© 2006 International Council for the Exploration of the Sea
The escape of juvenile farmed Atlantic salmon from hatcheries into freshwater streams in New Brunswick, Canada
Atlantic Salmon Federation PO Box 5200, St Andrews, New Brunswick, Canada E5B 3S8
*Correspondence to J. W. Carr: tel: +1 506 5291385; fax: +1 506 5294985. e-mail: jonwcarr{at}nbnet.nb.ca.
The escape of juvenile Atlantic salmon from freshwater hatcheries supplying the salmon farming industry may lead to interactions between wild and farmed fish. The scale of this problem, however, has not been examined in detail. We monitored temporal trends in the abundance of escaped juvenile farmed salmon in the Magaguadavic River and Chamcook Stream for several years. In addition, in 2004 we assessed more than 90% of the commercial hatcheries producing salmon smolts located next to freshwater streams in New Brunswick. Escaped juvenile fish were recorded in 75% of the streams electrofished close to hatcheries. Numbers varied by site and year. However, escaped juvenile salmon were found every year at sites near hatcheries in the Magaguadavic River and Chamcook Stream. In the Magaguadavic River, juvenile escapees outnumbered wild salmon parr in most years. These results highlight the need for implementation of a containment strategy for freshwater hatcheries to reduce escapes.
Keywords: escapes, farmed, hatcheries, salmon, wild
Received 21 October 2005; accepted 29 March 2006.
 |
Introduction |
|---|
 Top Introduction Study areaMethodsResultsDiscussionReferences |
|---|
The rapid expansion of the Atlantic salmon farming industry during the past two decades has increased the number of farmed fish escaping into the wild. There have been many reports of large farmed salmon escaping from sea cages and entering fresh water (Gudjonsson, 1991; Lund et al., 1991; Carr et al., 1997; Youngson et al., 1997; Fiske and Lund, 1999). The escape of juvenile farmed Atlantic salmon from freshwater hatcheries may also be a major cause of interactions between wild and farmed fish, but the scale of this problem has not been examined in detail. We documented the occurrence of escaped juvenile farmed salmon from freshwater hatcheries in rivers in New Brunswick, Canada.
|
Study area |
|---|
|
TopIntroductionStudy areaMethodsResultsDiscussionReferences |
|---|
The Magaguadavic River is the sixth largest river system in New Brunswick, with a drainage area of 1812 km2. A trap in the fish ladder at the hydroelectric dam located at the head of the tide allows the capture of all fish moving from the sea into the river (Figure 1). Since 1996, all escaped farmed salmon entering the river have been removed at the fish ladder after differentiation from wild fish based on external morphology and scale circuli characteristics (Carr, 1995; Carr et al., 1997). Three commercial salmon hatcheries that produce smolts for the salmon farming industry are located within the Magaguadavic River watershed (Figure 1). More than 70% of the Bay of Fundy sea cage facilities are situated within a 10 km radius of the river's mouth.
|
Chamcook Stream drains from Chamcook Lake, emptying into Passamaquoddy Bay after 0.4 km (Figure 1). A dam at the lake's outlet has blocked upstream fish passage for more than 100 years, and no indigenous anadromous Atlantic salmon have been reported in the stream. In 1974, a salmon genetics research and production facility was sited along the stream and operated until 2004. The hatchery's closure provided an interesting "natural experiment" to evaluate hatchery escapees as a source of fish to the stream.
Tay River and Kelly Creek are tributaries of the St John River watershed, the largest river in New Brunswick (Figure 1). Kelly Creek flows 2.3 km from its headwaters before emptying into the St John River near Mactaquac (Figure 1). A commercial salmon hatchery is located about 1.4 km from its confluence with the St John River (Figure 1). The Tay River empties into the Nashwaak River, which flows about 110 km from Nashwaak Lake to its confluence with the St John River (Jones et al., 2004). It accounts for about 28% of the St John River wild salmon production below Mactaquac dam (Jones et al., 2004). A salmon hatchery is situated on the banks of the Tay River about 15.1 km from its confluence with the Nashwaak River (Figure 1).
Buckman's Creek and Cripp's Stream are 2 km and 5.8 km long, respectively. Both streams drain into the Bay of Fundy in southwestern New Brunswick (Figure 1). Commercial salmon hatcheries are situated in the headwaters of each stream (Figure 1). An additional salmon hatchery is located at the mouth of Cripp's Stream (Figure 1). There are no records of indigenous Atlantic salmon inhabiting either stream.
|
Methods |
|---|
|
TopIntroductionStudy areaMethodsResultsDiscussionReferences |
|---|
Smolt sampling
A proportion of the smolt run from the Magaguadavic River was sampled each mid-April–June of 1998–2005 at the hydroelectric dam located at the head of the tide (Figure 1). From 1998 to 2003, smolts were trapped in the small stream that serves as the downstream fish bypass outlet for the dam (Carr et al., 2004). A fykenet with a holding trap (1998–2000) and a counting fence (2001–2003) were used to capture smolts. In 2004 and 2005, smolts were captured in the collection chambers of a newly constructed fish bypass facility, built in connection with a major upgrade at the hydroelectric dam. Additional smolts were collected 24 km upstream of the dam, using a rotary screw trap (E. G. Solutions Inc, OR, USA), operated in 1999–2003 and 2005 (Figure 1). A fykenet was also used to capture smolts near the outlet of Linton Stream in 1999 (Figure 1).
Parr and fry sampling
Accessible juvenile salmon rearing habitat was identified throughout the Magaguadavic River watershed, and open monitoring stations were established close to (within 2 km) and distant from (>7 km) commercial salmon hatcheries. From 1996 to 2005 (July–September), juvenile salmon were captured by electrofishing. The number of sites electrofished each year ranged from 6 to 26 distant from hatcheries and from four to nine close to hatcheries (Figure 1). In Chamcook Stream, four sites were selected along the length of the stream. Juvenile salmon were captured by electrofishing during September in 2002–2005.
In 2004, a list of all freshwater commercial salmon hatcheries licensed to operate in New Brunswick (n = 25) was obtained from the New Brunswick Department of Agriculture, Fisheries, and Aquaculture. Eighteen hatcheries were rearing juvenile salmon in 2004 (Figure 1). Of these, four were closed recirculating systems in which containment of fish should be 100%, five were situated next to seawater, and nine had some form of discharge into freshwater drainages. Of the nine hatcheries discharging into fresh water, one discharged into a reservoir and could not be sampled by electrofishing. Electrofishing surveys were conducted in streams next to the remaining eight hatcheries draining into fresh water to monitor for the presence of escaped farmed fish. Habitats electrofished were riffle areas with suitable rock substratum for juvenile Atlantic salmon (e.g. Symonds and Heland, 1978; Morantz et al., 1987).
Juvenile salmon densities
To a large extent, available labour and resources determined the intensity of electrofishing at the sites sampled. Density estimates (Zippen, 1958) were obtained at 30% of the sites during the ten-year study period, through repeated passes of the electrofishing equipment over the same area. At these sites, >56% of the fish captures occurred during the first electrofishing pass. Single pass electrofishing was conducted at an additional 31% of the sites during the ten-year period, providing minimum density estimates. Finally, no juvenile salmon were captured at 39% of sites where single pass electrofishing occurred during the ten-year period: 13 sites close to hatcheries and 62 sites distant from hatcheries. Electrofishing data were pooled into two groups, either close to or distant from commercial hatcheries. Mean density estimates were calculated for salmon parr for these two groups. These calculations included the zero values from sites both close to and distant from hatcheries. Farmed fish may have been more susceptible to capture by electrofishing than wild fish because of their larger body size and behaviour, although the extent of this effect is not known.
Classification of wild and farmed juvenile salmon
Sampled salmon parr and smolts were anaesthetized with a solution of clove oil diluted in water to a concentration of 40 mg l–1, weighed, measured (fork length), visually checked for the presence of fin erosion or shortened gill covers, and samples of scales and fin tissue were taken before release. The juvenile salmon were classified as being of wild or farmed origin, using morphological features such as fin erosion and shortened gill covers as well as size at age. Salmon with fin erosion and shortened gill covers were considered to be of farmed origin. Juvenile salmon with fork lengths >130 mm at age 1, >190 mm at age 2, and >230 mm at age 3 were classed as being of farmed origin because these lengths exceed the maximum size of wild juvenile salmon recorded previously in nearby rivers with similar temperatures (Stokesbury and Lacroix, 1997). One-year-old smolts were classed as being of farmed origin because wild smolts of Magaguadavic River origin are all two- and three-year-olds, except for a small number of one-year-old smolts resulting from the stocking of Magaguadavic River strain large parr in a restoration programme. These fish, however, had been adipose fin-clipped and could be identified readily.
|
Results |
|---|
|
TopIntroductionStudy areaMethodsResultsDiscussionReferences |
|---|
Escaped farmed smolts were detected at all sampling stations in the Magaguadavic River in 1998–2005. More than 92% of the smolts examined each year were captured at the lowest site (head of the tide dam, Figure 1). Escapee smolts outnumbered wild smolts in seven of the eight years (Figure 2). Escapee smolt numbers were lowest in 2004 and 2005. Fish stocked for restoration purposes outnumbered both wild and escapee smolts in 2005 (Figure 2). One-year-old escaped smolts accounted for more than 87% of the escapee sample in six of the eight years. Various levels of fin erosion were detected on more than 74% of the escaped farmed fish in each of the eight years.
|
Little or no fin erosion and no shortened gill covers were recorded on parr at any of the sites. All parr captured near commercial salmon hatcheries had exceptional sizes at age and were classified as escapees (Table 1; Stokesbury and Lacroix, 1997). Similarly, based on sizes at age and lack of external morphological deformities, all parr captured at sites distant from hatcheries were classified as wild. The mean lengths of all 0+ and 1+ parr differed significantly between escaped and wild parr in the four years for which sample sizes permitted comparisons between the two groups within age classes (p < 0.001) (Table 1).
|
Magaguadavic River wild salmon parr densities (number caught per 100 m2) were low during the ten-year study period (Table 2). The densities of escapee parr exceeded those of wild parr in eight of the ten years (Table 2). Escapee parr densities were also generally low, but varied from 0.52 to 14.71 fish per 100 m2 during the ten-year period. Wild salmon parr densities increased slightly in 2002–2005, and they outnumbered escaped farmed parr in two years (Table 2). The increase in "wild fish" numbers was principally the result of the stocking programme.
|
Escapee salmon parr were detected at 75% (six of eight) of the additional sites sampled in streams next to commercial salmon hatcheries throughout New Brunswick in 2004. The densities and age classes of the escaped fish varied from site to site (Table 3). No escaped farmed parr were found in Buckman's Creek or Tay River in 2004.
|
In Chamcook Stream, escaped parr were detected in all sample years (Table 3). Age 0+ escapees were captured in 2002–2004 (Table 3). Their densities ranged from 150 to 270 fish per 100 m2, and mean sizes ranged from 8.5 to 9.7 cm (Table 3). All fish were removed from the hatchery after its closure in November 2004. In 2005, only 1+ parr were detected and the density dropped to 60 parr per 100 m2 with a mean size of 12.2 ± 1.9 cm (Table 3). The large sizes of the 0+ fish and paucity of good over-wintering habitat in this stream probably account for the small numbers of 1+ fish captured in any year.
|
Discussion |
|---|
|
TopIntroductionStudy areaMethodsResultsDiscussionReferences |
|---|
Juvenile salmon that have escaped from freshwater commercial salmon hatcheries occur in the Magaguadavic River, and in some years, outnumber the wild salmon. Escapes of juvenile salmon occurred in streams next to at least 75% of the commercial salmon hatcheries in New Brunswick. These escapes may have a negative impact on wild salmon populations through competitive and genetic interaction (Fleming et al., 2002; McGinnity et al., 2003). Few of the escaped parr captured at different sites exhibited fin erosion. Because the body morphology of many of the escaped parr that we examined did not differ from that of wild fish, it seems likely that the scale of escapes of juvenile salmon has been underestimated by field workers in the past.
Genetic selection in farmed lines results in fast growth in the wild of both the escaped farmed salmon juveniles and their offspring. This provides a competitive advantage for the farmed fish over wild congeners (Fleming et al., 2000; Ferguson et al., 2002; McGinnity et al., 2003, 2004). Most salmon parr that we classified as escaped fish and were captured near hatcheries in the Magaguadavic River were much larger at a given age than wild parr caught at sites distant from hatcheries, consistent with reports of faster growth in farmed fish.
The wild fish population in the Magaguadavic River and other rivers in the commercial salmon farming region are severely depressed by critically low adult returns (DFO, 2003), and much of the suitable river habitat is unoccupied, or has fish present at low densities. Consequently, competition between wild and escaped farmed parr for food and space in the Magaguadavic River and other streams in this region, where escapes are occurring, is likely to be low at present. Ferguson et al. (2002) reported that the impact of deliberate and inadvertent introductions of non-native Atlantic salmon into a river was related to the density of wild parr in the river. They concluded that, where a river is below carrying capacity, the introduced fish might survive alongside wild fish, resulting in an overall increase in smolt and adult production. However, in a river that is below its carrying capacity, farmed salmon had lifetime fitness equivalent to 1% of that of wild fish because of poor marine survival. In contrast, in a river at carrying capacity, the lifetime success of farmed salmon increased to 2% of that of wild fish because the larger farmed juveniles displaced wild parr from favourable habitats and decreased wild fish survival.
Although competition between juvenile escaped farmed salmon and wild salmon may not be high in the Magaguadavic River, the risk of genetic introgression may increase if the escaped farmed parr attain sexual maturity in the river. Garant et al. (2003) conducted artificial stream matings of large salmon in the presence of controlled numbers of precocious male parr of both wild and farmed origin. They found that wild precocious male parr had only 25% of the reproductive success of farmed mature parr, and concluded that spawning by precocious farmed parr could be an important mechanism for the introduction of domesticated or non-native traits or both into wild salmon populations. The annual returns of wild adult salmon to the Magaguadavic River have declined to less than ten fish (mainly females) since 2002, and this may facilitate the spawning of precocious male parr.
The low number of farm-origin smolts captured leaving the Magaguadavic River in 2004 and 2005, compared with previous years, may be a sampling artefact resulting from the reconstruction of the hydroelectric dam. The dam was upgraded from a capacity of 3.5 MW to 15 MW, and a new downstream fish bypass facility was installed in 2004. Estimated smolt bypass efficiency fell to 3% and 0.9% in 2004 and 2005, respectively, compared with average estimated bypass efficiencies of 34% in 2000–2003. The decline in numbers of escaped farmed smolts post-2004 may also be explained by a furunculosis epidemic in 2003 that resulted in the removal of most farmed fish at one of the three hatcheries within the watershed, which may have reduced the number of juvenile salmon escapes.
The numbers of fish escaping from sea cages in the Bay of Fundy and Gulf of Maine region have generally declined in recent years, possibly as a result of the introduction of a Containment Management Plan by the salmon farming industry in Maine to improve husbandry practices and net pen security (Goode and Whoriskey, 2003). This study indicates that the frequency with which juvenile salmon escape remains high. The Containment Management Plan in Maine includes measures for freshwater hatcheries, and site-specific Hazard Assessment Critical Control Point (HACCP) plans have been developed to prevent the escape of juvenile salmon. Measures at hatcheries in New Brunswick must be implemented to prevent or at least minimize escapes, either by better containment (e.g. triple screening would be easy and cost-effective) or by HACCP-like changes in husbandry practices, or both. Such measures would be consistent with NASCO's Guidelines on Containment of Farm Salmon (NASCO, 2004).
In the past, surplus fish at some hatcheries may have been released intentionally into adjacent streams in the mistaken belief that this would benefit wild salmon populations. In such cases, education of farm workers could address the problem. At one site in New Brunswick, farmed smolts represented between 18% and 19% of the smolt captures at a counting fence in 1998 and 1999 (Jones et al., 2004). After federal officials from the Department of Fisheries and Oceans met with hatchery personnel in 2000 (R. Jones, pers. comm.), the frequency of smolts decreased to 6% and 0.5% of the smolts captured at the same counting fence in 2000 and 2001, respectively (Jones et al., 2004).
|
Acknowledgements |
|---|
Funding for this project was provided by the New Brunswick Environmental Trust Fund, "Your Environmental Trust Fund at Work", and the New Brunswick Wildlife Trust Fund. Special thanks to Mike Best, Gino Doucette, Steve Tinker, and Ellen Merrill for their assistance in the field; and to Thomas Moffat for preparation of the map.
|
References |
|---|
|
TopIntroductionStudy areaMethodsResultsDiscussionReferences |
|---|
-
Carr J. (1995) Interactions between wild and aquaculture Atlantic salmon in the Magaguadavic River, New Brunswick. MSc thesis, University of New Brunswick, Fredericton. 77 pp.
Carr J.W., Anderson J.M., Whoriskey F.G., Dilworth T. (1997) The occurrence and spawning of cultured Atlantic salmon (Salmo salar) in a Canadian River. ICES Journal of Marine Science 54:1064–1073.
Carr J.W., Whoriskey F., O'Reilly P. (2004) Efficacy of releasing captive reared broodstock into an imperilled wild Atlantic salmon population as a recovery strategy. Journal of Fish Biology 65:Suppl. A, 38–54.[CrossRef][Web of Science]
DFO. (2003) Atlantic salmon Maritime Provinces overview for 2002. Department of Fisheries and Oceans Canada, Gulf and Maritimes Region Science Stock Status Report, 2003/026.
Ferguson A., McGinnity P., Baker N., Cotter D., Hynes R., O'Hea B., Ó Maoiléidigh N., Prodöhl P., Rogan G. (2002) A two-generation experiment comparing the fitness and life history traits of native, ranched, non-native, farmed, and "hybrid" Atlantic salmon under natural conditions. ICES Document CM 2002/T:04.
Fiske P. and Lund R.A. (1999) Escapees of reared salmon in coastal and riverine fisheries in the period 1989–1998. NINA Oppdragsmelding 603:1–23 (In Norwegian with an English abstract).
Fleming I.A., Agustsson T., Finstad B., Johnsson J.I., Bjornsson B.T. (2002) Effects of domestication on growth physiology and endocrinology of Atlantic salmon (Salmo salar). Canadian Journal of Fisheries and Aquatic Sciences 59:1323–1330.
Fleming I.A., Hindar K., Mjølnerød I.G., Jonsson B., Balstad T., Lamberg A. (2000) Lifetime success and interactions of farmed salmon invading a native population. Proceedings of the Royal Society of London Series B 267:1517–1523.[Medline]
Garant D., Fleming I.A., Einum S., Bernatchez L. (2003) Alternate male life-history tactics as potential vehicles for speeding introgression of farm salmon traits into wild populations. Ecology Letters 6:541–549.[CrossRef][Web of Science]
Goode A. and Whoriskey F. (2003) Finding resolution to farmed salmon issues in Eastern North America. In Mills D. (Ed.). Salmon at the Edge(Blackwell Science, Oxford) pp. 144–158 307 pp.
Gudjonsson S. (1991) Occurrence of reared salmon in natural salmon rivers in Iceland. Aquaculture 98:133–142.[CrossRef][Web of Science]
Jones R. A., Anderson L., Goff T. (2004) Assessments of Atlantic salmon stocks in southwest New Brunswick, an update to 2003. CSAS Research Document 2004/019.
Lund R.A., Økland F., Hansen L.P. (1991) Farmed Atlantic salmon (Salmo salar) in fisheries and rivers in Norway. Aquaculture 98:143–150.[CrossRef][Web of Science]
McGinnity P., Prodöhl P., Ferguson A., Hynes R., Ó'Maoiléidigh N., Baker N., Cotter D., O'Hea B., Cooke D., Rogan G., Taggart J., Cross T. (2003) Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. Proceedings of the Royal Society of London Series B 270:2443–2450.[Medline]
McGinnity P., Prodöhl P., Ó'Maoiléidigh N., Hynes R., Cotter D., Baker N., O'Hea B., Ferguson A. (2004) Differential lifetime success and performance of native and non-native Atlantic salmon examined under communal natural conditions. Journal of Fish Biology 65:Suppl. A, 173–187.[CrossRef][Web of Science]
Morantz D.L., Sweeney R.K., Shirvell C.S., Longard D.A. (1987) Selection of microhabitat in summer by juvenile Atlantic salmon (Salmo salar). Canadian Journal of Fisheries and Aquatic Sciences 44:120–129.
NASCO. (2004) Resolution by the Parties to the Convention for the Conservation of Salmon in the North Atlantic Ocean to Minimize Impacts from Aquaculture, Introductions and Transfers, and Transgenics on the Wild Salmon Stocks(NASCO, Edinburgh) North Atlantic Salmon Conservation Organization (NASCO) Council Document, CNL(04)54 39 pp.
Stokesbury M.J. and Lacroix G.L. (1997) High incidence of hatchery origin Atlantic salmon in the smolt output of a Canadian river. ICES Journal of Marine Science 54:1074–1081.
Symonds P.E.K. and Heland M. (1978) Stream habitats and behavioural interactions of underyearling and yearling Atlantic salmon, Salmo salar. Journal of the Fisheries Research Board of Canada 35:175–183.[Web of Science]
Youngson A.F., Webb J.H., MacLean J.C., Whyte B.M. (1997) Frequency of occurrence of reared Atlantic salmon in Scottish salmon fisheries. ICES Journal of Marine Science 54:1216–1220.
Zippen C. (1958) The removal method of population estimation. Journal of Wildlife Management 22:82–90.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||