accessible to fish during periods of low water that the remote site. That fact that more Burrishoole fish were caught in the Erriff in the year of low water conditions suggests that the Erriff did not act as a general attractant for Burrishoole fish returning to the remote site. The numbers recorded were low and whether or not these fish would have remained in the system is unknown. Ricker (1975) showed that salmon may enter non-imprinted systems only to move out again prior to the spawning period.

It is evident that Burrishoole smolts can home to a non-native system with a high degree of accuracy and thus increasing the potential stock available to the rod. Whelan et al. 1994 showed in relation to ranching to the rod that increased rod catches provides for greater marketing potential for a fishery, which in turn resulted in increased levels of rod effort. Catch effort is a function of both abundance and catchability and CPUE assumes constant catchability, Bunt, (1990). The abundance of fish available to a rod fishery is determined by such factors as marine survival and coastal exploitation, whereas the catchability within a rod fishery will depend on factors such as environmental conditions, angler skill, fishing methods and in relation to reared fish, location of release. However rod catch only measures stock levels during the fishing season and its measure of stock is therefore dependent on the percentage of the total stock that is

7 available to anglers during the fishing season. In relation to this study availability of fish to the rod was shown to be particularly dependent on environmental conditions. Water flow has been shown to influence timing of entry of salmon to freshwater Smith et al. (1994) with increases in flow acting as an attractant to freshwater entry (Banks, 1969; Alabaster, 1970; Gee, 1980). Low water conditions occurred at both sites at the start of the angling season, during the first year of adult returns, however the effect on rod catch was different. On the Burrishoole system returning adults were able to access Lough Furnace, which is tidal, but were unable to migrate beyond the smolt release point and into Lough Feeagh. This resulted in a higher exploitation rate on ranched fish than in year two when higher water levels allowed access into Lough Feeagh. Mills and Piggins (1983) also reported higher exploitation rates of reared fish below the point of release at Burrishoole. They suggested that this might be related to an 'unsettled' behaviour of returning adults once past their homing point, reducing their catchability. Low water conditions at the remote site prevented access to the system at the start of the angling season and resulted in a lower rod catch in year 1. The fact that rod catch was lower in the year of the highest stock abundance as determined by returns to the Irish coastal fishery and to the Burrishoole trapping facilities indicate that rod catch data alone can give a false impression of overall stock survival. Harris (1988) suggests that rod catch data alone, without incorporating fishing effort is unreliable in estimating stock levels. Catch per unit effort is a more widely used measure (Ricker, 1975; Prouzet & Dumas, 1988) and when applied to the remote site also gives a false impression of overall stock survival. Therefore in relation to ranching to the rod good marine survival may not necessarily provide good angling. In the case of adults returning to the remote site the delay in entering freshwater reduced the availability of fish to the rod fishery and in addition resulted in these fish being more susceptible to the local draft nets. However in year 2 when water levels were sufficient to allow access to the remote site throughout the angling season it resulted in a greater rod catch and catch per unit effort of Burrishoole fish in the non-native system. On a system with a wild salmon population this increased rod effort could have an adverse effect. However, this study shows that the location of capture in the rod fishery of Burrishoole fish returning to a non native system is similar to that in its own system. The majority of fish were rod caught below the release location. Therefore careful consideration should be given to the location of smolt release as it could be used as a means of reducing exploitation rates on wild fish. At present on the Burrishoole system angling is only permitted in the lower lake which is below the point of the salmon smolt releases and a mandatory catch and release policy for wild fish is in operation. As a result the exploitation rate of wild fish has decreased from 24% to 0.3%.

In summary the behaviour of the Burrishoole ranch strain in the non-native system was similar to that in the native system. Although reduced growth rate at the remote site following transfer suggests increased stress levels, the return of the fish to the remote site suggests that it did not interfere with the imprinting process. Adult Burrishoole fish returning to the remote site showed a similar pattern of migration along the Irish coast to fish returning to Burrishoole. The rod catch data shows that a higher rod catch of Burrishoole fish is attainable in a non-native system. Therefore the behaviour of the Burrishoole stock ranched from other locations could be predicted with a high degree of certainty if local factors are taken into consideration.

References

8 Altukhov, Yu. P. and Salmenkova, E.A. (1994). Straying intensity and genetic differentiation in salmon populations. Aqua. Fish. Mgmt. 25 supp 2: 99-120. Anon. (1990-1994). Annual Reports of the Salmon Research Agency of Ireland. Salmon Research Agency, Newport, Co. Mayo. Banks, J.W. (1969). A review of the literature of the upstream migration of adult salmonids. J. Fish. Biol. 1: 85-136. Barton, B.A. and Peter, R.E. (1982). Plasma cortisol stress response in fingerling rainbow trout, Salmo gairdneri Richardson, to various transport conditions, anaesthesia, and cold shock. J. Fish. Biol. 20: 39-51. Barton, B.A. and Peter, R.E. and Paulenc, C.R. (1980). Plasma cortisol levels of fingerling rainbow trout, Salmo gairdneri , at rest and subjected to handling confinement, transport and stocking. Can. J. Fish. Aquat. Sci. 37: 805-811. Browne, J. (1982) First results from a new method of tagging salmon-the coded wire tag. Fishery leaflet,Depoartment of Fisheries and Forestry, Dublin, 1982. Bunt, D.A. (1990). Analysis of migratory rod catch effort data in Wales. In Catch Effort sampling Strategies: Their application in freshwater fisheries management ( I.G. Cowx ed.). Fishing News Books, Oxford. Byrne, C.J. Poole, R., Rogan, G. Dillane M. and. Whelan K. F.(2002) Temporal and environmental influences on the variation in salmon (Salmo salar L.) smolt migration in the Burrishoole system in the west of Ireland from 1970 – 2000. Gee, A.S. (1980). Angling success for Atlantic salmon Salmo salar in the river Wye in relation to effort and flows. Fish. Mgmt. 11 (3): 131-138. Hansen, L.P.,Jonnson, N. and Jonnson, B. (1993). Oceanic migration in homing Atlantic salmon. Anim. Behav. 45. Harris G.S. (1988). The status of exploitation of salmon in England and Wales. In Atlantic Salmon : Planning for the Future ( D. Mills and D. Piggins ed.). Croom Helm, London : 169- 190. Hasler, A.D. and Scholz A.T. (1983). Olfactory Imprinting and Homing of Salmon: investigations into the mechanism of imprinting process. Springer-Verlag, Berlin. Hvidsten, N.A., Jensen, A.J., Vivas,H., Bakke, O. and Heggberget, T.G. Nordic J. Freshw. Res. 1995 70:38-48. Keegan, B.F. and Mercer, J.P. (1986). An oceanographical survey of Killary Harbour on the west coast of Ireland. Proceedings of the Royal Irish Academy Section B. Biological, Geological and Chemical Science. 86: No. 1. Labelle, M. (1992). Straying patterns of coho salmon Oncorhynchus kisutch stocks from south east Vancouver Island, British Columbia. Can. J. Fish. Aquat. Sci. 49: 1843-1855. Maule, A.G., Schreck, C.B., Bradford, C.S. and Barton, B.A. (1988). The physiological effects of collecting and transporting emigrating juvenile chinook salmon past dams on the Columbia River. Trans. Am. Fish. Soc., 117: 245-261. Mills, C.P.R., Mahon, G.A.T. and Piggins, D.J. (1986). Influence of stock levels, fishing effort and environmental factors on angler's catches of Atlantic salmon, Salmo salar L., and sea trout, Salmo trutta L. Aqua. Fish. Mgmt. 17: 289-297. Morin, P.P. and Doving, K.B. (1992). Changes in the olfactory function of Atlantic salmon, Salmo salar in the course of smoltification. Can. J. Fish. Aquat. Sci. 49: 1704-1713. Neill, S.R. and Cullen, J.M. (1974). Experiments on whether schooling by their prey affects the

9 hunting behaviour of cephalopods and fish predators. J. Zool. (London) 172: 549-569. Nordeng, H. (1971). Is the local orientation of anadromous fishes determined by pheromones? Nature, 233 (5319): 411-413. Nordeng, H. (1977). A pheromone hypothesis for homeward migration in anadromous salmonids. Oikos, 28: 155-159. O’Connor, R. (1984). Economic Importance of Salmon in Ireland. ESRI Memorandum Series No. 165. O' Farrell, M.M., Whelan, K.F., Joyce, T. and Whelan, B. J. (1989). The performance of the river Erriff salmon fishery (1984-1988). In Proceedings of the Institute of Fisheries Management 20th Study Course (T.Cross and B. Ottway, ed.) O'Maoileidigh, N., Browne J., Cullen A., Mc Dermott T. and Keatinge M. (1994). Exploitation and survival of reared Burrishoole salmon released onto the Burrishoole river. Department of Marine, Fishery Leaflet 161. Palsson, J.O. and Einarsson, S.M. (1992). The effect of starvation on early maturation of arctic charr Salvelinus alpinus L. ICES C.M. 1992/M:7 Piggins, D.J. and Mills, C.P.R. (1985). Comparative aspects of the biology of naturally produced and hatchery reared Atlantic salmon smolts (Salmo salar L.). Aquaculture, 45: 321-333. Prouzet, P. and Dumas, J. (1988). Measurement of Atlantic salmon spawning escapement.In Atlantic Salmon : Planning for the Future (D. Mills and D. Piggins ed.). Croom Helm, London: 325-343. Quinn, T.P. (1993). A review of homing and straying of wild and hatchery produced salmon. Fisheries Research 18: 29-44. Refstie, T and Kittelsen, A. (1976). Effects of density on growth and survival of artificially reared Atlantic salmon. Aquaculture 8:319-326. Ricker, W.E. (1975). Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Bd. Can. 191, 382pp. Schreck, C.B., Solazzi, M.F., Johnson, S.L. and Nickelson, T.E. (1989). Transportation stress affects performance of coho salmon Oncorhynchus kisutch. Aquaculture, 82: 15-20. Skilbrei, O.T., Knut, E.J., Holm, M., Farestveit, E., Grimnes, A. and Aardal, L. (1994). A new release system for coastal ranching of Atlantic salmon Salmo salar, and behavioural patterns of released smolts. Nordic. J. Freshw. Res. 69: 84-94. Smith, G.W., Smith, I. P. and Armstrong, S.M. (1994). The relationship between river flow and entry to the Aberdeenshire Dee by returning adult Atlantic salmon. J. Fish. Biol. 45: 953-960. Specker, J.L. and Schreck, C.B. (1980). Stress responses to transportation and fitness for marine survival in coho salmon Onchorhynchus kisutch smolts. Can. J. Fish. Aquat. Sci. 37: 765-769. Went, A.E.J. (1970). Salmon movements around Ireland. Sonderdruck aus Bd. 21(1970) H. 1-4, S: 33-45. Whelan, K.F. (1991). Disappearing Sea Trout - Decline or collapse? The Salmon Net 23: 24-31 Whelan K., Rogan G. and Mantle P. (1994). Enhancement-stocking for catch. New England Atlantic Salmon Management Conference. In A Hard Look at some Tough Issues (S. Calabi and A. Stout, ed.) 134-150. Whelan, B.J. and Marsh, G. (1988). An economic evaluation of Irish angling. Report to the Central Fisheries Board, Dublin. Whelan, B.J., and Whelan, K.F. (1986). The Economics of Salmon Fishing in the Republic of Ireland: Present and Potential. Institute of Fisheries Management University of Ulster, Coleraine.