PREVENT ESCAPE Project Compendium by oceanográfica

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

ISBN: 978-82-14-05565-8

Table

of contents

Assessing the causes and developing measures to prevent the escape of fish from sea-cage aquaculture

Executive Summary

1. Escapes of fishes from European sea-cage aquaculture: environmental consequences and the need to better prevent escapes

2. A pan-European evaluation of the extent, causes and cost of escape events from sea-cage fish farming

3.1. Do farming environments and the inherent behaviours of fish predispose some species to higher rates of escape?

3.2. Factors affecting escape-related behaviours in Atlantic cod (Gadus morhua L)

3.3. The escape-related behaviour of European seabass (Dicentrarchus labrax) and gilthead seabream (Sparus aurata) 73 3.4. Escape behaviour of gilthead seabream and European seabass in commercial sea cages

www.preventescape.eu

3.5. General conclusions & recommendations for preventing and mitigating escape-related fish behaviour

4.1. The importance of identifying escaped fish from aquaculture and determining their post-escape behaviours for environmental and fisheries management

105

4.2. Methods to identify escaped Atlantic cod (Gadus morhua L.)

114

4.3. Methods to identify escaped seabass and seabream

123

4.4. Post-escape dispersal of Atlantic cod (Gadus morhua L.) and juvenile Atlantic salmon (Salmo salar L.)

139

4.5. Post-escape behaviours of farmed seabream and seabass

148

4.6. Recommendations for identification of escapees and conclusions from risk assessment analysis

160

5.1. A novel form of escape of fish from sea cages: the problem of â&#x20AC;&#x2DC;escape through spawningâ&#x20AC;&#x2122;

171

5.2. Escape through spawning by seabream (Sparus aurata)

178

5.3. Extent and ecological importance of spawning in sea-cages for Atlantic cod (Gadus morhua L.) 192 5.4. No evidence of egg escape from meagre (Argyrosomus regius) aquaculture under current practices

202

5.5. EGG Escape: General Conclusions & Recommendations

209

6.1. Preventing and mitigating escapes through research to underpin technological and operational improvements for seacage farming and recapture technologies

217

6.2. Damage to the net cage

220

6.3. Assessment of escape critical loads and damages from fish bite

240

6.4. Investigation of the properties of materials and technologies used in aquaculture

248

6.5. Sea-load exposure

260

6.6. Recapture of escaped seabass and seabream in the Mediterranean Sea

273

6.7. Recapture of escaped juvenile cod (Gadus morhua) in Northern Norway

281

6.8. â&#x20AC;&#x2DC;Escape through spawningâ&#x20AC;&#x2122;: solutions to reduce the escapes of viable fish eggs from sea-cages

288

7. Recommendations and guidelines for the design of fish farms, management and operation of equipment

295

ISBN: 978-82-14-05565-8 www.preventescape.eu

Assessing

the causes and developing measures to

prevent the escape of fish from sea- cage aquaculture Cite this article as: Dempster T, Jackson D, Noble C, Sanchez-Jerez P, Somarakis S, Jensen Ø (2013) Assessing the causes and developing measures to prevent the escape of fish from sea-cage aquaculture (Prevent Escape) – Executive Summary. In: PREVENT ESCAPE Project Compendium. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Tim Dempster1, Dave Jackson2, Chris Noble4, Pablo Sanchez-Jerez4, Stelios Somarakis5, Østen Jensen1 SINTEF Fisheries & Aquaculture, Norway, The Marine Institute, Ireland, 3 NOFIMA, Norway, 4 University of Alicante, Spain, 5 Hellenic Centre of Marine Research, Greece 1 2

Executive Summary

The escape of fish from sea-cage aquaculture is perceived as a threat to natural biodiversity in Europe's marine waters. Escaped fish may cause undesirable genetic effects in native populations through interbreeding, and ecological effects through predation, competition and the transfer of diseases to wild fish. Technical and operational failures of fish farming technology cause escapes. Cages break down in storms, wear and tear of the netting causes holes, and operational accidents lead to spills of fish. The Prevent Escape project conducted and integrated biological and technological research on a pan-European scale to improve recommendations and guidelines for aquaculture technologies and operational strategies that reduce escape events. Through research focused on sea-cages and their immediate surrounds, the Prevent Escape project determined that escape events are widespread throughout European sea-cage aquaculture. From 2007-2009, we documented 255 escape events across 6 countries involving Atlantic salmon, Atlantic cod, rainbow trout, seabream, seabass and meagre production. 9.2 million fish escaped from these 255 events, which mostly occurred due to structural failures during storms and the appearance of holes in nets. Seabream accounted for the highest number of escapes (74%) followed by Atlantic salmon (11.8%). On a pan-European scale, we estimate that this directly cost the industry â&#x201A;Ź47.5 million per year. Costs to the reputation of the industry were not able to be assessed, but were likely substantial. In addition to juvenile and adult fish escaping, both seabream and Atlantic cod mature and spawn in sea-cages. They produce viable eggs which flow out from fish farms and enter wild populations.

www.preventescape.eu

A detailed analysis of escapes in Europeâ&#x20AC;&#x2122;s largest industry, Atlantic salmon production in Norway, revealed that after the Norwegian technical standard (NS 9415) for the design, dimensioning and operation of sea-cage farms was implemented in 2006, the total number of escaped Atlantic salmon declined from >600 000 yrâ&#x20AC;&#x201C;1 (2001 to 2006) to <300 000 fish yrâ&#x20AC;&#x201C;1 (2007 to 2011), despite the total number of salmon held in sea-cages increasing by >50% during this period. Based on the success of this measure, to prevent escapes of juvenile and adult fish as sea-cage aquaculture industries develop, the Prevent Escape project recommends that policymakers introduce a technical standard for sea-cage aquaculture equipment, coupled with independent mechanisms to enforce the standard. Determining the behaviours of fish which may pre-dispose them to escape is a key step in designing farming strategies to reduce escape risks. Experiments identified that both seabream and Atlantic cod exhibit a range of behaviours that may pre-dispose them to high rates of escape through holes as they swim close to net walls and bite the netting. Manipulative experiments of the cage culture environment determined that cage management strategies could be effective in reducing escape risk. These strategies include maintaining well-fed fish, using environmental enrichment to distract fish from engaging in escape-related behaviours and maintaining clean nets. Fast, accurate and cost-effective tools for identifying escapees are central for assessments of the extent and consequences of fish escapes. The project tested a range of techniques to distinguish escaped fish within wild populations. Ultimately, the selection of suitable indicators depends on the final stakeholder. Farmers and consumers could use external appearance and morphometry for rapid assessment, however, trace elements in scales and fatty acid profiles are more useful for fisheries and environmental management applications. Where individuals need

to be discriminated with a high degree of accuracy to farmed or wild origin, we recommend genetic methods, scale features and trace elements of scales as suitable methods. To identify the specific farm where an escape event occurred, only genetic methods are currently capable. Re-capturing escapees is a possible method to mitigate escapes. Tracking of escaped cod, salmon, seabream, meagre, and seabass, indicated that unless recapture efforts are initiated within 24 hours after an escape incident, the potential for successful recapture of escapees is limited. Further, by-catch of wild marine fish was high during recapture efforts, implying that large-scale recapture efforts must be weighed against the possibility of affecting wild fish populations negatively. Wild fish that aggregated around fish farms where fish escapes were highly effective in predating escapees, thus maintaining healthy populations of predators around sea-cage facilities could assist in mitigating escapes. Sea-cage equipment is marketed and used across Europe, thus knowledge relevant to the culture of numerous species in diverse environments is required to produce robust equipment and implement risk-adverse operations. Prevent escape has delivered new knowledge on the materials and properties of sea-cage components and innovated new system designs to reduce the risks of specific escape events from occurring. Information from this component of the project will assist in benchmarking the performance of equipment during farming, improve operations and equipment production, and advance national and international standards for the design, construction and use of aquaculture equipment. These key pieces of information, when added to existing knowledge, have allowed a series of practical, implementable measures to prevent escapes and mitigate the effects of escapees. Ultimately, if prevention and mitigation are more successful, genetic and ecological impacts of escaped cultured fish on wild fish populations should diminish.

www.preventescape.eu

seeks to reduce the number of fish that escape from European aquaculture through research to improve fish farming techniques and technologies.

1. Escapes

fishes

from

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

European

sea- cage

aquaculture: environmental consequences and the need to better prevent escapes Cite this article as: Dempster T, Jensen Ă&#x2DC;, Fredheim A, Uglem I, Thorstad E, Somarakis S, Sanchez-Jerez P (2013) Escapes of fishes from European sea-cage aquaculture: environmental consequences and the need to better prevent escapes. In: PREVENT ESCAPE Project Compendium. Chapter 1. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Tim Dempster1, Ă&#x2DC;sten Jensen1, Arne Fredheim1, Ingebrigt Uglem2, Eva Thorstad2, Stelios Somarakis3, Pablo Sanchez Jerez4 SINTEF Fisheries and Aquaculture, Norway Norwegian Institute of Nature Research, Norway 3 Hellenic Centre of Marine Research, Greece 4 Department of Marine Science and Applied Biology, University of Alicante, Spain 1 2

www.preventescape.eu

The

rise of modern aquaculture

In 1971, famous marine explorer and ecologist Jacques Cousteau proclaimed that “We must plant the sea and herd its animals using the sea as farmers instead of hunters. That is what civilization is all about - farming replacing hunting”. Cousteau’s prediction has been borne out dramatically. Aquaculture is approaching wild fisheries as the major source of fish protein for humans (FAO 2011) through rapid domestication of marine species (Duarte et al. 2008). European aquaculture production has mirrored the global trend, rising from 1 million tons yr-1 in 1990 to >2 million tons y-1 in 2008, principally though producing fish in coastal waters in large net pens (hereafter sea-cages; Figure 1.1). This enormous expansion of aquaculture has multiplied its interactions with the environment in ways Cousteau could never have predicted. Chief among these are the interactions that occur when farmed fish escape from fish farms and enter wild populations.

Figure 1.1. Sea-cage fish farms: above the surface (left) and below (right). Sea-cages can be either square or rectangular and enclose volumes of water typically ranging from 10000-40000 m3. They may be moored individually or grouped together in multiple systems. Individual cages for many of the main fish species grown now contain anywhere between 10000-400000 fish.

Sea-cage aquaculture

and escaped farmed fish

Escapes of fish from sea-cage aquaculture have typically been thought of as referring to juvenile and adult fish. Such escapes have been reported for almost all species presently cultured around the world, including Atlantic salmon, Atlantic cod, rainbow trout, Arctic charr, halibut, seabream, seabass, meagre and kingfish (e.g. Soto et al. 2001, Naylor et al. 2005, Gillanders & Joyce 2005, Moe et al. 2007, Toledo Guedes et al. 2009; Figure 1.2). Recently,

a second form of escape has come into focus, involving the escape of viable, fertilized eggs spawned by farmed individuals from sea-cage facilities, or so called ‘escape through spawning’ (Jørstad et al. 2008). This phenomenon has forced a redefinition of the term ‘escapes from aquaculture’ to include the escapement of fertilized eggs into the wider marine environment.

Figure 1.2. Escaped seabream (Sparus aurata) beneath a sea-cage (left) and an escape attempt by a cod (Gadus morhua) through a test net mesh panel in Norway (right).

Escapees can have detrimental genetic and ecological effects on populations of wild conspecifics, and the present level of escapees is regarded as a problem for the future sustainability of sea-cage aquaculture (Naylor et al. 2005). For example, over 350 million Atlantic salmon are held in sea-cages in Norway at any given time (Jensen et al. 2010), which outnumbers the approximately 500 000 to 1 million salmon that return to Norwegian rivers from the ocean each year to spawn. In 2010, 291 000 farmed salmon were reported to escape from Norwegian farms, whereas the pre-fishery abundance of wild spawners was estimated at 480 000 salmon (Anon. 2011). A single fish farm cage may hold hundreds of thousands of cultured fish with over a million fish per site within multiple cages now common. Due to the large numerical imbalances of caged compared to wild populations, escapement raises important concerns about ecological and genetic impacts. Evidence of environmental effects on wild populations is largely limited to Atlantic salmon, as these interactions have been intensively studied, with more limited information for the other species farmed across Europe.

Environmental

consequences of escaped

Atlantic

salmon

In a comprehensive review of the effects of escaped Atlantic salmon on wild populations, Thorstad et al. (2008) concluded that while outcomes of escapee-wild fish interactions vary with environmental and genetic factors, they are frequently negative for wild salmon. As fish farm areas are typically located close to wild fish habitats, and escaped fish may disperse

www.preventescape.eu

over large geographic areas (e.g. Furevik et al. 1990, Whoriskey et al. 2006, Hansen 2006, Hansen & Youngson 2010), escaped salmon may mix with their wild con specifics and enter rivers tens to hundreds of kilometres from the escape site during the spawning period. The average proportion of escaped salmon in Norwegian rivers monitored close to the spawning period varied between 11 and 35% during 1989-2010 (13% in 2010), with the highest proportions during the late 1980s and early 1990s (Anon. 2011). Consequently, the potential exists for escapees to interact negatively with wild populations, through competition, transfer of diseases and pathogens, and interbreeding. Hindar & Diserud (2007) recommended that intrusion rates of escaped farmed salmon in rivers during spawning should not exceed 5% to avoid substantial and definite genetic changes of wild populations.

Transfer

of diseases and pathogens

Escape incidents may heighten the potential for the transfer of diseases and parasites, which are considered to be amplified in aquaculture settings (e.g. Heuch & Mo 2001, BjĂ¸rn & Finstad 2002, Skilbrei & Wennevik 2006, KrkoĹĄek et al. 2007). Escapees from salmon aquaculture in Norway have been identified as reservoirs of sea lice Lepeophtheirus salmonis in coastal waters (Heuch & Mo 2001). Newly-migrated post-smolts are particularly vulnerable for sea lice infestations, and salmon lice may represent a significant threat for some wild Atlantic salmon populations (Revie et al. 2009, Finstad et al. 2011, Gargan et al. 2012). In addition, 60 000 salmon infected with infectious salmon anaemia (ISA) and 115 000 salmon infected with pancreas disease (PD) escaped from farms in southern Norway in 2007, yet whether these precipitated infections in wild populations is unknown. The ability for escaped fish to transfer disease to wild fish depends on the extent of mixing between the two groups, which in turns varies with the life stage, timing and location of the escape (summarized by Thorstad et al. 2008). However, while escaped and wild fish mix, little direct evidence for disease transfer from escapees to wild salmon population has been documented, other than for the possible case of furunculosus, a fungal disease accidently introduced to Norway from Scotland in the 1990s with the transfer of stock and then believed to have been spread from farmed to wild populations by escapees (summarized in Naylor et al. 2005).

Interbreeding Successful spawning of escaped farmed salmon in rivers both within and outside their native range has been widely documented (see review by Weir & Grant 2006). The ability of escaped salmon to interbreed with wild salmon depends on their ability to ascend rivers, access spawning grounds and spawn successfully with wild partners. While the spawning success of farmed female salmon may be just 20-40% that of wild salmon and even lower for males (1-24%; Fleming et al. 1996, 2000), high proportions of escaped fish in many rivers can lead to a high proportion of farm x wild hybrids. Escaped female salmon may also interfere with wild salmon breeding through destroying the spawning redds of wild fish if they spawn later (Lura & SĂŚgrov 1991, 1993). Wild Atlantic salmon are structured into populations and meta-populations with little gene flow between them, and evidence for local adaptation in wild Atlantic salmon is compelling (reviewed by Garcia de Leaniz et al. 2007). Farmed salmon differ genetically from wild populations due to founder effects, domestication selection, selection for economic traits and genetic drift (reviewed by Ferguson et al. 2007). Hybridisation of farmed with wild salmon and later backcrossing of hybrids may change the level of genetic variability and the frequency and type of alleles present. Hence, hybridisation of farmed with wild salmon has the potential to genetically alter native populations, reduce local adaptation and negatively affect population viability and character (Ferguson et al. 2007). Several studies have shown that escaped farmed salmon breeding in the wild have changed the genetic composition of wild populations (e.g. Clifford et al. 1998, Skaala et al. 2006). Large-scale field experiments undertaken in Norway and Ireland showed highly reduced survival and lifetime success of farm and hybrid salmon compared to wild salmon (McGinnity et al. 1997, 2003, Fleming et al. 2000). The relative estimated lifetime success ranged from lowest for the farm progeny to highest for the local wild progeny with intermediate performance for the hybrids. Farmed salmon progeny and farm x wild hybrids may directly interact and compete with wild juveniles for food, habitat and territories. Farm juveniles and hybrids are generally more aggressive and consume similar resources in freshwater habitats as wild fish (Einum & Fleming 1997). In addition, they grow faster than wild fish, which may give them a competitive advantage during certain life stages. Invasions of escaped farmed salmon have the potential to impact the productivity of wild salmon populations negatively through juvenile resource competition and competitive displacement. Fleming et al. (2000) determined that invasion of a small river in Norway by escapees resulted in an overall reduction in smolt production by 28% due to resource competition and competitive displacement. Local fisheries could therefore suffer reduced catches as wild fish stocks decline (SvĂĽsand et al. 2007).

Competition

for food

Escaped salmon consume much the same diet as wild salmon in oceanic waters (Jacobsen & Hansen 2001, Hislop & Webb 1992) and could potentially compete for food with wild stocks. Substantial competitive interactions in the ocean, however, appear unlikely to occur as ocean mortality of salmon appears to be density-independent (Jonsson & Jonsson 2004), although limited information exists to assess if this is also the case for coastal waters (Jonsson & Jonsson 2006).

www.preventescape.eu

Environmental consequences of escaped Atlantic cod At present, little direct evidence exists for negative interactions of escaped and wild Atlantic cod juveniles or adults, despite predictions that negative consequences will result (Bekkevold et al. 2006). Cod farming is a relatively new industry, thus if negative consequences exist they may not have had sufficient time to manifest and/or be detected. Telemetry studies of simulated cod escapes have indicated that escapees, regardless of whether they originated from stocks of coastal or oceanic origin, mix with wild populations in fjord environments and can move to spawning grounds in the spawning season (Uglem et al. 2008, 2010). Behavioral studies have further indicated that escaped farmed cod are likely to hybridize with wild cod (Meager et al. 2009). However, farmed cod may have limited reproductive success in sperm competition with wild cod, which lowers the risk of genetic introgression from escapees (Skjæraasen et al. 2009). Other possible ecological effects of escaped farmed cod include increased predation pressure on out-migrating wild salmon smolt (Brooking et al. 2006) and transmission of pathogens and parasites to wild populations (Øines et al. 2006), although direct evidence for these effects is at present lacking. Recaptures of Atlantic cod escapees equipped with acoustic transmitters in local commercial and recreational fisheries in Norway are known to be high (approximately 40%; Uglem et al. 2008), indicating that local fisheries receive temporary increases after escape events and may be partially effective in reducing escaped cod numbers.

Possible

impacts of ‘escape through spawning’ of

Atlantic

cod

In the culture of Atlantic cod, some fish mature during the first year of culture, while a majority of farmed cod are believed to mature during the second year. This means that almost the entire culture stock in any particular farm has the potential to spawn in sea-cages before they are slaughtered. Spawning of Atlantic cod within a small experimental sea-cage containing 1000 farmed cod and dispersal of their spawned eggs in a fjord system has been demonstrated (Jørstad et al. 2008). In the proximity of this experimental sea-cage, 20-25% of the cod larvae in plankton samples were determined by genetic analyses to have originated from the 1000 farmed cod (Jørstad et al. 2008). Furthermore, preliminary results indicate that 4-6 % of juvenile cod (35-40 cm total length) caught in the area around the farm in following years were offspring of the farmed cod (van der Meeren & Jørstad 2009). This illustrates that if spawning occurs within commercial cod farms where numbers of farmed individuals are far greater, the contribution of ‘escaped’ larvae to cod recruitment within fjord systems may be substantial. Escape of large quantities of eggs from caged cod could lead to ecological and genetic effects in wild populations (Bekkevold et al. 2006, Jørstad et al. 2008) as; 1) coastal cod populations in some areas of Norway are presently weak, most likely due to overfishing (ICES 2008); 2) coastal cod have a high fidelity to specific spawning grounds (e.g. Wright et al. 2006); and 3) sea-cage cod farms are often located within short distances of known wild cod spawning grounds (Uglem et al. 2008). Recent research also suggests that cod eggs may be entrained in

the vicinity of the spawning grounds long after spawning (Knutsen et al. 2007). Therefore, there is considerable potential for larvae from escaped cod eggs to experience favourable conditions for survival and recruitment to coastal cod stocks if spawning in sea-cages occurs during the natural spawning season of wild cod.

Environmental

consequences of escaped seabream and seabass

For seabream and seabass, knowledge regarding how escapes might affect ecosystems is limited. Intentional releases of cultured seabream for stock enhancement have been reported from the southern Atlantic coast of Spain, and in the Bay of Cadiz (Sanchez-Lamadrid 2002, 2004). Released fish moved less than 10 km from the release point. Good growth rates and condition indices indicate that the released fish adapted to life in the wild and suggest that populations of wild fish could also be altered by released fish. For example, there is correlative evidence of a substantial increase in wild populations of seabream after fish farming began in the Messolonghi lagoon, Greece (Dimitriou et al. 2007). Dempster et al. (2002) found very few seabream near sea-cages in which seabream were being reared, which suggests either low levels of escape or that escapees move rapidly away from the farms to other habitats. Based on the ecology of seabream and the location of most fish farms in areas close to wild seabream habitats, it is probable that escapees would mix with their wild con specifics. Consequently, the potential exists for escapees to interact negatively with wild populations, through interbreeding, competition and transfer of diseases and pathogens. Escaped seabream were frequently recorded/recaptured in the most common natural habitats of this species and stomach analyses indicated that escapees may feed on natural prey from the first day after escape (Arechavala et al. 2012). Seabream are opportunistic feeders, adapting their diets to the food items available (Tancioni et al. 2003). Initially, escaped seabream eat mainly macrophytes and food pellets, but later also feed on common prey such as crustaceans and molluscs. Therefore, escaped seabream have dietary flexibility and feed well in wild environments shortly after they escape (Arechavala et al. 2011). The only published long-term data available on escapes of seabass indicate that when seabass cultured from western Mediterranean populations escaped in the eastern Mediterranean, they established and maintained distinct populations of the western Mediterranean phenotype

www.preventescape.eu

without introgressing with the local population (Bahri-Sfar et al. 2005). Escapes of seabass in some locations may be particularly problematic, principally where local populations are small or in areas outside the natural distribution of seabass. For example, seabass do not naturally occur in wild habitats around the Canary Islands and their recent appearance there in coastal waters is due to escapes from sea-cages (Toledo Guedes et al. 2009). Escape of seabass from fish farms in such areas is thus an introduction of a non-native species. The risk of transmission of pathogens through movements of escaped fish in the Mediterranean areas exists, although transmission has not been documented (reviewed by Arechavala et al. 2012a). Infected farmed fish that escape from cages failure might spread pathogens to other cages or farms, and also to both wild fish species that occur naturally in farming areas. Both seabream and seabass escapees have been observed to swim to adjacent fish farms post release and then disperse to natural habitats (Arechavala et al. 2011, 2012b), suggesting that they are capable of transmitting pathogens to wild populations should they carry them. The large variety of shared pathogens among wild and farmed fish species and the various pathways of pathogens transmission increase the possibilities of infection.

Possible

impacts of â&#x20AC;&#x2DC;escape through spawningâ&#x20AC;&#x2122; of seabream

In the Mediterranean region, information about spawning by fish kept in sea-cages is sparse. In Greece, the largest EU producer of seabream, both the number of fish farms and their production capacity increased over the past decade, accompanied by a substantial decrease in the price of seabream. This industrial development led to structural and functional changes in the rearing process. The time individual fish were farmed increased from just 12 to 18 months before 1995 (Petridis & Rogdakis 1996) to durations of up to 40 months after 1999 (Dimitriou et al. 2007). Gilthead seabream is a protandrous hermaphrodite species and the increased farming duration has resulted in the production of fish of a size compatible with that necessary for fish to reach the stage of sex inversion and female sexual maturation, normally observed at the age of 2-3 years in the wild. The changes in rearing processes have resulted in the presence of large gilthead seabream individuals (larger than 500g) in cages during the normal reproductive period of their wild counterparts (November-March: Bauchot & Hureau 1986). There is evidence that sex inversion and the production of both male and female gametes occur within cages under the present industrial rearing pattern (Dimitriou et al. 2007). A doubling of the population of wild seabream within the Messolonghi lagoon in Greece, based on standardised commercial fishing trap catch returns, correlates with the advent of farming sea-bream to large sizes in the region. Spawning within sea-cages is suspected to have led to greater recruitment to wild seabream stocks (Dimitriou et al. 2007). Ecological and economic consequences of this population shift have ensued as while more wild sea-bream are now available to the fishery, they are of much smaller mean size resulting in an overall lower economic return to local fishers.

The need to better prevent the escape of fish from sea-cage fish farms Improved physical containment at marine fish farming sites, through research and development of fish-farming technology, is a central recommendation of many international workshops and forums on the environmental impacts of escapees (Hansen & Windsor 2006). For example, the FP-6 EU Coordinating Action on the ‘Genetic Impact of Aquaculture Activities on Native Populations; GENIMPACT’ has concluded that efforts should be made to prevent escapes, as ‘instead of trying to protect wild populations from escapees, the best logical solution would be to try to prevent escapes. This will rely on technical improvements from the industry …’ (Triantafyllidis et al. 2007). A global report from the Salmon Aquaculture Dialogue on the incidence and impacts of escaped farmed Atlantic salmon in nature (Thorstad et al. 2008) concluded similarly: ‘The most important management issue at present is the need to reduce the numbers of escaped farmed salmon in nature.’ Further, it is a stated goal of both the Norwegian authorities and the Norwegian Fish Farmers Association to reduce escapes of fish to a level where they do not threaten wild populations (Norwegian Fisheries Directorate 2009). While far less is known about escapes in the Mediterranean Sea and their impacts, the 2007 CIESM Workshop on the impacts of mariculture in the Mediterranean concluded that ‘better information to document the extent of escapes in the Mediterranean is required, while improved methods to trace escapees and prevent escapes need development’ (CIESM 2007).

Prevention

and

Mitigation

The Prevent Escape project was specifically aimed at assessing the extent and causes of escapes and generating new knowledge through research to help mitigate the effects of escapees on wild populations on a pan-European scale. Solving technical and operational problems related to escapes is dependent on a combination of research into several technological disciplines and biological knowledge related to the behaviour of fish in sea-cages (Figure 1.3).

Figure 1.3. The research approach taken in the Prevent Escape project.

www.preventescape.eu

By conducting research focused on sea-cages and their immediate surroundings, we assessed the detailed technical and operational causes of escape incidents, assessed the extent of escapes of reproductive gametes, juveniles and large fish from sea-cages, determined the inherent biological mechanisms that pre-dispose certain species of fish towards behaviours within sea-cages that make escapes more likely, and documented the dispersal of escapees to better understand how they may be recaptured. Finally, through research on aquaculture structures, materials, designs and operational methods, we have developed new knowledge to prevent escapes and use technologies to recapture fish after they have escaped. Both of these results will assist efforts to mitigate the effects of escapes. Information from the various components of the project will help benchmark the performance of equipment under farming conditions and thereby improve national and international standards for the design, construction and use of aquaculture equipment. These key pieces of information, when added to existing knowledge, will allow determination of practical, implementable measures to prevent escapes and mitigate the effects of escapees. If prevention and mitigation are more successful, genetic and ecological impacts should diminish.

References

cited

Anon (2011) The status of Norwegian salmon stocks in 2011. Report from the Norwegian Scientific Advisory Committee for Atlantic Salmon Management no. 3, 285 pp. Vitenskapelig råd for lakseforvaltning, Trondheim. (In Norwegian) Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Bayle-Sempere JT, Sanchez-Jerez P (2011) Post-escape dispersion of farmed seabream (Sparus aurata L.) and recaptures by local fisheries in the Western Mediterranean Sea. Fish Res 121-122: 126-135 Arechavala-Lopez P, Sanchez-Jerez P, Bayle-Sempere JT, Uglem I, Mladineo I (2012a) Could wild fish and farmed escapees transfer pathogens among Mediterranean fish farming areas? Aquacult Environ Interact (in press) Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Bayle-Sempere JT, Sanchez-Jerez (2012b) Immediate post-escape behaviour of farmed sea-bass (Dicentrarchus labrax) in the Mediterranean Sea. J Appl Ichthyol 27: 1375-1378 Bahri-Sfar L, Lemaire C, Chatain B, Divanach P, Kalthoum Ben Hassine O, Bonhomme F (2005) Impact de l’élevage sur la structure génétique des populations méditerranéennes de Dicentrarchus labrax. Aquat Liv Resour 18: 71-76 (in French). Bauchot ML, Hureau JC (1986) Sparidae. In: Fishes of the North-Eastern Atlantic and Mediterranean (eds: Whitehead, PJ, Bauchot ML, Hureau JC, Nielsen J, Tortonese E), Vol. 2, pp. 883-907. UNESCO, UK Bekkevold D, Hansen MM, Nielsen EE (2006) Genetic impact of gadoid culture on wild fish populations: predictions, lessons from salmonids, and possibilities for minimizing adverse effects. ICES J Mar Sci 63: 198-208 Bjørn PA, Finstad B (2002) Salmon lice, Lepeophtheirus salmonis (Krøyer), infestation in sympatric populations of Arctic char, Salvelinus alpinus (L.), and sea trout, Salmo trutta (L.), in areas near and distant from salmon farms. ICES J Mar Sci 59: 131-139 Brooking P, Doucette G, Tinker S, Whoriskey FG (2006) Sonic tracking of wild cod, Gadus morhua, in an inshore region of the Bay of Fundy: a contribution to understanding the impact of cod farming for wild cod and endangered salmon populations. ICES J Mar Sci 63: 1364-1371 CIESM (2007) Impact of mariculture on coastal ecosystems: Executive summary. CIESM Workshop Monograph 32: 5-20. www.ciesm.org/online/monographs/lisboa07.pdf Clifford SL, McGinnity P, Ferguson A (1998) Genetic changes in Atlantic salmon (Salmo salar) populations of Northwest Irish rivers resulting from escapes of adult farm salmon. Can J Fish Aquat Sci 55: 358-363 www.preventescape.eu

Dimitriou E, Katselis G, Moutopoulos DK, Akovitiotis C, Koutsikopoulos C (2007) Possible influence of reared gilthead seabream (Sparus aurata, L.) on wild stocks in the area of the Messolonghi lagoon (Ionian Sea, Greece) Aquacult Res 38: 398-408 Duarte C, Marba N, Holmer M (2007) Rapid domestication of marine species. Science 316: 383-383 Einum S, Fleming IA (1997) Genetic divergence and interactions in the wild among native, farmed and hybrid Atlantic salmon. J Fish Biol 50: 634-651 Ferguson A, Fleming I, Hindar K, Skaala Ø, McGinnity P, Cross T.F., Prodöhl,P (2007) Farm escapes. In: The Atlantic salmon: Genetics, Conservation and Management (Verspoor, E., Stradmeyer, L. & Nielsen, J.L., eds.). Blackwell Publishing Ltd, pp. 357-398 Finstad B, Bjørn PA, Todd CD, Whoriskey F, Gargan PG, Forde G Revie CW (2011) The effect of sea lice on Atlantic salmon and other salmonid species. In Atlantic Salmon Ecology (Aas Ø, Einum S, Klemetsen A, Skurdal J eds) , pp. 253-276. Oxford: Wiley-Blackwell. Fleming IA, Jonsson B, Gross MR, Lamberg A (1996) An experimental study of the reproductive behaviour and success of farmed and wild salmon (Salmo salar). J Appl Ecol 33: 893-905 Fleming IA, Hindar K, Mjølnerød IB, Jonsson B, Balstad T, Lamberg A (2000) Lifetime success and interactions of farm salmon invading a native population. Proc Royal Soc London B 267: 1517-1523 Furevik D, Rabben H, Mikkelsen KO, Fosseidengen JE (1990) Migratory patterns of escaped farmed Atlantic salmon. ICES C.M. 1990/F:55, 19 pp Garcia de Leaniz C, Fleming IA, Einum S, Verspoor E, Jordan WC, Consuegra S, Aubin-Horth N, Lajus D, Letcher BH, Youngson AF, Webb J, Vøllestad LA, Villanueva B, Ferguson A, Quinn TP (2007) A critical review of inherited adaptive variation in Atlantic salmon. Biol Rev 82: 173-211 Gargan, PG, Forde G, Hazon N, Russell DJF, Todd CD (2012) Evidence for sea lice-induced marine mortality of Atlantic salmon (Salmo salar) in western Ireland from experimental releases of ranced smolts treated with emamectic benzoate. Can J Fish Aquat Sci 69: 343-353. Gillanders BM, Joyce TC (2005) Distinguishing aquaculture and wild yellowtail kingfish via natural elemental signatures in otoliths. Mar Freshwat Res 56:693-704 Hansen LP (2006) Migration and survival of farmed Atlantic salmon (Salmo salar L.) released from two Norwegian fish farms. ICES J Mar Sci 63: 1211-1217 Hansen LP, Windsor ML (2006) Interactions between aquaculture and wild stocks of Atlantic salmon and other diadromous fish species: Science and Management, Challenges and Solutions. An introduction by the conveners. ICES J Mar Sci 63: 1159-1161

Hansen, LP, Youngson AF (2010) Dispersal of large farmed Atlantic salmon, Salmo salar, from simulated escapes at fish farms in Norway and Scotland. Fish Manag Ecol 17: 28-32. Heuch PA, Mo TA (2001) A model of salmon louse production in Norway: effects of increasing salmon production and public management measures. Dis Aquat Org 45: 145-152 Hindar K, Diserud O (2007) Vulnerability analysis of wild salmon populations towards escaped farm salmon. NINA Report 244: 1-45 (In Norwegian with English summary) Hislop JRG, Webb JH (1992) Escaped farmed Atlantic salmon (Salmo salar) feeding in Scottish coastal waters. Aqua Fish Manag 23: 721-723 ICES (2008) Report of the ICES Advisory Committee, 2008. Book 6, North Sea. 332p. Jacobsen JA, Hansen LP (2001) Feeding habits of wild and escaped farmed Atlantic salmon, Salmo salar L., in the Northeast Atlantic. ICES J Mar Sci 58: 916-933 Jensen Ø, Dempster T, Thorstad EB, Uglem I, Fredheim A (2010) Escapes of fishes from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquacult Environ Interact 1:71-83 Jonsson B, Jonsson N (2004) Factors affecting marine production of Atlantic salmon (Salmo salar). Can J Fish Aquat Sci 61: 2369-2383 Jonsson B, Jonsson N (2006) Cultured Atlantic salmon in nature: a review of their ecology and interaction with wild fish. ICES J Mar Sci 63: 1162-1181 Jørstad KE, van der Meeren T, Paulsen OI, Thomsen T, Thorsen A, Svåsand T (2008) Escapes of eggs from farmed cod spawning in net pens: recruitment to wild stocks. Rev Fish Sci 16: 1-11 Knutsen H, Moland OE, Ciannelli L, Heiberg ES, Knutsen JA, Simonsen JH, Skreslet S, Stenseth NC (2007) Egg distribution, bottom topography and small scale cod population structure in a coastal marine system. Mar Ecol Prog Ser 333: 249-255 Krkošek M, Ford JS, Morton A, Lele S, Myers RA and Lewis MA (2007) Declining wild salmon populations in relation to parasites from farm salmon. Science 318: 1772-1775 Lura H, Sægrov H (1991) Documentation of successful spawning of escaped farmed female Atlantic salmon, Salmo salar, in Norwegian rivers. Aquaculture 98: 151-159 Lura H, Sægrov H (1993) Timing of spawning in cultured and wild Atlantic salmon (Salmo salar) and brown trout (Salmo trutta) in the River Vosso, Norway. Ecol Freshwat Fish 2: 167-172 Meager JJ, Skjæraasen JE, Fernö A, Karlse Ø, Løkkeborg S, Michalsen K, Utskot SO (2009) Vertical dynamics and reproductive behaviour of farmed and wild Atlantic cod Gadus morhua. Mar Ecol Prog Ser 389: 233-243

www.preventescape.eu

McGinnity P, Stone C, Taggart JB, Cooke D and others (1997) Genetic impact of escaped farmed Atlantic salmon (Salmo salar L.) on native populations: use of DNA profiling to assess freshwater performance of wild, farmed, and hybrid progeny in a natural river environment. ICES J Mar Sci 54:998–1008 McGinnity P, Prodohl P, Ferguson K, Hynes R and others (2003) Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. Proc Biol Sci 270: 2443–2450 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007) Technological solutions and operational measures to prevent escapes of Atlantic Cod (Gadus morhua) from sea-cages. Aquacult Res 38: 90-99 Naylor R, Hindar K, Fleming IA, Goldburg R, Williams S, Volpe J, Whoriskey F, Eagle J, Kelso D, Mangel M (2005). Fugitive salmon: assessing the risks of escaped fish from net-pen aquaculture. BioScience 55(5): 427-437 Norwegian Directorate of Fisheries Directorate (2009) Statistics for Aquaculture 2008. www. fiskeridir.no/fiskeridir/kystsone_og_havbruk/statistikk Øines Ø, Simonsen JH, Knutsen JA, Heuch PA (2006) Host preference of adult Caligus elongatus Nordmann in the laboratory and its implications for Atlantic cod aquaculture. J Fish Dis 29:167-174 Petridis D, Rogdakis Y (1996) The development of growth and feeding equations for seabream, Sparus aurata L., culture. Aquacult Res 27: 413-419 Revie C, Dill L, Finstad B, Todd CD (2009) Sea Lice Working Group Report. NINA Special Report 39: 1-17. Sánchez-Lamadrid A (2002) Stock enhancement of gilthead seabream (Sparus aurata, L.) assessment of season, fish size and place of release in SW Spanish coast. Aquaculture 210: 187-202 Sánchez-Lamadrid A (2004) Effectiveness of releasing gilthead seabream (Sparus aurata, L.) for stock enhancement in the bay of Cádiz. Aquaculture 231: 135-148 Skaala Ø, Wennevik V, Glover KA (2006) Evidence of temporal genetic change in wild Atlantic salmon, Salmo salar L., populations affected by farm escapees. ICES J Mar Sci 63: 1224-1233 Skilbrei O, V Wennevik (2006) Survival and growth of sea-ranched Atlantic salmon, Salmo salar L., treated against sea lice before release. ICES J Mar Sci 63: 1317-1325 Skjæraasen JE, Mayer I, Meager JJ, Rudolfsen G, Karlsen Ø, Haugland T, Kleven O (2009) Sperm characteristics and competitive ability in farmed and wild cod. Mar Ecol Prog Ser 375: 219-228

Soto D, Jara F, Moreno C (2001) Escaped salmon in the inner seas, southern Chile: facing ecological and social conflicts. Ecol App 11: 1750-1762 Svåsand T, Crosetti D, García-Vázquez E, Verspoor E (editors) (2007) Genimpact - Evaluation of genetic impact of aquaculture activities on native populations. http://genimpact.imr.no/__ data/page/7649/genetic_impact_of_aquaculture.pdf Tancioni L, Mariani S, Maccaroni A, Mariani A, Massac F, Scardia M, Cataudella S (2003) Localityspecific variation in the feeding of Sparus aurata L.: evidence from two Mediterranean lagoon systems. Estuar Coast Shelf Sci 57: 469-474 Thorstad EB, Fleming IA, McGinnity P, Soto D, Wennevik V, Whoriskey F (2008) Incidence and impacts of escaped farmed Atlantic salmon Salmo salar in nature. NINA Special Report 36: 1-110 Toledo Guedes K, Sanchez-Jerez P, Gonzalez-Lorenzo G, Brito Hernandez A (2009) Detecting the degree of establishment of a non-indigenous species in coastal ecosystems: seabass Dicentrarchus labrax escapes from sea cages in Canary Islands (Northeastern Central Atlantic). Hydrobiologia 623: 203-212 Triantafyllidis A (2007) Aquaculture escapes: new DNA based monitoring analysis and application on seabass and seabream. CIESM Workshop Monograph 32: 67-71 Uglem I, Bjørn PA, Dale T, Kerwath S, Økland F, Nilsen R, Aas K, Fleming I, McKinley RS (2008) Movements and spatiotemporal distribution of escaped farmed and local wild Atlantic cod (Gadus morhua L.). Aqua Res 39: 158-170 Uglem I, Bjørn P-A, Mitamura H, Nilsen R (2010) Spatiotemporal distribution of coastal and oceanic Atlantic cod (Gadus morhua L.) sub-groups after escape from a farm. Aquacult Environ Interact 1:11-20 van der Meeren T, Jørstad K (2009) Fanger torsk på vidvanke. Nytt fra havbruk 2009(2): 1 (In Norwegian) Weir LK, Grant JWA (2005) Effects of aquaculture on wild fish populations: a synthesis of data. Environ Res 13: 145-168 Whoriskey FG, Brooking P, Doucette G, Tinker S, Carr (2006) Movements and survival of sonically tagged farmed Atlantic salmon released in Cobscook bay, Maine, USA. ICES J Mar Sci 63: 1218-1223 Wright PJ, Galley E, Gibb IM, Neat FC (2006) Fidelity of adult cod to spawning grounds in Scottish waters. Fish Res 77: 148-158

www.preventescape.eu

seeks to reduce the number of fish that escape from European aquaculture through research to improve fish farming techniques and technologies.

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

2. A pan-European evaluation of the extent, causes and cost of escape events from sea-cage fish farming Cite this article as: Jackson D, Drumm A, McEvoy S, Jensen Ă&#x2DC;, Dempster T, Mendiola D, Borg JA, Papageorgiou N, Ioannis Karakassis (2013) A panEuropean evaluation of the extent, causes and cost of escape events from sea-cage fish farming. In: PREVENT ESCAPE Project Compendium. Chapter 2. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Dave Jackson1, Alan Drumm1, Sarah McEvoy1, Ă&#x2DC;sten Jensen2, Tim Dempster2, Diego Mendiola3, Joseph A Borg4, Nafsika Papageorgiou5, Ioannis Karakassis5 The Marine Institute, Ireland, SINTEF Fisheries and Aquaculture, Norway , 3 Azti Tecnalia, Spain, 4 University of Malta, Malta, 5 University of Crete, Greece. 1 2

www.preventescape.eu

Introduction Knowledge of the extent and causes of escape incidents from sea-cage fish farms varies greatly from country to country across Europe. Several countries, such as Norway, Scotland and Ireland, have legislated reporting requirements whereby farmers are obligated to report escape incidents, their size and cause, and when they occur. In contrast, Mediterranean countries have no such requirements, thus no statistics are available on the number of escapes or the underlying causes of escapes (Dempster et al. 2007). Norway has the most comprehensive record of escapes, dating back approximately 15 years for salmonids and 5 years for Atlantic cod. A total of 722,000 and 963,000 salmon and rainbow trout were reported to have escaped from Norwegian farms in 2005 and 2006, respectively (Norwegian Fisheries Directorate 2007). The real number of escapes has been estimated to be considerably greater (Torrissen 2007), because not all escape incidents are reported. Significant escape events of salmon have also occurred in other major salmonid producing countries, such as Scotland, Chile and Canada (Soto et al. 2001, Naylor et al. 2005). Over one million salmon were reported to have escaped from Scottish farms during the period from 2002-2006 (Thorstad et al. 2008). Over the decade from 2001 to 2009, 3.93 million Atlantic salmon, 0.98 million rainbow trout and 1.05 million Atlantic cod were reported to have escaped from Norwegian fish farms (Jensen et al. 2010). The proportion of Atlantic cod that escape is high in comparison to salmon (Moe et al. 2007a). While no official statistics on the extent of escapes exist for Mediterranean countries, data available from companies that insure fish farm businesses indicate that escapes are a significant component of economic losses claimed by farmers (EU FP-6 ECASA project; www.ecasa.org. uk). From 2001-2005, 76 claims accounting for 36% of the total value of all insurance claims made by fish farmers in Greece were due to stock losses from storms, while damage to farm equipment due to storms accounted for 19%. Further, 39 registered â&#x20AC;&#x2DC;predator attacksâ&#x20AC;&#x2122; resulted in claims of 10.4% of the total value of all insurance claims, although the proportion of this which relates to stock loss or cage damage is unknown. The existing evidence suggests that escapes are a relatively frequent occurrence on a pan-European scale.

Causes

of escapes

Escapes are caused by a variety of incidents related to farming equipment and their operation. Reports by fish farming companies to the Norwegian Fisheries Directorate following escape events during the period from 2001-2006, indicate that escapes can be categorized broadly into structural failure (52%), operational related failure (31%) and biological and other causes (17%). Structural failures may be generated by severe environmental forcing in strong winds, waves and currents, which may occur in combination with component fatigue or human error in the way farm installations have been installed or operated (Jensen et al. 2006). Operational related failures leading to escapes include

collisions with boats, incorrect handling of nets or damage to nets by boat propellers. The risks to farm installations from the marine environment largely come from exposure to waves and currents (Lader et al. 2007, 2008) and from collisions with seagoing vessels. The further offshore a farm is located, generally the more exposed it is to the elements, thus increasing the risk of escapes. For cod, there is gathering evidence that the reasons for escape differ from salmon. This stems from behavioural variations in captivity. Firstly, cod bite the net and might thus increase wear and tear and contribute to the creation of holes (Moe et al. 2007a). Secondly, cod show more pronounced exploratory behaviour than salmon and might thus have a higher probability of discovering small holes in the net (Hansen et al. 2008). Official statistics and other sources of information which apportion causality to escape events provide little explicit detail to support technological development that will improve farming equipment and modify operations to avoid mistakes that cause escapes. Categorization of causes may also be inaccurate, as causes are rarely investigated in detail (Valland 2005). Such detail only comes through thorough investigation of the causes of escape incidents on a case by case basis (e.g. Rist et al. 2004). Jensen (2006) visited 8 fish farms in northern Norway after severe storms caused damage to numerous farms in the region. While â&#x20AC;&#x2DC;stormâ&#x20AC;&#x2122; was listed as the official cause of these escapes, the specific circumstances behind each event varied widely. Storms may damage surface floaters, tear nets through net deformation or rubbing of the nets on net weights in the strong currents they generate, and overload the mooring structures that hold the farm in place. At a smaller scale, an understanding of how individual components perform in the mooring system (such as anchors, shackles, ropes, bolts and mooring coupling plates), the cage system (net material, cage ropes and cage weights) and the steel platform or polyethylene floaters is crucial to ensure each element is engineered to match the particular characteristics of each farm type and location. Standardised investigations of structural failures or accidents are common in other industries (e.g. construction and automotive industries), yet when a fish farm which contains millions of euros of fish crashes, no similar process exists.

www.preventescape.eu

The risk of an escape event is intricately linked to the location, containment methods, and handling equipment and methods used to culture the fish. This risk is compounded by farm management practices such as grading, treating, transferring of fish, towing of cages and harvesting. Culturing fish in sea-cages relies on using ‘hard’ cage structures to support ‘soft’ nets which contain the fish. Thus, correct cage and net design are of critical importance to minimize escapes (Moe et al. 2005). A variety of different cage designs are used in aquaculture currently, consisting of steel cages, plastic cages and rubber cages. These also differ in size and shape depending on the species and the stage of production of the fish, but typically range from 1000 to 20 000 m3 (Sunde et al. 2003). These cage and net structures rely on mooring and buoy systems to anchor the cages in situ.

Lack

Knowledge

Knowledge of the direct economic costs of escapes is an important element in understanding which technological and operational measures farmers are likely to implement to prevent escapes. Sea-cage fish farming is a highly competitive business, with profits made on the margin between the cost of production and sale price. Industry may resist implementation of particular preventative measures if they prove too costly. A full understanding of the economic cost of escapes, through equipment damage, operations to repair damages, loss of stock and loss of potential earnings from future growth of any lost stock, will enable the cost of escapes to be compared to the cost of mitigation actions and investments in improved containment technologies. Technical improvements to aquaculture facilities and operations are essential for preventing escapes and for curbing economic losses due to lost production volumes. As a first step towards preventing escapes, knowledge of how, why, when and where fish escape is critical for improving farm equipment and operations. A range of detailed information is needed to assess the complex interactions among particular culture technologies, fish farming operations and the characteristics of farming locations that lead to escapes. Culture methods and technologies within Europe vary considerably depending on the species farmed and geographic regions, therefore factors contributing to escapes need to be mapped and analysed in various locations to pinpoint the main causes of escape. A detailed picture

of how farms breakdown in storms, and which environmental and operational causes lie behind the formation of net tears and holes, need to be understood before improvements can be made successfully. Knowledge of the direct economic costs of escapes is a key aspect in understanding which technological and operational measures farmers are likely to implement. Work package 2 (MAP Escape), documents the extent and costs of escapes. In this report we present the biological, technical and operational causes giving rise to escapes of fish from seacage fish farms in marine waters throughout Europe over a three year period. We developed a specific methodology which we applied across all 6 countries (Ireland, UK, Norway, Spain, Greece, and Malta) in order to ensure comparability of results. The methodology was made up of the following components and actions: • Consult with industry and relevant agencies through a confidential questionnaire and follow up interviews to gather information on methodologies and technologies currently used in on-growing finfish in the marine environment. • Gather available existing information on the extent, size and knowledge of the causes of escapes from national reports and other published data. • Conduct detailed assessments of the explicit technical or operational causes of escapes at sea-cage fish farms throughout Europe by direct assessment of known escape events at industrial fish farms, by way of site visits and interviews. • Establish the total economic cost of escape events through a cost evaluation using both available data and through direct gathering of data by way of interview.

Results

of preliminary questionnaires:

The questionnaire was divided into 4 main sections: • • • •

Infrastructure Maintenance Escapes Environment

www.preventescape.eu

Section 1, Infrastructure, was designed to gather data relating to materials used and design of floater types, nets and mooring systems. Section 2, Maintenance, established if the site employed maintenance management systems for the infrastructure and how these maintenance systems were carried out. Section 3, Escapes, was used to establish if there were escape incidents and if so, how many and if there is further information available on the events. This section also required the farmers to give an estimate of the cost of the stock loss and clean up operations to the business. Finally, Section 4, Environment, was used to gather the environmental data available for the sites in question. A total of 242 escape incidents were identified through questionnaires, which were completed across the 6 countries and other data supplied by the Norwegian Fisheries Directorate and the Scottish Aquaculture Research Forum. The causes given for these events are shown below in Table 2.1. Some of the events were as a result of a combination of causes. The majority of escape incidents was related to net damage due to predator attacks and abrasion. Storm damage or weather was also a common cause. However, it was not clear from the responses obtained whether the storm losses were due to net, mooring or floater damage. Species

Total

Structural

No. of incidents

No. of escapes

No. of incidents

No. of escapes

No. of incidents

No. of escapes

113

820158

678279

6758

Cod

457005

16466

118974

Seabass

599600

540000

52100

Seabream

6846100

6181900

604000

Meagre

200,000

Totals

242

8922863

7616645

781832

Atlantic Salmon

Biological Operational External

Unknown No. of No. of incidents escapes

No. of incidents

No. of escapes

No. of No. of incidents escapes

88065

13194

33862

11839

180717

129009

7500

20000

25200

15000

119904

226611

177871

Table 2.1. General causes of escape incidents and numbers of escaped fish.

A total of 8,922,863 fish were reported to have escaped from the 242 incidents. Seabream accounted for the highest number of escapes at 76.7% followed by Atlantic salmon at 9.2%. Of the 6,846,100 bream reported to have escaped, two of the incidents accounted for 1.9 and 3.8 million fish, respectively. It should also be noted that three of the escape incidents relating to seabream had unknown numbers of fish reported. Of the 113 Atlantic salmon escape events, almost 75% were due to structure failure or operational error. Almost 50% of cod escape incidents were due to biological causes (e.g. biting of nets). One major incident involving a trawler accounted for 34% of all cod escapes over the selected period. The majority of escape incidents (Figure 2.1) relate to the enclosure netting, with biting of nets being most common. This net biting is a behavioural characteristic of both cod and seabream. While the type of predators causing net failure differs from the Atlantic (e.g. seals) to the Mediterranean (e.g. dolphins/wild fish), the outcome is the same. At the outset of the project many would have assumed that storm damage would rate very highly on the list of causes and, while it is a very significant cause, escapes due to human error are greater.

Figure 2.1. Underlying causes of escapes versus number of incidents.

Assessment

of the technical and operational causes of escape

events

The data collected through the MAP Escape questionnaire-based interview process was added to the official data available from Norway and the UK. Each partner was contacted and requested to identify 5 escape events in their region which were to be investigated in greater detail. It was suggested selecting the incidents which were due to either structural (cage break, mooring failure, mechanical net damage), biological (net biting, predator attack) or operational (harvesting, grading, and transport) failure. It was also indicated that the partners should focus on farms which had large losses and farms with regular losses. A standardised format for the assessments was documented, and qualified aquaculture engineers and aquaculture operations specialists oversaw the process to ensure standardisation of results. In some countries it was necessary to focus on a few companies which had encountered several escape events. While the data collated from the questionnaires proved very useful, the follow up investigations allowed for confirmation of the data supplied through the questionnaires. It also permitted for more detail to be gathered regarding inspection and maintenance, procedures. Although all of the producers visited had procedures for inspection and maintenance in some cases they were not as comprehensive as the data from the questionnaires had suggested.

www.preventescape.eu

Cages The majority of the producers who responded were using circular cages constructed of plastic. Square plastic cages (Figure 2.2) were being used for juvenile on-growing in sheltered sites in some locations before being transferred to larger circular cages in more exposed sites. Some companies which are sited in sheltered sites still prefer the steel constructed cages. However, where they are used in more exposed locations they have caused problems due to their less flexible nature (Figure 2.3). There is a general trend towards the use of circular plastic cages in all of the 6 countries in the study. Where circular plastic cages have failed during extreme weather events, it has usually been as a consequence of a failure in the mooring system. Figure 2.2. Square plastic cages for juveniles.

Figure 2.3. Steel cages with broken walkway.

Netting Information gathered through the questionnaire and from other sources, including the Norwegian Directorate of Fisheries, indicates that the vast majority of producers choose nylon as the preferred material for their nets. Table 2.2 gives the data acquired from the questionnaire. Although mesh sizes and breaking strengths used are similar for the species grown, construction of the nets is very much a personal choice. Most producers indicated that they would have their nets constructed to their own specification. These specifications were invariably developed through trial and error over the years. Some producers had encountered problems with quality of the netting material supplied for net construction. While they did have control over the specific construction of a net, in some cases they had little control over the source of the netting material. When asked if they would be in favour of an International/EU standard for netting materials, all were very much in favour. Many producers expressed an interest in the use of Dyneema速 nets but felt that they could not justify the additional cost of the material. Netting Materials Nylon

Nylon/Polyester

Ireland

Dyneema

Polypropylene

Unknown 1

Norway

Spain

Malta

Greece

Totals

Table 2.2. Net Material used in participating countries.

One area which raised some concerns was the net mending procedures. Some producers had poor systems when it came to marking areas of the net which required attention (Figure 2.4) and therefore some repairs were not complete. The length of time a net was in service varied considerably, from 3 to 10 years. Although the life span of a net is dependent on the materials used and the environment it has to function in, some of the nets observed were in poor condition. Many of the producers employed a series of systems to ensure effective maintenance and quality control of nets. The use of tags to mark areas for repair (Figure 2.5) and the subsequent removal of tags as the repairs were carried out, was a system employed in several major companies. The absence of tags on the repaired net ensures all identified areas for repair have been addressed. The routine testing of netting quality by a relatively simple and inexpensive system of strength testing, using a mesh strength tester gauge (Figure 2.6), between deployments was seen as best practice. However, a number of producers still relied on visual and manual strength tests i.e. pulling individual meshes with ones fingers.

www.preventescape.eu

Figure 2.4. Net with cable ties.

Figure 2.5. Net with tag.

Figure 2.6. Hydraulic net strength testing.

Net

weighting systems

From discussions with producers during the follow up interviews it became evident that this is one critical area which producers struggle with. Some producers are fully committed to the use of individual weights distributed evenly around the base of the net while others favour the ring weight system. While the stable platform provided by steel cage systems provides a structure for the suspension of corner weight systems, the flexible circular plastic cages provide a greater challenge. Many of the holes in netting have been caused by the sinker tubes or their suspension chain chafing the net. Incorrect weighting has also resulted in increased predator attack due to lack of tension of the net walls. Where there is preference for the ring weights, the producers feel it gives a better all-round tension to the net. However, where the ring is connected directly to the net it can lead to tearing of the mesh. Some producers suspend the ring using light rope so that should the pressure become great, then the rope will detach without tearing the net. For this very reason many producers prefer the individual weights (Figure 2.7) because, should the ring detach, the net has no weight and the net loses its shape.

Figure 2.7. Individual weight system.

www.preventescape.eu

Since the weighting of the net structure is so important, many producers felt that more research and advice on the weighting systems would be very useful. Most agreed that the system is site-specific and that it should be designed with the siteÂ´s environmental characteristics taken into account along with the cage structure. Information from Norway shows that sinker tubes are the favoured method of weighting on the larger circular plastic cages (Figure 2.8). The results on weighting systems collected by means of the questionnaire are presented in Table 2.3.

Figure 2.8. Ring weighting system.

Lead line

Lead line/weights

Weights 4

Ireland

Ring weight

Norway

Spain

Malta

Greece Totals

25 4

*Two sites in Spain reported using weights with ring weight

Table 2.3. Net weighting systems.

Unknown

Mooring

systems

The results of the questionnaires and follow up interviews indicate that many of the producers are relying on the experience and advice from the cage manufactures for their mooring systems. Because of the increase in the use of circular plastic cages many producers now employ the grid mooring system (Figure 2.9).

Figure 2.9. Grid mooring diagram.

While this mooring system provides greater safety since one component can compensate for a malfunction in another, it is only as good as the quality and standard of the components used. A key component of the mooring structure is the mooring plate and/or mooring ring (Figure 2.10). It is essential that these components are rated to withstand the expected loads as failure to one or more of these can lead to complete failure of the system. Figure 2.10. Broken mooring ring.

www.preventescape.eu

Maintenance

systems

All respondents indicated that they had maintenance systems in place for their cages, nets and mooring systems. Many carry out visual inspection on cages on a daily basis and visual inspections by divers on a regular basis. On most sites, when a diver comes across a hole or weakness it is marked with a plastic cable tie. In some countries the use of divers is quite costly so net inspections may not be carried out as frequently as desired. Most mooring systems are inspected on an annual basis by either divers or ROVâ&#x20AC;&#x2122;s.

Discussion Out of a total of almost 9 million escapees recorded in the study period, over 75% were accounted for by escapes of seabream. Of these, over 5 million occurred in two catastrophic escape incidents. The most significant factor in terms of number of fish escaping, (Figure 2.11) was mooring failure. In terms of numbers of fish escaping this factor accounts for over two thirds of all escapees recorded in the study. Percentage of escapees by number attributed to the 6 main underlying causes of escape incidents 4% 10% Inappropriate moorings

Predator Inappropriate cages

Unknown (hole in net) Other Storm event

9% 67%

Figure 2.11. Percentage number of escapees, attributed to 6 main underlying causes of escape incidents.

By far, the most significant cause in terms of numbers of escape incidents was a hole in the net due to either biting (16%), predator damage (14%) or other causes. When the causes of holes in the net are examined it can be seen that taken together, net biting and predator damage account for almost half (47%) of escape incidents due to a hole in the net (Figure 2.12).

Figure 2.12. Overall causes of hole in net (number of incidents as a percentage of total incidences of hole in net).

The number of escapees recorded in northern Europe was much lower than in the Mediterranean. The total number of salmon and cod recorded as escaping during the study period was 1.27 million or less than 15% of the total number of escapees recorded. This may in some measure be related to the more highly developed standards for equipment and structures including mooring arrays. The Norwegian government has enacted legislation called NYTEK, first version mandatory from 2006, updated version mandatory from 2012, with some parts not mandatory until 2013, in which it specifies the technical standard NS9415, first edition in 2003 revised in 2009 mooring systems and other components of the farm structure. This technical standard relates to cages, mooring systems and other components. Since the implementation of this legislation there has been a reduction in both the number of escape incidents and the numbers of fish escaping (Jensen et al. pers. comm.). The role of holes in the net in a large number of escape incidents points towards the need to improve surveillance of net integrity, preventative maintenance programmes and testing, and inspection of nets before deployment or redeployment. Where such programmes are widely employed, numbers of escapes are significantly lower.

Cost

Escapes

Introduction The cost of escapes from marine fish farms can be evaluated in a number of different ways. Depending on the starting point, the parameters and paradigm used to quantify costs can be very different. Many of the concerns held over the impacts of escapees relate to potential negative impacts on the surrounding environment. If such impacts were well described they could be assigned a cost, but doing so would be fraught with multiple assumptions based on very scant data. There is however a very pragmatic and relevant basis for assigning a cost to

www.preventescape.eu

aquaculture escapees; the measure of lost income at point of first sale due to loss of stock due to escape incidents. As part of the FP7 project Prevent Escape, an exercise to evaluate the cost of escapees in partner countries was undertaken. The basis of this exercise was to calculate the numbers of fish escaping and to assign them an appropriate value at point of first sale. A specific methodology was developed and applied across all of the countries in order to facilitate comparability of results. In the development of this methodology, cognisance had to be taken of the quality and extent of available data and information available. Where possible, published figures such as FAO statistics, together with nationally available official figures, were relied on as a basis for calculations. This data was combined with the outputs from the MAP Escape component of the Prevent Escape project to derive costs with a defined set of assumptions and limitations. The analysis was carried out for six countries: Ireland, Norway, Scotland (UK), Spain, Greece and Malta. For Ireland, Norway and Scotland, it was possible to obtain official figures for the total number of escapes. These were used as a basis for calculating the value at point of first sale of the escapees. The number of escapees per annum was multiplied by the average harvest weight in kilos for farmed salmon in each country and the result was multiplied by the average value per kilo of salmon sold, for that year in each country. Average weight at harvest and average first sale prices were obtained from representative organisations, national statistics and other recognised sources of such commercially sensitive information. The results are presented in Table 2.4. The total value of escaped salmon in terms of first sale value was estimated at 4.7 million per annum. Value of losses Norway 2007

2008

2009

Losses

246,488

76,387

180,407

Av price/kg (€)

3.24

3.3

3.76

Losses value per kg (€)

798,621

252,077

678,330

Total Value (€) at av size 5kg

3,993,105

1,260,385

3,391,650

8,645,140

Value of losses Scotland Losses

2007

2008

2009

136,891

56,941

88,044

Av price/kg (€)

3.64

3.27

3.2

Losses value per kg (€)

498,283

186,197

281,740

Total Value (€) at av size 5kg

2,491,415

930,985

1,408,704

4,831,104

Value of losses Ireland 2007

2009

Losses

35,000

Av price/kg (€)

5.35

Losses value per kg (€)

187,250

Total Value (€) at av size 3.67kg

687,208

Total

Table 2.4. Cost of losses Norway, Scotland and Ireland.

2008

687,208 14,163,452

For Malta, Spain and Greece official statistics on escapes are not compiled. In these countries, the results obtained in the questionnaire and supplemented by follow up investigations and interviews were used as the basis for calculating the number of escapees. The number of escapees recorded was taken as a representative subsample and the estimated total calculated by reference to the proportion of the total production sampled (Table 2.5). For example if 20% of the farm production was sampled the resulting figure was raised by a factor of 5 to give an estimate of the national total. Where the subsample was large, such as Malta (>60%) there is a higher confidence regarding the accuracy of the resulting estimate than where the sample represents a smaller proportion of the national production. FAO and Globefish statistics were used to calculate value at point of first sale and the size at point of sale was set at 500g. Over a three year period a total of 255 escape incidents were documented, representing a variable proportion of the farms in operation in each country. This percentage varied from a low of 20% to a maximum of 75%. In northern European countries, national statistics are available on total escapes due to mandatory reporting requirements. Where available, these were used as a basis for calculations. Results indicate that the cost to the industry, in terms of loss of sales revenue at point of first sale, is in terms of tens of millions of euro per annum. The partners are currently carrying out a scoping exercise to attempt to produce a validated figure for an average annual cost for the European industry, in terms of euro per tonne of licensed production. This figure could then act as a baseline to measure improvements in efficacy of containment against, and to derive cost-benefit metrics for, improvements in containment.

Spain

Greece*

Malta

No. of species from questionnaires

Escape Incidents

No. of fish escaped over 3 years

Escapes per annum

2007 price/ kg

Value (500g)

No. of sites

% overall sites

Total value of lost annual production

15 Bream

5,849,000

1,949,666

€4.30

€4,191,781

111

21%

€19,960,861

2 Bass

520,000

173,333

€4.98

€431,599

108

23%

€1,876,521

1 Meagre

200,000

66,666

10%

15 Bream

909,200

303,066

€3.57

€540,972

210

2.8%

€19,320,428

9 Bass

65,100

21,700

€5.12

€55,552

130

3.8%

€1,461,894

22 Bream

87,900

29,300

€4.33

€63,434

75%

€84,579

4 Bass

14,500

4,833

€14.71

€35,546

67%

€53,053

TOTAL

€42,757,336

Market size average for both Seabream and Seabass 500g * 340 is the combined number of farms in 2006. FAO Globefish 2007 Used the species volume ratio for site numbers.

Table 2.5. Cost of losses Spain, Greece and Malta.

www.preventescape.eu

Conclusions • Significant numbers of fish escape in all sectors of the finfish industry in the areas studied (>9 million fish over the period of the study). • The cost of escapes at point of first sale is very significant in terms of lost income (€ 47.5 million p.a.). • Implications of escapes have been shown to have negative effects on the viability of individual commercial concerns. • The public perception of the aquaculture industry has also been adversely affected by publicity surrounding high profile escape incidents. • The number one cause of escape incidents was due to net biting and the number one cause of large escape numbers was mooring failure. • There was a large variation in the level of awareness of the necessity of both training of staff and procedures or Standard Operating Procedures on containment-related issues. • Two key drivers towards reducing escapes identified by the industry were standards for materials and site-specific procedures and processes to ensure the use of appropriate equipment and its maintenance.

Recommendations • International/EU Standard for netting materials. • Development of a Standard Operating Procedure for net strength testing and maintenance on sites. • Increased research on the weighting systems to be employed. • Increased research and trials of new materials for netting. • Development of a Containment Training Module for farm staff. • Development of an accurate mandatory reporting system for escape incidents within the EU. • Further research into the prevention of net biting, e.g. new materials, net sealants. • Develop guidelines on evaluation and design factors of mooring systems related to a range of exposed environments.

References

cited

Dempster T, Moe H, Fredheim A, Sanchez-Jerez P (2007) Escapes of marine fish from sea-cage aquaculture in the Mediterranean Sea: status and prevention. CIESM Workshop Monograph 32: 55- 60. www.ciesm.org/online/monographs/Lisboa.html Hansen LA, Dale T, Uglem, I, Aas K, Damsgård B, Bjørn P A (2008). Escape related behaviour of Atlantic cod (Gadus morhua L) in a farm situation. Applied Animal Behavioural Science (in press). Jensen Ø (2006) Assessment of escape causes from Norwegian fish farms during two storm periods in January 2006. SINTEF Report SFH80 A066056. ISBN 82-14-03953-8. Jensen Ø, Dempster T, Thorstad EB, Uglem I, Fredheim A (2010) Escapes of fishes from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquacult Environ Interact 1:71-83 Lader PF, Olsen A, Jensen A, Sveen JK, Fredheim A, Enerhaug B (2007) Experimental investigation of the interaction between waves and net structures - Damping mechanism. Aquaculture Engineering 37(2): 100-114 Lader PF, Fredheim A (2007) Dynamic properties of a flexible net sheet in waves and current – A numerical approach. Aquaculture Engineering 35(3): 228-238 Lader P, Dempster T, Fredheim A, Jensen Ø (2008) Current induced net deformations in fullscale seacages for Atlantic salmon (Salmo salar). Aquaculture Engineering 38(1): 52-65 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007a) Technological solutions and operational measures to prevent escapes of Atlantic Cod (Gadus morhua) from sea-cages. Aquaculture Research 38: 90-99 Moe H, Gaarder R, Sunde LM, Borthen J, Olafsen K (2005) Escape-free nets for cod. SINTEF Fiskeri og havbruk Report SFH A 054041, Trondheim, Norway (In Norwegian) Naylor R, Hindar K, Fleming IA, Goldburg R, Williams S, Volpe J, Whoriskey F, Eagle J, Kelso D, Mangel M (2005) Fugitive salmon: assessing the risks of escaped fish from net-pen aquaculture. BioScience 55(5), 427-437 Norwegian Fisheries Directorate (2007) Statistics for Aquaculture 2007 (in Norwegian). http://www.fiskeridir.no/fiskeridir/kystsone_og_havbruk/statistikk Rist T, Skjeggedal K, Haga B, Monsen B-RH, Rysjedal J, Vad J, Åsvang H (2004) Fisken rømmer. En risikoanalyse av driftsrelaterte årsaker. Aqua Man AS, 35 pp. (In Norwegian)

www.preventescape.eu

Soto D, Jara F, Moreno C (2001) Escaped salmon in the Inner Seas, Southern Chile: facing ecological and social conflicts. Ecological Applications 11(6): 1750-1762 Sunde LM, Heide MA, Hagen N, Fredheim A, ForĂĽs E, Prestvik Ă&#x2DC; (2003) Review on technology in the Norwegian aquaculture industry. SINTEF Fiskeri og havbruk Report STF 80 A034002, Trondheim, Norway. 32 pp Thorstad EB, Fleming IA, McGinnity P, Soto D, Wennevik V, Whoriskey F (2008) Incidence and impacts of escaped farmed Atlantic salmon Salmo salar in nature. Report from the Technical Working Group on Escapes of the Salmon Aquaculture Dialogue, January 2008. 108 pp Torrissen OJ (2007) Status report for Norwegian aquaculture 2007. Kyst og Havbruk 2007: 11-12 (in Norwegian) Valland A (2005) The causes and scale of escapes from salmon farming. In: Interactions between aquaculture and wild stocks of Atlantic salmon and other diadromous fish species: science and management, challenges and solutions. ICES/NASCO Bergen 18-21 October 2005, pp. 15

www.preventescape.eu

seeks to reduce the number of fish that escape from European aquaculture through research to improve fish farming techniques and technologies.

3.1. Do

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

farming environments and the inherent

behaviours of fish predispose some species to higher rates of escape?

Cite this article as: Noble C, Damsgård B, Hedger RD, Uglem I, Papadakis VM, Papadakis IE, Glaropoulos A, Kentouri M, Smith CJ, Zimmermann E, Fleming IA, Høy E, Dempster T (2013) Do farming environments and the inherent behaviours of fish predispose some species to higher rates of escape? In: PREVENT ESCAPE Project Compendium. Chapter 3.1. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Chris Noble1, Børge Damsgård1, Richard D. Hedger2, Ingebrigt Uglem2, Vassilis M. Papadakis3, Ioannis E. Papadakis3, Alexios Glaropoulos3, Maroudio Kentouri3, Christopher J. Smith4, Emily Zimmermann5, Ian A. Fleming5, Erik Høy6, Tim Dempster6 Nofima, Norway, Norwegian Institute for Nature Research, Norway, 3 University of Crete, Greece, 4 Hellenic Centre for Marine Research, Greece, 5 Ocean Sciences Centre, Memorial University of Newfoundland, Canada, 6 SINTEF Fisheries & Aquaculture, Norway 1 2

www.preventescape.eu

Escape-related

fish behaviours in cod, seabass and seabream

Although the majority of escape events are caused by structural and operation failures (see Jensen et al. 2010), Atlantic cod, European seabass and gilthead seabream may exhibit behaviours that lead them to bite net walls and also escape through holes in sea-cages (see ICES 2006, Moe et al. 2007, Hansen et al. 2009, Papadakis et al. 2012). For example, Atlantic cod form loose and unsynchronised shoals in cages (e.g. Huse 1991) and frequently display investigatory and exploratory behaviours that bring them in close proximity to the net wall (Rillahan et al. 2011). Gilthead seabream are a largely benthic species that live in rocky or sea grass habitats, so it is likely that they will exhibit similar exploratory behaviour towards the cage walls (see Figure 3.1.1 for a seabream escape attempt through a hole). Seabream have also been reported to bite the net in specific places (ICES 2006) and this may weaken netting and lead to the formation of holes. European seabass can shoal and may also exhibit exploratory behaviour when held in sea cages, and their behaviour can be modified by common farm practices such as feeding (Andrew et al. 2002, SarĂ et al. 2010; see Figure 3.1.2 for a seabass escape attempt through a hole). Other farmed species such as Atlantic salmon (Salmo salar) do not exhibit these behaviours: they generally avoid the cage walls and net bottom, and demonstrate schooling and shoaling behaviours (Juell 1995). These speciesspecific differences in behaviour can have a considerable effect upon a fishâ&#x20AC;&#x2122;s propensity to escape, and it has been suggested that Atlantic cod are 10 to 20 times more likely to escape from sea-cages than Atlantic salmon (Norwegian Directorate of Fisheries 2005).

Figure 3.1.1. Gilthead seabream (Sparus aurata) escape attempt (Photo: C. Smith, HCMR).

Figure 3.1.2. European seabass (Dicentrarchus labrax) escape attempt (Photo: C. Smith, HCMR).

The biological and husbandry factors that may promote escape-related fish behaviours are poorly understood. Previous work has suggested a lack of feed may increase the risk of escapes in Atlantic cod, by increasing the frequency of net inspection and biting behaviours (Moe et al. 2007, Hansen et al. 2009; Figure 3.1.3). However, the effects of other farm management factors such as net cleanliness or stocking densities upon escape-related fish behaviours is poorly known for numerous European farm species.

Figure 3.1.3. Atlantic cod (Gadus morhua) biting a loose thread on the net wall (Photo: C. Noble, Nofima).

Determining the biological mechanisms and risk factors that may promote escape-related behaviours in commercial aquaculture The lack of knowledge on the biological mechanisms and risk factors that may promote escape-related behaviours in commercial aquaculture situations restricts our ability to implement mitigative actions aimed at reducing or preventing fish escapes from sea-cage facilities. Basic questions such as the effects of common husbandry practices upon escaperelated behaviours across a range of relevant European farmed species must therefore be experimentally addressed. The PRE-Escape work package (WP) of Prevent Escape aimed to answer these questions by using Atlantic cod as a model species for aquaculture in northern Europe, and seabass and gilthead seabream as model species for aquaculture in southern Europe. We evaluated the potential biological risk factors and mitigation measures for escapes at the species level, and then identified synergies between risk factors and potential interventions for the three species. Our research focused on the biotic and abiotic factors

www.preventescape.eu

fthat could induce escape-related behaviours. The PRE-Escape WP provides new biological knowledge for formulating technological and biological strategies of preventing escape of farmed fish.

Objectives The main objective of the Prevent Escape WP ‘PRE-Escape’ was to assess and determine the biological motivation for the expression of ‘escape-related behaviours’ in a variety of European farmed fish species: Atlantic cod (Gadus morhua), European seabass (Dicentrarchus labrax) and gilthead seabream (Sparus aurata). The work consisted of intensive experimental studies in both laboratory tanks and sea-cage facilities. The specific objectives of the PRE-Escape WP were: • To study and evaluate the biological risk factors for promoting escape-related behaviours in Atlantic cod • To assess and indentify biological risk factors for promoting escape-related behaviours in European seabass • To identify biological risk factors for promoting escape-related behaviours in gilthead seabream • To evaluate generic biological risk factors for escape-related behaviour in all three species and devise mitigative strategies for reducing or preventing behaviours that promote escapes.

References

cited

Andrew JE, Noble C, Kadri S, Jewell H, Huntingford FA (2002) The effect of demand feeding on swimming speed and feeding responses in Atlantic salmon Salmo salar L gilthead seabream Sparus aurata L and European seabass Dicentrarchus labrax L in sea cages. Aquaculture Research 33:501–507 Hansen LA, Dale T, Damsgård B, Uglem I, Aas K, Bjørn P-A (2009) Escape-related behaviour of Atlantic cod, Gadus morhua L., in a simulated farm situation. Aquaculture Research 40:26–34 Huse I (1991) Culturing of cod (Gadus morhua L.). In: Finfish Aquaculture. Ed: McVey JP. Handbook of Mariculture, Vol 2/CRC Press ICES WGEIM (2006) Risk analysis of the potential interbreeding of wild and escaped farmed seabream. Annex 8: 144–155 Jensen Ø, Dempster T, Thorstad EB, Uglem I, Fredheim A (2010) Escapes of fish from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquaculture Environment Interactions 1:71–83 Juell J-E (1995) The behaviour of Atlantic salmon in relation to efficient cage culture. Reviews in Fish Biology and Fisheries 5:320–335 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007) Technological solutions and operational measures to prevent escapes of Atlantic Cod (Gadus morhua) from sea-cages. Aquaculture Research 38:90–99 Norwegian Directorate of Fisheries (2005) Statistics for Aquaculture 2005. www.fiskeridir.no/ fiskeridir/kystsone_og_havbruk/statistikk Papadakis VM, Papadakis IE, Lamprianidou F, Glaropoulos A, Kentouri M (2012) A computervision system and methodology for the analysis of fish behavior. Aquaculture Engineering 46:53–59 Rillahan C, Chambers MD, Howell HW, Watson WH (2010) The behavior of cod (Gadus morhua) in an offshore aquaculture net pen. Aquaculture 310:361–368 Sarà G, Oliveri A, Martino G, Campobello D (2010) Changes in behavioural response of Mediterranean seabass (Dicentrarchus labrax L.) under different feeding distributions. Italian Journal of Animal Science 9:e23

www.preventescape.eu

3.2. Factors affecting escape-related in Atlantic cod (Gadus morhua L)

behaviours

Cite this article as: Noble C, Evensen TH, Jakobsen R, Hedger RD, Uglem I, Zimmermann E, Fleming I, Izquierdo-Gomez D, Høy E, Damsgård B (2013) Factors affecting escape-related behaviours in Atlantic cod (Gadus morhua L). In: PREVENT ESCAPE Project Compendium. Chapter 3.2. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Chris Noble1, Tor H. Evensen1, Ronny Jakobsen1, Richard D. Hedger2, Ingebrigt Uglem2, Emily Zimmermann3, Ian A. Fleming3, David Izquierdo-Gomez4, Erik Høy5 & Børge Damsgård1 Nofima, The Norwegian Institute of Food, Fisheries and Aquaculture Research, Muninbakken 9-13, P.O. Box 6122, NO-9291 Tromsø, Norway 2 Norwegian Institute for Nature Research, Sluppen, NO-7485 Trondheim, Norway 3 Ocean Sciences Centre, Memorial University of Newfoundland, St. John's, Newfoundland, A1C 5S7, Canada 4 Department of Marine Science and Applied Biology, University of Alicante, 03080 Alicante, Spain 5 SINTEF Fisheries and Aquaculture, Brattørkaia 17C, 7010 Trondheim, Norway 1

Introduction The escape of farmed fish is a major challenge to the overall environmental and economic sustainability of the aquaculture industry. Whilst the major causes of escape events in cage aquaculture are associated with structural and operation failures (Jensen et al. 2010), numerous European farmed species including Atlantic cod (Gadus morhua L), gilthead seabream (Sparus aurata L) and European seabass (Dicentrarchus labrax L) can express behaviours that increase the risk of escapes. Major technical or operational failures can lead to large-scale losses, but escapes through small holes have also been frequently reported (Jensen et al. 2010). With regard to Atlantic cod, cage-held fish form unsynchronised shoals (Huse 1991) and display exploratory behaviours that include swimming close to the net wall (Rillahan et al. 2011). They also exhibit a tendency to inspect and bite the wall of aquaculture cages (Moe et al. 2007). Net inspection behaviour may increase the likelihood of fish discovering a hole in a net and escaping (Hansen et al. 2009), whereas net biting behaviour may increase the risk of fish damaging the cage netting materials (Høy et al. 2011).

Atlantic

cod farming in

Norway

Although recent data show that the number of Norwegian cod farms and their production tonnage has decreased since 2010, production showed a steady yearly increase prior to this point, doubling from 10,370 to 20,620 tonnes per year between 2007 and 2010 (Norwegian Directorate of Fisheries 2012a). This increase in production tonnage did not equate to an increase in the number of escaped farmed cod. In fact, although the absolute number of

escapes from Norwegian cod aquaculture facilities peaked in 2008, it has decreased since then, culminating in a rapid fall in the total number of cod escapes in 2011 to 650 fish (see Figure 3.2.1, based upon data from Norwegian Directorate of Fisheries 2012b). Figure 3.2.1. Number of Atlantic cod (Gadus morhua) escapes from Norwegian aquaculture facilities from 2007 â&#x20AC;&#x201C; 2011. Data from the Norwegian Directorate of Fisheries, 2012b.

This reduction in the number of escapes may be related to fewer fish being farmed during the same corresponding period (a reduction from 2.5 million to 1.1 million from 2008 - 2010). Percentage data on cod escapes between 2008 - 2010 supports this, as escapes were a relatively constant 1.1 - 1.5% of the total farmed stock (see Figure 3.2.2, data calculated from Norwegian Directorate of Fisheries 2012a). As of March 2012, there is currently no data available from the Norwegian Directorate of Fisheries on the percentage of farmed stock that escaped during 2011, but as only 650 individuals have been reported as escaping cod farming facilities, the percentage of current cod escapes may be markedly less than 2010 figures. Even with this marked reduction in escapes, cod farmers and regulatory bodies are still diligently striving to realise their target of zero escapes from Norwegian aquaculture (see for example, the Norwegian Aquaculture Escapes Commission http://www.rommingskommisjonen.no/). The Prevent Escape project aims to provide these stakeholders and industry regulators with robust, quantified data on escape risks and mitigation strategies to help them achieve this goal.

Figure 3.2.2. The percentage of escapes in relation to the number of Atlantic cod (Gadus morhua) farmed in Norwegian seacages from 2007 â&#x20AC;&#x201C; 2010. Calculated using data from the Norwegian Directorate of Fisheries, 2012b.

www.preventescape.eu

Objectives The factors that encourage the expression of escape-related behaviours in Atlantic cod are poorly understood. Moe et al. (2007) and Hansen et al. (2009) suggested that hunger may increase the incidence of net biting and inspection behaviour, but the effects of other farm management factors on the manifestation of these behaviours is lacking or unpublished. The main objectives of our experiments were to assess the key husbandry factors that can affect the incidence of escape-related behaviour in Atlantic cod and increase the risk of escape. The work consisted of intensive experimental studies in both laboratory tanks and sea-cage facilities and was broken down into three sections: • Tank experiments that focused on two key risk factors that assessed whether a) the presence/absence of a small hole in a net, and b) different types of net repairs affect escape-related behaviour in relation to feeding motivation. The severity and location of net damage was also examined at the end of the study. • Sea cage experiments that investigated the effect of the following four risk factors: a) different types of net repair, b) short-term feed restriction, c) algal biofilms and macrophytic biofouling, and d) short-term changes in stocking density upon the expression of escape-related behaviour. • A final sea cage experiment that evaluated a possible mitigation strategy of providing cageheld cod with enrichment materials to reduce their net wall inspection and biting behaviour.

Methods – Section 1 The behavioural motivation of individual cod to inspect and bite cage netting materials was examined in four replicate multi-patch experiments carried out at the Aquaculture Research Station (Tromsø, northern Norway). Individual cod (a total of 160 fish, weight 560 ± 171 g., mean ± SD) were used to test how cod interact with either a hole in the net or holes that were mended with different repair techniques. These interactions were evaluated in both the presence and absence of feed outside the net panel.

The experiment was conducted in a large indoor observation tank (diameter 5.0 m, depth 1.2 m) that was supplied with unfiltered seawater at ambient water temperatures for approximately 2 weeks. A hexagonal tank wall was constructed within the experimental tank. This tank wall presented the cod with six separate 50 x 50 cm net panels constructed of white, square meshed (measuring 20 mm by 20 mm bar length), untreated nylon netting commonly used in salmon and cod farming. These nets consisted of either a) a pair of plain control nets with no damage or net repairs, or b) two pairs of treatment panels that had four holes (ca. 10 cm in length), three of which were mended with different repair techniques including short and loose thread ends and contrasting colour repair threads (see Figure 3.2.3). To test the effect of feeding motivation a feed source (a net bag with fish pellets) was placed outside two of the net panels for one hour. All panels were filmed using six underwater cameras that recorded the frequency of inspection and biting behaviour for three hours a day from 09.00 - 12.00. Inspection behaviour was defined as a distinct and directional interest towards the net. Biting behaviour was defined as a clear bite or nibble on the net. The location of these behaviours in relation to the net panel, net hole or specific net repair was also recorded. Fish were not fed for the duration of the two week study.

Figure 3.2.3. Examples of the how Atlantic cod (Gadus morhua) biting behaviour can damage repairrs to the net panel a) knots nearly undone with a severely frayed loose thread, b) one knot undone but with little visible damage to the loose black thread, c) open hole with minor fraying of net twine around the hole, and d) all knots undone and the repair thread completely absent. Net panel twines also have frayed ends. Photo: SINTEF

Each net panel was visually examined for damage at the end of the experiment according to the method described by Moe et al. (2009). Damage to both the netting material and the repair thread were assessed using the following categories: i) no visible wear; ii) partly fluffed ends and abrasion; and iii) severely fluffed ends and abrasion. Any knot damage was also assessed using the categories i) no visible damage, ii) one knot undone, or iii) two knots undone. A total score was then calculated for each repair.

www.preventescape.eu

Results â&#x20AC;&#x201C; Section 1 To a large extent the cod focused their inspection and biting behaviour on loose thread repairs and the net hole, rather than on the net panels themselves. This meant that the type of net repair significantly affected both inspection behaviour and biting behaviour, and fish primarily directed these behaviours at nets repaired with black thread (see Figure 3.2.4). The frequency of inspection and biting behaviours were significantly affected by the presence of a hole and motivation to access feed outside the hole; they increased if there was a feed source present outside the net wall (see Figure 3.2.5). In summary, 50.5% (n = 3355) of the total number of inspections and 83.6% (n = 459) of the total number of bites occurred in the treatment where the net panels had a hole, various net repairs and feed present outside the net wall. In the treatment where the net panels had a hole, net repairs and no feed present, this percentage dropped to 28.9% and 13.7% for inspections and bites, respectively. The control treatment only attracted 20.6% and 2.7% of the total number of inspections and bites, respectively. These differences were significant, with the number of inspections and bites being significantly greater in the Hole + Feed treatment > Hole treatment > Control.

Figure 3.2.4. The frequency of a) inspection and b) biting behaviours directed at different types of net repair in tank-held Atlantic cod (Gadus morhua).

Interestingly, in the feeding motivation treatment, 62.5% of the inspections and 86.8% of the bites occurred during the presence of the feed, in comparison to 12.5 and 3.1% before, and 25.0 and 10.1% after removal of the feedbag, respectively. Fish were capable of causing severe net damage, which in turn was significantly affected by different types of net repair. Control panels and the area around unrepaired holes exhibited little damage at the end of the experiment. Repairs made with white thread and loose ends were more damaged than all the other repairs and the open hole. Repairs made with the black thread and loose ends were also significantly more damaged than the net hole and the white thread repair that had no loose ends (perfect repair). Holes mended with a perfect repair had little or no visible damage.

Figure 3.2.5. The frequency of a) inspection and b) biting behaviours around a meal in tank-held Atlantic cod (Gadus morhua). Different treatments were either i) an undamaged net panel, ii) a panel with repairs and a small hole, or iii) a panel with repairs, a small hole and a feeding stimulus on the other side of the net.

www.preventescape.eu

Methods – Sections 2 & 3 The motivation of groups of cage-held cod to inspect and bite cage netting materials was examined in four further experiments carried out at the Aquaculture Research Station sea farm (Tromsø, northern Norway). Experiments were conducted between October and March, where and fish were held in 5.3 m x 5.3 m x 6 m square cages, and exposed to ambient water temperatures and light levels. Each cage was fed via an automated feeding system, which delivered a fixed ration between 11.00 - 12.00 daily. • To see if the type of net repair affected fish interactions with the net wall, each cage (six cages in total) had a single 50 cm x 50 cm net panel constructed of white, untreated nylon square mesh (measuring 20 mm by 20 mm bar length and of the type commonly used in salmon and cod farming) attached to the existing net wall at a depth of 1 m. These net panels had two holes (ca. 10cm in length) that were repaired with either a black or white thread that had loose thread ends (ca. 3 cm in length). Each panel was filmed for 2 hours daily between 10.30 and 12.30 using underwater video cameras connected to a mobile video recording system. From this footage, the frequency of inspection and biting behaviour for 10 mins before, during and after a meal were investigated. Inspection behaviour was defined as a distinct and directional body orientation towards the net. Biting behaviour was defined as a clear bite or nibble on the net. The location of these behaviours in relation to the net panel, specific net repair or net frame (where the panel was attached to the existing net) was also recorded. • To test the effect of short-term feed restriction upon the frequency of interactions with the net, six cages of cod (ca. 700 g average weight, 550 fish per cage) were split into two treatments for a total of 42 days. The first three cages were fed a 100% ration according to commercial feed tables, whilst a further three treatment cages were fed a 100% ration for the first 14 days, 50% daily rations for the next 14 days, and 100% for the final 14 days of the study. • To test the effect of net cleanliness and biofouling upon the frequency of interactions with the net, two cages of cod (ca. 1040 g average weight, 600 fish per cage) were subject to a 10 day period with “clean” net panels that had been submerged in seawater for 30 days during winter, but exhibited no macro-biofouling and only 30 days exposure to algal

biofilm. After 10 days, these net panels were replaced with “biofouled” panels of the same size and type, but which had been submerged in seawater for 140 days and exhibited extensive fouling with algal biofilm. In addition, a 20 cm x 20 cm section of this panel was artificially fouled with a mix of two different live seaweeds: knotted wrack (Ascophyllum nodosum L.) and kelp (Laminaria spp.). These seaweeds are found on and around the sea farm, and can often break off and foul the cage nets. The daily frequency of net inspection and biting behaviour around meal times was studied before and after the change in the level of biofouling, for a total of 20 days. • To investigate the effect of short-term changes in stocking density upon the frequency of interactions with the net, an additional cage of cod was stocked with 600 fish (ca. 1030 g average weight). After a 10 day period, an additional 1800 fish were transferred to the cage, increasing stocking density from a low (3.7 kg m-3) to a medium (14.7 kg m-3) density (current stocking density thresholds are 25 kg m-3 for Atlantic cod in Norway). The daily frequency of net inspection and biting behaviour around meal times was studied before and after this change in stocking density for a total of 20 days. • To investigate whether environmental enrichment could be used as a mitigation strategy to reduce the frequency of cod interactions with the net wall, six cages were stocked with 600 cod (ca. 960 g. average weight) for 1.5 months. The first three cages received no environmental enrichment for the duration of the experiment and acted as control cages. A further three treatment cages were subject to a) 10 days without environmental enrichment, b) 30 days of enrichment, followed by c) no enrichment for the remainder of the study. Enrichment material consisted of one piece of commercial artificial kelp, ca. 3.5 m in length whose strands (ca. 50 cm in length) are manufactured from reinforced PE fabric. A section of this seaweed measuring 2.5 m was submerged 1 m from the net wall opposite the net panels. Additional enrichment was provided by a manufactured PE tube rack (32 tubes in an 8 x 4 tube rack, tube diameter 10 cm and tube length 50 cm), that was also submerged 1m from the net wall opposite the net panels (see Figure 3.2.6). The daily frequency of net inspection and biting behaviour around meal times in relation to whether fish were subject to environmental enrichment was studied in addition to any interaction with cage enrichment materials.

www.preventescape.eu

Figure 3.2.6. Photograph showing how environmental enrichment materials (an 8 x 4 tube rack or artificial kelp) were deployed in seacage containing Atlantic cod (Gadus morhua). Picture also shows cod inspecting the tube rack. Photo: Nofima

Results – Sections 2 & 3 • With regard to the components of the net panels that attracted the attention of groupheld cod, results from the sea cage trials somewhat differed from the findings of the tank studies. For example, in the control cages of the feed restriction experiment, inspection behaviour was primarily directed at the net panel > net frame > net repair (see Figure 3.2.7). In the sea cage studies there was no clear relationship between the components of the net panel and the frequency of biting behaviour. • Data from the short-term feed restriction experiment suggests that a 2-week period of 50% reduced rations during autumn/winter had little effect upon net panel inspection behaviour, aside from inspection behaviour remaining relatively stable during feeding when 100% rations were restored, compared to a decrease in inspection behaviour in corresponding control cages during the same period (see Figure 3.2.8). In fact, the only relationship that short-term feed restriction had upon the expression of escaperelated behaviour was an increase in the frequency of net biting during feeding, but this biting frequency decreased when 100% rations were restored.

The frequency of net inspection and biting behaviour decreased during a meal regardless of treatment. In addition, there was a long-term decline in net biting and inspection frequency as winter approached, irrespective of whether fish were subject to reduced rations or not. • There was a clear relationship between biofouling and the number of interactions with the net. The frequency of inspection and biting behaviours increased when fish were exposed to a dirty net (see Figure 3.2.9). This increase was specifically directed at biofouling with seaweed, and long-term biofouling with algal biofilms appears to have had little or no effect upon net inspection frequency (see Figure 3.2.10) which also applied to net biting behaviour. The frequency of net inspection and biting behaviour also decreased during a meal. • There was no clear relationship between stocking density and the expression of net inspection and biting behaviours. Although the total number of net inspections increased when stocking density was quadrupled, this appeared to be primarily a function of the increase in the number of fish, rather than stocking density per se. When data were adjusted to account for this increase in fish number, no clear relationship between stocking density and cod inspection behaviour could be found, aside from after a meal when the frequency of inspections was marginally higher in fish held at low densities (see Figure 3.2.11). • Environmental enrichment had no clear effect upon the frequency of net inspection or biting behaviour in sea cages and there was high variability between cage replicates within each treatment. Exposure to enrichment materials did not significantly reduce the number of net wall inspections in relation to control cages, (see Figure 3.2.12). When cod did interact with the enrichment materials, their interactions were primarily directed at the artificial seaweed, rather than the tubes that were provided to offer potential shelter for the cod (see Figure 3.2.13). No cod were observed either swimming through the tubes or resting in them.

www.preventescape.eu

Figure 3.2.7. The frequency of a) inspection and b) biting behaviours directed at different types of net repair in cage-held Atlantic cod (Gadus morhua) fed full satiation rations during autumn and winter (data represents mean + SD).

Figure 3.2.8. The frequency of a) inspection and b) biting behaviours around a meal in cage-held Atlantic cod (Gadus morhua) subjected to a shortterm period of feed restriction. Different coloured lines represent the different 2 week periods prior-, during and postfeed restriction. Dashed lines show data from corresponding time periods in the control cages that received 100% rations for the duration of the study. SD bars are omitted to aid clarity.

Figure 3.2.9. The frequency of a) inspection and b) biting behaviours around a meal in cage-held Atlantic cod (Gadus morhua) exposed to clean or dirty nets (mean Âą SEM).

Figure 3.2.10. The frequency of inspection behaviours (mean + SD) directed at different types of net repair and biofouling materials with a) clean nets and b) dirty nets in cage-held Atlantic cod (Gadus morhua). NB there was no seaweed on the nets during part a).

www.preventescape.eu

Figure 3.2.11. The frequency of inspection and biting behaviours around a meal in cage-held Atlantic cod (Gadus morhua) in relation to short-term changes in stocking density. SD bars are omitted to aid clarity.

Figure 3.2.12. The frequency of net inspection behaviours around a meal in cage-held Atlantic cod (Gadus morhua) in relation to whether the cages had environmental enrichment. Different coloured lines represent the different periods pre-, during and post- enrichment. Dashed lines show data from corresponding time periods in the control cages that received no enrichment for the duration of the study. SD bars are omitted to aid clarity.

Figure 3.2.13. The frequency of a) inspections and b) bites with different types of environmental enrichment in cage-held Atlantic cod (Gadus morhua) exposed to either artificial kelp or a tube rack (mean Âą SEM).

www.preventescape.eu

Discussion We have demonstrated that a number of common aquacultural factors can increase the expression of escape-related behaviours in Atlantic cod and have also demonstrated the efficacy of measures to reduce and mitigate these risks. In controlled studies carried out in tanks, feeding motivation emerged as a strong risk factor for inspecting and biting a net when fish were deprived of feed (Damsg책rd et al. in press), and this data supports previous studies that suggested complete feed deprivation increases net inspection and biting in cod (Moe et al. 2007, Hansen et al. 2009). These behaviours were primarily directed at poorly repaired areas of the net wall, specifically where a contrasting coloured thread was used to repair net damage. In cages, the overall frequency of net inspection and biting was relatively low, and decreased further as winter approached. Results from the sea cage trials differed somewhat from the tank studies in relation to what aspects of the net attracted cod interactions. Specifically, when fish were fed a full satiation ration during autumn/winter in sea cages, their inspection behaviour was primarily directed at the net panel and net frame, rather than net repairs. There was no clear relationship between the components of the net panel and the frequency of biting behaviour. The majority of the sea cage studies showed that the number of inspections and bites decreased during feeding which suggests that the presence of feed in the water column distracts the fish from browsing the net wall. These results were markedly different from the earlier tank studies, where the presence of feed outside the net wall affected the frequency of interactions with the net. In the sea cage studies, fish could access feed when it was presented to them, thus reducing the frequency of their interactions with the net. Appetite varies within and between days for many farmed fish species (see for example, Noble et al. 2008). However, numerous farmers feed their fish a fixed daily ration according to commercial feed tables that do not account for short-term changes in appetite levels, meaning there is a risk of farmers under- or over-feeding their fish. Our experiments show that a short-term period of feed restriction (not complete deprivation of feed) does not emerge as a robust risk factor for stimulating the expression of escape-related behaviour in Atlantic cod during autumn and winter. The only risk is an increase in the already low frequency of biting behaviour during meals under the feed-restricted regime. Biofouling in the form of algal biofilms does not emerge as a risk factor for net inspection or net biting in cage-held Atlantic cod during spring. However, macro biofouling with seaweeds does increase the frequency of net biting and inspection behaviour, and farmers should attempt to reduce this if high volumes of seaweeds get snagged on the net wall. There is no clear relationship between short-term changes in stocking density and the frequency of net inspection behaviours. This means that when farmers transfer fish to holding cages during transport and grading procedures, the increase in density per se does not increase the expression of escape-related behaviours. Providing environmental enrichment with artificial seaweed and tubing materials for potential shelter was not a robust mitigation strategy for reducing the expression of net inspection behaviours in sea cages during winter/spring. In fact, there was no significant effect of

enrichment upon either inspection or biting behaviour, although there was high variability between cage replicates in each treatment. When cod did interact with enrichment materials they did so in different ways: the majority of interactions with the enrichment materials were directed at the artificial kelp rather than the tube rack and no cod were observed swimming through or resting in the tubes.

Recommendations • Complete feed deprivation increases the risk of cod inspecting and biting the net panel. Farmers should follow their existing strategies of limiting the time they starve fish during common husbandry practices e.g. during size grading or fish transports. • Short-term feed restriction during winter does not appear to be a robust risk factor for stimulating the expression of inspection behaviours in cage-held Atlantic cod, but does increase net biting frequency during feeding. This means that even if farmers miscalculate the daily feed requirements of their fish for short periods, it may not be as severe as first thought. • Short-term changes in stocking density, such as when fish are transferred to a central holding cage, do not appear to affect net inspection frequency before or during feeding. • Biofouling with long stranded seaweeds does increase the risk of cod biting and inspecting the net. Farmers should attempt to remove these fouling objects as and when they build up. • Environmental enrichment by suspending artificial kelp and tubes within a cage during winter/spring is not a robust mitigation strategy for reducing the expression of inspection and biting behaviours at the net wall.

www.preventescape.eu

References

cited

Damsgård B, Høy E, Uglem I, Hedger RD, Izquierdo-Gomez D, Bjørn PA (in press) Net-biting and escape behaviour in farmed Atlantic cod, Gadus morhua: Effects of feed motivation and net traits. Aquaculture Environment Interactions Hansen LA, Dale T, Damsgård B, Uglem I, Aas K, Bjørn PA (2009) Escape-related behaviour of Atlantic cod, Gadus morhua L., in a simulated farm situation. Aquaculture Research 40:26–34 Huse I (1991) Culturing of cod (Gadus morhua L.). In: Finfish Aquaculture. Ed: McVey JP. Handbook of Mariculture, Vol 2/CRC Press Høy E, Volent Z, Moe-Føre H, Dempster T (2011) Loads applied to aquaculture nets by the biting behaviour of Atlantic cod (Gadus morhua). Aquacultural Engineering 47:60-63 Jensen Ø, Dempster T, Thorstad EB, Uglem I, Fredheim A (2010) Escapes of fish from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquaculture Environment Interactions 1:71–83 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007) Technological solutions and operational measures to prevent escapes of Atlantic Cod (Gadus morhua) from sea-cages. Aquaculture Research 38:90–99 Moe H, Gaarder RH, Olsen A, Hopperstad SO (2009). Resistance of aquaculture net cage materials to biting by Atlantic cod (Gadus morhua). Aquacultural Engineering 40:126–134 Noble C, Kadri S, Mitchell DF, Huntingford FA (2008). Growth, production and fin damage in cage-held 0+ Atlantic salmon pre-smolts (Salmo salar L.) fed either a) on-demand, or b) to a fixed satiation-restriction regime: data from a commercial farm. Aquaculture 275:163–168 Norwegian Directorate of Fisheries (2012a) Escape statistics of farmed cod. http://www. fiskeridir.no/english/statistics/norwegian-aquaculture/aquaculture-statistics/cod Norwegian Directorate of Fisheries (2012a) Escape statistics of farmed cod. http://www. fiskeridir.no/statistikk/akvakultur/oppdaterte-roemmingstall Rillahan C, Chambers MD, Howell HW, Watson WH (2010) The behavior of cod (Gadus morhua) in an offshore aquaculture net pen. Aquaculture 310:361–368

3.3. The escape-related behaviour seabass (Dicentrarchus labrax) seabream (Sparus aurata)

of and

European gilthead

Cite this article as: Papadakis IE, Papadakis VM, Glaropoulos A, Kentouri M (2013) The escape-related behaviour of European seabass (Dicentrarchus labrax) and gilthead seabream (Sparus aurata). In: PREVENT ESCAPE Project Compendium. Chapter 3.3. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Ioannis E. Papadakis, Vassilis M. Papadakis, Alexios Glaropoulos, Maroudio Kentouri* * University of Crete, Greece

Introduction Seabass (Dicentrarchus labrax) and seabream (Sparus aurata) aquaculture is a major industry in the Mediterranean (130 000 and 170 000 tonnes, respectively; FAO 2010), and involves growing juvenile fish to harvest-sized adults in sea cages. Fish escapes from sea cage facilities can be a serious problem, and threatens the sustainability of the aquaculture industry and viability of wild fish populations. The main causes of escapes are severe environmental conditions, such as strong winds and storms, and fish predators attacking the farmed population (Jensen et al. 2010). Fish farmers also regularly report escape incidents of the main farmed species (seabream and seabass) through holes in the net of sea cages, and fish escape has been associated with species-specific interactions with sea cage nets (Jensen et al. 2010, Papadakis et al. 2012). Such fish interactions include general net inspection behaviour, such as overt net biting that can lead to structural weakness and holes in the net, and exploratory or risk taking behaviour exhibited when fish pass though holes in the net. An understanding of the conditions that increase the propensity for aquaculture fish species to swim through holes in sea cage nets is of key importance for helping both researchers and farmers develop mitigation measures to reduce escape risk.

www.preventescape.eu

This study aimed to assess the main husbandry factors that could increase the propensity of both seabass and seabream to escape from sea cages. Seabass studies focused on the factors that may increase exploratory or risk taking behaviour, and could increase the risk of fish swimming through holes in the net (such as food deprivation, light conditions and visual acuity). Seabream studies assessed the factors that can cause fish to bite the net wall, in addition to those factors that can cause fish to swim through holes in the net. Hence, the effects of fish density, prolonged periods of food deprivation and also the presence of biofouling on the net surface were assessed. The results of these studies are then used to provide recommendations and key operational strategies for reducing the prevalence of escape-related fish behaviour.

Materials

and methods

â&#x20AC;&#x201C;

seabass

The first set of experiments were performed in nine tanks (100 L) under controlled experimental conditions. A total number of 135 seabass individuals, of approximately the same weight (30.93 Âą 1.88 g) and length (13.61 Âą 0.44 cm) were used to test the effect of different feeding conditions (well-fed fish, fish fed reduced rations and starved fish) on their escape behaviour over 20 days. Feeding conditions were tested in triplicates for statistical purposes. A plastic frame with a net panel was placed into the tanks, separating the tank into two areas. Fish were confined to the left side of the tank (holding area). The net panel contained a hole (5 cm) that the fish could swim through (to the other side of the tank) to simulate an aquacultural escape event. A single camera was mounted in front of each tank to observe fish behaviour. Escape behaviour was quantified by recording the number of incidents that a fish swam through the hole in the net panel to the other side of the tank. Video acquisition used an advanced computer vision system (Papadakis et al. 2012), that was able to record video footage during the day and night for the entire experiment.

The second set of experiments were performed in a 30m3 volume tank. Six handmade sea cages (length 80 cm, depth 60 cm and width 60 cm) were constructed and placed in the tank. A common white-color aquaculture net was used to construct the sea cage net pen. A hole was created in one side of the cage net, allowing fish to escape. 35 seabass were confined in each of the sea cages (6 x 35). Two cage groups were created and each group differed according to the position of the hole. These groups consisted of i) an ‘internal’ group, where the hole faced the Figureof 3.3.1. up of the second of sea bassii) experiments. front wall theExperimental tank (at set a distance <60setcm) and an ‘external’ group, where the hole faced The clear position of the ''internal'' and ''external'' sea cage group and the the open view area of the tank (3.5 m from the farthest wall of the tank, see Figure 3.3.1). cameras is demonstrated. Additional information also refers to the distances between the cage and the wall as well as from the bottom of the tank.

Figure 3.3.1. Experimental set up of the second set of seabass experiments. The clear position of the ''internal'' and ''external'' sea cage group and the cameras is demonstrated. Additional information also refers to the distances between the cage and the wall as well as from the bottom of the tank.

The second set of experiments consisted of two successive experiments that were performed with two different sub groups of the initial large population. • In the first experiment (henceforth referred to Experiment A), the hole of the internal cages faced the proximate black wall of the tank, while the hole in the external cages faced the open area of the tank. • In the second experiment (henceforth referred to Experiment B), the same experimental setup was repeated but an extra net pen was placed in front of the hole on the external cages (at a distance <60 cm), to act as a visual obstacle. Experiment A and B were compared to define how the presence of an obstacle in front of the external cages influenced the number of escapes of D. labrax. The number of fish that remained in the external cages was divided by the number of fish remaining in internal cages for each day during each experiment. This provided a standardized proportion, or ratio, of fish escapes (i.e. the number fish remaining in external cages compared to internal cages) each day. We then compared this between Experiments A and B.

www.preventescape.eu

To observe fish behaviour, one camera was mounted above each cage. Escape behaviours referred to the number of incidents where a fish swam through the hole and exited the cages. Video acquisition was performed through a multi-camera frame grabber (GV-1120, Geovision), able to record video simultaneously from all cameras.

Results â&#x20AC;&#x201C;

seabass

In the first set of experiments (tanks), seabass individuals showed no interactions with the aquaculture net (either via inspection behaviour or biting the net). Escapes were observed immediately (<5 min) after a hole was made in the net pen. However, the chosen fish density (15 fish/tank) and the amount of food supplied had no influence on the escape activity of this species (Figure 3.3.2). Escapes occurred in a linear fashion, with a number of fish following the first escapee in a direct line out of the hole. In the second set of experiments (small sea cages) the proximity of neighboring cages or the tank wall in front of the point of escape clearly effected the frequency of escapes. After fish were confined to the cages they showed low activity, but became more active as time progressed (<10 min) inspecting the surrounding area, locating the hole and then escaping (<15 min). Escapes also occurred in a linear fashion, with a number of fish following the first escapee in a direct line out of the hole. The presence of the wall of the tank (solid obstacle) had an influence on the escape behavior of D. labrax. A significantly higher number of fish escaped from externally facing pens than from internally facing pens (p < 0.05, Figure 3.3.3). However, the installation of an extra net pen in front of the point of escape reduced the escape rate from the external cages, and the number of fish that escaped did not differ between the internal and the external cages after a net pen was installed (p > 0.05, Figure 3.3.4). Comparisons of the ratio of individuals remaining in external/internal groups between Experiment A and Experiment B further demonstrate that the presence of an obstacle in front of the point of escape significantly reduced the number of escapes. The ratio of the fish remaining in external cages compared to internal cages were significantly lower in Exp A. than in Exp B. Significant differences were observed in all experimental days (Figure 3.3.5). In particular, on day 2 the ratio in Exp. A was 0.081, while in Exp. B it was 0.849. Additionally, on day 5 the ratio in Exp. A. was 0.061 while in Exp. B it was 0.873. Lastly, by the end of the experimental period, in Exp. B. the ratio of the external/internal groups was measured to be 0.319, while in Exp. A it was zero, since no fish remained in the external cages.

Number of escapes

Figure 3.3.2. Changes in fish escape (upper) and return (lower) incidents of seabass between days (mean value ± SE).

Number of escapes

Time (days)

Number of fish remaining in cages

Figure 3.3.3. Changes in the mean (± SE) number of fish remaining in internal and external cages over a 17 day period in Experiment A. Statistically significant differences between the two-cage groups are indicated with an asterisk (*).

Time (days)

Number of fish remaining in cages

Figure 3.3.4. Changes in the mean number (± SE) of fish remaining in internal and external cages over a 17 day period in Experiment B. External cages had a net pen placed in front of the escape hole in this experiment. Statistically significant differences between the two-cage groups are indicated with an asterisk (*).

Time (days)

www.preventescape.eu

Figure 3.3.5. The proportion of individual seabream remaining in the external cages relative to the number of fish remaining in internals cages in Experiment A and B over a 17 day experimental period.

Discussion –

seabass

The morphology of seabass (Volckaert et al. 2008) and its swimming activity (Pickett and Pawso 1994) may lead to increased exploratory behaviour and lead it to swim through holes in the aquaculture net. Similar behaviour has also been observed in other experiments, where intense activity was an indicator of a fish’s willingness to escape from a hole that appeared in fishing gear (Brown 2001). In the seabass studies, the lack of significant responses may be related to the feed restriction period being relatively short (20 days). It could also be related to the low volume of each net enclosure (100 L). Thus, instead of interacting with the net (inspecting it or biting it), seabass may shoal and then rapidly escape when given the opportunity. In the experiments presented here, the behavior of seabass individuals changed when fish discovered the hole in the experimental net pen. Instead of remaining within the cage, fish increased their escape behavior, possibly as an attempt to exploit more space or get into a more familiar environment, such as the open volume of the tank. The fish groups used in the experiment were previously raised in plastic tanks and then confined in handmade sea cages. This intense escape activity could be due to the attraction of the seabass to a familiar and previously experienced open tank environment. The net wall of the cages was the first visible obstacle that fish inspected during the early stages of each experiment. The wall of the tank simply looked like a solid object blocking their path of escape, whereas in the external cages, there was a small area of open water beyond the net hole. In addition, in internal cage treatments, fish could see the wall of the tank through the mesh and the hole in the net, which may have been a familiar object from their prior life history. Thus, their familiarity with the view of the wall of the tank may have resulted in a lower number of fish escaping from the internal cages compared to external cages. Several studies also indicate that the previous experience has a significant influence on the current behavior of fish (Coves et al. 2006, Rubio et al. 2003).

Materials

and methods

–

seabream

Experiments were performed in nine tanks (100 L) under controlled experimental conditions. A total number of 405 seabream individuals, of approximately the same weight (30.93±1.88 g) and length (13.61 ± 0.44 cm) were used in three different series of experiments. Three sub groups (3 x 135) from an initial large group were selected and used in the three experiments. We firstly tested three different fish densities, according to the size of fish. Three groups of seabream were confined in the tanks, in group sizes of 10, 15 and 20 individuals, respectively. Fish densities were tested in triplicates for statistical purposes. A net wall divided the tank into two areas – a holding and an escape area. Fish were confined into the holding area while a hole in the net wall allowed fish to swim through to the escape area.

www.preventescape.eu

Fish inspection rates per individual refers to the number of frames that one fish spent in proximity the net panel. Fish biting rates per individual refers to the number of bite incidents that occurred on the net panel. The number of crossings per individual refers to the number of incidents where one fish swam through the hole in the net panel to the escape area. In the second experiment, fish escape behaviour was examined under different feeding conditions, including daily satiation that satisfies fish energy requirements (2% of the total initial body weight), prolonged food deprivation (1% of the total initial body weight) and starvation (0% total initial body weight). The rest of experimental design and the assessment of behavioural traits were identical to the experiment on stocking density outlined above. In the third experiment, fish interactions (inspection or biting) with the aquaculture net were examined in relation to a thin film of biofouling on the net surface. In addition, we tested whether the colour and/or the presence of biofouling was the major factor that attracted fish individuals. In accordance with the experimental design described previously, fish inspection and biting behaviour was measured and compared between a biofouled net pen, a colored net pen and a normal white net pen. In this experiment, no hole was made on the net pen as we did not observe any escape activity. The observation of fish behaviour (fish inspection and biting the net, escape incidents) was done with nine colour digital CCD cameras (Fire-i, Unibrain), with each camera recording a single tank. Acquisition was performed with an advanced version of previous custom-made software, specially developed for the experiments (Papadakis et al. 2012). The requested frame rate was set to 9 frames per second, allowing for a smooth observation of fish movements. The cameras recorded data continuously from 8:00 until 20:00 daily, for the entire duration of each experiment. Every time that the system encountered a fish close to the net pen (at a distance < 2 cm) it recorded the event and a photo was extracted by the system. Remote, real-time observations of fish behaviour in all tanks was also possible through the use of a web-server (LabView, Web Publishing Tool) installed together with the acquisition software. Thus, the influence of human presence on fish behaviour was minimal.

Results â&#x20AC;&#x201C;

seabream

The escape behaviour of seabream was found to be related to fish density, food availability and the condition of the net panel. A large number of escape incidents were observed during the three experiments. Elevated fish densities caused an increase in the time that fish spent inspecting the net and increased the number of net bites. In addition, fish density was found to be proportional to escape activity (Figure 3.3.6). In particular, the number of escapes/individual/day was 11 times greater in the high density treatment compared to the mid density treatment. It was also 55 times higher activity than the low density treatment. Food supply also had a clear impact upon fish escape behaviour. Fish fed a restricted ration showed the highest propensity to escape when a hole was present on the net wall (Figure 3.3.7). Furthermore, fish fed a restricted ration were more attracted by damage on the net structure in comparison with the starved and satiated fish. Finally, seabream interactions towards the aquaculture net were closely related to both its condition and its colour. In particular, the highest (p < 0.05) amount of fish inspection was observed on the white net wall. In contrast, the lowest rate of inspection occurred in the coloured net pen (Figure 3.3.8a, b). No statistically significant difference was observed on fish biting the net between the three net conditions.

Figure 3.3.6. Mean number (Âą SE) of crossings through a hole in an aquaculture net per individual seabream per day under different stocking densities. Statistically significant differences are indicated by different letters (a,b). The level of significance was set to P < 0.05.

www.preventescape.eu

Figure 3.3.7. Mean number (Âą SE) of seabream crossing through a hole in an aquaculture net over 21 experimental days under different feeding conditions. Statistically significant differences are indicated by an asterisk (*). The level of significance was set to P < 0.05.

Figure 3.3.8a. Mean number (Âą SE) of camera frames that seabream performed exploratory behaviour on an aquaculture net. Statistically significant differences are indicated by an asterisk (*). The level of significance was set to P < 0.05.

Figure 3.3.8b. Mean number (Âą SE) of biting incidents on an aquaculture net panel by seabream in the experiment. Statistically significant differences are indicated by an asterisk (*). The level of significance was set to P<0.05.

Discussion â&#x20AC;&#x201C;

seabream

Food deprivation and high stocking density have been considered as two major stress factors in aquaculture. They can result in size-related effects on fish growth performance and on specific behavioural traits of farmed species. According to previous studies, stocking density can be a stress factor in aquaculture, clearly affecting fish behaviour (Pickering and Pottinger 1989, Papadakis et al. 2012). Furthermore, food deprivation has been associated with increased fish activity and intense prey searching behaviours (Yacoob and Brownman 2007). Little is known about the potential effect of biofouling on the behaviour of fish. It is believed that biofouling filaments may be the reason for increased fish biting the net pen and the eventual creation of holes in the net. Further research is needed to clarify this. Based on our results, we recommend farmers take into account stocking density, feed deprivation and biofouling when assessing escape risks from escape-related fish behaviours, as they were found to be significantly associated with the escape behaviour of seabream. Frequent control and cleaning of the net pens and appropriate feeding schedules are required to minimize the escape risk on sea cage installations.

www.preventescape.eu

Published

link

Papadakis VM, Papadakis IE, Lamprianidou F, Glaropoulos A, Kentouri M (2012) A computervision system and methodology for the analysis of fish behavior. Aquacultural Engineering 46:53-59

References

cited

Brown, C., 2001. Familiarity with the test environment improves escape responses in the crimson spotted rainbow fish, Melanotaenia duboulayi. Anim Cogn. 4, 109-113. Coves D, Beauchaud M, Attia J, Dutto G, Bouchut C, Begout M (2006) Long-term monitoring of individual ﬁsh triggering activity on a self-feeding system: an example using European seabass (Dicentrarchus labrax). Aquaculture 253:385–392 FAO, 2010. Synthesis of the Mediterranean Marine Finfish Aquaculture - A marketing and promotion strategy.vol.88, Rome. Jensen Ø, Dempster T, Thorstad E, Uglem I, Heide MA, Fredheim A (2010) Escapes of fish from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquaculture Environment Interactions 1:71-83 Pickett and Pawso (1994) Seabass Biology, Exploitation and Management. London: Chapman and Hall Pickering AD, Pottinger TG (1989) Stress responses and disease resistance in salmonid fish: effects of chronic elevation of plasma cortisol. Fish physiological biocemistry 7: 253-258 Rubio VC, Sánchez-Vázquez FJ, Madrid JA (2003) Nocturnal feeding reduces seabass (Dicentrarchus labrax, L.) pellet-catching ability. Aquaculture, 22: 697–705 Volckaert FAM, Batargias C, Canario A, Chatziplis D, Chistiakov D, Haley C, Libertini A and Tsigenopoulos C (2008) European seabass. In: Genome Mapping and Genomics in Animals (Ed. by T.D. Kocher & C. Kole), Volume 2: 117–33. Springer-Heidelberg, Berlin. Yacoob SY, Brownman HJ (2007) Prey extracts evoke swimming behaviour in juvenile Atlantic halibut (Hippoglossus hippoglossus). Aquaculture 270: 570-573

3.4. Escape behaviour of gilthead seabream European seabass in commercial sea cages

and

Cite this article as: Smith C (2013) Escape behaviour of gilthead seabream and European seabass in commercial sea cages. In: PREVENT ESCAPE Project Compendium. Chapter 3.4. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Chris Smith1 1

Hellenic Centre for Marine Research, Greece

Introduction European seabass (Dicentrarchus labrax) and gilthead seabream (Sparus aurata) are two of the primary marine aquaculture fish species in the Mediterranean region (Rana 2005). Qualitative evidence from researchers and aquaculturists suggests seabass are not prone to net biting, but they are thought to be much more opportunistic about escaping through any potential net openings; with the majority of escapes coming primarily from technical and operational failures (Dempster et al. 2007). It is quite well known that caged seabream have a net biting behaviour that focuses on visual imperfections, whether it is fouling on the net, attachment areas, repairs, or worn and loose threads (ICES, 2006). The fish will also repeatedly re-visit an area of interest and gradually open up a hole in the net material or weaken the net structure. At some point when the opening is large enough they may start to escape.

www.preventescape.eu

Objectives The aim of this cage farm experimental work was to take some of the important issues from the laboratory experiments on seabass and seabream carried out at the University of Crete and investigate escape-related behaviour on a Mediterranean commercial sea cage farm. With regard to seabass, three major investigations were undertaken, including an inspection of cage nets to look at possible damage from biting behaviour, and response experiments with respect to loose threads on the cage wall and a hole in the cage wall. With regard to seabream, two major investigations were undertaken: an inspection of cage nets to look at possible damage from biting behaviour and response experiments with respect to a hole in the cage wall.

Cage Farm Experiments - Seabass The cage farm experimental work was carried out on a mixed species farm on the north west coast of the island of Crete. To observe evidence of potential biting damage on the inside of cage nets, a Remotely Operated Vehicle (ROV) was used for visual observation in cages. Multiple vertical transects were carried out on the inside of the cages with observation stops at various depths from the surface to the bottom of the net. This was carried out in separate cages of seabass of 170 g and 400 g size. Biting would have been evident from holes, unbraided or loose thread or simply visible white patches from the removal of dark red antifouling coating on the net. However for all the observed transects, there was no evidence of any biting on the inside of the cages for this species. In the second experiment, a divided cage system was used to observe the behaviour of 170 g seabass to loose threads on the net (Figure 3.4.1 and 3.4.2). A divider net was introduced daily to a 12 m x 12 m cage with a camera on the net lined up to observe a loose thread tied to the net (5 cm thread of the same material as the net). The net was deployed down the side of the cage and then gently pulled away from the side of the cage to a position approximately 3 m away, to divide the cage and confine the fish to approximately 60% of the cage (leaving a smaller area of the cage remaining with no fish). Video examination noted the incidents of a number of specific behaviours, including: fish inspecting the thread, contacting (touching the thread with their snout), biting or tearing at the thread. From the first to the second day there was a large increase in the level of interaction with the net. Generally speaking the major part of the seabass shoal kept away from the net with only a few individuals cruising along the net wall. On the first day the number of individuals inspecting the thread was constant but the number of thread contacts decreased as the day progressed. On the second day the pattern was very similar, although there was generally a higher level of interaction. The number of fish inspecting the thread doubled and was again at a constant level through the day. The number of contacts with the thread declined towards midday, whilst contact with the net, apart from the thread, was low and did not change through the day.

The breakdown of contacts was very similar for the two observation periods with approximately 28% of behaviours involving a simple contact with the snout on the thread, 43% involving a small bite with the jaws closing on the thread and 28% a tear, where the individuals grabbed the thread and used the body to turn away and try and tear the thread away from the net.

Figure 3.4.1. Schematic of the split net system.

Figure 3.4.2. Photograph of the split net system.

In conclusion: • Seabass were shown to bite on loose thread on the cage wall • Seabass showed a constant level of interest in the thread, but contact with the thread declined through the morning. • Contact with the thread was primarily through simple biting, but also with more physical tearing at the thread. • There were differences in activities between days.

www.preventescape.eu

In the third and final experiment the same experimental procedure as described above was used to investigate the behaviour of 150 - 200 g seabass to a hole in the net over 3 days. In place of the thread, a hole was made in the net by cutting mesh in a vertical line, 10 cm long and representing roughly 2 body heights of the seabass. Video examination noted the incidents of a number of specific behaviours, including: shoaling distance from the net, proximity of individual fish to the hole, inspection of the hole, contact inspection, exiting the confined area, entering the confined area. Seabass exhibited a relatively tight shoaling behaviour in close proximity to the net, and the shoal seemed to be quite sensitive to disturbance, quickly moving away if disturbed, but just as quickly moving back to the net. A few independent fish were noted along the edge of the cage. From day to day, shoaling patterns showed some differences, with the shoal gradually getting closer to the net towards midday for two of the days, allowing greater interaction with the net. Overall there were higher response levels on the third day, which may be due to the fish acclimatising to the presence of the net. The pattern of proximity and hole inspection was relatively constant but largely increasing on Day 3. This was perhaps reflected by the shoaling behaviour, where the shoal would often turn to point at the cage wall and consequently higher numbers of individuals seemed to be inspecting the hole. Seabass exhibited very little contact with the hole (snout contact). In terms of passing through the hole, activity was high, but unfortunately the number of entrances was much higher than the number of exits/escapes. A significant number of individuals had passed around the divider net (an escape response in itself), although this group was extremely small in size compared to the â&#x20AC;&#x153;confinedâ&#x20AC;? group. The number of exits increased into the afternoon, whilst the number of entrances peaked at midday. From general observations, entering the hole seemed to be involve fish swimming in a straight line from the edge of the hole into the cage (the approach could not be seen), but escapes/exits were more likely to be associated with the fish approaching the hole from a distance, followed by a pause where the seabass inspected at the hole, followed by the fish swimming directly through. Figure 3.4.3 shows some of the seabass and net hole interactions.

Figure 3.4.3. Cage with seabass, shoal away from the cage wall, shoal up at the cage wall with an individual escaping and entering the net.

In conclusion: â&#x20AC;˘ Seabass are willing to escape through a hole in the net, and the incidence of this escape behaviour is both high and direct. â&#x20AC;˘ Exit of fish through the hole seemed to be dependent on proximity of the shoal

General Discussion - Seabass Prior to the start of the experiments, the perceived wisdom on the behaviour of seabass was that they were not known as net biters, but did exhibit a propensity to escape through holes in the net. In terms of net biting, the experiments undertaken in this study had mixed results. Whilst the cage nets used on the farm for seabass showed no indication of damage that could be related to biting, the loose thread experiment showed that there could be a high level of biting with this species. The reason for seabass biting on the thread may be that the thread was an attractant on an otherwise homogenous net surface. The loose thread was quite obvious (5 cm long). The danger of the loose thread was that firstly it attracted biting and secondly tearing that transferred strain away from the immediate bite. Seabass were also seen to inspect and contact the net independently of the thread. It is thought that for seabass, interaction with the net was highly dependent on shoaling behaviour. When the fish densely shoaled away from the net there was little chance of interaction with the net. If the shoal was in close vicinity to the net, then the number of interactions increased. Movement of the shoal was mostly parallel to the cage walls especially when the shoal was away form the cage wall, but was also directly toward and away from the net in response, for example, to the presence of person on the walkway (movement away from the person, returning to the net when the person moved away). Seabass seemed to be very reactive to external stimuli, with the shoal often seen to suddenly change direction. The responses of seabass to the thread and hole were different. They would make contact with the thread, but would only pass cleanly through a net and were not observed to come into contact with the periphery of the hole (either inspecting or biting the edge of the hole). Passing through the hole was a straight swimming action with an approach from a distance.

www.preventescape.eu

Cage Farm Experiments - Seabream The cage farm experimental work was carried out on a mixed species farm on the northwest coast of the island of Crete. To observe evidence of biting on the inside of cage nets, a Remotely Operated Vehicle (ROV) was used for visual observation in cages. Multiple vertical transects were carried out on the inside of the cages with observation stops at various depths from the surface to the bottom of the net. This was carried out in separate cages of fish of 200 g and 400 g size. Biting would have been evident from holes, unbraided or loose thread or simply visible white patches from the removal of dark red antifouling coating on the net. For the smaller sized seabream there was no evidence of biting, but for the large seabream biting evidence was seen on the inside of the cages. Biting was easily evident through removal of the dark red antifouling coating, leaving exposed white patches and areas. There were three types of bite interactions: a) On the inside side wall of the net in contrast to normal net sections (Figure 3.4.4) there were small patches of white net consisting of a few strands to small patches of several centimetres diameter (Figure 3.4.5) â&#x20AC;&#x201C; it was assumed that these cleaned areas were caused by large bream biting the netting material, leading to a removal of antifouling agent. This removal of red antifouling may have provided the seabream with a colour contrast stimulus that could increase biting on that particular area. b) There were some previously repaired tears in the net, and the tears had been repaired by drawing together the edges from the outside of the net then using tie-wraps (cable-ties) to close the opening, i.e. the seam was on the outside of the net (Figure 3.4.6). Where some threads or loose ends were visible on the inside these ends showed white frayed thread or small white patches where potential biting had removed the antifouling (Figure 3.4.7). c) Toward the bottom of the net there were several cases of small raised white patches (one to two mesh diameters in size), that appeared to be indentations on the inside of the net. It was assumed that these were caused by hooking the net onto the cage stanchions during maintenance and general husbandry procedures. The weight of the net on this small focal point may have led to stretching that produced this â&#x20AC;&#x153;plugâ&#x20AC;?. When the net was dropped back into the water, the stretch plug was retained and may have attracted seabream biting/inspection behaviour as evident by the removal of antifouling and viewed as raised white areas (Figure 3.4.8a,b), which could be quite dense in number (Figure 3.4.9a,b).

Figure 3.4.4. Normal net

Figure 3.4.5. Cleaned patch of net

Figure 3.4.7. Repair with bitten seam

Figure 3.4.6. Normal Repair

Figure 3.4.8b. Bitten net ‘plug’

Figure 3.4.8a. Bitten net ‘plug’

Figure 3.4.9a. Frequent net ‘plugs’ in white circles

Figure 3.4.9b. Frequent net ‘plugs’ in white circles

www.preventescape.eu

Clear areas and bite marked repairs were distributed generally around the cage, whilst cleaned ‘plug’ marks were found generally below 9 m near the cage bottom (see Figure 3.4.10).

Figure 3.4.10. Concentration of bite areas around the cages.

In conclusion: • Markings on the cage wall of large seabream may be evidence of net biting on loose threads and removal of the antifouling coating. • Targeted areas included unblemished parts of the net, loose threads from repairs and raised portions of the net. • There were no overt indications of potential biting damage from juvenile seabream. In the second experiment, a divided cage system was used to observe the behaviour of 150 - 200 g seabream over 3 days in relation to a hole the net. A divider net was introduced daily to the 12 m x 12 m cage with a camera lined up to observe a hole cut into the net. The hole was approximately 10 cm long and represented roughly 1.5 body heights of fish – under tension the hole opened out into an oval shape. The net was deployed down the side of the cage, then gently pulled away from the side of the cage to a position approximately 3 m away, to divide the cage in two, where fish were confined to an area consisting of 60% of the cage and a smaller area with no fish. Video examination noted the incidents of a number of specific behaviours, including: shoaling distance from the net, proximity of individual fish to the hole, inspection of the hole, contact inspection, escaping from the confined area, re-entering the confined area.

Generally speaking, the number of interactions (inspection, contact, exiting) were highest by day 3, probably related to the shoal moving closer to the net wall and the fish consequently having a greater possibility of interacting with the net. Shoaling behaviour of seabream was generally seen as a loose aggregation of fish that did not approach the cage wall very often. The shoal was mostly distant in the early hours of the day getting slightly closer to the net wall towards the afternoon. There were a number of independent fish cruising along the net, but these fish did not approach the hole, and most of them took avoiding action. Individual proximity to the hole increased each day, particularly towards midday. The number of fish inspecting the hole, which was independent of proximity, followed a similar pattern. This is also reflected in the number of contacts with the hole where the fish came up to the hole and touched it with their snout or bit the edge of the hole but did not go through it (the most common practice). In terms of escapes, there were very few escape incidents and these were mostly on the final day and mostly after midday. The fish that passed through the hole approached it in a straight line from a distance and went cleanly through the hole. Fish gave the impression that passing through the hole was accidental, as they just swam through a clear line/space in front of them. Figure 3.4.11 shows some seabream interactions with the net. Incidental observations made for seabream indicated that they kept up a constant level of nibbling/biting/tearing contact with a loose thread from another experiment in the background of the camera view during the hole experiment.

Figure 3.4.11. Cage with seabream, shoal away from the cage wall with a fish nibbling on the opening, shoal up at the cage wall with an individual escaping and re-entering the net.

In conclusion: • Seabream do escape through a hole in the net, but the incidence of escapes were low. • Interactions with the hole were not as dependent on the position of the shoal in relation to the net as in seabass, as there were a number of independently moving fish along the wall of the cage. • Most close interactions with the net would not lead to escape.

www.preventescape.eu

General Discussion - Seabream Prior to the start of these experiments, the perceived wisdom on the behaviour of seabream was that they were net biters but were reluctant escapers. In terms of net biting, the experiments undertaken in this task have confirmed that seabream are indeed biters, both from evidence of bite marks on the commercial cage nets and also in the background of the hole experiment. In the previous section of this report it is suggested that seabass interactions with the net are related to their shoaling behaviour, which could be complex. Seabream also had a complex shoaling with similar tight and loose formation, but in contrast would in some cases move towards a person on the walkway (perhaps linking the presence of people around the nets with feeding). Seabream were less reactive than seabass and also exhibited more independent behaviour, with individual fish sometimes swimming along the cage wall separate from the shoal. They exhibited an interest in the hole with a high incidence of contact or biting at the hole edge, but had few escapes.

Recommendations

for

European Seabass

Net biting behaviour may lead to net damage and a reduction in net strength, which in turn may lead to the formation of holes from strain on weakened net parts. Holes will lead to escapes. In light of the seabass experiments, some recommendations can be made concerning farming practices that can reduce the chances of hole forming, which include: â&#x20AC;˘ Ensuring that there are no loose threads inside the net â&#x20AC;˘ Frequent inspection of nets for seabass, as they can be highly responsive to a hole in the net. â&#x20AC;˘ If a hole in the net is suspected, personnel should stay above the site until the net can be inspected and repaired, as seabass seemed to move away from the net when personnel were present.

Key Recommendations

for

Gilthead Seabream

In light of the seabream experiments, some recommendations can be made concerning farming practices that can reduce the chances of hole formation which include: • If the net is repaired out of the water, ensure all seams, knots and loose threads are on the outside of the net • All repaired seams should have smooth finishes inside the net • In general there should be no loose threads inside the net • Tie-wraps/cable-ties are very useful for quick repairs until the net can be removed during scheduled maintenance. • If a hole in the net is suspected, personnel should stay away from the site until the net can be quickly inspected and repaired, as seabream seem to move towards the net if personnel were present. • If the same colour antifouling is frequently used, it may be better to have the initial nets in the same colour so there is no colour contrast if the anti-fouling is removed.

www.preventescape.eu

References

cited

Dempster, T.; Moe, H.; Fredheim, A.; Jensen, Q.; Sanchez-Jerez, P., 2007: Escapes of marine fish from sea-cage aquaculture in the Mediterranean Sea: status and prevention. In: Impact of mariculture on coastal ecosystems. F. Briand (Ed.). Lisboa, 21–24 Feb., CIESM Workshop Monographs. No. 32, pp. 86. ICES, 2006. ICES Mariculture Committee, 2006. Report of the Working Group on Environmental Interactions of Mariculture (WGEIM), pp. 144–145. http://www.ices.dk/reports/MCC/2006/ WGEIM06.pdf. Rana, K.J. 2005. Regional review on aquaculture development 6. Western-European Regions – 2005. FAO Fisheries Circular No. 1017/6. FIMA/C1017/6. 55pp.

3.5. General

conclusions

recommendations

for preventing and mitigating escape-related fish behaviour Cite this article as: Noble C, Hedger RD, Uglem I, Papadakis VM, Papadakis VM, Papadakis IE, Glaropoulos A, Kentouri M, Smith CJ, Zimmermann E, Fleming IA, Høy E, Damsgård B (2013) General conclusions & recommendations for preventing and mitigating escape-related fish behaviour. In: PREVENT ESCAPE Project Compendium. Chapter 3.5. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Chris Noble1, Richard D. Hedger2, Ingebrigt Uglem2, Vassilis M. Papadakis3, Ioannis E. Papadakis3, Alexios Glaropoulos3, Maroudio Kentouri3, Christopher J. Smith4, Emily Zimmermann5, Ian A. Fleming5, Erik Høy6 & Børge Damsgård1 Nofima, Norway, Norwegian Institute for Nature Research, Norway, 3 University of Crete, Greece, 4 Hellenic Centre for Marine Research, Greece, 5 Ocean Sciences Centre, Memorial University of Newfoundland, Canada 6 SINTEF Fisheries & Aquaculture, Norway 1 2

Fish escapes from aquaculture can be economically and environmentally detrimental to the aquaculture industry and its surrounding environs. In farmed species that exhibit a propensity to inspect and bite the cage wall, determined efforts should be undertaken to understand a) the motivation behind these behaviours, b) the factors that can increase the risk of these behaviours occurring, and c) the measures that can be taken to reduce or mitigate against these risks. The Pre-Escape work package of the Prevent Escape project was tasked with these specific objectives. Three key European aquaculture species that can exhibit escape-related behaviours: Atlantic cod (Gadus morhua L.), European seabass (Dicentrarchus labrax L.) and gilthead seabream (Sparus aurata L.) were used to assess the factors that can increase or mitigate against these behaviours. These species are primarily farmed in northern Europe (Atlantic cod) and southern Europe (seabass and

www.preventescape.eu

seabream) and mean the work package addresses the problem of behaviourally driven escapes on a pan-European scale. Prior to the Prevent Escape project, the biological and husbandry factors that may drive and augment the risk of fish exhibiting escape-related behaviours were poorly understood. The Pre-Escape work package was designed to address this shortfall.

Recommendations for preventing behaviours in Atlantic cod

and mitigating escape-related

With regard to Atlantic cod (see Figure 3.5.1), the Pre-Escape Work Package showed that there were a number of husbandry factors that can increase the frequency of escape-related behaviours. We showed that Atlantic cod do interact with, inspect and bite the net wall. This behaviour has also been noted in previous studies (Moe et al. 2007, Hansen et al. 2009) and net biting does have the potential to damage cage netting materials (Høy et al. 2011).

Figure 3.5.1 Atlantic cod (Gadus morhua) in a sea cage. Photo: Frank Gregersen, Nofima

Findings from the Atlantic cod experiments undertaken during the Pre-Escape WP included: • Feed deprivation increased net inspection and biting frequency (Damsgård et al. submitted, see also Moe et al. 2007, Hansen et al. 2009). Tank studies also showed that poorly repaired areas of the net wall, specifically where a contrasting coloured thread was used to repair net damage, also increases net inspection and biting frequency.

• In sea cages where cod were fed a full satiation ration during autumn/winter, fish directed their inspection behaviour at the net panel and the net frame, rather than at different types of repair. • The frequency of escape related behaviours may have a seasonal aspect - the overall number of net inspections and bites decreased during autumn and winter when water temperature and daylength decreased. • The majority of the sea cage studies showed that the number of inspections and bites decreased during feeding. This suggests the presence of feed in the water column distracts the fish from browsing the net wall. • Short-term, 50% feed restriction did not increase the frequency of net inspections. However, low level net biting frequency did increase during feeding when fish were fed reduced rations • Biofouling with a long-term algal biofilm did not emerge as a risk factor for net inspection or net biting behaviour during spring. However, biofouling with seaweeds did increase the frequency of both net biting and inspection behaviour. • There was no clear relationship between short-term changes in stocking density and the frequency of net inspection behaviours. To reduce the prevalence of net inspection and biting behaviour in Atlantic cod farmers should: • Limit the length of time that they deprive cod of feed. Short-term feed deprivation is a common husbandry practice when cod are being graded and transported. Farmers should be aware that cod will increase net inspection and biting in the days during and following these periods of feed deprivation. • Although a short-term, 50% feed restriction did not emerge as a risk for increasing net inspection behaviour, farmers should be aware that net biting can increase if fish are underfed. • If seaweeds (or other macro net fouling materials) build up around the net walls, farmers should attempt to remove these as quickly as possible as long seaweed strands increase net inspection and biting behaviour in Atlantic cod. • Net repairs should be carried out diligently, with no loose thread left exposed at the end of each repair. Where possible, the colour of the repair thread and net colour should be matched.

www.preventescape.eu

Recommendations for behaviours in European

preventing and mitigating escape-related seabass

With regard to European seabass, work carried out during the Pre-Escape WP also identified a number of husbandry factors that promote net biting and inspection behaviours, and also increase the prevalence of fish swimming through a hole in the net. Findings from the seabass experiments carried out within the WP included: • Studies carried out on commercial farms showed that seabass had a propensity to bite loose thread on the cage wall. Contact with this thread was primarily through simple biting, but also with more physical tearing at the thread. However, this biting behaviour was only observed in commercial and not experimental studies. • The level of interaction with the cage wall and loose thread can change both within and between days on a commercial farm. • Reduced rations or short-term starvation had no effect on the escape activity of seabass. • The proximity of neighbouring cage structures reduced the frequency of escapes. • Exit of fish through the hole seemed to be dependent on proximity of the shoal to the net wall. To reduce the prevalence of net inspection, biting or escape-related behaviour in European seabass, farmers should: • Ensure that there are no loose threads inside the net wall, on commercial farms. This will markedly reduce the frequency of bass biting and interacting with the net wall, and reduce the risk of further net damage. • Seabass seemed to move away from the net when personnel were present. If a hole in the net is suspected, personnel should stay above the site until the net can be inspected and repaired. • After a net has been damaged, the close proximity of neighbouring cage structures may decrease the risk of seabass swimming through a hole in the net. If farmers wish to reduce the potential number of escapes through a hole in the net wall, they could install a floating net near or around the cages until the hole is repaired.

100

Recommendations

for preventing and mitigating escape-related

behaviours in gilthead seabream

With regard to gilthead seabream, both small and large scale experiments carried out during the Pre-Escape WP also identified a number of husbandry factors that promote net inspection, biting (Figure 3.5.2) and the propensity for bream to swim through a hole in the net (Figure 3.5.3).

Figure 3.5.2 Photo-sets showing gilthead seabream biting the net. Photograph a) fish approaching the net wall, b) fish bends to the right and bites the twine , c) - g) fish repeatedly pulls the twine, h) fish terminates the biting and begins to leave the area.

Figure 3.5.3 Photo-sets showing the escape sequence of gilthead seabream. Photograph a) shows the fish locating the hole, b) and c) shows the fish entering and exiting the hole and d) shows the fish on the other side of the net. Total time required for escaping through the hole was less than 333 ms. Photo-set: University of Crete.

www.preventescape.eu

101

Findings from the seabream experiments undertaken during the Pre-Escape WP included: • Stocking density does increase the risk of seabream inspecting a net wall and escaping through a hole (Papadakis et al. 2012). This contrasts with our findings on Atlantic cod, and shows the need for research to be carried out on escape risks at a species-specific level. • Both reduced rations and feed deprivation increased net inspection and the frequency of bream escaping through a hole in the net (Glaropoulos et al. in press). • White netting material increased the number of net inspections in comparison to dark coloured nets in seabream. • Biofouling also increased the frequency of net biting behaviour in seabream. • Studies carried out in commercial cages showed that the incidence of escapes can be low when compared with seabass. • However, larger seabream do appear to browse upon and damage the net wall, as markings on the cage wall may be evidence of net biting on loose threads and removal of the antifouling coating. • Commercial cage studies showed that interactions with a hole in the net were not very dependent on the position of the shoal in relation to the net; a number of fish were observed to be moving along the wall of the cage, independent of the shoal. • Close interactions with a net hole did not often lead to an escape in cage-held seabream on a commercial farm. To reduce the prevalence of net inspection, biting or the frequency of gilthead seabream swimming through net holes, farmers should: • Be diligent about seabream increasing the frequency of their escape-related behaviours when held at higher stocking densities. • Match feed delivery to appetite, and neither underfeed nor completely starve fish for prolonged periods. • Maintain net cleanliness. If biofouling materials build up on or around the net walls, farmers should attempt to remove these as quickly as possible.

102

• Seabream are attracted by loose threads and damage the net wall (Glaropoulos et al. in press). If the net is repaired out of the water, ensure all seams, knots and loose threads are on the outside of the net. Further, all repaired seams should have smooth finishes inside the net and there should be no loose threads inside the net. Tie-wraps/cable-ties are very useful for quick repairs until the net can be removed during scheduled maintenance. • Seabream seemed to move towards the net if personnel were present. If a hole in the net is suspected, personnel should stay away from the site until the net can be quickly inspected and repaired.

References

cited

Damsgård B, Høy E, Uglem I, Hedger RD, Izquierdo-Gomez D, Bjørn PA (in press) Net-biting and escape behaviour in farmed Atlantic cod, Gadus morhua: Effects of feed motivation and net traits. Aquaculture Environment Interactions Glaropoulos A , Papadakis VM, Papadakis IE, Kentouri M (in press) Escape-related behavior and coping ability of seabream due to food supply. Aquaculture International Hansen LA, Dale T, Damsgård B, Uglem I, Aas K, Bjørn P-A (2009) Escape-related behaviour of Atlantic cod, Gadus morhua L., in a simulated farm situation. Aquaculture Research 40:26–34 Høy E, Volent Z, Moe-Føre H, Dempster T (2011) Loads applied to aquaculture nets by the biting behaviour of Atlantic cod (Gadus morhua). Aquacultural Engineering 47:60-63 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007) Technological solutions and operational measures to prevent escapes of Atlantic Cod (Gadus morhua) from sea-cages. Aquaculture Research 38:90–99 Moe H, Gaarder RH, Olsen A, Hopperstad SO (2009). Resistance of aquaculture net cage materials to biting by Atlantic cod (Gadus morhua). Aquacultural Engineering 40:126–134 Papadakis VM, Papadakis IE, Lamprianidou F, Glaropoulos A, Kentouri M (2012) A computer-vision system and methodology for the analysis of fish behavior. Aquacultural Engineering 46:53-59

www.preventescape.eu

103

104

seeks to reduce the number of fish that escape from European aquaculture through research to improve fish farming techniques and technologies.

4.1. The

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

importance of identifying escaped fish from

aquaculture and determining their post-escape behaviours for environmental and fisheries management Cite this article as: Sanchez-Jerez P, Arechavala-Lopez P, Fernandez-Jover D, Uglem I, Black K, Somarakis S, Ladoukakis M, Haroun R & Dempster T (2013) The importance of identifying escaped fish from aquaculture and determining their post-escape behaviours for environmental and fisheries management. In: PREVENT ESCAPE Project Compendium. Chapter 4.1. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Pablo Sanchez-Jerez1, Pablo Arechavala-Lopez1, Damian Fernandez-Jover1, Ingebrigt Uglem2, Kenny Black3, Stelios Somarakis4, Manolis Ladoukakis5, Ricardo Haroun6, Tim Dempster7 University of Alicante, Spain Norwegian Institute of Nature Research, Norway 3 Scottish Association of Marine Science, Scotland 4 Hellenic Centre of Marine Research, Crete, Greece 5 University of Crete, Crete, Greece 6 University of Las Palmas de Gran Canarias, Spain 7 SINTEF Fisheries & Aquaculture, Norway 1 2

www.preventescape.eu

105

Natural

fish populations mixed with escapees from aquaculture:

how can we detect the escapees?

Farmed fish may differ both in morphology and condition from wild fish, which likely affects their behaviour, competitive ability and spawning success compared with wild populations. These changes to biological characteristics are both environmental, due to the conditions of the culture, and genetic in origin. Escapes have been documented for almost all of the main species produced along European coastlines, such as Atlantic salmon (Salmon salar), Atlantic cod (Gadus morhua), seabass (Dichentrarchus labrax), seabream (Sparus aurata) and meagre (Argyrosomus regius). Escapes have the potential to mix with wild populations and negatively affect survival and fitness (Chapter 4.3). After an escape event, farmed fish may swim to natural habitats and mix with wild stocks. If this takes place on feeding grounds and the escapees compete for food resources used by wild fish, the availability of food may decrease, particularly if the carrying capacity of the ecosystem is near its upper limit. In some cases, cannibalism can be also an important impact, with increasing mortality of young individuals through predation by escapees. Reproduction of wild fish populations can also be affected if escapes can spawn successfully in areas where wild conspecifics exist, resulting in interbreeding. After successful interbreeding, the behavioural and life history characteristics of the resulting wild-farmed hybrids may be altered, potentially reducing their performance in the wild. Therefore, invasions of escaped farmed fish have the potential to negatively impact the productivity of wild fish populations through resource competition and competitive displacement. While the outcome of interactions between farm and wild fish will be context-dependent, varying with a number of environmental and genetic factors, they will frequently be negative for wild fish (Thorstad et al. 2008). For fisheries biology, the analysis of population structure for a species is of primary importance in developing an optimal strategy for efficient management. Distinguishing escapees from natural individuals can be of great importance in this context. In addition to distinguishing between wild and farmed fish, determining the spatial distribution of escapees in wild populations and the movements and migratory patterns of escapees when they reach natural environments is clearly important. For example, measures of growth, survival and reproductive success all assume that a single wild population is being studied. Population mixing with escapees confounds such measures, introducing noise into assessments of wild populations. In areas where aquaculture is intensively practiced, managers must

106

include the “immigration” and potential establishment of escaped congeners in their decision-making. Therefore, they need information on the abundance, size structure, reproductive success, trophic behaviour and spatial distribution of both wild and escaped populations. For this purpose, managers need to know the biological differences between escapees and wild populations, and they will need tested tools for identifying intraspecific units or stocks of a species to enable better management.

Objectives

of the

“Post-escape”

work package.

The “Post-escape” work package aimed to develop efficient methods to identify escaped Atlantic cod, seabream, seabass and meagre based on analyses of morphological variation, growth patterns in scales and otoliths, trace element profiles of scales, tissue fatty acid profiles, and genetic indicators. Further, the work package determined the immediate postescape behaviour and the long-term spatiotemporal distribution of escapees from fish farms to assess the prospects and develop optimal protocols for recapture of escapees. The work package also evaluated the ecological functions of escapees and assessed the impact of escapees on local professional and recreational fisheries.

How

can escapees be identified?

Fisheries biology has developed different approaches to identify and classify fish stocks, because understanding stock structure is vital to design appropriate management regulations in fisheries (Begg and Waldman 1999). Based on this principle, escapees can be treated as a specific stock within a species, with particular features that can be used for clear distinction from wild congeners.

Morphological changes In hatcheries, fish grow faster and experience a different environment than their wild counterparts. This phenomenon has been utilized to distinguish between wild and reared salmonids with a relatively high degree of certainty (Fleming et al. 1994, Fiske et al. 2005). Differential relative growth of body parts conditioned by environmental factors is a common feature of fish development (Osse and van den Boogaart 1995, Loy et al. 1999). In several

www.preventescape.eu

107

species, developmental modifications may also be closely linked to ontogenetic changes in resource use (Sagnes et al. 1997, Ward-Campbell and Beamish 2005). Such different developmental modifications may exist between wild and farmed fish given that they experience large differences in feeding regime and environment. For example, farmed fish are fed on feed pellets and confined to fish cages. Geometric analysis of scales and otoliths provide sources of variation amenable to morphometric and other forms of analysis (e.g. Richards and Esteves 1997). Features containing stock-specific information such as annuli spacing are biologically interpretable (i.e. related to age and growth), whereas other features such as perimeter shape are not easily interpretable (Begg and Waldman 1999).

Fatty acid profiles as biomarkers Aquaculture feeds have a very specific formulation with respect to their fatty acid constituents, with terrestrial vegetable oils and meals now heavily used to replace oils and meals derived from marine fish sources. Much research has focused on determining the optimal proportions for the substitution of fish meal and fish oil by plant products, without compromising fish growth or health (Turchini et al. 2009). However, vegetable oils, such as soybean, rapeseed, linseed, or palm oils, cause a distinct change in the fatty acid profile of aquacultured fish away from their wild counterparts. This is because aquafeeds are rich in saturated acids such as palmitic (16:0) or stearic acid (18:0), monounsaturated fatty acids such as oleic acid (18:1n9), and polyunsaturated fatty acids (PUFA), especially linoleic acid (18:2n-6). Fatty acids are not inert compounds. They accumulate over time and represent an integration of dietary intake over days, weeks, or months, depending on the organism and its energy intake and storage rates (Iverson 2009). Fatty acids have often been used as dietary markers (Iverson et al. 2004) and can be applied for monitoring the influence of fish farming on the environment (Fernandez-Jover et al. 2011), because changes in the fatty acid composition can be easily and unequivocally identified in fish tissues.

Changes to otolith and scales elemental composition Elemental signatures contained in some body elements, such as otoliths or scales, reflects the exposure of an individual fish to both the environment and its own physiology. These signatures typically differ among groups of fish, which have experienced different environmental histories, whether or not the groups come from the same population (Campana 2005). The presence of different signatures can be used to infer the existence of groups of fish that have remained separate for a certain period of time. The presence of variations in water temperature and chemistry inside fish farms can result in different otolith and scale chemical composition, and suggests that elemental fingerprints should discriminate well among escapees and wild fish (Campana et al. 2000).

108

Genetic tools The strength of genetic differences among stocks is positively associated with time since divergence of stocks (mediated by generation time, with shorter generation time accelerating genetic differentiation), and their degree of isolation (i.e., reproductive exchange between stocks eroding genetic differences, Adkinson 1996). Fish aquaculture uses wild caught broodstock from different regions, which can result in immediate differences between wild and farmed fish. In addition, the selective breeding of aquaculture fish currently underway for many species will increase divergence between farmed and wild populations. Therefore, the potential for detectable genetic differences between farmed and wild stocks is high. Some loss of genetic diversity, in comparison with wild stocks, had been reported in farmed stock, probably arising from genetic drift. The scale of genetic differences could be used as a precise tool to discriminate escapees.

Escapees

in the wild: where do they go?

Farmed fish are confined in a reduced water volume, with typical densities ranging from 5 to 25 kg m-3. When an escape event occurs through storm-induced damage or human error (Chapter 2.2), escapees can move freely around the farm structure or migrate to natural habitats. Understanding the movements and dispersal of escapees at scales of days to years after an escape event is important to management and conservation of natural resources in numerous ways (Metcalfe et al. 2008), although obtaining the necessary information is frequently difficult. It is only by understanding the movements and behaviour of individuals over short (hour-days), medium (days-weeks) and long (seasons and years) time-scales that managers can reveal the potential to recapture escapees and determine how escapees use natural resources such as food and compete with wild congenerics. Therefore, in the Post-escape work package, we studied fish movements with a focus on determining practical, implementable management options. Mixed behavioural strategies within species and populations (Dingle 1996), selection pressure imposed by fishing (Law 2000) and predators (Sanchez-Jerez et al. 2008), and differences in habitat suitability and oceanographic conditions result in large variations in the spatial distribution in a region where an escape event has happened. Detailed knowledge of the movement behaviour of a fish species is required to decide which proportion of suitable habitat needs to be protected to positively affect the exploited stock and to quantify this effect to provide a convincing argument for the closure of an area to fishing.

www.preventescape.eu

109

Direct underwater observation of escapes could detect specific interactions between escapees and wild pelagic and benthic fish, but only over a short-time scale (e.g. predation, cage reentering behaviour, schooling with wild population, the use of shelter). For studies longer than the scale of hours, tagging methods that involve marking the fish with a marker (e.g. plastic tag) that allows the identification of the fish either visually or with the aid of a detection device are required (Dempster and Kinsgford 2003). Commercial or sport fisheries can also recapture tagged fish. Mark and recapture is widely used in the study of dramatic, largescale, often transoceanic, movement patterns of commercially valuable species. However, mark and recapture data are of limited value when used to predict home range behaviour, as they typically contain only two positions occupied during the life of a fish. Thus, the spatial resolution in such studies is too coarse to quantify the use of space accurately (Kerwath 2005). Attaching an acoustic transmitter capable of emitting a signal that can be recorded by a passive acoustic receiver is an alternate technique (Uglem et al. 2009, Arechavala-Lopez et al. 2010). Acoustic telemetry techniques enable the tracking of small-scale movements of marine fishes, because the presence and depth of a tagged fish can be recorded through an array of receivers. Continuous tracking of escapees for extended time periods is therefore possible.

Conclusions The process of detecting escapees from aquaculture is essential for effective fisheries management and to promote mitigation actions. Methods to detect Atlantic salmon escapees are largely established (e.g. Glover et al. 2010), but the development of new tools and technologies to discriminate escapees continues. The most optimal indicator(s) for distinguishing escaped Atlantic cod, seabass, seabream or meagre was poorly described prior to the Prevent Escape project. Further, knowledge regarding their post-escape behaviours was lacking, making assessments of whether to invest time and resources in recapture efforts and how to undertake recapture without a knowledge basis. The following chapters (Chapter 4.2 â&#x20AC;&#x201D; 4.7, this compendium) detail the indicators that can be used to distinguish escapees and wild fish of these species and document their behaviour once they enter the natural environment as escapees, with the principal aim to develop optimal strategies to detect and mitigate the environmental effects of escapees.

110

Figure 4.1.1. Example of a mistakenly labelled wild fish: farmed escaped gilthead seabream labelled and priced as wild fish in a Spanish fish market.

Figure 4.1.2. A single hole in the net can lead to escapes of fish from aquaculture. Tracking the movement of fish after an escape can provide information to organize recapture efforts.

www.preventescape.eu

111

References

cited

Adkinson MD (1996) Population differentiation in Pacific salmon: local adaptation, genetic drift, or the environment. Can J Fish Aquat Sci 52: 2762-2777 Arechavala-Lopez P, Uglem I, Sanchez-Jerez P, Fernandez-Jover D, Bayle-Sempere JT, Nilsen R (2010) Movements of grey mullet Liza aurata and Chelon labrosus associated with coastal fish farms in the western Mediterranean Sea. Aquacult Environ Interact 1: 127-136 Begg GA, Waldman JR (1999) An holistic approach to fish stock identification. Fish Res 43: 35-44 Campana SE (2005) Otolith elemental composition as a natural marker of fish stocks. In: Stock identification methods. Application in Fishery Science. Ed: Cadrin SX, Friedlan, KD, Waldman, JR, Elsevier Campana SE, Chouinard GA, Hanson M, FrĂŠchet A, Brattey J (2000) Otolith elemental fingerprints as biological tracers of fish stocks. Fish Res 46: 343-357 Dempster T, Kingsford MJ (2003) Attraction of pelagic fishes to fish aggregation devices (FADs): the role of sensory cues. Mar Ecol Prog Ser 258: 213-222 Dingle H (1996) Migration. The biology of life on the move. Oxford University Press, New York. 474 p Fernandez-Jover D, Arechavala-Lopez P, Martinez-Rubio L, Tocher DR, Bayle-Sempere JT, Lopez-Jimenez JA, Martinez-Lopez FJ, Sanchez-Jerez P (2011) Monitoring the influence of marine aquaculture on wild fish communities: benefits and limitations of fatty acid profiles. Aquacult Environ Interact 2: 39-47 Fiske P, Lund RA, Hansen LP (2005) Identifying fish farm escapees. In Stock Identification Methods: Applications in Fishery Science. Ed: Cadrin SX, Friedland KD and Waldman JR. Elsevier, Amsterdam: 659-680 Fleming IA, Jonsson B, Gross MR (1994) Phenotypic divergence of sea-ranched, farmed, and wild salmon. Can J Fish Aquat Sci 51: 808-2824 Glover KA (2010) Forensic identification of farmed escapees: a review of the Norwegian experience. Aquacult Environ Interact 1:1â&#x2C6;&#x2019;10 Iverson SJ (2009) Tracing aquatic food webs using fatty acids: from qualitative indicators to quantitative determination. In: Arts MT, Brett MT, Kainz M (eds) Lipids in aquatic ecosystems. Springer Science-Business Media, New York, NY, p 281-306

112

Iverson SJ, Field C, Bowen WD, Blanchard W (2004) Quantitative fatty acid signature analysis: a new method of estimating predator diets. Ecol Monogr 74: 211-235 Kerwath SE (2005) Empirical studies of fish movement behaviour and their application in spatially explicit models for marine conservation. Ph.D. Rhodes University. 227 pp Law R (2000) Fishing, selection, and phenotypic evolution. J Mar Sci 57(3): 659-668 Loy A, Boglione C, Cataudella C (1999) Geometric morphometrics and morpho-anatomy: a combined tool in the study of seabream (Sparus aurata, Sparidae) shape. Appl Ichthyol 15: 104-110 Metcalfe J, Arnold G, McDowall R (2008) Migration. In: Fish Biology and Fisheries (Vol 1) Ed: Hart PJB and Reynolds JD., Blackwell Science, Oxford Osse JWM, van den Boogaart, JGM (1995) Fish larvae, development, allometric growth, and the aquatic environment. ICES Marine Sciences Symposium 201: 21-34 Richards RA, Esteves C (1997) Stock-specific variation in scale morphology of Atlantic coast striped bass. Trans Am Fish Soc 126: 908-918 Sanchez-Jerez P, Fernandez-Jover D, Bayle-Sempere J, Valle C, Dempster T, Tuya F, Juanes F (2008) Interactions between bluefish Pomatomus saltatrix (L.) and coastal sea-cage farms in the Mediterranean Sea. Aquaculture 282: 61-67 Sagnes P, Gaudin P, Statzner B (1997) Shifts in morphometry and their relation to hydrodynamic potential and habitat use during grayling ontogenesis. J Fish Biol 50: 846-858 Thorstad EB, Fleming IA, McGinnity P, Soto D, Wennevik V, Whoriskey F (2008) Incidence and impacts of escaped Atlantic salmon Salmo salar in nature. Report from the Technical Working Group on Escapes of the Salmon Aquaculture Dialogue. January, 2008. 113 pp Turchini GM, Torstensen BE, Ng WK (2009) Fish oil replacement in finfish nutrition. Rev Aquacult 1: 10-57 Uglem I, Dempster T, Bjorn PA, Sanchez-Jerez P & Okland F (2009) High connectivity of salmon farms revealed by aggregation, residence and repeated movements of wild fish among farms. Mar Ecol Prog Ser 384: 251-260 Ward-Campbell BMS, Beamish WH (2005) Ontogenetic changes in morphology and diet in the snakehead, Channa limbata, a predatory fish in western Thailand. Env Biol Fish 72: 251-257

www.preventescape.eu

113

4.2. Methods to identify (Gadus morhua L.)

escaped

Atlantic

cod

Cite this article as: Uglem I, Black K, Berg M, Varne R, Nilsen R, Mork J, Bjørn PA (2012) Methods to identify Atlantic escaped cod (Gadus morhua L.). In: PREVENT ESCAPE Project Compendium. Chapter 4.2. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu

authors: Ingebrigt Uglem1, Kenny Black2, Marius Berg1, Rebecca Varne3, Rune Nilsen4, Jarle Mork3 & Pål Arne Bjørn4 Norwegian Institute of Nature Research, Tungasletta 2, NO-7485 Trondheim, Norway Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, Oban PA34 1QA, UK 3 Trondheim Biological Station, Department of Biology, Norwegian University of Science and Technology, N-7491 Trondheim, Norway 4 Nofima Marin, Muninbakken 9-13, P.O. Box 6122, N-9291 Tromsø, Norway 1 2

Introduction To map the distribution and possible ecological impacts of escaped farmed cod, escapees need to be traced in the wild. Simple, reliable, and fast methods for determining the origin of cod are required. The importance of reliable determination of the origin of cod caught in the wild is illustrated by frequent reports in Norwegian media during recent years regarding catches of abnormal and assumed escaped farmed cod. In many of these cases, it has hitherto been difficult to verify if such fish were of farm origin, because genetic samples were not taken. Identification of escapees from fish farms may be done in several ways. For instance, recent developments within genetics have led to increasingly more efficient and less costly ways for distinguishing between farmed and wild cod (e.g. Glover 2010). In addition, analyses of scales and body morphology can distinguish farmed and wild salmonids with a high degree of certainty (e.g. Fiske et al. 2006). Trace element compositions in scales and otoliths are also effective in distinguishing between wild and farmed salmon (Adey et al. 2009). Furthermore, variation in fatty acid composition in body tissues has been suggested as a tool for determining

114

wild or farmed origin because commercial fish feeds used in aquaculture affect the fatty acid composition of farmed fish compared to wild fish which feed on natural organisms (Fernandez-Jover et al. 2007).

Study

species

The Atlantic cod is an epibenthic-pelagic species, which is widely distributed in a variety of habitats, from the shoreline down to the continental shelf. Cod are distributed along the North American coast; around Greenland and Iceland, and along the coasts of Europe from the Bay of Biscay to the Barents Sea. Cod may grow up to 2 m in length and more than 90 kg, although this is extremely seldom today. In commercial fisheries, sizes typically range from 2 to 20 kg. Cod are omnivorous; they feed at dawn or dusk on invertebrates and fish, including young cod. They spawn pelagic eggs in batches once a year, usually during late winter and early spring. A 5 kg wild female may spawn around 2.5 million eggs during a spawning season. The most important stocks are the Norwegian Arctic stock in the Barents Sea and the Icelandic stock. Populations around Greenland and Newfoundland have declined dramatically, whereas the stock in the Barents Sea remains healthy. The Atlantic cod is en economically important species and it is marketed fresh, dried or salted, smoked and frozen.

Objective We evaluated whether analyses of scales, body morphology, fatty acids and trace elements have the potential to rapidly and accurately separate escapees from wild Atlantic cod. We did not focus on genetic methods, as reliable and effective methods to separate escaped and wild cod based on genetic variation already exist (Glover 2010).

Methods We evaluated several methods for identifying escaped cod by sampling fish from several commercial cod farms, situated along the Norwegian coast, and wild cod caught in the vicinity of these farms. First, the potential for using scales and fish morphology to separate between escapees and wild fish was assessed by analysing digital images of fish and their scales using computer-based image

www.preventescape.eu

115

analyses. Three measurements of the fish scales were used to separate wild and farmed cod; mean circuli breadth, length-adjusted scale radius and length-adjusted circuli number per scale (see Uglem et al. 2011 for details). The radius was measured from the centre to the edge of the scale, and the distances between individual circuli were measured along the same axis (Figure 4.2.1). We selected morphological features on the basis that they would: 1) clearly differentiate between farmed and wild fish; and 2) be easy to measure in the field. Morphological codes were then assigned to these features (see Figure 4.2.1 for a clear illustration). Fatty acid variation was examined in samples from two commercial cod farms and also in a sample of wild cod captured at a spawning ground 80 km away from the nearest fish farm. Fatty acids were measured in ovary and liver tissue using a capillary gas chromatograph (Perkin Elmer, Autosystem XL, USA), with methods according to KjĂ¸rsvik et al. (2009). In addition, trace elements were measured in fish scales by ICPMS, after cleaning and complete dissolution, using the method of Adey et al. (2009). Samples were analysed together with certified reference materials and blanks. The applicability of morphological, scale, fatty acid and trace element measurements for identifying escapees was analysed using multivariate statistical analyses (Uglem et al. 2011), and the results presented as MDS plots (R open source statistical software).

Figure 4.2.1. Principle sketch showing morphological measures and scale measures. The abbreviations are described in Table 3. The area from where the scale samples were taken is indicated under the last dorsal fin.

116

Results Analysis showed that the mean scale circuli breadth and the length-adjusted scale radius measurements differed significantly between wild and farmed cod, whereas the lengthadjusted circuli number did not. Discriminant analyses indicated that the mean scale circuli breadth measurement correctly classified wild and farmed fish 86% of the time, while the length-adjusted scale radius measurement was successful in 80% of cases (Table 4.2.1). Correct classified (%) Model

Eigenvalue

Wild

Farmed

Mean circuli breadth and length-adjusted 1 scale radius

0.51

<0.001

86.1

80.0

Morphology, PC1, 2, 3

3.16

<0.001

Morphology, LJ, HA, and FA

4.03

<0.001

100

Fatty acids, Ovaries, PC1, 2, 3, 4

37.3

<0.001

100

3.8

<0.001

Fatty acids, Liver PC1, 2, 3

34.0

<0.001

1.4

<0.001

100

Table 4.2.1. Data from discriminant analysis for scale and morphological parameters, and also fatty acids, including the proportion of wild and farmed Atlantic cod being correctly classified. Original classification and cross-validation is identical. The codes for the morphological parameters are described in detail in Figure 4.2.1.

The extent to which morphological variation could be used to separate between escapees and wild cod was assessed in two ways. Firstly, principal component analysis (PCA) was used to reduce the variation among the morphological features (size-adjusted) into three principal components (PCs), and these explained 73% of the variation. A discriminant analysis of individual PC scores showed that 97% of the wild and 96% of the farmed fish were classified correctly (Table 4.2.1). Subsequently, three morphological features were selected to assess the viability of using a few simple measurements to discriminate between farmed and wild cod. The primary selection criterion was that these parameters would be easy to measure in the field, and the parameters FA, LJ and HA were thus selected (Figure 4.2.1). A discriminant analysis with FA, LJ, and HA showed that 100% of the wild fish were classified correctly, while 95% of the farmed fish were classified correctly (Table 4.2.1). Fatty acids profiles varied among wild and farmed fish and between farms (Figure 4.2.2, 4.2.3) PCA reduced the number of variables that represented the variation in fatty acids between farmed and wild fish. Discriminant analysis of the significant PCs, for both the liver and the ovary samples, showed that the fish could be correctly classified with respect to origin 100% of the time using this technique (Table 4.2.1).

www.preventescape.eu

117

The metal profiles from farmed and wild fish from TromsĂ¸ (Figure 4.2.4b) and Trondheimsfjord (Figure 4.2.4c) can be distinguished with very high confidence. When plotted together (Figure 4.2.4a) it is clear that the 95% confidence limits from some populations overlap. For example, this reveals that wild TromsĂ¸ scales are much more similar to farmed Trondheim fish than to farmed fish from the same fjord. Farmed fish from the three farms are separated with high confidence as do the two wild populations. These results indicate that fish scale chemistry has the potential to discriminate between farmed fish and local wild populations of Norwegian cod with high confidence, and could provide a useful tool in determining the origin of suspected escapees. However, at present, we do not know whether the significant difference found in the metal profiles of wild and farmed fish would diminish with increasing time after escape.

118

Figure 4.2.2. Individual factor scores and loadings for ovaries samples for PC1 and PC2 in relation to Atlantic cod (Gadus morhua) origin.

Figure 4.2.3. Individual factor scores and loadings for liver samples for PC1 and PC2 in relation to Atlantic cod (Gadus morhua) origin.

www.preventescape.eu

119

Figure 4.2.4. MDS plots of cod scale chemistry profiles from (right to left) a) Trondheimsfjord (farmed and wild), TromsĂ¸ (farmed and wild), and Tysfjord (farmed); b) TromsĂ¸ (farmed and wild) and; c) Trondheimsfjord (farmed and wild). Ellipses indicate 95% confidence intervals.

120

Discussion We have shown that analyses of scales, morphology, fatty acid profiles and trace metals all have the potential to distinguish between wild and farmed Atlantic cod, and that in the future these tests might be useful as management tools. For instance, our results indicate that a high proportion of Atlantic cod can be correctly classified as farmed or wild, based on three simple morphological features, which can be measured either from digital images or from live fish, following anaesthesia. Often a field-based determination of fish origin is an advantage over more labour and time intensive laboratory techniques, such as genetic discrimination. A standardized methodology based on a few simple traits for identifying escaped farmed cod, would represent a cheap and simple approach for evaluating the origin of cod, complementing more advanced methods. This series of experiments can be regarded as a proof-of-concept study. We recommend that the methods tested be verified through blind tests before they are applied in a real life situation, i.e. testing datasets not originally used to develop the statistical models. Furthermore, there is a need to examine additional farmed and wild fish populations, as well as several year classes, ages and diets, before functional and reliable methodologies for discrimination between wild and farmed cod can be developed. The precision of the various measures could be verified by simultaneously carrying out genetic analyses.

Recommendations â&#x20AC;˘ We have shown that analyses of scales, morphology, fatty acid profiles and trace metals all have the potential to distinguish between wild and farmed Atlantic cod â&#x20AC;˘ However, our results are based on proof-of-concept studies and considerable efforts are needed to develop standardized, reliable and functional methods that can be used as operational management tools

www.preventescape.eu

121

Links

to published documents:

Uglem I, Berg M, Varne R, Nilsen R, Mork J, Bjørn PA (2011) Discrimination of wild and farmed Atlantic cod (Gadus morhua) based on morphology and scale-circuli pattern. ICES Journal of Marine Science 68, 1928-1936, DOI: 10.1093/icesjms/fsr120

References

cited

Adey EA, Black KD, Sawyer T (2009) Scale microchemistry as a tool to investigate the origin of wild and farmed Salmo salar. Ma Ecol Prog Ser 390: 225-235 Fernandez-Jover D, Jimenez JAL, Sanchez-Jerez P (2007) Changes in body condition and fatty acid composition of wild Mediterranean horse mackerel (Trachurus mediterraneus, Steindachner, 1868) associated to sea cage fish farms. Mar Environ Res 63:1-18 Fiske P, Lund RA, Hansen LP (2006) Relationships between the frequency of farmed Atlantic salmon, Salmo salar (L.), in wild salmon populations and fish farming activity in Norway, 19892004. ICES J Mar Sci 63:1182-1189 Glover KA, Dahle G, Westgaard JI, Johansen T, Knutsen H, Jørstad KE (2010) Genetic diversity within and among Atlantic cod (Gadus morhua) farmed in marine cages: a proof-of-concept study for the identification of escapees. Anim Genet 41, 515-522 Kjorsvik E, Olsen C, Wold PA, Hoehne-Reitan K, Cahu CL, Rainuzzo J, Olsen AI, Oie G, Olsen Y (2009) Comparison of dietary phospholipids and neutral lipids on skeletal development and fatty acid composition in Atlantic cod (Gadus morhua). Aquaculture 294:246-255 Uglem I, Berg M, Varne R, Nilsen R, Mork J, Bjørn PA (2011) Discrimination of wild and farmed Atlantic cod (Gadus morhua) based on morphology and scale-circuli pattern. ICES J Mar Sci 68:928-1936

122

4.3. Methods

to identify escaped seabass and

seabream Cite this article as: Arechavala-Lopez P, Sanchez-Jerez P, Fernandez-Jover D, Bayle-Sempere JT, Black KD, Ladoukakis E, Somarakis S, Dempster T (2013) Methods to identify escaped seabass and seabream. In: PREVENT ESCAPE Project Compendium. Chapter 4.3. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Pablo Arechavala-Lopez1, Pablo Sanchez-Jerez1, Damiรกn Fernandez-Jover1, Just T. BayleSempere1, Kenny Black2, Emmanuel Ladoukakis3, Stelios Somarakis4 & Tim Dempster5 University of Alicante, Spain; Scottish Marine Institute, UK; 3 University of Crete, Greece; 4 Hellenic Centre for Marine Research, Greece; 5 NOFIMA, Norway. 1 2

Introduction Fish escape from farm facilities as a consequence of technical and operational failures, leading to the escape of millions of seabream and seabass into the wild each year (Prevent Escape Compendium Chapter 2). Once in the wild, escaped seabream and seabass can survive for months, swimming away from their cages to nearby fish farms, fishing grounds, coastal habitats and local harbours, where they feed on natural prey and compete for natural resources with wild populations (Chapter 4.5, this compendium). Thus, escapees could have negative ecological and genetic consequences on wild fish populations and nearby farmed stocks through the spread of pathogens, interbreeding or resources competition (Naylor et al. 2005). Moreover, there is a large interaction between aquaculture and local fisheries where the latter benefit from farmaggregated wild fish and farmed escapees. An increase of escapees in fisheries landings has been recorded during the last years in Mediterranean coastal areas, which is accompanied by a decrease both in price and mean size of individuals (Dimitriou et al. 2007).

www.preventescape.eu

123

To better understand these potential negative effects, it is imperative to distinguish and quantify the number of individuals that escape from sea cages. Effective tools which differentiate farmed individuals within wild stocks are required. These tools will improve scientific knowledge of escape events, help to assess potential genetic and ecological risks of escapees on wild populations, and help assess their contribution to fisheries landings. Furthermore, within the production chain, similar fish products can arise from different points of origin and there is potential for fraud due to product mislabelling. The temptation to label farmed fish as wild fish by fish merchants, retailers and restauranteurs is significant because of the price premium commanded by wild fish. Thus, verifiable and robust methods to distinguish farmed from wild fish are required for consumer confidence and for local authority enforcement purposes, to combat mislabelling and conform to legislation (Bell et al. 2007, Morrison et al. 2007).

Tools

to identify escapees

Several techniques have been applied to classify salmon according to their farmed or wild origin, based on the assumption that wild and cultured salmon experience large differences in growth, feeding regimes and environments. This process started in the 1980s when, for a simple and quick identification, a combination of several techniques typically used for stock identification in fisheries were used routinely to survey the amount of farmed escapees in wild catches of salmon (Fiske et al. 2005). At present, a wide range of techniques (genetic, chemical characteristics, fatty acid composition, trace element levels, presence of pollutants, stable isotopes, morphology and organoleptic characteristics) are used to distinguish between wild and escaped farmed fish.

Objective We tested a broad suite of tools to discriminate between the farmed or wild origin of seabream and seabass in the Mediterranean Sea. Our aim was to assess which technique could be effectively applied to accurately assess the level of escapee intrusion into wild populations and fisheries landings.

124

Identification

of seabream and seabass escapees

Different techniques were applied to differentiate between wild and farmed seabream and seabass from Spain and Greece, from the easiest and cheapest methods to the most labourintensive or expensive.

External

appearance and morphology

External appearance and morphological characteristics reflect to some degree the life history of the fish, since external characteristics in farmed fish are affected by culture conditions, such as stocking density and feeding strategy (Grigorakis 2007). Apart from the usual body malformations easily observable in some farmed fish (e.g. Sola et al. 1998, Loy et al. 1999), wild gilthead seabream exhibit lower body height than farmed seabream, which have a sharper snout and a more squat and compact shape (i.e. being shorter, wider and higher; Figure 4.3.1; Flos et al. 2002, Grigorakis et al. 2002). Farmed seabream had small, rounded and less developed teeth compared to the bigger, sharper teeth in their wild counterparts (Grigorakis et al. 2002). Skin characteristics also differ between wild and farmed fish; farmed seabream have thinner skin which is much darker in the dorsal and head areas. The characteristic iridescent colours of this species are much duller, which is suggested to be related to lack of access to natural, rather than commercial food (Grigorakis et al. 2002). Moreover, farmed seabream have a smaller belly and sharper dorsal fins and a higher degree of erosion in the caudal and pectoral fins (Grigorakis et al. 2002), which are strongly related to the stocking density and swimming behaviour of farmed seabream. In contrast, external differences in European seabass among wild and cultured fish are not as pronounced, and identification cannot rely on shape, colour, general appearance or fin erosion (Eaton 1996). However, we detected significant morphometric differences (through Truss Network System analysis; Figure 4.3.2) in the cranial and body regions of seabream and seabass relating to their farmed or wild origin. Furthermore, a higher condition index, which is a typical indicator of good dietary condition, was found for cultured fish compared to wild fish for both seabream and seabass. Other indices, such as relative profile for seabream and cephalic index for seabass, are good indicators of fish origin. Thus, wild and cultured seabream and seabass show external differences, which cannot only be used to indicate dietary condition and history

www.preventescape.eu

125

(Grigorakis 2007), but might also be easy, cheap and reliable tools to discriminate escapees from wild fish shortly after escape incidents have occurred. Therefore, we recommend morphological features as a first step to determining escapees. Such evaluations of fish origin could be made by farmers, fishermen, merchants, consumers, scientists or management agencies.

Figure 4.3.1. Wild (top) and farmed (bottom) gilthead seabream from the Spanish coast.

126

Figure 4.3.2. Example of landmarks and distances measured on seabream (top) and seabass botoom) used to assess morphometric differences (Truss Network System) in wild and farmed fish.

www.preventescape.eu

127

Scales

and otoliths shape

High fish densities within cages (5 - 10 kg m-3) lead to greater numbers of physical collisions and thus high friction among individuals. Pronounced levels of scale loss occur through abrasion (Eaton 1996), aggressiveness, lymphocystis or handling (Fiske et al. 2005). Logically, a high rate of scale detachment in farmed fish leads to a high rate of scale replacement. In comparison, scale loss and replacement in wild fish are relatively rare. We detected higher average levels of replacement scales in farmed seabream than wild fish; this can be used to classify seabream specimens to farmed or wild origin. Although fish can completely replace a scale within a year of its loss, wild fish scales are still distinguishable from those of farmed fish, which will never regenerate the nucleus (Figure 4.3.3). In addition, growth patterns are also reflected in scale patterns. Wild fish have a slower growth rate and show more annuli, and fewer esclerites, on their scales than farmed fish. This is a result of the more constant environmental conditions under which the fish are reared, combined with a regular food supply, which greatly reduces the effects of seasonality on the growth patterns of the fish (Figure 4.3.3). Therefore, scale characteristics are the easiest and quickest way to identify escapees. They can be used directly in the field by non-experts, without expensive equipment or labour intensive methods. Environmental conditions, food supply and genetic dissimilarities can influence the shape of the sagittal otoliths (Pannella 1971), which could help distinguish the farmed or wild origin of seabream and seabass. In general, seabream sagittal otoliths had a pentagonal to elliptical shape with serrate margins, while seabass sagittal otoliths are fusiform to oblong in shape (Figure 4.3.4). In both species, the use of image analysis techniques, such as shape descriptors or elliptical Fourier analysis, and further discriminant analysis, detected differences based on wild or farmed origin with high accuracy, and also detected differences between geographical origins for fish of larger sizes. This methodology is a more objective, reliable method that the use of external morphometric traits, as they are not affected by short-term variations in fish physiological condition, standard tissue preservation techniques, or by geographical differences in morphology.

128

Figure 4.3.3. Scale characteristics of wild and farmed seabream and seabass.

Figure 4.3.4. Sagittal otoliths from wild and farmed seabream and seabass.

www.preventescape.eu

129

Fatty

acid profiles in flesh and liver tissues

In marine aquaculture, the high lipid content of the diet and the intensive feeding regime affect the chemical composition of the fish, resulting in a higher fat content (Lopparrelli et al. 2004). Fish lipids are well known to be rich in long-chain n-3 polyunsaturated fatty acids (LC n-3 PUFA), especially eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), which together with arachidonic acid (ARA), are considered to be essential dietary fatty acids. The incorporation and storage of FAs in fish tissues strongly depend on the FA profile of the diet (Sargent et al. 2002). Hence, the current practice of substituting fish oils with other vegetable lipid sources in farmed marine fish diets leads to notable changes in lipid composition and FAs profiles, due to the presence of FAs of terrestrial origin such as oleic acid (OA), Îą-linolenic acid (LNA) or linoleic acid (LA), which are usually found in low levels in marine fish. Therefore, differences between wild and farmed dietary FA profiles have been widely used to discriminate farmed and wild fish origin. Several authors have studied the differences in FA composition of different tissues, such as muscle, liver, skin or brain, in both seabream and seabass. Generally, these studies showed

SEABREAM

Figure 4.3.5. Proportion of Linoleic acid and Arachidonic acid in farm and wild seabream muscle.

130

pronounced differences between wild and farmed counterparts according to LA, which is found in higher proportion in farmed fish, and ARA, which is found in lower levels for adult cultured seabass and seabream (Figures 4.3.5 and 4.3.6). However, different patterns have been established for the rest of individual FAs between wild and farmed fish because of the high variability and high standard deviations within the reviewed results. Such high heterogeneity is strongly affected by the dietary history and exhibits strong seasonality (Grigorakis et al. 2007, Yildiz et al. 2008). To strengthen the reliability of the analysis of FA profiles, a multivariate approach (e.g. Principal Component Analysis), which takes into account the whole FA profile, is effective for differentiating wild and farmed seabream and seabass. We used a multivariate analysis to discriminate wild and cultured fish species according to their FA profile and this technique proved to be a strong tool. The most important source of concern regarding this technique is the â&#x20AC;&#x2DC;wash-outâ&#x20AC;&#x2122; of fatty acids due to the ability of escaped fish tofeed in the wild (Chapter 4.5, this compendium) and therefore substitute FAs from terrestrial origin with a FA from natural diets. The fatty acid signature may clearly reveal the origin of an individual if it has recently escaped from a fish farm after being feed with commercial pellets for a period of time. However, the FA profile of an escapee will change if a natural diet is consumed over time. In this case, other techniques need to be applied. SEABASS

Figure 4.3.6. Proportion of Linoleic acid and Arachidonic acid in farm and wild seabass muscle.

www.preventescape.eu

131

Trace

elements in scales and otoliths

Marine fish will incorporate different trace elements from the environment into their skeletal tissues and organs, either present in seawater or the diet, forming a chemical signature that will reflect the length of time that a fish has inhabited a particular water body (Lal 1989). Hence, trace element profiles are likely to be unique to a given population that inhabits one given location, and aquaculture creates a special situation in which the normally roaming species become fixed in one specific location with also unique environmental conditions. Wild populations of seabass and seabream in the Mediterranean are known to roam between different zones, and for this reason it could be difficult to find differences in trace elemental signatures among otoliths from different populations of wild seabass and seabream. We studied trace elements in scales and otoliths and detected clear differences between wild and farmed seabream and seabass for specific elements. For instance, higher values of Mn and Ba, and lower levels of Sr, were found in farmed fish than in wild fish for both species (Figures 4.3.7 and 4.3.8). Despite the fact that other studies gave contrasting results, most were able to distinguish wild and farmed fish with great accuracy, but only through a multivariate approach which accounted for a wide range of elements. No simple diagnostic elemental concentration or ratio exists. This indicates that the trace elemental profile might be more appropriate than the presence or absence of a specific quantity of an element (Figures 4.3.7 and 4.3.8).

Figure 4.3.7. MDS plot of elemental analysis of scales from putatively wild (W) and farmed (F) seabream from 2 farms in Mediterranean Spain (Alicante). Stress factors are shown in corner boxes. Ellipses delimit 95% confidence intervals. Elements are ranked in terms of their contribution to the observed distribution.

132

Figure 4.3.8. MDS plot of elemental analysis of scales from putatively wild (W) and farmed (F) seabass from 2 farms in Mediterranean Spain (Alicante). Stress factors are shown in corner boxes. Ellipses delimit 95% confidence intervals. Elements are ranked in terms of their contribution to the observed distribution.

Genetic

differences

Identifying escapees with genetic methods may prove quick, reliable and applicable in different ontogenetic levels of fish. Molecular markers could be applied to characterize the genetic structure of wild and farmed populations. Provided that farmed and wild populations are genetically distinguishable and well-characterized, individual fish taken at random can be allocated in one or the other population with some probability based on their genetic profiles. We used a set of 16 polymorphic microsatellite loci for seabream and another 16 loci for seabass to study the genetic profile of farmed and wild populations from Spain and

www.preventescape.eu

133

Greece. All populations for both species showed a deficit in Hardy-Weinberg equilibrium. Gene diversity was slightly higher in wild than in farmed populations for both species. This is expected as farmed populations contain a small proportion of the total diversity of the wild populations. Average inbreeding coefficient (FIS) was smaller for seabream than for seabass. It was significantly higher than zero for all populations, except the Greek farmed population for both species. Pairwise FST coefficient was significantly higher than zero for all populations for both species. This suggests that all populations are genetically different from each other. There are three different explanations for the genetic differentiation between farmed populations and the nearby wild populations. First, farmed populations have evolved under different conditions from the wild populations from which they were originated. This explanation is highly unlikely because there has not been enough time for such genetic differentiation and because there is continuous gene flow towards farmed populations as the breeders come from wild populations. Particularly for gilthead seabream, a sequential protandrous hermaphrodite species, there is a continuous need to introduce wild males into aquaculture since most individuals turn into females after their second year of reproduction. Second, breeders from the wild that are used in aquaculture are a small and non-random sample of the wild populations. Third, breeders or fry, which are used in a location, have been imported from another location. All three explanations might hold simultaneously, but with different significance for each of them. The genetic differentiation of farmed from wild population is also shown using Bayesian clustering. This method makes possible the identification of potential escapees to the environment (Glover et al. 2009). The probability of allocation of an individual to a certain population depends on the genetic differentiation of the populations. This probability was higher in the Greek populations for both species than in the Spanish populations (Figure 4.3.9 and 4.3.10).

Figure 4.3.9. Bayesian clustering of farmed and wild seabream for Greek and Spanish populations.

134

Discussion To assist in evaluating the potential risk on wild stocks due to farm escapees, we have identified several useful tools to identify the wild or farmed origin of fish. Techniques such as morphology, external appearance or fatty acid profiles are valuable methods to distinguish between the farmed and wild origins of seabream and seabass in a short-term period. However, how long these differences last once a farmed fish enters the wild, which will influence the accuracy of these methods over time, remains unknown. Scale characteristics are the quickest and cheapest way to identify escapees with a high accuracy in field studies, and are easy to use for farmers, fishermen, merchants, consumers, scientists or managers. While effective tracing using these various methods will assist in identifying the levels of escapee intrusions into wild populations, an alternative is to mark all farmed fish in hatcheries before they are transferred to sea cages. Marking may enable, for the first time, direct tracing of escapees back to the farm and even cage of escape, which will assist in pinpointing how the escape incident first occurred and inform escape prevention methods (Chapter 2, this compendium). Marking of fish in ways that ensure mark retention throughout the life cycle of the fish, without adverse effects on growth and welfare, is an important area for future research.

Figure 4.3.10. Bayesian clustering of farmed and wild seabass for Greek and Spanish populations. populations.

www.preventescape.eu

135

Recommendations â&#x20AC;˘ The presence of a regenerated nucleus in seabream scales and the lack of annual rings on seabass scales are the easiest and quickest way to identify an escapee. We recommend these methods for applications where rapid assessments are required in the immediate days to weeks after an escape event has occurred. â&#x20AC;˘ A multivariate approach using chemical and/or molecular characteristics represents a highly accurate method for distinguishing farmed seabream and seabass from their wild counterparts. We recommend these methods are used by management agencies and scientists for applications where a high degree of accuracy is required and escapes need to be detected many months after an escape incident has occurred.

136

Links

to published documents

Arechavala-Lopez P, Sanchez-Jerez P, Bayle-Sempere JT, Sfakianakis DG, Somarakis S (2012a) Morphological differences between wild and farmed Mediterranean fish. Hydrobiologia 679:217–231 Arechavala-Lopez P, Sanchez-Jerez P, Bayle-Sempere JT, Sfakianakis DG, Somarakis S (2012b) Discriminating farmed fish from wild Mediterranean stocks through scales and otoliths. J Fish Biol (in press) Arechavala-Lopez P, Sanchez-Jerez P, Izquierdo-Gomez D, Toledo-Guedes K, Bayle-Sempere JT (2012c) Fin erosion on wild and farmed Sparus aurata (L.) and Dicentrarchus labrax (L.). J Appl Ichthyol (in press)

References

cited

Bell JG, Preston T, Henderson RJ, Strachan F, Bron JE, Cooper K, Morrison DJ (2007) Discrimination of wild and cultured European seabass (Dicentrarchus labrax) using chemical and isotopic analyses. J Agri Food Chem 55:5934-5941 Dempster T, Moe H, Fredheim A, Jensen Ø, Sanchez-Jerez P (2007) Escapes of marine fish from sea-cage aquaculture in the Mediterranean Sea: status and prevention. CIESM Workshop Monogr 32:55–60 Dimitriou E, Katselis G, Moutopoulos DK, Akovitiotis C, Koutsikopoulos C (2007) Possible influence of reared gilthead seabream (Sparus aurata L.) on wild stocks in the area of the Messolonghi lagoon (Ionian Sea, Greece). Aquacult Res 38:398–408 Eaton DR (1996) The Identification and Separation of Wild-caught and Cultivated Seabass (Dicentrarchus labrax). MAFF Fisheries Research Technical Report, vol. 103. MAFF, Lowestoft, 15 pp Fiske P, Lund RA, Hansen LP (2005) Identifying fish farm escapees. In Stock Identification Methods: Applications in Fishery Science Ed: Cadrin, S. X., K. D. Friedland & J. R. Waldman / Elsevier, Amsterdam. pp 659–680 Flos R, Reig L, Oca J, Ginovart M (2002) Influence of marketing and different land-based system on gilthead seabream (Sparus aurata) quality. Aquacult Int 10:189–206 Gillandres BM, Sanchez-Jerez P, Bayle-Sempere JT, Ramos-Esplá A (2001) Trace elements in otoliths of the two-banded bream from a coastal region in the south-west Mediterranean: are there differences among locations? J Fish Biol 59:350–363

www.preventescape.eu

137

Glover KA, Hansen MM, Skaala O (2009) Identifying the source of farmed escaped Atlantic salmon (Salmo salar): Bayesian clustering analysis increases accuracy of assignment. Aquaculture 290:37–46 Grigorakis K (2007) Compositional and organoleptic quality of farmed and wild gilthead seabream (Sparus aurata) and seabass (Dicentrarchus labrax) and factors affecting it: a review. Aquaculture 272:55–75 Grigorakis K, Alexis MN, Taylor KDA, Hole M (2002) Comparison of wild and cultured gilthead seabream; composition, appearance and seasonal alterations. Int J Food Sci Tech 37:477–484 Lal SP (1989) Minerals. In Fish nutrition. Ed: J. E. Halver /Academic Press San Diego. pp 220–257 Loy A, Boglione C, Cataudella S (1999) Geometric morphometrics and morpho-anatomy: a combined tool in the study of seabream (Sparus aurata, Sparidae) shape. J Appl Ichthyol 15:104–110 Morrison J, Preston T, Bron JE, Henderson RJ, Cooper K, Strachan F, Bell JG (2007) Authenticating production origin of gilthead seabream (Sparus aurata) by chemical and isotopic fingerprinting. Lipids 42:537–545 Naylor R, Hindar K, Fleming IA, Goldburg R, Williams S, Volpe J, Whoriskey F, Eagle J, Kelso D, Mangel M (2005) Fugitive salmon: assessing the risks of escaped fish from net-pen aquaculture. BioScience 55(5):427–437 Pannella G (1971) Fish otoliths: daily growth layers and periodical patterns. Science 173:1124–1127 Sargent JR, Tocher DR, Bell JG (2002) The lipids. In: Fish nutrition. Ed: Halver JE, Hardy RW / Academic Press, San Diego, CA Sola L, De Innocentiis S, Rossi AR, Crosetti D, Scardi M, Boglione C, Cataudella S (1998) Genetic variability and fingerling quality in wild and reared stocks of European seabass. In Bartley, D. & B. Basurco (eds). Cah Opt Méd 34:273–280 Yildiz M, Sener E, Timur M (2008) Effects of differences in diet and seasonal changes on the fatty acid composition in fillets from farmed and wild seabream (Sparus aurata L.) and seabass (Dicentrarchus labrax L.). Int J Food Sci Tech 43:853–858

138

4.4. Post-escape dispersal of Atlantic cod (Gadus morhua L.) and juvenile Atlantic salmon (Salmo salar L.) Cite this article as: Uglem I, Llinares Serra RM, Mork J, Nilsen R, Økland F, Varne R, Rikardsen A, Bjørn PA (2013) Post-escape dispersal of Atlantic cod (Gadus morhua L.) and juvenile Atlantic salmon (Salmo salar L.). In: PREVENT ESCAPE Project Compendium. Chapter 4.4. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Ingebrigt Uglem1, Rosa Maria Llinares Serra2, Jarle Mork3, Rune Nilsen2, Finn Økland1, Rebekka Varne3, Audun Rikardsen4 & Pål Arne Bjørn2 Norwegian Institute of Nature Research, Tungasletta 2, No-7485 Trondheim, Norway Nofima Marin, Muninbakken 9-13, P. box 6122, No-9291 Tromsø, Norway 3 Trondhjem Biological Station, Department of Biology, Norwegian University of Science and Technology, No-7491 Trondheim, Norway 4 University of Tromsø, No-9037 Tromsø 1 2

Introduction Understanding the post-escape dispersal pattern of fish farm escapees would be useful for predicting their possible ecological impacts as well as for improving recapture efficiency. While a range of studies have been carried out to assess the post-escape behaviour of adult Atlantic cod, Gadhus morhua (Uglem et al. 2008, 2010) and adult Atlantic salmon, Salmo salar (Skilbrei 2010a, b, et al. 2010), little is known about the post-escape behaviour of juvenile farmed cod and salmon. Therefore, we focused mainly on the behaviour and movements of juvenile cod and salmon following experimental escape events. Post-escape behaviour and dispersal were studied by tracking fish in their natural environment, using acoustic transmitters and traditional mark-recapture methodologies. For the latter, fish were marked with external, visible tags, released from participating fish farms, and their movements were assessed on the basis of recapture event, either in organised recapture efforts, or following capture within commercial and recreational fisheries.

www.preventescape.eu

139

Study

species

The Atlantic cod is an economically important species marketed in fresh, dried, salted, smoked and frozen form. It is widely distributed in a variety of habitats, from the shoreline to the continental shelf, along the North American coast, around Greenland and Iceland, and along the Europe coastline from the Bay of Biscay to the Barents Sea. The most important stocks are the Norwegian Arctic stock, in the Barents Sea, and the Icelandic stock. The populations around Greenland and Newfoundland have declined dramatically, while the Barents Sea stock remains healthy. It is an epibenthic-pelagic species that may grow up to 1.5 m in length and weigh more than 50 kg, although fish of this size are rarely encountered today. Cod caught within commercial fisheries typically range in weight from a couple of kg to around 20 kg. They are omnivorous; feeding at dawn or dusk on invertebrates and fish, including young cod. They spawn in batches once a year, usually during late winter and early spring, and have pelagic eggs. A 5 kg wild female may spawn around 2.5 million eggs during a spawning season. The Atlantic salmon is another important commercial species and is a target for both recreational and commercial fisheries. However, wild stocks have declined significantly during the last few decades, mainly due to anthropogenic impacts. Salmon is now one of the most important aquaculture species in the world, but salmon aquaculture is thought to have serious negative impacts on wild salmon stocks. The Atlantic salmon is an anadromous species; i.e. it reproduces in rivers while the juveniles migrate to the ocean for feeding before they return to the rivers as adults. Before the juvenile salmon migrate from freshwater to sea water they undergo a physiological transition that enables them to survive in high salinity. Juvenile salmon are referred to as “parr” before this transition and “smolt” afterwards. Salmon are pelagic in the ocean and have a similar geographical distribution to cod. They can grow to more than 1.2 m in length and weigh more than 30 kg.

Objectives We examined the spatio-temporal distribution of juvenile farmed Atlantic cod and Atlantic salmon after simulated escape incidents using acoustic telemetry, and the long-term dispersal and movements of both adult and juvenile cod using mark-recapture techniques.

140

Methods Acoustic telemetry Small acoustic transmitters were implanted into the abdominal cavity of juvenile cod (n=24) and salmon (n=97) prior to the simulation of an escape incident. Cod were implanted and then released from commercial cod farms in GildeskĂĽl in county Nordland and juvenile salmon from a land-based commercial smolt farm in the Trondheimsfjord in county TrĂ¸ndelag. Two groups of salmon were examined; pre-smolt fish i.e. parr (n=50), which have not developed complete sea water tolerance, and smolts (n=47), which are adapted to saltwater. The shortterm dispersal behaviour of the fish was tracked by an array of automatic receivers positioned at the farms and also throughout the fjord system (see Figure 4.4.1 for the receiver array for cod). Receivers deployed in the sea were attached to ropes at 3 m depth, while receivers in rivers were deployed on the river bed. All receivers recorded the transmitter identification code, and the date and time of detection when a tagged fish was within the receiver range. Range tests showed that the receiver ranges varied between 200 and 600 m in radius. Figure 4.4.1. Study area and configuration of the telemetry receiver array. The positions of the receivers are indicated with black circles.

www.preventescape.eu

141

Capture-mark-recapture The long-term behaviour and distribution of escaped juvenile and adult cod was studied by tagging 3377 farmed cod with external spaghetti tags and subsequently releasing them from three large-scale commercial farms to simulate escape incidents. Juvenile cod were tagged and released at one farm in the GildeskĂĽl area in county Nordland, while adult cod were tagged and released at two farms in the Trondheimsfjord area in county TrĂ¸ndelag. Three groups of juvenile cod were tagged and released from the same farming location in Nordland. The first group was released for the purpose of studying long-term dispersal, and was deliberately released 3 weeks prior to initiation of an organized recapture program (gill nets and commercial cod pots). The two other groups were released only 24 hours before the organized recapture programs started. Recaptures were also recorded in both target recapture fisheries, as well as recreational and commercial fisheries.

Results Acoustic telemetry Nearly all of the cod remained within the fjord where they had been released (Figure 4.4.1), but most of the juvenile cod moved away from the release site (the cod farm) within the first week (Figure 4.4.2); the majority left along the shore line (Figure 4.4.3). Some of the fish were continuously detected around the release site during this period, and two were observed at another cod farm located in the same fjord system. One third of the salmon pre-smolt, equipped with acoustic transmitters, died within the immediate vicinity of their release location, compared with only 8.5% mortality of the smolts. The surviving parr dispersed away from the release site, at the fish farm, after two to three days, and, as with the juvenile cod, they moved predominantly along the shore line (Figure 4.4.3). In contrast, most of the surviving smolts left the farm area during the first day and around half moved away from the shore, adopting a more pelagic distribution than the pre-smolt. Both pre-smolt and smolt appeared to have a similar movement pattern and speed following departure from the release site close to the smolt farm. The number of surviving fish recorded in the fjord decreased throughout the study period, possibly due to fish migrating out of the fjord (Figure 4.4.4), although none of the fish migrated up into freshwater. Compared to existing knowledge on movements of released hatchery-reared smolts during spring, our results indicate a less directional and slower migration pattern during autumn.

Capture-mark-recapture Recapture rates of the three groups of juvenile cod were low. No recaptures from the first group were reported. However, several of these tags were found in the stomach of a small sample of adult saithe (Pollachius virens) and cod caught in the proximity of the fish farm immediately after release. As a result of this observation, we caught 160 potential predators around the release site during the following 3-4 weeks and checked their stomachs for tagged fish. A total of 105 tags from the first group of tagged juveniles were recovered accounting

142

for 10.1% of the tagged fish. As the abundance of large predators around salmon farms may be very high (e.g. Dempster et al. 2009), it can be assumed that the first release group suffered high mortality due to predation during their first days after release. The other two release groups were larger in size than the first group and probably too large to be exposed to the same intense predation pressure. The recapture rate for the second release was 5% (44 fish), with 39 being caught in organized recapture efforts and the 5 by local fisheries. Five fish were found within the stomachs of captured predators. The third group of juvenile cod was released about 6 months after the first group and one fish from this group was recaptured in organized gillnet fishery, while 4 fish were reported in commercial fisheries during the following months. All of the juvenile cod were recaptured less than 3 kilometers from the release site (Figure 4.4.5). As the organized recapture fisheries accounted for the majority of the recaptures and since this fishery was carried out close to the farms immediately after release, it is difficult to generalize with respect to the long term dispersal pattern of juvenile cod after escape. The recapture of adult tagged cod varied between the two farms in TrĂ¸ndelag (Figure 4.4.6), with higher recapture rate for fish released from the farm located in the inner part of the fjord (N = 500, recapture rate: 4.5%) compared to the outermost farm (N = 250, recapture rate: 0.5%). Only one of the adult fish released at the farm located close to the coast was recaptured, three months after release. The adult fish from the innermost farm fish were recaptured throughout the entire fjord, up to 70 km away from the farm. The difference in recapture rates between these two locations may be because fish released closest to the coast rapidly dispersed to areas with a lower intensity of both commercial and recreational fishery. The recaptures of fish from the inner part of the fjord took place during a 7 month period, with the highest recapture rate from day 100 to day 160 after release. This period corresponds to July and August, or mid-summer in Norway with high water temperatures and a high fishing pressure due to recreational fishing during the summer holidays. Figure 4.4.2. Cumulative proportion of juvenile farmed cod departing from the release site at the fish farm within the first week following the simulated escape.

www.preventescape.eu

143

Figure 4.4.3. Direction for the first departure from the release location for Atlantic salmon pre-smolt/parr and smolt following simulated escape from a land based hatchery.

Figure 4.4.4. Proportions of surviving Atlantic salmon pre-smolt/parr and smolts detected in other areas than the hatchery area throughout the five week study period.

144

Figure 4.4.5. Study area for juvenile cod with release and recapture locations.

www.preventescape.eu

145

Figure 4.4.6. Study area for adult cod with release and recapture locations.

Discussion Following a simulated escape incident from an Atlantic cod farm, juveniles initially showed rapid dispersal, however, the escapees did not appear to move far from the source farm. The long-term recapture of tagged juveniles was close to zero, and smaller juveniles experienced high predation pressure due to the abundance of large predatory fish in the vicinity of the farms, as evidenced by the presence of tags in captured predator stomachs. These results indicate that the immediate mortality for juvenile fish following escape may be high and that their escape from sea cages may be of a limited ecological importance compared to escape at late life stages. Previous research has shown that adult cod also disperse rapidly from farms, but that adult escapees may be recaptured relatively far from the farms (Uglem et al. 2008, 2010). However, both the results from Uglem et al. (2008; 2010) and our results on adult recaptures suggest that the recapture period is relatively limited. This may indicate that escaped adults also suffer substantial mortality following escape, albeit over a longer timeframe than the juveniles. The results were similar for juvenile Atlantic salmon, which showed considerable mortality of pre-smolt. However, the mortality of seawater-adapted smolts was significantly lower. Our study, which was conducted in autumn, indicates a less directional and slower migration pattern of released hatchery-reared smolts compared with other studies where the smolts were released during spring. In general, the rapid dispersal from farms for both cod and salmon indicate that the potential for recapture of escapees following an escape incident is limited.

146

References

cited

Dempster T, Uglem I, Sanchez-Jerez P, Fernandez-Jover D, Bayle-Sempere J, Nilsen R, Bjørn PA (2009) Coastal salmon farms attract large and persistent aggregations of wild fish: an ecosystem effect. Marine Ecology Progress Series 385, 1-14. Skilbrei OT (2010a) Reduced migratory performance of simulated escaped Atlantic salmon post-smolts during autumn. Aquaculture Environment Interactions 1: 117–125 Skilbrei OT (2010b) Adult recaptures of farmed Atlantic salmon post-smolts allowed to escape during summer. Aquaculture Environment Interactions 1: 147–153 Skilbrei OT, Holst JC, Asplin L, Mortensen S (2010). Horizontal movements of simulated escaped farmed Atlantic salmon (Salmo salar L.) in a western Norwegian fjord. ICES J Mar Sci 6, 1206-1215 Uglem I, Bjørn PA, Dale T, Kerwath S, Økland F, Nilsen R, Aas K, Fleming I, McKinley RS (2008) Movements and spatiotemporal distribution of escaped famed and local wild Atlantic cod (Gadus morhua L.) in a Norwegian fjord. Aquacult. Res. 39, 158-170. Uglem I, Bjørn PA, Mitamura H, Nilsen R (2010) Spatiotemporal distribution of oceanic and coastal Atlantic cod sub-groups after escape from a farm. Aquaculture Environment Interaction. 1, 11-19

www.preventescape.eu

147

4.5. Post-escape

behaviours of farmed seabream

and seabass Cite this article as: Arechavala-Lopez P, Sanchez-Jerez P, Fernandez-Jover D, Bayle-Sempere JT, Uglem I, Dempster T (2013) Post-escape behaviours of farmed seabream and seabass. In: PREVENT ESCAPE Project Compendium. Chapter 4.5. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Pablo Arechavala-Lopez1, Pablo Sanchez-Jerez1, Damiรกn Fernandez-Jover1, Just T. BayleSempere1, Ingebrigt Uglem2 & Tim Dempster3. University of Alicante, Spain; Norwegian Institute of Nature Research, Norway; 3 SINTEF, Norway. 1 2

148

Introduction

of the problem/task

Escape of farmed fish from sea-cages is considered as one of the main environmental problems caused by aquaculture and it is perceived as a threat to natural biodiversity in marine waters (Dempster et al. 2007, Prevent Escape Compendium Chapter 1). Escaped fish may cause undesirable ecological effects in native populations through interbreeding, competition for food or habitats, as well as transfer of pathogens to wild fish or other farmed stocks (Fiske et al. 2005, Dempster et al. 2007, Thorstad et al., 2008). Escape events of gilthead seabream Sparus aurata and European seabass Dicentrarchus labrax have been sporadically recorded in the Mediterranean Sea (Dempster et al. 2002, Boyra et al. 2004, Vita et al. 2004, Tuya et al. 2005, 2006, Valle et al. 2007, Fernandez-Jover et al. 2008, Toledo-Guedes et al. 2009). Escapes occur mainly as a result of technical and operational failures in terms of cage breakdown due to extreme weather, holes caused by wear and tear of the netting and operational accidents (Prevent Escape Compendium Chapter 2). Farmed fish experience non-natural high densities within the cages, with an abundance of food and typically an absence of predators. This may affect the development of behaviour and possibly also entail artificial selection on traits that adapt the farmed fish to a life in captivity, and may also influence the fitness of escaped fish compared to their wild conspecifics (Huntingford et al. 2006, Salvanes and Braithwaite 2006). The post-escape behaviour of several other aquaculture fish species has been studied, for instance in Atlantic salmon Salmo salar (Thorstad et al. 2008, Skilbrei et al. 2009, 2010), and Atlantic cod Gadus morhua (Moe et al. 2007, Uglem et al. 2008, 2010, Hansen et al. 2009, Prevent Escape Compendium Chapter 4.4). These studies have suggested escapees are a potential risk to natural populations. However, few corresponding studies have been carried out for escaped fish from fish farms in the Mediterranean.

Importance

of seabream and seabass

Gilthead seabream have traditionally been cultured in the Mediterranean Sea and is a popular target species for fisheries which is regularly present in the fish markets. The world capture production of gilthead seabream is relatively constant across years, fluctuating from 5000 to 8000 t y-1 (total world capture was 7889 t in 2009). However, cultured seabream production has increased over the past 2 decades and now accounts for 95% of total seabream production (APROMAR 2011). The wild gilthead seabream is a subtropical fish that inhabits seagrass beds, sandy bottoms and the surf zone, commonly to depths of about 30 m, but adults may occur at 150 m depth. They occur naturally in the Mediterranean and the Black Sea (rare), and in the Eastern Atlantic, from the British Isles, Strait of Gibraltar to Cape Verde and around the Canary Islands. It is reported as a sedentary fish, though migrations are likely to occur on the Eastern Atlantic coast, from Spain to the British Isles. Seabream are present as either solitary individuals or in small aggregations. It is a euryaline species and moves in early spring towards protected coastal waters in search of abundant food and milder temperatures (trophic migration). In late autumn, they return to the open sea for breeding purposes, being very sensitive to low temperatures (lower lethal limit is 2째C). They are mainly carnivorous (shellfish, including mussels and oysters) and occasionally herbivorous (Froese and Pauly 2006).

www.preventescape.eu

149

The European (or common) seabass is also a fish with high commercial interest for both aquaculture and fisheries. Captures fisheries production in the Mediterranean Sea and Atlantic Ocean remain relatively constant among years at 8000 – 12000 t y-1. Like seabream, aquaculture production now accounts for 90% of total seabass production (APROMAR 2011). The European seabass is a gonochoristic species demersal fish that inhabits coastal waters down to about 100 m depth, but is more common in shallow waters. They occur in coastal waters of the Atlantic Ocean from South of Norway (60°N) to the Western Sahara (30°N) and throughout the Mediterranean Sea and the Black Sea. They are found on various kinds of bottoms on marine coastal waters, estuaries, lagoons and occasionally rivers. Young fish form schools, but adults are less gregarious. Juveniles feed on invertebrates (shrimps and molluscs), but consume more fish in their diets as they grow. European seabass is a gonochoristic species. Spawning takes place in winter in the Mediterranean Sea (December to March) and in spring (up to June) in the Atlantic Ocean. Eggs and larvae disperse widely during the first 3 months of life and adults migrate over several hundreds of kilometres (Froese and Pauly 2006).

Objective We simulated small and large-scale escape incidents with fish equipped with acoustic transmitters or tagged with external tags to: (i) assess the dispersion and survival of farmed seabream after an escape incident; (ii) test for connectivity among farms and local fishing areas through escaped seabream and seabass movements; (iii) study the habitat use and feeding habits of escapees; and (iv) evaluate to what extent escapees are recaptured by local fisheries.

Post-escape

behaviour of farmed seabream and seabass

Simulated escapes of seabream and seabass were made from fish farms in a coastal bay in the southeast of Spain (UTM: 30S 0710736 4219249; Figure 4.5.1). In this bay, there are four floating-cage fish farms located 3 - 4 km offshore on soft muddy bottoms at depths ranging from 23 - 30 m. Distances among farms varied from 1 - 5 km. Farms F1, F3 and FR (the farm where the experimental releases were carried out) grow approximately 1200 t y-1 each, mainly of seabream but also seabass. Farm F2 was an inactive farm whose floating structures remain without fish and nets (Figure 4.5.1). Artisanal fisheries usually work at a local scale, i.e. near the coast and around farm facilities, only moving some kilometres away from their local harbours. Eleven artisanal vessels operate in the study area from the port of Guardamar del Segura, around 30 artisanal vessels from Santa Pola, 4 artisanal vessels from Tabarca and 9 vessels from Torrevieja (Figure 4.5.1). Most artisanal fishers were trammel-netters, with some long-line boats. Furthermore, a marine protected area with special fishing restrictions is located around Tabarca Island (Forcada et al. 2009).

150

Figure 4.5.1 Farming area and acoustic receiver array (Vemco速, model: VR2W) where the seabream and seabass behavioural assessment was carried out. F: fish-farm facility; White pushpins: acoustic receivers sited at farms; Black pushpins: receivers surrounding the release farm (FR).

Survival and behaviour at farm facilities Tagging of seabream with ultrasonic tags allowed evaluation of their survival and movements after the simulated escape event. Escaped seabream survived up to 4 weeks at farm facilities and a high proportion of them remained at the release/escape farm for the first 5 days. The escaped seabream that remained next to farms for long periods tended to stay close to the surface during the night time, while they descended to greater depths in the morning (Figure 4.5.2). This may be a result of the natural diurnal variation in behaviour and vertical movements of escaped seabream, or may have been influenced by activity and feeding time (typically 06:00 to 14:00) at farms, since waste feed from farms becomes available at depths below 10 m for escapees and wild fish that reside outside the cages. Some escapees visited farms in the bay other than their release farm, indicating the existence of connectivity among farm facilities through their repeated movements among farms (Figure 4.5.3). Similar movement patterns were found for escaped seabass; several individuals moved quickly and repeatedly among several fish farms in the bay (Figure 4.5.3). Fast and repeated movements among fish farms might represent a vector for disease transmission (Uglem et al. 2009), since an infected escapee could spread pathogens to nearby cages.

www.preventescape.eu

151

A high proportion of both escaped seabream and seabass dispersed away from farms within the first week and recorded mortality rates were high (50 - 60%). Toledoâ&#x20AC;&#x201C;Guedes et al. (2009) suggested that escapees show a high degree of site fidelity or, in contrast, low site fidelity but high mortality rates. High densities of piscivorous predators occur around farms in this specific area (Fernandez-Jover et al. 2008, Arechavala-Lopez et al. 2010), such as bluefish Pomatomus saltatrix, which are attracted to the farms due to an abundance of smaller prey fish. These predators occasionally enter cages and feed on the caged seabream and seabass (Sanchez-Jerez et al. 2008). Hence, it is likely the released seabream and seabass experienced high predation pressure during the first few days after escape. Further, it is likely that some of the tagged fish were caught by local fishermen and not reported. However, it cannot be ruled out that tagging in some way affected the behaviour and survival of the escapees.

Figure 4.5.2 Vertical distribution (bars) and daily detections (dots) of tagged seabream around fish farms during the period of a day. Bars show mean depth values with standard error lines. Plotted line shows mean number of daily detections. Dark area: night time; white area: day time; striped area: feeding time at farms.

152

Figure 4.5.3 Movement sequences of escapees at farming area. Example of 2 tagged seabream individuals (A, B) and 2 tagged seabass individuals (C, D) within the receivers array visiting different farms (rectangles). Arrows represent a simulated direction of the tagged fish following the detections. F: acoustic receiver at farm; A-G: receivers surrounding release farm (FR).

www.preventescape.eu

153

Dispersion, habitat use and feeding habits A high number of farmed seabream (>2100) and seabass (>1200) were externally tagged (Hallprint速, model: T-bar) and released at farm facilities, simulating escape events, to assess the dispersion of escapees. Altogether, a total of 159 tagged seabream (7%) and 15 tagged seabass (1%) that dispersed from farm facilities were recaptured by local fisheries. Professional trammel-netters (gill-nets) caught most escaped seabream (72%), while recreational fishermen (fishing rods) caught all of the recaptured escaped seabass (100%). For seabream, most of the tagged and recaptured individuals were caught over seagrass and sandy bottoms, where their wild conspecifics live, but some were also captured in Guardamar harbour. The recorded recapture rates should be regarded as minimum estimates due to possible tag loss (Sanchez-Lamadrid 2001) and capture under-reporting. It is remarkable that 75% of total recaptures were caught within the first week after escape (Figure 4.5.4). This indicates a high intensity of fishing on the released seabream, mainly during the first days after release, and presumably before the escapees had adapted to their new environment (Sanchez-Lamadrid et al. 2004). Moreover, the last 2 recaptures were reported 45 and 60 days after release in a nearby coastal fishing ground, and the two farthest recaptures were reported >20 km (the 3rd day) and >15 km (the 4th day) distance, both south and north, from the release farm. Moreover, a significant number of wild fish were caught together with escapees, even up to 20 km from the release farm. Analysis of the stomach content of recaptured seabream illustrated their ability to feed on the most common natural prey after a short period of time, as the feeding habits of escapees approached those of wild seabream one week after escape. Initially, escaped seabream fed on macrophytes and food pellets from farms. Later, they began to prey on echinoderms and crustaceans mostly associated with benthic habitats beneath farms. Finally, their diet shifted to more common natural prey from the fifth day after release, consisting mainly of molluscs and crustaceans. For seabass, individuals were recaptured gradually over 3 months at the same site (Figure 4.5.4), Guardamar harbour. Wild seabass are also commonly caught in this estuarine environment by local recreational fishermen. However, huge shoals of farmed individuals occur in local harbours after escape incidents at farms (Lopez pers. comm.). Stomach content analyses of recaptured seabass did not reveal as clear a pattern as for seabream. Empty stomachs, particulate organic matter, food pellets and natural preys (decapods, echinoderms and fish) were observed throughout the study period (from day 7 to 69). Further studies with a greater number of samples are necessary for a better understanding of feeding habits and habitat use of seabass escapees. Thus, the fact that feeding grounds and habitat use of both seabream and seabass escapees overlapped with their wild counterparts, since escapees occur together with their wild conspecific and are able to switch to natural prey within short time periods, reflects their potential for survival and interaction with wild fish populations.

154

Figure 4.5.4 Percentage of recaptured tagged seabream and seabass from total recaptures throughout the study time.

Discussion Escaped seabream and seabass can remain for long periods around farm facilities, making repeated movements among farms, which increases the risk of pathogen transmission to other farmed stocks. Moreover, they disperse from farm facilities to natural habitats and fishing grounds where they co-occur with wild conspecifics, and are able to utilize natural food resources within a short time frame. Therefore, the potential for interactions of escapees with wild populations is evident from our results. The initial mortality of farmed fish following small-scale escape incidents may be high, possibly as a result of predation, but vary according to season and the occurrence of predators around the farms. The recapture of seabream escapees from small-scale escape incidents that survive the first days may also be substantial, even without specific recapture fisheries. In addition, successfully recapturing escaped seabass appears more difficult than for seabream, due to their better swimming capability and dispersion. Our data do not allow long-term evaluation of the survival and ecological effects of escapees, but it is possible that the cumulative impact of escapees may be substantially reduced by the high rate of mortalities after escape. However, as known from other farmed species, even a modest survival rate after escape might entail negative ecological consequences for wild populations (Thorstad et al. 2008). A greater knowledge basis regarding the survival and movements of escapees following large-scale escape incidents is required to evaluate the potential for negative ecological impacts due to escaped farmed seabass. This will improve the management for recapture of escapees and thereby reduce the potential risk of intermingling and compete with natural populations.

www.preventescape.eu

155

Recommendations • There is a high connectivity among farms through movements of escapees. This increases the potential risk of transfer of pathogens among farming sites. • Escapees are able to survive, disperse to nearby natural habitats, and feed on natural prey. These aspects increase the potential genetic and ecological risk of escapees to wild populations, since they could interbreed with wild conspecifics, prey upon wild food sources and thus compete for food and habitat, and share pathogens with wild stocks. • Predation of escapees immediately after escape incidents by wild fish closely associated to fish farms was likely significant. Thus, measures to maintain healthy populations of wild fish predators in the near vicinity of fish farms will assist in mitigating the ecological effects of escapees.

156

Links

to published documents

Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Bayle-Sempere JT, Sanchez-Jerez P (2011) Immediate post-escape behaviour of farmed seabass (Dicentrarchus labrax) in the Mediterranean Sea. J Appl Ichthyol 27:1375-1378 Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Bayle-Sempere JT, Sanchez-Jerez P (2012) Post-escape dispersion of farmed seabream (Sparus aurata L.) and recaptures by local fisheries in the Western Mediterranean Sea. Fisheries Research (in press). Sanchez-Jerez P, Fernandez-Jover D, Uglem I, Arechavala-Lopez P, Dempster T, Bayle-Sempere JT, Valle C, Izquierdo-Gomez D, Bjørn PA, Nilsen R (2011) Coastal fish farms as Fish Aggregation Devices (FADs). In: Artificial Reefs in Fisheries Management. Ed: Bortone S. et al., Taylor and Francis/CRC Press

References

cited

APROMAR (2011) La acuicultura marina en España. Asociación Empresarial de Productores de Cultivos Marinos de España). http://www.apromar.es/ Informes/S. Arechavala-Lopez P, Uglem I, Sanchez-Jerez P, Fernandez-Jover D, Bayle-Sempere JT, Nilsen R (2010) Movements of grey mullets (Liza aurata and Chelon labrosus) associated with coastal fish farms in the western Mediterranean Sea. Aquacult Environ Interact 1:127–136 Boyra A, Sanchez-Jerez P, Tuya F, Espino F, Haroun R (2004) Attraction of wild coastal fishes to Atlantic subtropical cage fish farms, Gran Canaria, Canary Islands. Environ Biol Fish 70:393–401 Dempster T, Sanchez-Jerez P, Bayle-Sempere JT, Gimenez-Casualdero F, Valle C (2002) Attraction of wild fish to sea-cage fish farms in the south-western Mediterranean Sea: spatial and shortterm variability. Mar Ecol Prog Ser 242:237–252 Dempster T, Moe H, Fredheim A, Jensen Ø, Sanchez-Jerez P (2007) Escapes of marine fish from sea-cage aquaculture in the Mediterranean Sea: status and prevention. CIESM Workshop Monogr 32:55–60 Fernandez-Jover D, Sanchez-Jerez P, Bayle-Sempere JT, Valle C, Dempster T (2008) Seasonal patterns and diets of wild fish assemblages associated to Mediterranean coastal fish farms. ICES J Mar Sci 65:1153–1160

www.preventescape.eu

157

Fiske P, Lund RA, Hansen LP (2005) Identifying fish farm escapees. In Stock Identification Methods: Applications in Fishery Science Ed: Cadrin, S. X., K. D. Friedland & J. R. Waldman / Elsevier, Amsterdam: 659–680 Forcada A, Bayle-Sempere JT, Valle C, Sanchez-Jerez P (2009) Habitat continuity effects on gradients of fish biomass across marine protected area boundaries. Mar Environ Res 66:536–547 Froese R, Pauly D (2006) Fish Base. http://www.fishbase.org. Hansen LA, Dale T, Damsgård B, Uglem I, Aas K, Bjørn PA (2009) Escape-related behaviour of Atlantic cod, Gadus morhua L., in a simulated farm situation. Aquacult Res 40:26–34 Huntingford FA, Adams C, Braithwaite VA, Kadri S, Pottinger TG, Sandøe P, Turnbull JF (2006) Review: Current issues in fish welfare. J Fish Biol 68:332–372 Lloris D (2002) A world overview of species of interest to fisheries. Chapter: Dicentrarchus labrax. FIGIS Species Fact Sheets. Species Identification and Data Programme-SIDP, FAO-FIGIS, 3p. http//:www.fao.org/figis/servlet/species?fid=2291. Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007) Technological solutions and operational measures to prevent escapes of Atlantic cod (Gadus morhua) from sea cages. Aquacult Res 38:91–99 Salvanes AGV, Braitwaite V (2006) The need to understand the behaviour of fish reared for mariculture or restocking. ICES J Mar Sci 63:346–354 Sanchez-Lamadrid A (2001) Effectiveness of four methods for tagging juveniles of farm-reared gilthead seabream, Sparus aurata, L. Fish Manag Ecol 8:271–278 Sanchez-Lamadrid A (2004) Effectiveness of releasing gilthead seabream (Sparus aurata, L.) for stock enhancement in the bay of Cádiz. Aquaculture 231:135–148 Skilbrei OT, Holst JC, Asplin L, Holm M (2009). Vertical movements of “escaped” farmed Atlantic salmon (Salmo salar)—a simulation study in a western Norwegian fjord. ICES J Mar Sci 66:278– 288 Skilbrei OT, Holst JC, Asplin L, Mortensen S (2010) Horizontal movements of simulated escaped farmed Atlantic salmon (Salmo salar) in a western Norwegian fjord. ICES J Mar Sci 67:1206– 1215 Thorstad EB, Fleming IA, McGinnity P, Soto D, Wennevik V, Whoriskey F (2008) Incidence and impacts of escaped farmed Atlantic salmon Salmo salar in nature. NINA Special Report 36, 110 pp.

158

Toledo-Guedes K, Sanchez-Jerez P, González-Lorenzo G, Brito Hernandez A (2009) Detecting the degree of establishment of a non-indigenous species in coastal ecosystems: seabass Dicentrarchus labrax escapes from sea cages in Canary Islands (Northeastern Central Atlantic) Hydrobiologia 623:203–212 Tuya F, Boyra A, Sanchez-Jerez P, Haroun R (2005) Non-metric multivariate analysis of the demersal ichthyofauna along soft bottoms of the Eastern Atlantic: comparison between unvegetated substrates, seagrass meadows and sandy bottoms under the influence of seacage fish farms. Mar Biol 147:1229–1237 Tuya F, Sanchez-Jerez P, Dempster T, Boyra A, Haroun R (2006) Changes in demersal wild fish aggregations beneath a sea-cage fish farm after the cessation of farming. J Fish Biol 69:682–697 Uglem I, Dempster T, Bjørn PA, Sanchez-Jerez P (2009) High connectivity of salmon farms revealed by aggregation, residence and repeated movements of wild fish among farms. Mar Ecol Prog Ser 384:251–260 Uglem I, Bjørn PA, Mitamura H, Nilsen R (2010) Spatiotemporal distribution of coastal and oceanic Atlantic cod Gadus morhua sub-groups after escape from a farm. Aquacult Environ Interact 1:11–19 Valle C, Bayle-Sempere JT, Dempster T, Sanchez-Jerez P, Gimenez-Casualdero F (2007) Temporal variability of wild fish assemblages associated with a sea-cage fish farm in the southwestern Mediterranean Sea. Est Coast Shelf Sci 72:299–307 Vita R, Marín A, Madrid JA, Jiménez-Brinquis B, Cesar A, Marín-Guirao L (2004) Effects of wild fishes on waste exportation from a Mediterranean fish farm. Mar Ecol Prog Ser 277:253–261

www.preventescape.eu

159

4.6. Recommendations for identification of escapees and conclusions from risk assessment analysis Sanchez-Jerez P, Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Black K, Somarakis S, Ladoukakis M, Haroun R, Borg J & Tim Dempster T (2013) Recommendations for identification of escapes and conclusions from risk assessment analysis. In: PREVENT ESCAPE Project Compendium. Chapter 4.6. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Pablo Sanchez-Jerez1, Pablo Arechavala-Lopez1, Ingebrigh Uglem2, Damian FernandezJover1, Kenny Black3, Stelios Somarakis4, Manolis Ladoukakis5, Ricardo Haroun6, Joseph Borg7, Tim Dempster8 University of Alicante, Spain Norwegian Institute of Nature Research, Norway 3 Scottish Association of Marine Science, Scotland 4 Hellenic Centre of Marine Research, Crete, Greece 5 University of Crete, Crete, Greece 6 University of Las Palmas de Gran Canarias, Spain 7 University of Malta, Malta 8 SINTEF Fisheries & Aquaculture, Norway 1 2

Introduction Aquaculture production is likely to increase and with it, the potential for escapes of farmed stock. The increasing demand for fishery products and the technical and commercial opportunities now available will drive increased aquaculture production, highlighting the need for clear policies to minimize escapes and their impacts. Average annual per capita supply of food fish from aquaculture for human consumption has increased by ten times, from 0.7 kg in 1970 to 7.8 kg in 2008, at an average rate of 6.6% per year. Seaweed and freshwater fish account for the main production at world level, but marine aquaculture is increasing its importance, especially in Europe. Aquaculture using seawater (in the sea and also in ponds) accounts for 32.3% of world aquaculture production by quantity and 30.7% by value, with a total production of marine fishes of 1.8 million tonnes in 2008 (FAO 2010). In Europe, the main products from aquaculture are fish and molluscs. The species contributing to the largest production volumes in 2011 include Atlantic salmon (1 009 331 t), Atlantic cod (20 923 t; Norwegian Directorate of Fisheries www.fiskeridir.no), seabream (96 419 t), seabass (57 004 t) and turbot (Scophthalmus maximus; 9 168 t; APROMAR 2011). Furthermore, farmers are presently diversifying into new species such as meagre, red porgy (Pagrus pagrus), blackspot seabream (Pagellus bogaraveo), and greater amberjack (Seriola dumerili).

160

The majority of European finfish aquaculture is, and will continue to be, located in sheltered areas within 3 km of shore in marine cages. For most species this introduces a risk of escape and subsequent consequences for wild fish populations. Empirical evidence demonstrates that: 1) escapes are difficult to prevent completely; and 2) open net pen systems appear to carry the greatest risk of fish escaping. Therefore, it is important to better understand the environmental impact of escapes, to have good methods for distinguishing escaped from wild fish, and to develop effective mitigation actions. Risk assessment is a process that provides a flexible framework within which the risks of adverse consequences resulting from a course of action can be evaluated in a systematic science-based manner. It permits a defendable decision to be made on whether a particular risk is acceptable or not, and the means to evaluate ways to reduce a risk from an unacceptable level to one that is acceptable. For example, risk assessment could be an estimation of the likelihood of the occurrence of genetic harm becoming realized following exposure to a genetic hazard. After an escape or release from a culture system, direct genetic harm may flow from the cultured stock interbreeding with reproductively compatible populations in the receiving ecosystem, and could include loss of adaptation in natural populations, introgression of new genetic material into speciesâ&#x20AC;&#x2122; gene pools and, in the extreme case, loss of locally adapted populations (Hallerman 2008).

Objective Here, we summarise the main results obtained from research in the Prevent Escape project on methods to effectively detect escapees in wild populations and determining their survival and distribution in wild populations post-escape. Firstly, we detail a cost-benefit analysis following a Delphi approach, to select the best method to detect escapees under various circumstances. Secondly, we conduct a risk analysis, both genetic and environmental, based on the information obtained from the Prevent Escape projectâ&#x20AC;&#x2122;s experiments and the existing literature. Finally, we provide recommendations related to methods to trace escapes and mitigate their effects in wild populations.

www.preventescape.eu

161

Recommended

indicators for the identification of escapees

A Delphi method was used to select a series of indicators for identifying escapees. This method is based on the opinion of a group of experts determined by a series of intensive questionnaires (Linstone et al. 2002). There are three major points of the Delphi exercise that underline the objectivity of this method (Dalkey 1969): (1) the scientific views of the team members are given by filling official questionnaires (maintaining anonymity of the participants); (2) the questions for the experts are posed during one or more rounds, during which the experts are informed of the general results of the previous round (feedback control); and (3) the statistical analyses determine the general response of the specialist team. In this way, the opinion of the group is an aggregation of individual opinions and each of them is represented in the final result. An expert panel was asked to provide their judgement on the suitability of the indicators used to differentiate farmed and wild fish, based on the cost and benefit involved in the use of a single or a group of indicators. 1. Cost-Benefit (which indicator can give most information with least cost): COD

SEABASS

SEABREAM

Morphometry

Condition index

External appearance

0.93

Morphometry

0.93

Morphometry

0.96

Condition index

0.88

Fatty acid profiles

0.89

Scale features

0.96

Scale features

0.80

Trace elements in scales

0.86

External appearance

0.92

Fatty acid profiles

0.78

Genetic methods

0.81

Fatty acid profiles

0.89

Trace elements in scales

0.64

Scale features

0.80

Genetic methods

0.86

2. Best Indicator for rapid assessment after an escape event:

162

COD

SEABASS

SEABREAM

Faster

morphometry, external appearance

condition index, morphometry

Fast

otolith and scale features, fin erosion

otolith and scale features, fin erosion and fatty acid profiles

3. Duration of the usefulness of an indicator in detecting escapees after an escape: COD

SEABASS

SEABREAM

Long term

Genetic methods and Trace element in otoliths

Medium term

Trace element in scales

4. Best Indicator to identify a single individual as escaped or wild: COD

SEABASS

SEABREAM

Highly recommended

Condition index

Genetic methods

Scale morphology

Recommended

External appearance and fish erosion

Morphometry

Genetic methods and morphometry

5. Best Indicator to identify the original farm an escapee came from: COD

SEABASS

SEABREAM

Highly recommended

Genetic methods

Recommended

Fatty acid profiles and morphometry

Analysis of trace elements Analysis of trace elements of otoliths and scales of otoliths and scales

6. Sector where technique can be applied: COD

SEABASS & SEABREAM

Fisheries management

otolith features, fin erosion, fatty acid profile and trace element in scales

morphometry, external appearance, condition index, fin erosion and fatty acid profiles

Environmental management

scale features and otolith features, fin erosion, fatty acid profiles and trace element in scales

morphometry, external appearance, condition index, fin erosion, scale features and fatty acid profiles

Farmer

morphometry, external appearance and scale feature

condition index, morphometry , external appearance, fin erosion

Seller and consumers

external appearance and morphometry and condition index

condition index, external appearance and fin erosion

www.preventescape.eu

163

Summary

of indicators

Summary of recommendation on the use of indicators for identifying escapees. Navy indicates most suitable indicator, light blue indicates medium suitability and white indicates least suitable. COD External Morphometry appearance

Condition index

Otolith features

Scale features

Fin erosion

Fatty acid profiles

Trace element scales

Trace element otoliths

Genetic methods

Fin erosion

Fatty acid profiles

Trace element scales

Trace element otoliths

Genetic methods

Cost-Benefit Quick response Temporal persistence Fisheries management Sellers & Consumers Farmers Environmental Management Identification single individual Original farm stock

SEABASS External Morphometry appearance Cost-Benefit Quick response Temporal persistence Fisheries management Sellers & Consumers Farmers Environmental Management Identification single individual Original farm stock

164

Condition index

Otolith features

Scale features

SEABREAM External Morphometry appearance

Condition index

Otolith features

Scale features

Fin erosion

Fatty acid profiles

Trace element scales

Trace element otoliths

Genetic methods

Cost-Benefit Quick response Temporal persistence Fisheries management Sellers & Consumers Farmers Environmental Management Identification single individual Original farm stock

Risk assessment of escapes from aquaculture Risk analysis has been widely applied in many fields, including aquaculture. It can help understand the physical and biological impacts that are related to safety and public health, pathogens, genetic issues, exotic organisms, environment, introduced marine species and financial risk (Bondad-Reantaso et al. 2008). Escapees can entail environmental and genetic effects for marine populations and ecosystems. Because of the intrinsic complexity of biological and ecological interaction, managers must estimate their potential ecological (Nash et al. 2008) and genetic risks (Hallerman 2008). Risk was evaluated in three dimensions. First, it is necessary to define the severity of the change in the affected ecosystem or species, the geographical extent of the change and the temporal duration. Severity can be ranked as catastrophic, high, moderate, low or negligible (ICES 2006). Second, based on previous experience, it is necessary to assess scientific knowledge, and using expert judgment, the probability of the risk being expressed (from high to negligible). Third, the assignment of probabilities to particular risks is a critical part of the risk analysis process; there will inevitably be a degree of imprecision and uncertainty in the final assigned probability (valuated from high to low uncertainty).

www.preventescape.eu

165

With respect to genetic risk assessment, the logical model steps below are based on those used by ICES (2006): 1) Fish farms are established in coastal waters. 2) There are phenotypic differences between the wild and cultured populations. 3) These phenotypic differences arise primarily for genetic rather than environmental reasons. 4) The primary route for genetic interaction (interbreeding) between cultured and wild fish is through escapes of fish from cages. 5) Cultured fish escape from cages. 6) Cultured fish interbreed with wild populations. 7) The progeny of this interbreeding (hybrids) show reduced fitness. 8) Sufficient gene flow exists to affect survival rates of fish in individual fisheries management units, i.e. the population structure of wild fish is such that the rate of interbreeding is sufficient to affect population fitness, at the population or meta-population levels. 9) Genetic interaction causes decline in local, evolutionarily significant units (populations), i.e. genetic interaction between wild and populations of escaped cultured fish causes significant declines in survival in wild populations. 10) Gene flow is pervasive and persistent enough to affect fitness at the level of species or meta-population, i.e. escapes of cultured fish cause significant decreases in wild/feral stocks. For environmental risk, the logic model steps are as follows: 1. Fish farms are established in coastal waters. 2. There are phenotypic differences between the wild and cultured fish populations. 3. Cultured fish escape from cages. 4. Cultured fish survive for a sufficient period to entail ecosystem effects. 5.1. Escapees prey on conspecifics or other species to such a degree that potential prey species are negatively affected. 5.2 Escapees are able to successfully compete with conspecifics or other species over limited resources like food, habitats and mates. 5.3. Escaped fish spread diseases and parasites that may affect wild organisms negatively.

166

For cod, the final rating of the revised risk evaluation regarding genetic impacts is the same as for the previous rating (ICES 2006), i.e. low intensity and probability and high uncertainty. The assessments of other ecological risks indicate that the uncertainty in general is high, while the probability is low for escapees to be detrimental predators or for escapees to spread diseases. A rapid and considerable up-scaling of the cod farming industry was assumed to be highly probably when the previous risk evaluation concerning genetic effects was developed. It is now clear that this did not happen. The cod farming industry has contracted substantially since 2008, and, at present, it is uncertain if the cod farming industry will continue. Hence, the intensity of potential impacts is evaluated as low and the extent to which cod farming may impose severe negative ecological impacts is now considerable less than was anticipated in the previous risk evaluation. Paradoxically, escape of farmed cod through spawning in cages may have a positive ecological impact by increasing wild cod numbers, particularly if local adaptations are of minor importance and fitness predominantly is related to phenotypic plasticity. Increased recruitment may, however, lead to greater numbers of smaller cod with lower growth rates due to density dependent competition. Nevertheless, the many environmental conditions in which farmedâ&#x20AC;&#x201C; wild hybrid fishes may be at a disadvantage in the wild, could mean that potential benefits of farmedâ&#x20AC;&#x201C;wild gene flow will be outweighed by its costs in many natural situations, perhaps most notably in the long term (Hindar et al. 1991, Naylor et al. 2005, Hutchings and Fraser 2008). The final rating of the revised risk evaluation for seabream and seabass is similar to the previous rating (ICES WGEIM 2006). For seabass, our evaluation indicates medium uncertainty in spite of high uncertainty regarding the reduction of fitness of hybrids. For seabream, the revised risk evaluation is the same as for the previous rating (ICES WGEIM 2006), i.e. low intensity and probability and medium uncertainty. Aquaculture of seabass and seabream is relatively mature and many escapes have been happening from the early 1990s, without serious evidence for deleterious effects on wild populations to date. However, the general lack of evidence for or against deleterious effects could provide an illusion of low risk intensity. The lack of evidence could be purely an indication of uncertainty and not the absence of effects but, because of the research effort carried out during the Prevent Escape project, we can attribute a medium uncertainty. The relative magnitude of the impact on wild populations will be related to the proportion of escapes in relation to the number of wild individuals. Because of competition for trophic resources and use of the same habitat (Arechavala et al. 2012), mitigation actions are important to reduce the abundance of escapees in the wild. It has been reported that escape seabream are a substantial proportion of some local fishery catches (Izquierdo-Gomez et al. 2011) Environmental risk assessment of seabass and seabream can be estimated as low intensity, low probability and high uncertainty. Competition for resources or predation on wild populations will be highly dependent on the spatial concentration of fish farms, magnitude of escapes events and mortality due to predation and fisheries. Because Mediterranean fish stocks are overexploited, and fisheries captures of seabass and seabream are very low compared with aquaculture production, the magnitude of the environmental impact may increase progressively over time. Although the landings of seabream from fisheries in Mediterranean and Atlantic ports is relatively constant year to year (between 5000 â&#x20AC;&#x201C; 8000 t yr-1; FAO www. globefish.org), aquaculture produces 95% of total production, with 139 925 t in 2010. Seabass production is similar, with 118 931 t grown in 2010 (FEAP 2010). Seabass fisheries land

www.preventescape.eu

167

8000 â&#x20AC;&#x201C; 12000 t yr-1, representing only 10% of total annual production. Therefore, wild populations are highly likely to be significantly smaller than farmed populations, especially in areas where farms are concentrated. Predation on conspecifics or other fish species will be more likely for escaped seabass. Juveniles feed on invertebrates, taking increasingly more fish with age and adults are piscivorous (trophic level = 3.8; www.fishbase.org). Seabream are found in seagrass beds and sandy bottoms as well as in the surf zone, commonly to depths of about 30 m, but adults may occur to 150 m depth, being a sedentary fish, either solitary or in small aggregations. It is mainly carnivorous, feeding on shellfish, including mussels and oysters (trophic level = 3.3; www.fishbase.org). If predation of seabream on other fish is unlikely, the main problem of seabream escape will be competition for food in natural habitats, such as seagrass meadows. This problem will depend on the carrying capacity of natural ecosystems. Problems may arise after large escapes in areas where wild populations are reaching the natural carrying capacity of the ecosystem.

Conclusions Management of escapes in wild ecosystem should be a priority in areas where aquaculture in coastal fish farms is developed extensively and escapes might produce harmful effects on wild populations. This is especially important if the cultured species is considered exotic to the area. When farmed species are native, the ecological risks may be less dramatic, but are still potentially serious. If a farmed population genetically diverges from a wild population (e.g. through selective breeding programs), interbreeding can compromise the genetic fitness and integrity of the wild population. This is particularly true when the wild population has already been reduced to low abundance. To better assess and ameliorate these risks, it is important to account for both the potential for fish to escape and the ensuing ecological effects. Organized monitoring programs are required to detect escapes around points of escape, in preferred natural habitats, areas of conservation concern (e.g. marine protected areas) and in fisheries catches. Based on the results of the Delphi study, we strongly recommended the use of a combination of indicators, depending on the objectives of management. Escapes can affect natural populations by competition. Competition will be a strongly densitydependent process linked to population structure. Experiments on fish competition (Post et al. 1999) have demonstrated that growth of the larger fish classes is density-dependent and driven primarily by exploitative competition. Future extensive research is needed to define more accurately the ecological risk of seabass and seabream escapes. The uncertainty for other ecological impacts to occur is assessed as being high, but the overall lack of knowledge regarding effects of escapees in the wild greatly reduces the certainty of these risk assessments. In this context, it is important to emphasize that the present lack of evidence for effects is primarily a consequence of a lack of adequate studies. Lack of evidence is an indication of high uncertainty, rather than an absence of effects, emphasizing the need to reduce this uncertainty through targeted studies. In addition, evaluation of potential effects at the species or meta-population levels may disguise elevated probabilities of deleterious effects for small, local stocks in areas where fish farming is particularly intense.

168

References

cited

APROMAR (2011) La acuicultura marina en España 2011. 77 pp (www.apromar.es) Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Bayle-Sempere JT, Sanchez-Jerez P (2012) Post-escape dispersion of farmed seabream (Sparus aurata L.) and recaptures by local fisheries in the Western Mediterranean Sea. Fish Res (in press) Bondad-Reantaso, MG, Arthur JR, Subasinghe RP (2008) Understanding and applying risk analysis in aquaculture. FAO Fisheries and Aquaculture Technical Paper. No. 519. Rome, FAO. 304 p Dalkey NC (1969) The Delphi method: an experimental study of group opinion. Prepared for United States Air Force project RAND FAO (2010) The state of world fisheries and aquaculture 2010. FAO Fisheries and Aquaculture Department, Rome. 218 pp Hallerman, E (2008) Application of risk analysis to genetic issues in aquaculture. In: Understanding and applying risk analysis in aquaculture. Ed: Bondad-Reantaso MG, Arthur JR, Subasinghe RP. FAO Fisheries and Aquaculture Technical Paper. No. 519. Rome, FAO. 47-66 pp. Hindar K, Ryman N, Utter F (1991) Genetic effects of cultured fish on natural fish populations. Can J Fish Aquat Sci 48: 945-957 Hutchings JA, Fraser DJ (2008) The nature of fisheries- and farming-induced evolution. Mol Ecol 17: 294-313 ICES (2006) Report of the Working Group on Environmental Interactions of Mariculture (WGEIM), 24-28 April 2006, Narragansett, Rhode Island, USA. ICES CM 2006/MCC:03. 195 pp Izquierdo-Gómez D, Arechavala-Lopez P, Bayle-Sempere JT, Sanchez-Jerez P, Valle C (2011) Captures of seabream (Sparus aurata) escapes form fish cages on artisanal fisheries at the Southeast of Spain. Proceedings of Aquaculture Europe Conference (EAS, Rhodes, Greece, 1821 October 2011) Linstone H, Turoff M, Helmer O (2002) The Delphi Method: Techniques and applications. Adison-Wesley Naylor R, Hindar K, Fleming IA, Goldburg R, Williams S, Volpe J, Whoriskey F, Eagle J, Kelso D, Mangel M (2005) Fugitive salmon: assessing the risks of escaped fish from net-pen aquaculture. BioScience 55: 427-437 Post JR, Parkinson EA, Johnston, NT (1999) Density-dependent processes in structured fish populations: interaction strengths in whole-lake experiments. Ecol Monogr 69:155–175 www.preventescape.eu

169

170

seeks to reduce the number of fish that escape from European aquaculture through research to improve fish farming techniques and technologies.

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

5.1. A novel form of escape of fish from sea cages: the problem of ‘escape through spawning’ Cite this article as: Somarakis S, Uglem I, Dempster T (2013) A novel form of escape of fish from sea cages: the problem of ‘escape through spawning’. In: PREVENT ESCAPE Project Compendium. Chapter 5.1. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Stelios Somarakis1, Ingebrigt Uglem2 & Tim Dempster3 Hellenic Centre for Marine Research, Crete, Greece, Norwegian Institute for Nature Research, Norway, 3 SINTEF Fisheries & Aquaculture, Norway 1 2

www.preventescape.eu

171

Fish farming in sea cages is increasing worldwide, driving an increase in both the volumes of fish produced and the numbers of fish species being cultured. During the last decade, farming of species that may reproduce within sea cages has become more common. Examples of such species within European aquaculture are Atlantic cod (Jørstad et al. 2008) and seabream (Dimitriou et al. 2007). However, knowledge of the extent and ecological effects of reproduction of farmed fish within commercial sea cages is sparse. The lack of knowledge on the extent and consequences of egg escape in commercial culture restricts our ability to predict the need to implement mitigative actions aimed at reducing escape of fertilised eggs from sea cages. We require a greater understanding of: The lack of knowledge on the extent and consequences of egg escape in commercial culture restricts our ability to predict the need to implement mitigative actions aimed at reducing escape of fertilised eggs from sea cages. We require a greater understanding of • how frequently spawning occurs within sea cages; • how many eggs escape and survive; • whether wild and farmed fish spawn at the same time; • where escaped eggs and larvae disperse to; and • if escaped eggs and larvae mix with those of wild conspecifics. The EGG Escape WP aimed to build knowledge in these areas using a major ‘industrial’ species, seabream (Sparus aurata), a rapidly ‘emerging’ species, Atlantic cod (Gadus morhua), and a relatively new ‘emerging’ species, meagre (Argyrosomus regius). Answers to these fundamental questions will be central for evaluating the need for mitigation through such measures as inhibiting spawning, recommendations for feeding strategies to reduce the quality of spawned eggs so that egg quality is poor, developing technologies to stop eggs from escaping or recommendations for siting of farms to areas that are distant from spawning areas of wild fish.

172

Escape

of cod through spawning

In Atlantic cod farms, some fish mature during the first year of culture, while the majority are believed to mature during the second year of culture. This means that almost the entire culture stock in any particular farm has the potential to spawn in sea cages. Efforts to prevent maturation, mainly through manipulation of the light regime in sea cages, have so far been unsuccessful. Although timing of maturation can be changed, it is difficult to inhibit maturation completely (e.g. Hansen et al. 2001, Taranger et al. 2006). The use of hybridization, sterilization and polyploidy to counter this problem are possibilities, but problems such as initially higher mortality, greater fingerling costs, poorer growth and consumer acceptance need to be resolved before they are taken up by industry (Triantafyllidis et al. 2007). Consequently, farmed cod are capable of producing viable eggs and larvae within sea cages, which subsequently mix with larvae from wild cod in the areas around cod farms and survive until one year of age (Jørstad et al. 2008, van der Meeren and Jørstad 2009). Thus, cod farming has the potential to result in unfavourable genetic changes in wild cod populations in the same way as is found for Atlantic salmon (Hindar et al. 2006), but not only through escape of adult fish. Jørstad et al. (2008) demonstrated that Atlantic cod held within an experimental sea cage are capable of spawning eggs that then dispersed into a nearby fjord system. In the same study, larvae from genetically marked farmed cod were found in plankton net samples up to 8 km away from an experimental farm during the natural spawning season of cod. Furthermore, in the proximity of this farm, 25% of the cod larvae in plankton samples were determined by genetic analyses to have originated from the 1000 farmed cod. This indicates that if spawning occurs within cod farms where numbers of animals are far greater, the contribution of ‘escaped’ larvae to cod recruitment within fjords may be substantial. However, most current farms are located at more exposed locations than the experimental farm studied by Jørstad et al. (2008). The escape of eggs will likely be a persistent occurrence in cod culture. The escape of large quantities of eggs from caged cod may be problematic as: 1) coastal cod populations in parts of Norway have decreased considerably over several decades, principally due to overfishing; 2) coastal cod have a high fidelity to specific spawning grounds (e.g. Wright et al. 2006); and 3) sea cage cod farms are often located within short distances of known wild cod spawning grounds (Uglem et al. 2008). Recent research also suggests that cod eggs may be entrained in the vicinity of the spawning grounds long after spawning (Knutsen et al. 2007). Therefore, it is possible for larvae from escaped cod eggs to experience favourable conditions for survival and recruitment to coastal cod stocks. The extent and effect of spawning within commercial sea cages is largely unknown, even though it is unquestionable that farmed cod have the potential for producing large numbers of “escaped” egg and larvae. The effects of spawning in cages will depend not only on the numbers of egg and larvae which escape, but also on their quality and survival, which in turn is a result of both innate and environmental factors, such as broodstock nutrition, timing of hatching and egg/larval drift following spawning. Further, the ecological effect of this type of escape will depend on the degree to which local adaptations exist in wild cod stocks.

www.preventescape.eu

173

Escape

of seabream through spawning

In the Mediterranean region, information about spawning by fish kept in sea cages is limited. In Greece, the largest EU producer of seabream, a spectacular increase in both the number of fish farms and their production capacity took place over the past decade, accompanied by a substantial decrease in the price of seabream. This industrial development led to structural and functional changes in the rearing process. Farming durations increased from just 12 to 18 months before 1995 to durations of up to 40 months after 1999 (Dimitriou et al. 2007). Gilthead seabream is a protandrous hermaphrodite species and the increased farming duration has resulted in the production of fish of a size that can reach the stage of sex inversion and female sexual maturation, normally observed at the age of 2 – 3 years in the wild (Mylonas et al. 2011). The aforementioned changes in rearing processes have resulted in the presence of large gilthead seabream individuals (larger than 500 g) in cages during the normal reproductive period of their wild counterparts (November – March: Mylonas et al. 2011). There is evidence that sex inversion and the production of both male and female gametes occur within cages under the present industrial rearing pattern (Dimitriou et al. 2007). A doubling of the population of wild seabream within the Messolonghi lagoon in Greece, based on standardised commercial fishing trap catch returns, correlates with the advent of farming sea-bream to large sizes in the region. Spawning within sea cages is suspected to have led to greater recruitment to wild seabream stocks (Dimitriou et al. 2007). Ecological and economic consequences of this population shift have ensued. While more wild sea-bream are now available to the fishery, they are of much smaller mean size resulting in an overall lower economic return to local fishers. The quality of the eggs produced by caged fish and the eventual survival of the larvae is an important yet unknown feature of egg escapes. Cultured cod are known to have different body fat content and fatty acid distributions than their wild counterparts and it is possible that the quality of eggs produced by cultured fish differ from wild fish (Salze et al. 2005). In contrast to Atlantic cod, various fatty acid groups have not been found to differ significantly between wild and cultured seabream, mainly due to the high variability reported in the literature (Grigorakis 2007). However, it has been well demonstrated that cultured fish have higher levels of fat deposits (both perivisceral and peritoneal) and linoleic acid (18:2n-6) when compared to wild fish, the latter being richer in arachidonic acid (Grigorakis 2007, Saglik et al. 2003). How this effects egg production and quality is unknown.

Meagre –

an emerging species of interest

A species of emerging interest for sea cage aquaculture in the Mediterranean, the meagre (Argyrosomus regius), also has the potential to reproduce during culture in sea cages, as it is typically grown to a commercial size of 40 – 60 cm (1 – 3 kg). If spawning occurs within cages, recruitment of juveniles to wild populations may result. Meagre is uncommon in many areas of the Mediterranean Sea, thus large increases in their abundance in specific coastal areas could lead to ecological changes.

174

Objectives The main objective of the ‘EGG Escape’ work package was to assess the extent and ecological importance of escape through spawning in sea cages and suggest possible mitigation actions. The work consisted of both an extensive field program and a modelling component. In the field program, farmed fish were randomly collected at selected Atlantic cod (Gadus morhua), seabream (Sparus aurata) and meagre (Argyrosomus regius) farms in Norway, Greece and Spain, respectively. The three species were selected to include: a) a major industrial species (seabream) with increasing potential to spawn within cages; b) a species (cod) initially predicted to make a transition from an emerging to a major industrial species within a few years, which is known to spawn within cages; and c) a new emerging species (meagre) that was suspected to spawn within cages. The specific objectives of WP5 were to: • evaluate the extent and timing of spawning within sea cage fish farms at an industrywide scale; • assess the quality of released eggs; • assess the survival and distribution of escaped eggs; and • evaluate the need for implementation of mitigative strategies for reducing or preventing escape of eggs

www.preventescape.eu

175

References

cited

Dimitriou E, Katselis G, Moutopoulos DK, Akovitiotis C, Koutsikopoulos C (2007) Possible influence of reared gilthead seabream (Sparus aurata, L.) on wild stocks in the area of the Messolonghi lagoon (Ionian Sea, Greece). Aquac Res 38:398-408 Grigorakis K (2007) Compositional and organoleptic quality of farmed and wild gilthead seabream (Sparus aurata) and seabass (Dicentrarchus labrax) and factors affecting it: A review. Aquaculture 272:55-75 Hansen T, Karlsen Ø, Taranger GL, Hemre GI, Holm JC, Kjesbu OS (2001) Growth, gonadal development and spawning time of Atlantic cod (Gadus morhua) reared under different photoperiods. Aquaculture 203:51-67 Hindar K, Fleming IA, McGinnity P, Diserud O (2006) The genetic and ecological effects of salmon farming on wild salmon: modelling from experimental results. ICES J Mar Sci 63, 1234-1247 Jørstad KE, van der Meeren T, Paulsen OI, Thomsen T, Thorsen T, Svåsand T (2008) “Escapes" of eggs from farmed cod spawning in net pens: Recruitment to wild stocks. Rev Fish Sci 16:285295 Knutsen H, Moland OE, Ciannelli L, Heiberg ES, Knutsen JA, Simonsen JH, Skreslet S, Stenseth NC (2007) Egg distribution, bottom topography and small scale cod population structure in a coastal marine system. Mar Ecol Prog Ser 333:249-255 Mylonas CC, Zohar Y, Pankhurst N, Kagawa H (2011) Reproduction and broodstock management. In: Pavlidis MA, Mylonas CC (eds) Sparidae: Biology and Aquaculture of gilthead seabream and other species. Blackwell Publishing, pp. 95-131 Saglik S, Alpaslan M, Gezlin T, Cetlinturk K, Tekinay A, Guven KC (2003) Fatty acid composition of wild and cultivated gilthead seabream (Sparus aurata) and seabass (Dicentrarchus labrax). Eur J Lipid Sci Technol 105:104-107 Salze G, Tocher DR, Roy WJ, Robertson DA (2005) Egg quality determinants in cod (Gadus morhua L.): egg performance and lipids in eggs from farmed and wild broodstock. Aquac Res 36:1488-1499 Taranger GL, Aardal L, Hansen T, Kjesbu OS (2006) Continuous light delays sexual maturation and increases growth of Atlantic cod (Gadus morhua L.) in sea cages. ICES J Mar Sci 63:365-375 Triantafyllidis A, Karaiskou N, Bonhomme F, Colombo L, Crosstie D, Danancher D, GarcíaVázquez E, Gilbey J, Svåsand T, Verspoor E, Triantaphyllidis C (2007) Management options to reduce genetic impacts of aquaculture activities. In: Genetic Impact of Aquaculture Activities on Native Populations. GenImpact Compendium (http://genimpact.imr.no/compendium)

176

Uglem I, Bjørn P-A, Dale T, Kerwath S, Økland F, Nilsen R, Aas K, Fleming I, McKinley RS (2008) Movements and spatiotemporal distribution of escaped famed and local wild Atlantic cod (Gadus morhua L.) in a Norwegian fjord. Aquac Res 39:158-170 van der Meeren T, Jørstad KE (2009) Fanger torsk på vidvanke. Nytt fra havbruk 2009(2), 1 (In Norwegian) Wright PJ, Galley E, Gibb IM, Neat FC (2006) Fidelity of adult cod to spawning grounds in Scottish waters. Fish Res 77:148-158

www.preventescape.eu

177

5.2. Escape through (Sparus aurata)

spawning

seabream

Cite this article as: Somarakis S, Saapoglou C, Tsigenopoulos C, Pavlidis M (2013) Escape through spawning by seabream (Sparus aurata). In: PREVENT ESCAPE Project Compendium. Chapter 5.2. Commission of the European Communities, 7th Research Framework Program. www. preventescape.eu ISBN: 978-82-14-05565-8

authors: Stelios Somarakis1, Christina Saapoglou1,2, Costas Tsigenopoulos1 & Michalis Pavlidis2 1 2

Hellenic Centre for Marine Research, Crete, Greece, Department of Biology, University of Crete, Greece

Introduction In Greece, the largest EU producer of seabream, both the number of fish farms and their production capacity increase spectacularly over the past decade, accompanied by a substantial decrease in the price of seabream. This industrial development has led to structural and functional changes in rearing processes. Farming durations increased from just 12 to 18 months before 1995 to durations of up to 40 months after 1999 (Dimitriou et al. 2007). The increased farming duration of gilthead seabream (Sparus aurata), a protandrous hermaphrodite, has resulted in the production of fish large enough to reach the stage of sex inversion and female sexual maturation, normally observed at an age of 2 â&#x20AC;&#x201C; 3 years in the wild (Mylonas et al. 2011). Changes in rearing processes have resulted in the presence of large gilthead seabream individuals (larger than 500 g) in cages during the normal reproductive period of their wild counterparts (November â&#x20AC;&#x201C; March: Mylonas et al. 2011). There is evidence that sex inversion and the production of both male and female gametes occur within cages under the present industrial rearing pattern (Dimitriou et al. 2007). However, in the Mediterranean region, information about spawning by fish kept in sea cages is sparse.

178

Objectives: The objectives of our work on gilthead seabream were to: • investigate whether large fish (i.e. fish beyond the size of sex reversal and onset of female maturation) produce eggs in sea cages, using the Greek seabream industry as a case study; • evaluate the quantity and quality of eggs released in gilthead seabream broodstock held in sea cages; and • assess whether eggs produced within sea cages disperse and survive in surrounding coastal waters.

Evaluation

of sexual maturation and egg release in seabream

reared in sea cages

Gilthead seabream have a complex reproductive biology that includes sex reversal, indeterminate annual fecundity and an extended spawning period. In general, sparid fishes are multiple spawners with asynchronous oocyte development (Mylonas et al. 2011). To obtain and analyse seabream gonads and estimate daily and annual egg production, we sampled fish each month from five farms in Greece (Table 5.2.1). A total of 1262 large gilthead seabream were sampled from November 2009 to April 2010. Farms 1 – 3 were located in the Ionian Sea, whereas farms 4 – 5 were located in the Aegean Sea. Mean fish weight ranged from 0.820 to 1.75 kg and the percentage of females varied among farms from 38 to 81% (Table 5.2.1). We removed and weighed gonads and took subsamples for subsequent histological analysis and fecundity measurements. The monthly development of the mean gonadosomatic index (gonad weight/fish weight or GSI) is illustrated in (Figure 5.2.1) for female fish. With the exception of farm 1, GSIs remained high up until February. Thereafter, gonadal regression occurred from February to April. In farm 1, the lowering of temperatures to below 13oC in January likely resulted in earlier gonadal regression. Changes in histological maturity stages of both males and females (e.g. males in Figure 5.2.2) matched closely with monthly variations in GSI. Histological scoring of ovaries included the stage of development of the most advanced oocytes, the presence and histological characteristics of postovulatory follicles (POFs) and the incidence/ prevalence of atresia (Hunter and Macewitz, 1985). The most prominent characteristic of females with yolked oocytes (spawning capable fish) was a relatively high prevalence of atresia (absorption) of yolked oocytes. Alpha-stage atresia involves the resorbing of yolk and chorion (Figure 5.2.3a) which we quantified in all spawning

www.preventescape.eu

179

capable females by measuring the percentage of oocytes affected. We measured batch fecundity (number of eggs produced in a spawning event) by counting the number of hydrated oocytes (Figure 5.2.3b) in pre-weighed subsamples from the ovary (Hunter et al. 1985). After ovulation and egg release, post-ovulatory follicles (POFs) appear in the ovary (Figure 5.2.3c) which soon start to degenerate (Hunter and Macewitz 1985). POFs are an unambiguous marker of recent spawning and can readily be identified in the ovaries as long as their lumen is still visible (Hunter and Macewitz 1985). The POFs we observed in seabream ovaries were in a moderate degree of degeneration and always had a lumen. Given the temperature regimes in the farms sampled and information from POF degeneration experiments (Alday et al. 2008), we considered them to represent a single night’s (previous night’s) spawning. We collected females with hydrated oocytes and/or POFs, (i.e. actively spawning females) from December to February (Figure 5.2.4), depending on the farm, and used them to define the approximate duration of the spawning period (SP), i.e. the number of months with incidence of actively spawning females (Table 5.2.2). We used the average fraction of females with hydrated ovaries and ovaries with POFs to estimate S, i.e. the fraction of females spawning each day (Table 5.2.2). Farm 4 was excluded from the analysis due to small sample sizes. We then derived mean relative batch fecundities in the four farms (Table 5.2.2) as the marginal means from the general linear model: log(RF) = Intercept + FARM + MONTH + ATRESIA, n = 37, p < 0.001, adj. r2: 0.846, where RF: relative batch fecundity (eggs kg-1) and ATRESIA: prevalence of alpha stage atresia (% of oocytes affected). Estimates of daily female-specific fecundity (number of eggs produced daily per kg of females in the cage, DFSF = S × RF) and annual female-specific fecundity (number of eggs produced annually per kg of females, AFSF: DFSF × SP; Table 5.2.2), decreased exponentially with sex ratio, R (Figure 5.2.5). Figure 5.2.1. Monthly development of the gonadosomatic index (GSI) of female seabreams in the five farms.

180

Figure 5.2.2. Microphotographs of seabream functional testes. (A) Developing. (B) and (C) Spawning. (D) Regressing. (E) and (F) Regenerating. Note the ovarian tissue in (E). Green arrows: spermatogonia. Red arrows: spermatozoa. SC: spermatocytes. Sp: spermatids. Scale bar: 0.1 mm in (A) and (B); 0.2 mm in C-F.

Figure 5.2.3. Key histological features of spawning capable female seabreams. (A) Ovary with yolked oocytes in alpha atretic stage (a). (B) Ovary with hydrated oocytes (H). (C). Post-ovulatory follicle. Scale bar: 0.2 mm in A-B; 0.1 mm in C.

www.preventescape.eu

181

Figure 5.2.4. Upper panel: Fraction of females with yolked oocytes that had hydrated oocytes. Lower panel: Fraction of females with yolked oocytes that had postovulatory follicles.

Number of fish sampled

Sex ratio

Average fish weight (kg)

farm 1

300

0.81

1.182

farm 2

300

0.61

1.435

farm 3

300

0.38

0.820

farm 4

120

0.63

1.410

farm 5

242

0.72

1.752

Table 5.2.1. Summarized information on seabream sampling. Sex ratio: number of females / total number of fish. Farms 1-3 were located in the Ionian Sea; farms 4-5, in the Aegean Sea.

farm

DFSF

AFSF

0.81

0.145

4000

579

17368

0.61

0.096

14256

1362

81740

0.38

0.186

16982

3158

284196

0.72

0.103

8110

833

49963

Table 5.2.2. Estimates of mean spawning fraction, S; mean relative batch fecundity, RF; sex ratio, R; daily female-specific fecundity (number of eggs produced daily per kg of female in the population, DFSF = S xRF, DFSF; duration of the spawning period in days approximated by the number of months with incidence of hydrated females and/or postovulatory follicles, SP, annual female-specific fecundity, (number of eggs produced annually per kg of female in the population, AFSF: DFSF x SP), AFSF.

182

Figure 5.2.5. Relationships between daily specific fecundity (DFSF: number of eggs produced daily per kg of female in the cage); annual specific fecundity (AFSF: number of eggs produced annually per kg of female in the cage) and sex ratio.

Monitoring

egg releases in seabream sea cages

To evaluate the quantity and quality of eggs released in gilthead seabream held in sea cages, we designed, manufactured and installed an egg collector in a seabream cage. We ran a study in a private fish farm (Karpasia) located in the island of Evia, Greece. It consisted of four phases: (a) design of an egg collector, (b) manufacturing of the collector, (c) installation on the net-pen sea cage, and (d) egg release monitoring. The egg collector was constructed in 2010 (Figure 5.2.6) and we performed a test experiment in January – February 2011. To ensure the best in situ managerial conditions for spawning, we ran the experiment with an ideal sex ratio (1:1), at a low stocking density and a number of fish similar to that used in commercial seabream hatcheries. Sixty fish at a sex ratio of 1:1 (30 females and 30 males of a mean body weight of 990 and 577 g, respectively) were placed in a net pen cage (diameter 40 m, depth 10 m) in the middle of January 2011. Fish were left undisturbed for two weeks after which the egg collector was placed in the cage (Figure 5.2.6d and Figure 5.2.7a). We monitored eggs (Figure 5.2.7b and c) on a daily basis for two consecutive weeks (11 – 25 February 2011). There was a small daily release of eggs (1250 – 3250 eggs per day) apart from two consecutive days where a larger amount of eggs were collected (14/02/11: 30 375 eggs; 15/02/11: 13 250 eggs). We also monitored eggs for one-week in the middle of March, but no eggs were collected. Following collection, we placed eggs in a 10 L bucket and transported them to land for quality evaluation. We then checked the quality and viability of floating eggs using a stereoscope. The measures of egg quality we used were: egg shape, number of oil droplets and symmetrical embryonic divisions. Egg viability was less than 5%.

www.preventescape.eu

183

Figure 5.2.6. (A) Phase 1. Construction of the upper part of the egg collector. Material used: non-toxic polyethylene; height of the construction: 1 m. (B) Completion of the construction. (C) Phase 2: construction of the egg collectorâ&#x20AC;&#x201C;net complex. Size of the net 12 mm; depth: 4 m. (D) Installation of the egg collector on the sea cage.

184

Figure 5.2.7. (A) Experimental net-pen cage provided with the manufactured egg collector. (B) Procedure for the collection of eggs â&#x20AC;&#x201C; phase 1. (C) Procedure for the collection of eggs â&#x20AC;&#x201C; phase 2.

www.preventescape.eu

185

Assessment

of egg dispersal and mortality around seabream

sea cages

To determine if eggs produced within sea cages disperse and survive in surrounding coastal waters, we carried out a fine-scale study of the distribution and abundance of sparid embryos around sea cages using vertical plankton tows. Sampling took place in the vicinity of a farm known to have large fish that could potentially spawn. We carried out the site intensive ichthyoplankton survey for three consecutive days in 20 – 22 January 2011. Sampling took place onboard the 26 m Research vessel “PHILIA”. On each day, vertical hauls of a standard WP2 sampler (200 microns mesh-sized net) were done at 19 fixed stations to collect fish eggs and larvae (Figure 5.2.8). The seabream farm in the area contained a cage with large (>1.5 kg) seabream and another experimental cage with large fish that was used for the egg collector experiment (see above). We sorted plankton samples onboard immediately after capture and all fish eggs potentially belonging to the family Sparidae (Divanach 1985) were staged and put individually in 96% alcohol for subsequent identification to the species level using molecular techniques. We then subjected eggs from the plankton survey as well as 40 eggs collected with the experimental egg collector (see above) to molecular analysis. We extracted total genomic DNA and amplified a region of cytochrome oxidase subunit I (COI; ~670 bp) mitochondrial gene via PCR using universal primers FishF2 and FishR1 (Ward et al. 2005). PCR products were purified and we performed a single stranded sequencing with primer FishF2. We also used additional sequences for Sparus aurata and Diplodus vulgaris to align the sequences obtained in the present study and to assist molecular species identification. We assigned staged sparid eggs to three age-classes: age-0 eggs (<1 day old), age-1 eggs (1-2 days old), and age-2 eggs (>2 days) based on a temperature–stage–age key devised from information in Divanach (1985), Polo et al. (1991) and Koumoundouros (1993). Finally, we estimated the weighted mean abundance of age-0, age-1 and age-2 eggs for each sampling day. We used station weighting factors that were proportional to their representative area. The difference in weighted mean abundance of age-1 eggs in sampling day 2 from age-0 eggs in sampling day 1 (the fraction age-1/age-0) was an estimate of daily eggs survival. Based on the molecular identifications, we identified all sparid eggs to species level (Table 5.2.3). The great majority belonged to Sparus aurata and Diplodus vulgaris. The later was a wild species, so we considered it useful to compare the distribution, abundance and mortality of its eggs with those of S. aurata. Most gilthead seabream eggs were collected on day 1, whereas on day 2 and 3 their abundance was very low. Most D. vulgaris eggs were also collected on day 1. Furthermore, with the exception of two eggs that belonged to D. vulgaris, the remainder (95%) eggs collected with the egg collector (see above) were Sparus aurata, implying that spawning takes place inside the cages.

186

The fraction (age-1 eggs on day 2)â&#x20AC;&#x201C;onâ&#x20AC;&#x201C;(age-0 eggs on day 1) (Table 5.2.4), expressed as a percentage, was 10% for S. aurata and 90% for D. vulgaris. Assuming that most eggs of both species had been spawned inside or very near to the cages (see below) and that no significant numbers of age-1 D. vulgaris eggs had been advected from outside the surveyed area, the survival of S. aurata eggs, most likely produced inside the cages, was very low compared to the wild population of D. vulgaris. Plots of mean standardized egg abundance against distance from the cages (Figure 5.2.9) clearly showed that the abundance of age-0 eggs for both S. aurata and D. vulgaris was high in the close vicinity of the farm whereas age-1 eggs were more abundant away from the cages. These findings suggest that the fish farm was a site of increased egg production of both (farmed) gilthead seabream and other wild populations of the same family that were attracted close to the cages. The eggs were subsequently dispersed in the adjacent coastal habitat by the prevailing currents in the area.

Species

day 1

day 2

day 3

eggs m

eggs m2

Sparus aurata

58.00

21.05

1.84

26.32

2.11

Diplodus vulgaris

58.00

22.63

73.68

10.00

31.58

6.32

Diplodus sargus

5.26

0.21

15.80

1.05

5.26

0.21

21.05

1.05

Boops boop

Table 5.2.3. Assignment of sparid eggs to species based on results of the molecular analyses. N = number of individuals collected in each day; % = per cent frequency of occurrence; eggs m2= average abundance in the 19 stations.

Sampling day 1

Age class

Sparus aurata

Diplodus vulgaris

age-0

3.22

6.92

age-1

1.86

2.99

age-2

1.94

age-0

0.17

1.33

age-1

0.31

6.21

age-2

age-0

0.99

0.39

age-1

1.70

age-2

Table 5.2.4. Weighted mean abundance (eggs m-2) of eggs per age class and sampling day. NA: no eggs were caught in this class. Note the difference between age-1 eggs on day 2 and age-0 eggs on day 1.

www.preventescape.eu

187

Figure 5.2.8. Map of the ichthyoplankton study area. Triangles indicate sampling stations. The location of the seabream farm is also indicated (red arrow).

Figure 5.2.9. Plots of mean standardized egg abundance (eggs m2) against distance from the farm (m). The abundance of eggs in each day was standardized by dividing the individual station abundances with the maximum abundance recorded during that day. Subsequently, for each station, we averaged the abundance of eggs for the three days of sampling. 188

Discussion The analysis of seabream gonads provides the first data showing that female seabream reach final maturation, ovulation and spawning in sea cages during the normal spawning period of the species. The fecundity estimates presented in Table 5.2.2 indicate that actual egg production is very low compared to commercial broodstock (Mylonas et al. 2011). Egg production seems to be down-regulated by atresia and decreases with sex ratio (i.e. with fish size in the cage). The molecular identifications showed that the eggs caught inside the cage by the seabream egg collector were 95% Sparus aurata, which confirms the findings of the reproductive study. Egg collections show that eggs are produced inside the cages. Therefore, the production of S. aurata eggs caught in the plankton survey was most likely due to the farmed fish inside the cages rather than wild fish or escapees, although this possibility cannot be ruled out. Day-to-day variability in egg production during the plankton survey was very high in Sparus aurata, which reflected our findings from the egg-collector experiment. This was also the case for Diplodus vulgaris which was attracted to and spawned in the close vicinity of the cages. Egg production was particularly high during the first day of the ichthyoplankton sampling and was much lower in the next two days, especially in S. aurata. As the owner of the fish farm informed us later, the fish in the cages were not fed at all for two days prior to the first day of plankton sampling, which may explain the high egg production recorded on day 1. S. aurata and D. vulgaris eggs showed similar patterns of distribution and dispersal around the sea cages, with most eggs being recorded next to the cages and thereafter dispersing in the surrounding coastal waters. However, the estimate of daily egg survival calculated for S. aurata was very low (10%) compared to D. vulgaris (90%). Although there is a high degree of uncertainty in field estimates of egg mortality, the large difference in mortality estimates for S. aurata and D. vulgaris might imply that the eggs originating from farmed seabream can be of inferior quality. This is further supported by the low survival rates recorded in the egg collector experiment.

Recommendations The results of this study demonstrate that seabream cultivated in sea cages beyond the size of sex reversal can reach female maturation, ovulation and release eggs during the normal spawning period of the species. However, egg production is very low, decreases considerably with sex ratio (i.e. with fish size in the cage) and varies considerably from day to day. In addition, the survival of fertilized eggs is likely to be low. These findings imply that the escape of eggs from seabream farms and subsequent ecological consequences might be low, depending on the size and intensity of farming within specific regions.

www.preventescape.eu

189

There are two cases in which the probability of producing more eggs or that larvae may recruit to the wild populations are increased: â&#x20AC;˘ sex ratio in cages is balanced (close to 1:1); and â&#x20AC;˘ farms are sited in areas where seabream can complete its life cycle, e.g. close to lagoons. Available information on gilthead seabream ecology is sparse, but indicates that it may be an estuary-dependent species. The use of a curtain-like egg collector, such as used in the present study, could prevent some dispersion of eggs away from the cages. A more precautionary mitigation measure would be to restrict the culture of large seabream (of sizes beyond that of sex reversal) in areas close to known nursery grounds of seabream, such as lagoons.

190

References

cited

Alday A, Uriarte A, Santos M, Martín I, Martinez de Murguia A, Motos L (2008) Degeneration of postovulatory follicles of the Bay of Biscay anchovy (Engraulis encrasicolus L.). Sci Mar 72:565-575 Dimitriou E, Katselis G, Moutopoulos DK, Akovitiotis C, Koutsikopoulos C (2007) Possible influence of reared gilthead seabream (Sparus aurata, L.) on wild stocks in the area of the Messolonghi lagoon (Ionian Sea, Greece). Aquac Res 38:398-408 Divanach P (1985) Contribution a la connaissance de la biologie et de l’ élevage de 6 sparides Méditerranéens: Sparus aurata, Diplodus sargus, Diplodus vulgaris, Diplodus annularis, Lithognathus mormyrus, Puntazzo puntazzo (Poissons téléostéens). PhD Dissertation, Université de Sciences et Techniques du Languedoc, Montpellier, France Hunter JR, Macewitz B (1985) Measurement of spawning frequency in multiple spawning fishes. In: Lasker R (ed) An Egg Production Method for Estimating Spawning Biomass of Pelagic Fish: Application to the Northern Anchovy, Engraulis mordax. NOAA Tech Rep NMFS 36:79-93 Hunter JR, Lo NCH, Leong RJH (1985) Batch fecundity in Multiple Spawning Fishes. In: Lasker R (ed) An Egg Production Method for Estimating Spawning Biomass of Pelagic Fish: Application to the Northern Anchovy, Engraulis mordax. NOAA Tech Rep NMFS 36:67-77 Koumoundouros G (1993) Biology of the development of the gilthead seabream (Sparus aurata), Linnaeus, 1758, Percoidea, Sparidae) in intensive culture. MS Dissertation, University of Crete, Greece Mylonas CC, Zohar Y, Pankhurst N, Kagawa H (2011) Reproduction and broodstock management. In: Pavlidis MA, Mylonas CC (eds) Sparidae: Biology and Aquaculture of gilthead seabream and other species. Blackwell Publishing, pp. 95-131 Polo A, Yúfera M, Pascual E (1991) Effects of temperature on egg and larval development of Sparus aurata L. Aquaculture 92:367-375 Ward RD, Zemlak TS, Innes BH, Last PR, Hebert PDN (2005) DNA barcoding Australia's fish species. Philos Trans R Soc Lond B 360:1847-1857

www.preventescape.eu

191

5.3. Extent

and

ecological

spawning in sea cages for

morhua L.)

importance

Atlantic cod (Gadus

Cite this article as: Uglem I, Knutsen Ø, Kjesbu OS, Hansen ØJ, Mork J, Bjørn PA, Varne R, Nilsen R, Ellingsen I, Dempster T (2012) Extent and ecological importance of spawning in sea cages for Atlantic cod (Gadus morhua L.). In: PREVENT ESCAPE Project Compendium. Chapter 5.3. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu

authors: Ingebrigt Uglem1, Øyvind Knutsen2, Olav Sigurd Kjesbu3, Øyvind J Hansen4, Jarle Mork5, Pål Arne Bjørn3,4, Rebekka Varne5, Rune Nilsen3,4, Ingrid Ellingsen2 & Tim Dempster6,7 Norwegian Institute of Nature Research, Tungasletta 2, 7485 Trondheim, Norway SINTEF Fisheries and Aquaculture, PO Box 4762 Sluppen, 7465 Trondheim, Norway 3 Institute of Marine Research, PO Box 1870 Nordnes, 5817 Bergen, Norway 4 NOFIMA Marin, Muninbakken 9-13, PO Box 6122, 9291 Tromsø, Norway 5 Trondheim Biological Station, Department of Biology, Norwegian University of Science and Technology, 7491 Trondheim, Norway 6 Sustainable Aquaculture Laboratory – Temperate and Tropical (SALTT), Department of Zoology, University of Melbourne, 3010 Victoria, Australia 7 Centre for Research-based Innovation in Aquaculture Technology (CREATE), SINTEF Fisheries and Aquaculture, PO Box 4762 Sluppen, 7465 Trondheim, Norway 1 2

Introduction As the demand for protein increases to support a growing global human population, the production of fish in sea cage aquaculture systems has expanded. While modern, industrial sea cage aquaculture had been built around salmonids, which do not spawn in sea cages, the culture of fish species that are marine, pelagic broadcast spawners and that may reproduce within sea cages has recently increased. These include Atlantic cod (Gadus morhua; Jørstad et al. 2008) and seabream (Pagrus major; Dimitriou et al. 2007). In Atlantic cod farms, some fish mature during the first year of culture, while a majority of fish mature during the second year of culture. Consequently, almost the entire culture stock in any particular farm has the potential to spawn in the sea cages. Larvae from genetically marked farmed cod that spawned in a sea cage have been found up to 8 km from farms during the natural spawning season of cod and survive to at least until one year of age (Jørstad et al. 2008, van der Meeren and Jørstad 2009). If spawning occurs within commercial cod farms where numbers of animals are high, the contribution of ‘escaped’ larvae to cod recruitment within a particular fjord may be substantial. The extent and effect of spawning within commercial sea cages is largely unknown, even though it is unquestionable that farmed cod have the potential for producing large numbers of “escaped” egg and larvae. The effects of spawning in sea cages will depend not only on the numbers of egg and larvae which escape, but also on their quality, dispersal and survival.

192

Study

species

The Atlantic cod is an epibenthic-pelagic species, which is found in a wide variety of habitats, from the shoreline down to the continental shelf. It is distributed along the North American coast, around Greenland and Iceland, and along the coasts of Europe from the Bay of Biscay to the Barents Sea. Individuals may grow up to 2 m in length and more than 90 kg, though this is extremely rare at present. In the commercial fisheries cod size typically ranges from approximately 2 kg up to around 20 kg. Cod are omnivorous, feeding at dawn or dusk on invertebrates and fish, including young cod. They spawn in batches once a year, usually during late winter and early spring, producing pelagic eggs. A 5 kg wild female may spawn around 2.5 million eggs during a spawning season. The most important stocks are the Norwegian Arctic stock in the Barents Sea and the Icelandic stock. The populations around Greenland and Newfoundland have declined dramatically, whereas the stock in the Barents Sea still is healthy. The Atlantic cod is an economically important species and it is marketed fresh, dried or salted, smoked and frozen.

Objective We examined the extent and ecological importance of escape through spawning in sea cages for Atlantic cod by: • evaluating the amount, frequency and timing of spawning in commercial cod culture; • analysing the quality of eggs released from farms using fatty acid profiles as proxy indicators; • modelling the distribution of eggs and larvae from a commercial cod culture site; and • assessing the post-escape survival of eggs through summarizing existing knowledge on survival rates of different life stages.

www.preventescape.eu

193

Methods The amount, frequency and timing of spawning of Atlantic cod in commercial cod culture was examined by sampling fish from seven farms distributed from western to northern Norway, prior to or during the spawning season of wild cod, from late 2009 to early 2010 and in February and March 2011 (Figure 5.3.1). Fish that had been in sea cages for both 1 and 2 years were sampled to determine sex ratio, reproductive status, gonad mass, fecundity as standing stock of vitellogenic oocytes and timing and duration of spawning season. Differences in egg quality between farmed and wild cod were studied by analysing the variation in fatty acid profiles from eggs sampled from two farms and from wild fish in an adjacent natural spawning ground. To evaluate if escaped cod eggs and larvae mix with those of wild cod we developed a three-dimensional hydrodynamic model that simulated egg and larval drift from two spawning areas for wild fish and from a cod farm in the Trondheimsfjord. Finally, survival of eggs spawned by cod in sea cages was predicted by developing two simple scenario models on basis of existing knowledge on survival of various ages of cod in the wild, as well as results from the current study on maturation ratios and fecundity of farmed cod.

Figure 5.3.1. Locations where Atlantic cod (Gadus morhua) were farmed (black circle) and location for capture of wild fish for fatty acid analysis (black star).

194

Results The maturation rates for females examined during the first year in the sea (first spawning season) varied from 46.5 to 86.2 % for four farms without artificial light. Maturation rates for males from the same farms were higher with more than 87% of the males being mature in three of these farms. Artificial light was applied to only one farm and no mature fish were found in the samples from this farm. Almost all fish that were sampled during the second spawning season were mature. By analysing gonadal developmental stage and oocyte size for samples from three farms, we found that that spawning in sea cages mainly occurred from March to June in the two northern farms, and also during February in the southernmost farm (Figure 5.3.2). Fecundity varied among farms, but the overall mean number of eggs estimated to be spawned per kg female was about 630Â 000 eggs. Figure 5.3.2. Estimated initiation of time of spawning for two-year old female Atlantic cod (Gadus morhua) at three farms by date. The average temperature was assumed to be 5 ÂşC.

Fatty acid profiles of eggs from wild and farmed cod differed significantly, most likely as a result of their different diets (Figure 5.3.3). For instance, cod eggs from farmed females contained 3 â&#x20AC;&#x201C; 4 times more linoleic acid, originating from vegetable oils added to the fish feed, than eggs from wild fish. This suggests that eggs of farmed cod are affected by parental diet. Furthermore, we found that wild cod eggs had higher levels of DHA and a greater DHA/EPA ratio compared with eggs from farmed cod. There was, however, no variation in aracidonic acid (AA), which is believed to be very important for egg and larval quality, or EPA/AA ratio between wild and farmed cod. Thus, our results indicate that eggs from wild cod might be of better quality than farmed cod, even though there was no difference in the levels of AA between wild and farmed cod (Salze et al. 2005). The lack of variation in AA may on the other hand suggest that the viability of eggs and larvae of farmed cod are not critically inferior to that of wild

www.preventescape.eu

195

fish. The ability to conclude with respect variation in viability between eggs and larvae from wild and farmed cod is nevertheless restricted because other biochemical components in cod eggs not measured in the current study also influences egg quality, such as vitamins and the proportion of major lipid classes (Salze et al. 2005). 1.5 1.0 0.5 0 -0.5 -1.0 -1.5 -2.0 -2.5 -2.0

-1.0

1.0

2.0

PC1 (44.7%)

Figure 5.3.3. Individual factor scores and loadings from principal component analyses of the variation in fatty acid profiles among different groups of Atlantic cod (Gadus morhua). PC1 explained 44.7 % of the variation and had an Eigenvalue = 8.49, while PC2 explained 24.2% of the variation (Eigenvalue = 4.60).

Simulations derived from the 3D dispersal model indicated that escaped eggs and larvae from a cod farm mixed with wild counterparts in the fjord system (Figure 5.3.4). If the distributions of eggs and larvae from wild and farmed cod overlap in time and space, it is reasonable to assume that farmed and wild cod eggs and larvae would experience similar early life history

196

environments. As long as egg and larval quality is similar, comparable survival rates between wild and farmed eggs may be expected. Potential may exist to reduce the degree of mixing of farmed and wild eggs if farming locations can be established whereby any escaped eggs are rapidly transported out of the fjord system. The two scenario models developed for predicting the survival of escaped eggs showed that survival is likely to be highly variable under different conditions. In one of the models, the number of 3 yr old cod that result from spawning in a “typical sea cage” with 60 000 cod is 1140 individuals (Table 5.3.1). In the second model, which is based on higher survival rates, spawning in a “typical” cage results in 17 280 fish (Table 5.3.1). The survival estimates from the two scenario models should, however, be regarded as coarse estimates of possible survival rates of eggs and larvae until 3 yr of age due to the many sources for variation in the underlying data. For instance, the proportions of mature females will vary among farms and if artificial light has been used, which will influence whether eggs hatch during periods where food for the developing larvae is present. Fecundity will also vary among farms and ages. The accuracy of estimated survival rates for species with pelagic larvae are, in general, limited by numerous sources of variation and error. In addition, survival may be overestimated as firsttime spawners generally produce fewer eggs of lower quality than older spawners, and it is likely that the survival of larvae from first season spawning would be lower than for second season spawning (Solemdal et al. 1993). Sperm quality and subsequent fertilisation success is reduced for farmed males compared to wild males which could also lead to overestimates of survival (Skjæraasen et al. 2009, Butts et al. 2011). Nevertheless, in fjord systems or individual fjords where cod farming is significant and where catches of wild cod may be in the order of 10s to 100s of t, the impact of spawning in sea cages could be considerable. Our scenarios suggest that a typical sea cage with 60 000 fish under optimal conditions may produce tons of first generation farmed cod through spawning in sea cages, which is significant when compared to known wild cod biomasses in specific fjord systems (e.g. Masfjord: total estimated wild cod biomass = 28 ton, Salvanes and Ulltang 1992).

Number of 3 yr cod farm fish-1

Biomass of 3 yr cod farm fish-1

Number 3 yr escapees "typical" sea cage-1

Biomass (kg) 3 yr escapees "typical" sea cage-1

Spawning 1st year

0.003

0.004

180

219

Spawning 2nd year

0.016

0.019

960

1169

Spawning 1st year

0.048

0.058

2880

3508

Spawning 2nd year

0.240

0.292

14400

17539

Model 1

Model 2

Table 5.3.1 Overview of results from different scenario models. Number and biomass of escapees originating from spawning in cages is presented per farm fish (both sexes) and for “typical” sea cages, i.e. a sea cage holding 60 000 fish.

www.preventescape.eu

197

Figure 5.3.4. Spread of simulated escaped and wild spawned Atlantic cod (Gadus morhua) eggs after 2 and 4 days (upper), 10 and 15 days (second row), 20 and 28 days (third row), 56 and 126 days (lower). Eggs spawned on natural spawning grounds are marked as green or blue, while farmed eggs are red. The farmed eggs become distributed effectively over most of the fjord, while the naturally spawned eggs have a significantly more restricted distribution in the beginning.

198

Discussion We have demonstrated that a large proportion of farmed Atlantic cod mature in industrial sea cages both the first and second year of sea-based culture and that the fecundity of farmed cod is high. In addition to these findings, we found that the fatty acid composition of farmed and wild cod eggs differed, but not to a critical degree. Hence, farmed cod eggs are likely to produce viable larvae within sea cages. Egg dispersal modelling suggested that farmed and wild cod eggs may mix extensively in a fjord ecosystem, indicating that farmed and wild cod eggs will likely experience similar larval environments provided the farmed eggs are not shed off-season due to introduction of various protocol manipulations. Farms examined during this study were spread over a broad latitudinal range of cod farming, thus the results are relevant at an industry-wide scale. Taken together with previous studies that have indicated that escaped farmed cod eggs can result in juveniles recruiting to coastal areas (van der Meeren and Jørstad 2009), our data show that spawning and egg release from sea cages is a significant process which may have ecological repercussions for these type of ecosystems, unless the extent of spawning within sea cages is reduced.

Recommendations • Growth rates of fish in commercial cod farming have increased during the last 2 – 3 yr due to selective breeding and improved production methods. As a result of this, it is now common for harvesting of fish to occur before the second potential spawning season. Thus, a simple, realistic and profitable action to drastically reduce the risk of ecological effects that may occur as a result of egg escape is to make harvesting before the second spawning season mandatory. • In addition, photoperiod manipulation will decrease maturation ratio and fecundity, and delay the spawning time of farmed cod with several months (Hansen et al. 2001, Taranger et al. 2006, Trippel et al. 2008). As spawning during the first season in the sea produced 4 – 5 times fewer eggs than spawning during the second season, photoperiod manipulation will further reduce the potential for unwanted ecological effects of spawning in sea cages as the eggs hatch outside the presumed optimal season. • Furthermore, recent research has shown that production of triploid Atlantic cod may practically eliminate the risk of egg escape, as the gamete production of triploid females are delayed and dramatically lowered compared to diploid females (e.g. Feindel et al. 2011). However, problems such as initially higher mortality, greater fingerling costs, maturation of triploid males and consumer acceptance need to be solved before production of triploid fish is taken up by industry (Triantafyllidis et al. 2007, Feindel et al. 2010).

www.preventescape.eu

199

References

cited

Butts IAE, Trippel EA, Ciereszko A, Soler C, Słowińska M, Hadi Alavi SM, Litvak MK, Babiak I (2011) Seminal plasma bioche mistry and spermatozoa characteristics of Atlantic cod (Gadus morhua L.) of wild and cultivated origin. Comp Biochem Physiol 159: 16-24 Dimitriou E, Katselis G, Moutopoulos DK, Akovitiotis C, Koutsikopoulos C (2007) Possible influence of reared gilthead seabream (Sparus aurata, L.) on wild stocks in the area of the Messolonghi lagoon (Ionian Sea, Greece). Aquacult Res 38: 398-408 Feindel NJ, Benfey TJ, Trippel EA (2010). Competitive spawning success and fertility of triploid male Atlantic cod (Gadus morhua). Aquacult Environ Interact 1: 47–55 Feindel NJ, Benfy TJ, Trippel EA (2011). Gonadal development of triploid Atlantic cod Gadus morhua. J Fish Biol 78: 1900-1912 Hansen T, Karlsen Ø, Taranger GL, Hemre GI, Holm JC, Kjesbu OS (2001) Growth, gonadal development and spawning time of Atlantic cod (Gadus morhua) reared under different photoperiods. Aquaculture 203: 51-67 Jørstad KE, van der Meeren T, Paulsen OI, Thomsen T, Thorsen T, Svåsand T (2008) “Escapes" of eggs from farmed cod spawning in net pens: Recruitment to wild stocks. Rev Fish Sci 16: 285-295 Salze G, Tocher DR, Roy WJ, Robertson DA (2005) Egg quality determinants in cod (Gadus morhua L.): egg performance and lipids in eggs from farmed and wild broodstock. Aquacult Res 36: 1488-1499 Salvanes AGV, Ullvang O (1992) Population parameters, migration and exploitation of the cod (Gadus morhua L. ) in Masfjorden, western Norway. Fish Res 15: 253-289 Solemdal P, Bergh Ø, Dahle G, Falk-Petersen IB, Fyhn HJ, Grahl-Nielsen O, Haaland O, Kjesbu OS, Kjørsvik E, Løken S, Opstad O, Pedersen T, Skiftesvik AB, Thorsen A (1993) Size of spawning Arcto-Norwegian cod (Gadus morhua L.) and the effects on their eggs and early larvae. ICES CM 1993/G:41 Skjæraasen JA, Mayer I, Meager JJ, Rudolfsen G, Karlsen O, Haugland T, Kleven O (2009) Sperm characteristics and competitive ability in farmed and wild cod. Mar Ecol Prog Ser 375: 219–228 Taranger GL, Aardal L, Hansen T, Kjesbu, OS (2006) Continuous light delays sexual maturation and increases growth of Atlantic cod (Gadus morhua L.) in sea cages. ICES J Mar Sci 63: 365-375

200

Thorsen A, Kjesbu OS (2001) A rapid method for the estimation of oocyte size and potential fecundity in Atlantic cod using computer-aided particle analysis system. J Sea Res 46: Triantafyllidis A, Karaiskou N, Bonhomme F, Colombo L, Crosstie D, Danancher D, GarcíaVázquez E, Gilbey J, Svåsand T, Verspoor E, Triantaphyllidis C (2007) Management options to reduce genetic impacts of aquaculture activities. In: Genetic Impact of Aquaculture Activities on Native Populations. GenImpact Compendium (http://genimpact.imr.no/ compendium) Trippel EA, Benfey TJ, Neil SRE, Cross N, Blanchard MJ, Powell F (2008) Effects of continuous light and triploidy on growth and sexual maturation in Atlantic cod, Gadus morhua. Cybium 32: 136-138 van der Meeren T, Jørstad KE (2009) Fanger torsk på vidvanke. Nytt fra havbruk 2009(2), 1 (In Norwegian)

www.preventescape.eu

201

5.4. No evidence of egg escape from (Argyrosomus regius) aquaculture under

meagre current

practices Cite this article as: Montero D, Ramírez B, Sanchez-Jerez P, Bayle-Sempere JT, Fernandez-Jover D, Fernández-Palacios H, Haroun R (2013) No evidence of egg escape from meagre (Argyrosomus regius) aquaculture under current practices. In: PREVENT ESCAPE Project Compendium. Chapter 5.4. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Daniel Montero1, Besay Ramírez1, Pablo Sanchez-Jerez2, Just T. Bayle-Sempere2, Damian Fernandez-Jover2, Hipólito Fernández-Palacios1 & Ricardo Haroun1. University of Las Palmas de Gran Canaria, Spain, University of Alicante, Spain.

1 2

Introduction Meagre (Argyrosomus regius; Figure 5.4.1) are distributed throughout the Mediterranean Sea, although they are uncommon in many parts of their range, although has not been very common in the wildness until it was farmed and some reiterate escape events increased its presence in the commercial catches during the last 5 years (Sánchez-Jerez et al. 2011). In the Atlantic, meagre are distributed from the southern coast of Sweden to the Guinean gulf, including Madeira and the Canary Islands. They are a large predatory fish and can grow up to 2 m and reach more than 50 kg. Meagre is an emerging species in Mediterranean aquaculture, due to its faster growth rate than other common Mediterranean aquaculture species such as European seabass or gilthead seabream. In 2010, 3887 t were produced in the European Union, with Spain growing 98% of this amount. Production is expected to increase in the future (FEAP 2011). Meagre are typically grown to a harvest size of 1.5 kg in sea cage aquaculture.

202

Fig. 5.4.1. Meagre (Argyrosomus regius)

Meagre are a gonochoric species, with sexual maturation in the wild at 4 kg in males and 7.5 kg in females (Prista at al. 2007). In early spring, meagre migrate into estuarine water to spawn (QuĂŠmĂŠner at al. 2002). Juvenile fish remain in coastal waters for 2 or 3 years, before migrating to deeper waters as adults. Natural spawning of meagre broodstock in captivity has not been reported. Production of viable eggs has only been achieved through artificial reproduction with administration of luteinizing hormone releasing hormone agonist (LHRHa). Since meagre are an uncommon or rare species in some areas of the coastline where intensive culture is being developed, concern existed about the possibility that fertile, viable eggs could be released from sea cages. This could lead to increased recruitment rates and therefore increases in natural populations, which may have other cascading effects in coastal ecosystems. Further, other marine species mature precociously in aquaculture settings (e.g. Atlantic cod) at sizes much smaller than is usual in the wild.

Objective We evaluated whether meagre were capable of spawning in sea cages under current day production conditions in fish farms in both Mediterranean Spain and the Canary Islands.

www.preventescape.eu

203

Methods To evaluate the degree of gonad development in industrial cultured meagre, we sampled meagre from sea cages in two locations where aquaculture of this species is increasing: the Canary Islands and the south-east coast of Spain. The potential for gonad maturation and spawning within sea cages was assessed based on whether we detected gonad maturation in market-size fish within sea cages.

Canary Islands study: A total of 138 cultivated meagre were collected from an industrial sea cage farm during the known spawning season of wild meagre (April – June). For each fish, size, weight and gonad weight and stage were recorded. In both studies, the extent of gonadal maturation was calculated using the gonadosomatic index (GSI). Gonads were classified macroscopically as immature (I), resting (II), ripe (III), ripe and running (IV), and spent (V; Holden and Raitt 1975). In addition, gonads collected during the Canary Islands study were removed and samples were fixed in 10% buffered formalin, dehydrated in a graded ethanol series and embedded in paraffin. Serial 4 µm sections were stained with haematoxylin and eosin (H&E) and slides produced from the gonads were analysed by microscopy.

Mediterranean study A total of 138 cultivated meagre were collected from an industrial sea cage farm during the known spawning season of wild meagre (April – June). For each fish, size, weight and gonad weight and stage were recorded. In both studies, the extent of gonadal maturation was calculated using the gonadosomatic index (GSI). Gonads were classified macroscopically as immature (I), resting (II), ripe (III), ripe and running (IV), and spent (V; Holden and Raitt 1975). In addition, gonads collected during the Canary Islands study were removed and samples were fixed in 10% buffered formalin, dehydrated in a graded ethanol series and embedded in paraffin. Serial 4 µm sections were stained with haematoxylin and eosin (H&E) and slides produced from the gonads were analysed by microscopy.

204

Results All fish sampled in sea cages at the Canary Islands had immature gonads in previtellogenic stages (Figure 5.4.2). Sampled fish were on average 1106 Âą 318 g (mean Âą SD), and gonad weights were particularly low (0.19 Âą 0.1 g) indicating the juvenile status of these harvestsized fish. Sexual maturation was not detected in any of the individuals studied. No correlation was found between the gonadosomatic index and fish weight mainly due to the lack of sexual maturation within this size range (Figure 5.4.3). For the Mediterranean location, all sampled meagre had immature gonads, with gonadal maturity stage ranging between 1.03 (for fish 30 cm in body length) to 2.62 (for fish 77 cm in body length, Figure 5.4.4). The gonadosomatic index reflected the lack of gonad development in sea cages at commercial size (Figure 5.4.5).

Fig. 5.4.2. Gonads from a female meagre in Stage I (previtellogenic). Fish body weight = 1.7 kg.

www.preventescape.eu

205

Fig. 5.4.3. Relationship between total fish weight and the gonadosomatic index for meagre (Argyrosomus regius) held in sea cages to commercial size in the Canary Islands. A polynomial regression line is included.

Fig. 5.4.4. Gonads of a meagre of 43 cm body length.

Fig. 5.4.5. Scatterplot of the gonadosomatic index (GSI) of meagre from two commercial Mediterranean farms.

206

Discussion The results obtained in both studies indicate that meagre have immature gonads in sea cages at present-day harvest sizes. Our study provides strong evidence that release of eggs from meagre produced to 1.5 kg in sea cages does not occur. Thus, we conclude that there no risk of egg escapes from meagre sea cage farms under current production practices. On occasion, large meagre (> 6 kg in body weight) are held in inland experimental conditions or found in industrial fish farms, either as part of normal production (Mediterranean Sea) or held as broodstock (Canary Institute of Marine Sciences). These fish have gonads in the final stage V of gonadal development (Figure 5.4.6 and 5.4.7). The existence of these larger animals in sea cages is an exceptional situation and may result from poor commercial management in specific farms and conditions. While spawning of these larger fish in sea cage environments has never been detected and appears unlikely given the need to chemically induce spawning in meagre broodstock, as production of larger fish becomes normal in the future, the possibility that large, mature fish can spawn in sea cages will require further assessment.

Fig. 5.4.7. Female gonads from a large meagre (71 cm body length).

Fig. 5.4.6. Gonads from a female meagre in Stage V (Vitellogenic). The female was kept in experimental conditions at the Canary Institute of Marine Sciences. Fish body weight = 10.4 kg.

Recommendation Even though spontaneous spawning has never been recorded in captivity, and natural spawning of meagre requires a combination of brackish and estuarine waters, a conservative recommendation to avoid any potential for â&#x20AC;&#x2DC;escape through spawningâ&#x20AC;&#x2122; to occur for meagre is to limit culture to fish smaller than 5 kg body weight in sea cages.

www.preventescape.eu

207

References

cited

FEAP (Federation of European Aquaculture Producers) (2011) www.FEAP.info Holden MJ, Raitt DFS (1975) Manual of fisheries science. Part 2. Methods of resource investigations and their application. FAO Fish. Tech. Pap. 115: 1-214. Prista N, Costa JL, Costa MJ, Jones CM (2007) New methodology for studying large valuable fish in data poor situations: commercial mark-recapture of meagre Argyrosomus regius in the southern coast of Portugal. ICES CM 2007/O:43. ICES, 18p. Quéméner L (2002) Le Maigre Commun (Argyrosomus regius) – Biologie, Pêche, Marché et Potentiel Aquacole. Ifremer, Plouzané. 31 pp. Sanchez-Jerez P, Izquierdo, D, Arechavala-Lopez, P, Bayle-Sempere, JT, Fernandez-Jover, D, Valero-Rodriguez, JM, Dempster, T (2011). Escapes of Argyrosomus regius in the mediterranean sea from fish farms: helping a rare species to be abundant in coastal areas. Proceedings of Aquaculture Europe Conference (EAS, Rhodes, Greece, 18-21 October 2011).

208

5.5. EGG Escape: General Conclusions & Recommendations Cite this article as: Somarakis S, Uglem I, Montero D, Sanchez-Jerez P, Nilsen R, Dempster T (2013) Escape through spawning: general conclusions and recommendations. In: PREVENT ESCAPE Project Compendium. Chapter 5.5. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Stelios Somarakis1, Ingebrigt Uglem2, Daniel Montero3, Pablo Sanchez-Jerez4, Rune Nilsen5 & Tim Dempster6 Hellenic Centre for Marine Research, Crete, Greece, Norwegian Institute for Nature Research, Norway, 3 University of Las Palmas de Gran Canaria, Spain, 4 University of Alicante, Spain, 5 I nstitute of Marine Research, TromsĂ¸, Norway, 6 SINTEF Fisheries & Aquaculture, Norway 1 2

Potential spawning of farmed fish in sea cages and subsequent dispersal of viable eggs and larvae in surrounding waters, where they can develop normally and ultimately mix with the wild populations, will only be possible if the species is reared beyond the size or age of puberty. In the particular case of sequential hermaphrodites, like the gilthead seabream (Sparus aurata), release of fertilized eggs will only be possible if rearing duration exceeds the time needed for fish to reach the size of sex reversal and onset of female maturation (Mylonas et al. 2011). In aquaculture, harvesting the fish beyond the size/age of first reproduction, i.e., giving the fish the chance to reproduce inside the cages, represent a potential problem that is opposite to that in fisheries in which harvesting the fish prior to size of first reproduction is undesirable. To predict the possibility that a reared species will actually reproduce inside the cages, the process of reproduction within the aquaculture setting has to be first understood. Attainment of first maturation/sex change, timing of gonadal growth and regression and final oocyte maturation/ovulation are complex physiological processes that may differ from those of wild populations. They are influenced by environmental factors, trophic regimes and genetic backgrounds whereas egg release and successful fertilization may require conditions suitable for performing a required spawning behavior. Size/age at puberty is a phenotypically variable trait often correlated with somatic growth rate (Taranger et al. 2010). The higher growth rates typically seen in fish farming normally result in puberty occurring both at an earlier age and at a smaller body size, compared to wild populations. For example, growth and age at puberty varies between wild Atlantic cod stocks and is affected by prey availability and temperature in their habitat (Drinkwater 2002). The Northeast Arctic cod usually spawns at an age between 4 and 8 years, while Norwegian

www.preventescape.eu

209

coastal cod recruits to the spawning stock at 3 years and older (Berg and Albert 2003). In farming, cod spawn at an age of 1 or 2 years under current growing conditions. The proportion of mature 1 year old fish varies among farms and sexes whereas almost all 2 years old fish mature under farming conditions. The phenotypic responses to improved growth conditions and feed availability, often with associated higher adiposity and energy stores, are considered to be the major causes for the early maturation commonly observed in many farmed fish species (Taranger et al. 2010). In fish with marked seasonality in egg production, the reproductive cycle is controlled and synchronized by seasonal changes in environmental conditions in relation to local climatic and productivity regimes. A suite of environmental factors have been shown to synchronize the reproductive cycle with the seasonal cycle (Bromage et al. 2001), however, in temperate regions, photoperiod and/or temperature variations are the main agents controlling the reproductive cycle. Temperature thresholds for spawning have been defined for many fish species (Bromage et al. 2001). For example, in European seabass, ovulation and spawning are blocked when temperature is >17ยบC, even when oocyte maturation has been completed (Zanuy at al. 1986, Devauchelle and Coves 1988, Zanuy at al. 1995). European seabass spawn in winter when water temperatures are the lowest. In an experiment in which European seabass was held at >17ยบC from October to February spawning was delayed by two months, and only occurred when temperatures decreased below 17ยบC (Carrillo et al. 1995). The temperature thresholds for spawning are of particular importance with regard to temperature rise due to climatic change and species introductions as aquaculture species to new areas. For instance, the Canary Islands is a geographical area that is out of the thermal limit for reproduction of European seabass because of water temperatures >17ยบC in winter. Hence, this species is not expected to produce eggs in sea cages around the Canary Islands. Furthermore, escapees from aquaculture are unlikely to establish self-sustaining populations in the Canaries. The extent to which the production of fertilised eggs from aquaculture cages will result in successful recruitment of additional individuals to naturally occurring wild stocks will largely depend on the timing of spawning. The timing of spawning in the wild is believed to be the result of some sort of selection process as a response to a range of abiotic and biotic

210

environmental factors like temperature and abundance of suitable prey. For instance it is believed that timing of spawning in many marine broadcast spawners with a pelagic egg stage, like the Atlantic cod, is likely to be an adaptation to maximize recruitment; too early or too late spawning may mean that the first feeding larvae will miss the seasonal zooplankton bloom (e.g. Ellertsen et al. 1981, Taggart and Frank 1990, Pepin and Myers 1991, Sundby 2000). If the first feeding larvae miss the zooplankton bloom their potential for growth and survival might be considerably affected. In aquaculture, the possibility to manipulate the spawning period for the cultured species might be utilised to reduce potential ecological impacts of spawning in cages. Typically, populations of cod spawn over a period of <3 months (Brander 1994, Chambers and Waiwood 1996, Kjesbu et al. 1996) and initiation of spawning may depend on environmental or genetic variation (Otter책 et al. 2006). Furthermore, photoperiod manipulation of farmed cod will, in addition to decrease maturation ratio and fecundity, also delay the spawning time of farmed cod by several months (Hansen et al. 2001, Taranger et al. 2006, Trippel et al. 2008). Hence, manipulation of spawning period of farmed cod is clearly possible and will involve that perhaps the bulk of the spawning within sea cages occur after the spawning of wild cod. If and to what extent delayed spawning in farms will affect the survival of first feeding larvae in the wild is at present unknown, but it is reasonable to assume that spawning outside the natural spawning season will reduce the potential for growth and also long-term survival. When taking place spontaneously, egg production in sea cages might be normal in certain species like cod, i.e. fecundities and egg quality may be comparable to those of wild populations, whereas in others, like the gilthead seabream, egg production is low. Gilthead seabream have a complex reproductive biology and conditions within the cages are likely to be suboptimal for spawning probably because of unbalanced sex ratios and increased stress due to overcrowding, which is likely to impede the normal performance of spawning behavior (Mylonas et al. 2011). Sterility, achieved by, for example, triploidization, can be an effective means to inhibit puberty and the associated negative effects on growth, feed utilization, health and welfare, but also the potential genetic effects on wild stocks after release of fertilized eggs into natural ecosystems. Furthermore, the production of monosex stocks can diminish the release of fertilized eggs while such stocks can exploit sex dimorphic growth patterns (Taranger et al. 2010). However, all these techniques have limitations under commercial farming conditions.

www.preventescape.eu

211

In cases of increased probability that fertilized eggs might be released into the natural environment, a precautionary approach to prevent potential negative effects on wild populations would be to put spatial restrictions as to where such farms can be located. It would be recommended that farms with fish larger than size at first reproduction be placed away from areas where the wild population of the same species can normally reproduce and close its life cycle, e.g. in the vicinity of known nursery areas. Protection of existing marine protected areas (MPAs) from potential negative effects of egg escapes would also be desirable. The eventual recruitment of â&#x20AC;&#x2DC;escapedâ&#x20AC;&#x2122; eggs into wild populations may affect intra- and interspecific competition and alter the size structures of wild populations and communities, especially in MPAs where fishing mortality is low (Harmelin-Vivien et al. 2008). Site selection of farms to mitigate negative effects from escaped eggs would be facilitated by the detailed knowledge of species ecology, habitat use and migration routes as well as local hydrodynamics that determine egg and larval advection in relation to location of the fish farms and sensitive coastal areas like MPAs or nurseries of wild fish.

212

References

cited

Berg E, Albert OT (2003) Cod in fjords and coastal waters of North Norway: distribution and variation in length and maturity at age. ICES J Mar Sci 60:787–797 Brander KM (1994) The location and timing of cod spawning around the British Isles. ICES J Mar Sci 51:71-89 Bromage N, Porter M, Randall C (2001) The environmental regulation of maturation in farmed finfish with special reference to the role of photoperiod and melatonin. Aquaculture 197:63-98 Carrillo M, Zanuy S, Prat F, Cerdá M, Ramos J, Mañanos E, Bromage N (1995) Seabass. In: Bromage NR, Roberts R J (eds) Broodstock management and egg and larval quality. Blackwell, Oxford, UK, pp. 138-168 Devauchelle N, Coves D (1988) Seabass (Dicentrarchus labrax) reproduction in captivity: gametogenesis and spawning. Aquat Living Resour 1:215-222 Drinkwater KF (2002) A review of the role of climate variability in the decline of northern cod. In: McGinn, NA (Ed) Fisheries in a Changing Climate. American Fisheries Society Symposium 32, Bethesda, MD, USA, pp. 113–130 Ellertsen B, Solemdal P, Sundby S, Tilseth S, Westgard T, Øiestad V (1981) Feeding and vertical distribution of cod larvae in relation to availability of prey organisms. Rapp Pro-v Reun Cons Inter Expl Mer 178:317-319 Hansen T, Karlsen Ø, Taranger GL, Hemre GI, Holm JC, Kjesbu OS (2001) Growth, gonadal development and spawning time of Atlantic cod (Gadus morhua) reared under different photoperiods. Aquaculture 203:51-67 Harmelin-Vivien M, Le Diréach L, Bayle-Sempere J, Charbonne E, García-Charton JA, Ody D, Pérez-Ruzafa A, Reñones O, Sanchez-Jerez P, Valle C (2008) Gradients of abundance and biomass across reserve boundaries in six Mediterranean marine protected areas: Evidence of fish spillover? Biol Conserv 141:1829-1839 Kjesbu OS, Solemdal P, Bratland P, Fonn M (1996) Variation in annual egg production in individual captive Atlantic cod (Gadus morhua). Can J Fish Aquat Sci 53:610-620 Mylonas CC, Zohar Y, Pankhurst N, Kagawa H (2011) Reproduction and broodstock management. In: Pavlidis MA, Mylonas CC (eds) Sparidae: Biology and Aquaculture of gilthead seabream and other species. Blackwell Publishing, pp. 95-131

www.preventescape.eu

213

Otterå H, Agnalt A-L, Jørstad KE (2006) Differences in spawning time of captive Atlantic cod from four regions of Norway, kept under identical conditions. ICES J Mar Sci 63:216-223 Pepin P, Myers RA (1991) Significance of egg and larval size to recruitment variability of temperate marine fish. Can J Fish Aquat Sci 48:1820-1828 Sundby S (2000) Recruitment of Atlantic cod stocks in relation to temperature and advection of copepod populations. Sarsia 85:277-298 Taggart CT, Frank KT (1990) Perspectives on larval fish ecology and recruitment processes. Probing the scales of relationships. In: Sherman K, Alexander LM, Gold BD (eds) Large Marine Ecosystems. American Association for the Advancement of Science, Washington, USA, pp. 151-164 Taranger GL, Aardal L, Hansen T, Kjesbu, OS (2006) Continuous light delays sexual maturation and increases growth of Atlantic cod (Gadus morhua L.) in sea cages. ICES J Mar Sci 63:365375 Taranger GL, Carrillo M, Schulza RW, Fontaine P, Zanuy S, Felip A, Weltziene F-A, Dufour S, Karlseng Ø, Norberg B, Anderssona E, Hansenh T (2010) Control of puberty in farmed fish. Gen Comp Endocr 165:483-515 Trippel EA, Benfey TJ, Neil SRE, Cross N, Blanchard MJ, Powell F (2008) Effects of continuous light and triploidy on growth and sexual maturation in Atlantic cod, Gadus morhua. Cybium 32:136-138 Zanuy S, Carrillo M, Ruiz F (1986) Delayed gametogenesis and spawning of seabass (Dicentrarchus labrax L.) kept under different photoperiod and temperature regimes. Fish Physiol Biochem 2:53-63 Zanuy S, Prat F, Carrillo M, Bromage N (1995) Effects of constant photoperiod on spawning and plasma 17b-oestradiol levels of seabass (Dicentrarchus labrax). Aquat Living Resour 8:147-152

214

www.preventescape.eu

215

216

seeks to reduce the number of fish that escape from European aquaculture through research to improve fish farming techniques and technologies.

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

6.1. Preventing and mitigating escapes through research to underpin technological and operational improvements for seacage farming and recapture technologies Cite this article as: Jensen Ă&#x2DC; (2013) Preventing and mitigating escapes through research to underpin technological and operational improvements for seacage farming and recapture technologies. In: PREVENT ESCAPE Project Compendium. Chapter 6.1. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Ă&#x2DC;sten Jensen1 1

SINTEF Fisheries & Aquaculture, Norway

www.preventescape.eu

217

Introduction Much of the state-of-the-art of knowledge for ensuring that fish farms do not fail structurally, and that operations do not cause escapes, is contained in technical standards. Such standards are well developed in some European countries, but non-existent in others. Even though Norway has a comprehensive technical standard, escapes still occur at a level which is regarded by many to be detrimental to wild stocks, indicating that the knowledge which forms the basis of technical standards is in constant need of improvement. Work is ongoing to develop an ISO standard that can apply to all sea-cage fish farm installations. The main objective of WP6: Prevent Escape was to extend the knowledge on culture operations and technology, to improve technical and operational guidelines, to prevent fish escape and to increase the recapture success of escaped fish on the basis of integrated biological and technological research. The work consisted of full scale tests, physical and numerical modelling of farms and structures, material and component testing as well as developing recommendations and guidelines for the design an operation of marine fish farms. The specific objectives of WP6 were: (1) integrate biological and technological research to facilitate knowledge-based development of robust containment technology, (2) generate fundamental knowledge on the properties of component aquaculture technologies to help improve design and production of sea-cage equipment components, (3) develop guidelines for the design and use of sea-cage technologies and equipment to minimize the risk of escapes, (4) test recapture technologies to improve recapture rates of escapees based on knowledge of the post escape behaviour of fish, and (5) disseminate the results to fish farmers, aquaculture technology manufacturers and suppliers, standards organisations, government agencies and the wider scientific community. Net cages used to farm cod, salmon, seabass and seabream were inspected for holes throughout a production cycle. A higher number of holes were found in nets used to farm seabream and cod compared to nets used to farm salmon and seabass, due to the cod and seabream biting the

218

net. Measurements performed during laboratory tests confirmed that even small to medium sized cod were able to pull the net with a force sufficient to tear a single nylon filament in the net. To avoid the farmed fish creating holes in the net, the strength of a single filament should be higher than the strength of the fish. Numerous tests have been performed on new virgin material as well as on net materials used to farm cod, salmon, seabass and seabream. Different mesh types and different materials have been tested using different procedures to determine the strength, flexibility, resistance to abrasion, weakening due to washing and other operations on the net. The net strength declines with use, but it was not possible to directly relate the remaining strength of the material to the age or number of productions cycles for the net. Some of the tests performed (high pressure cleaning, cyclic tests, creep tests) had a minor if any effect on the strength of the material whereas others (e.g. washing in tumble washers and abrasion) had larger impact on the net strength. Environmental conditions along the coast of Spain (mainland and Canary Islands), Ireland, and Norway were determined and served as valuable input for numerical and physical modelling of commonly used aquaculture structures. Several methods for recapturing escaped fish were tested, but none of them gave recapture rates at a satisfactory level. In addition the by-catch was significant and often orders of magnitude larger than the recapture. Thus effort should be placed on preventing fish from escaping. However, natural- and human-induced mitigation occurs in the Mediterranean Sea, because predation pressure could reduce survival as well as the capture during the next days by commercial and sport fisheries at scale of kilometres around the fish farms. The results from the project has served as valuable input in the process of developing a Scottish Technical Standard as well as a ISO standard for fish farms, SINTEF Fisheries and Aquaculture has been involved in the development of both.

www.preventescape.eu

219

6.2. Damage

to the net cage

Cite this article as: Moe Føre H, Olsen A, Jackson D, Drumm A, Mendiola D, Gabiña G (2013) Damage to the net cage. In: PREVENT ESCAPE Project Compendium. Chapter 6.2. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Heidi Moe Føre1, Anna Olsen1, Dave Jackson2, Alan Drumm2, Diego Mendiola3 & Gorka Gabiña3 SINTEF Fisheries and Aquaculture, Norway, The Marine Institute, Ireland, 3 Azti Technalia, Spain. 1 2

Introduction Technical standards contain up-to-date information to ensure that fish farms do not fail structurally and that their operations and procedures do not lead to fish escapes. Such standards are well developed in some European countries, but poorly developed, or non-existent, in others. That escapes still occur at a level regarded by many to be detrimental to wild stocks (e.g. Naylor et al. 2005), indicates that technical standards are in constant need of improvement. As shifts in the fish-farming industry occur that result in changes in the type of technology used, the physical environments in which farms are located and practical aspects related to use of the equipment of ever-increasing size, research is constantly required to update the knowledge upon which standards are based. Objective testing and documentation of how fish-farming technology (netting, floaters, mooring systems) interacts with the diverse range of physical (waves, currents) and biological (cultured species, biofouling) environments in Europe is essential to prevent escape. The net plays an important role in most escape episodes, either directly through holes in the sea-cage from over-loading or contact with other objects, or indirectly by transferring major loads to the fish farm (Jensen 2006, Moe et al. 2007, Norwegian Fisheries Directorate 2007). The mechanical properties of nets and net materials, and how they change over time with use and wear, are poorly understood. Current knowledge and testing, is mainly limited to unused netting materials (Figure 6.2.1; Moe et al. 2007). How net material properties change with damage from wear, UV-radiation, fatigue, abrasion and biting by the caged fish is largely unknown. A wide range of netting materials are available in the marketplace, from simple uncoated and coated nylon nets to Dynema, yet how the strength and durability of these varying materials changes with practical use is poorly understood (Moe et al. 2005).

220

Objective We identified and assessed the range of critical operations, conditions and circumstances that can lead to escape episodes at sea farms. The study included aquaculture sites from three countries: Norway, Spain, and Ireland. Culture operations for species typical grown in each country were covered, including Atlantic salmon (Salmo salar) and Atlantic cod (Gadus morhua) in Norway and Ireland, and European seabass (Dicentrarchus labrax) and Gilthead seabream (Sparus aurata) in Spain. Farming sites were monitored over time to detect the nature and extent of damages to nets, document general operational procedures, and detect escape and near-escape episodes during a production cycle from smolt to harvest. Findings from the observations, supported by experience from the fish farmers, have been analysed with respect to the level of risk for fish escapes. Through extensive analysis of past escape incidents and practical experience, the net is known as one of the weakest points in the sea-cage construction (Jensen et al. 2010). Approximately 60% of escapes, measured both in the number of incidents and the numbers of escaped fish, occur due to holes that develop in the net. We inspected cage nets regularly through a production cycle to document: the number of damages to the nets; development of damage (time and size/extent); location of damage; type of damage; and possible cause of damage.

Figure 6.2.1. New undamaged net with antifouling surface treatment.

Fish farmers were also interviewed to determine the critical moments of risk to the cage net integrity in handling operations, such as; splitting and grading, transfer of fish between cages, parasite treatment, and smolt and slaughter-fish delivery. In Norway, data was collected from two salmon and two cod producers, encompassing three farms for each species. Interviews were also conducted with three Irish salmon producers. For seabream and seabass farms, four fish farmers from mainland Spain, one farmer from Gran Canary Island and one from Tenerife, were interviewed.

www.preventescape.eu

221

Based on the interview results, a risk analysis of critical operations for fish escapes was carried out using qualitative risk analysis (Bahr 1997). The initial results include an overview of the critical damage that can occur in fish-farm nets, which potentially lead to fish escapes, for a number of sea-cage designs. In additional, the interviews provided information on general fish farm operations, plus information to assist in the formulation of guidelines for the design, operation, inspection and repair of nets. These guidelines should help fish-farms to avoid or limit future net damages. Learning from fish farmers about their sea-cage operations, equipment, and previous escape incidents, enables us to better understand the technical/operational causes of fish escapes. This information constitutes an important contribution to current knowledge to the improvements which can be made that will benefit the entire industry.

Questionnaires The questionnaires were completed by direct interviews with the producers. The questions were divided into three sections that specified different causes of fish escapes: technical/ mechanical failure, biological factors and human errors. The questions aimed to ascertain the most important issues arising within the three categories, as well as to gather more detailed information concerning day-to-day fish farm management. The following issues were raised: Topics Net handling Wear and tear Feeding regimes Biofouling Boat traffic Predators Table 6.2.1. Relevant topics in the questionnaire.

Please note that these are the subjective opinions of a selected group of people. We assume that these opinions will represent the majority of fin-fish producers along the coasts of Norway, Ireland and Spain, however, we recognise that individual opinions may differ from those recorded here.

222

Qualitative

risk analysis

The risk of an event can be defined in a number of ways. The common definition of risk is “the prediction of what may happen in the future”, whereas the more academic definition is the following (ISO14121-1:2007): “Risk is a combination of the probability of occurrence of harm and the severity of that harm”. Once the interview results were collated and the risks identified, the probability of an escape occurring, and its consequences, were placed under four categories (see Table 6.2.2).

Probability Category 1

Rare event is defined as event which is expected to occur once in 100 years or not expected to occur, exceptional circumstances only.

Category 2

Possible event is defined as event which is expected to occur once in 10 years.

Category 3

Likely event happens at least once every second year (at least once during the production cycle).

Category 4

Almost certain event happens at least once a month (expected in most circumstances on the daily basis) Severity of consequences

Category 1

Negligible consequences: Individual escapes less than 100 fish.

Category 2

Minor consequences: Moderate escapes 100 to 10 000 fish.

Category 3

Major consequences: Large escape 10 000 to 100 000 fish.

Category 4

Severe consequences: Catastrophic escape over 100 000 fish. Table 6.2.2. Categories of probability and severity of consequences.

www.preventescape.eu

223

The categories were adapted for this project based on the particulars of the aquaculture industry (Helle 2011). The relevant answers were then assigned within a risk matrix (Table 6.2.3).

CONSEQUENCES

PROBABILITY

4 3 2 1 High risk. Immediate actions should be applied to reduce the risk of escape. Moderate risk. Correcting actions should be considered. Low risk. Actions are implemented from other considerations than escape risk.

Table 6.2.3. Risk matrix.

224

Net

damage monitoring

We monitored three cages for net damage for each of the following fish species: Atlantic salmon; Atlantic cod; Gilthead seabream, and European seabass, in Norway, Ireland, and Spain. Relationships were established with fish farmers, their net servicers and their divers. Through this collaboration we gained access to their facilities and equipment documentation. The following topics were covered: • Description of the cage net, material choices, weight and mooring systems (see Figure 6.2.2), the age and condition of the net, antifouling treatments and the breaking strength of the mesh. • Description of damages to the net (noting the location, size and assumed cause). • Biomass parameters. • Description of environmental loads (where available).

Figure 6.2.2. Illustration of fish farm and net cage components.

This data, and supporting information, was gathered at four stages during the production cycle: stocking; during net changes; at approximately half-way between net change and harvest, and at harvest time.

www.preventescape.eu

225

Whenever possible, additional information was gathered from the divers about net conditions between the assigned inspection times. The divers were asked to pay extra attention to the following: • Specify the exact number of damages found and repaired, and the dimensions and position of the damaged areas on the net. The position should be stated as the depth and the proximity to the ropes, either vertical or horizontal. • Specify the position of the damages with respect to the other cage equipment such as chains, weights, cables and mooring lines. • Specify the position of the damages with respect to the main current direction if possible. • If it is possible to recognise wear of the netting and ropes without apparent holes or tear, specify dimensions and position. • Suggest the cause of the damages (wear/fish bites/predators/mechanical damages).

226

To ensure consistent terminology, classifications for net and rope damages were used (Tables 6.2.4 & 6.2.5).

Nomenclature

Description

Illustration

Broken twine

A single broken twine in the netting

Hole

Several broken twines forming a hole in the netting

Tear

Several broken twines in a row

Wear

Netting damaged by abrasion against ropes, chains, floater, weight etc

Fraying

Singular or small bundles of filaments (fibres) are pulled out of the twine

Torn seam

Seam in netting that has been torn

Table 6.2.4. Classification of net damages.

www.preventescape.eu

227

Nomenclature

Description

Broken rope

Illustration Not available

Wear

Rope damaged by abrasion against other ropes, chains, propeller, floater, weights etc

Torn seam

Torn seam (loosened rope)

Table 6.2.5. Classification of rope damages.

Questionnaire The fish-farm managers were also interviewed and a questionnaire completed. The respondents gave their opinions on what they felt were the most critical reasons for fish escapes (Figure 6.2.3). The majority of the Spanish fish farmers interviewed indicated that their main area for concern was technical failure, e.g. inadequate equipment and cage structures that were unable to withstand the harsh marine environment. The Norwegians indicated that both human error and biological factors were the main risks, while the Irish considered human error, net handling and predator attacks to be equally important.

Figure 6.2.3. Most important reason for escapes, sorted by country.

228

Figure 6.2.4 shows the same results categorized by fish species (e.g. salmon producers from Norway and Ireland came under the same category). The main causes of escape of seabream and seabass, from Spanish fish farms, were technical. In contrast, cod producers attributed their losses to biological factors, often mentioning fish behaviour. Salmon producers, on the other hand, were most concerned about net failures due to human error.

Figure 6.2.4. Most important reason for escapes, sorted by species.

During the interview the respondents selected the fish farm activities most likely to lead to escapes (Figure 6.2.5 shows activities sorted by countries, and 6.2.6 by species). Irish farmers indicated that equipment wear and tear and predators were their greatest concern, whereas feeding regime and boat traffic did not represent any risk for escapes (Figure 6.2.5). In contrast, the Norwegian and Spanish farmers did not highlight any particular activities as being more likely to lead to escapes.

Figure 6.2.5. Summary of factors which represent a risk of escape, sorted by countries.

www.preventescape.eu

229

Sorting the data by species revealed that feeding behaviour is an important escape risk factor for Atlantic cod (Gadus morhua) and Gilthead seabream (Sparus aurata), but of no concern for Atlantic salmon (Salmo salar) and European seabass (Dicentrarchus labrax) (Figure 6.2.6). The other risk factors were represented more or less uniformly across species.

Figure 6.2.6. Summary of factors which represent a risk of escape, sorted by species.

230

Figure 6.2.7. Critical factors which represent a risk of escape, sorted by countries.

Risk

assessment

The fish farmers also evaluated the frequency of occurrence of critical net-damage events and their consequences. Each answer was assigned to one of the four categories from the lists of ‘probability of occurrence’ and ‘severity of consequences’ (see lists in Table 6.2.2). The questions were also divided into the three groups mentioned before i.e. technical failure, human error and biological factors. Some of the questions could be placed under more than one group. For instance, some of the net handling operations, such as changing and lifting of nets, are regarded both as an event with risk for human error and mechanical failure. This assumption is considered to be valid since several of the respondents explicitly mentioned it in their answers. In the risk assessment for technical causes, the natural environment lies partly in the red zone, and this represents a high risk of escape (Table 6.2.6 Charts 1-5.a). In Spain, their high energy environment, which is subject to frequent storms, represents a high risk of damage to the critical components of cage structures such as as net pens and the mooring system (Table 6.2.6 Charts 4-5.a). In contrast, Norwegian and Irish sites are mostly semi-exposed and experience less wave loads on the structure. The respondents from these countries gave less weight

www.preventescape.eu

231

to the environment in the risk assesment, moving it to a moderate level (Table 6.2.6 Charts 1-3.a). The issue of cleaning the nets with high pressure cleaning discs was raised on a number of occasions. The respondents reported that the extensive pressure exerted on the nets initiates damage to the netting. Figure 6.2.8, Charts 1-5.b illustrates the risk of human error occurring when net handling operations and boat traffic are included in the analysis. Several Spanish fish farmers mentioned that boat navigation around their facilities led to an increased risk of damage occurring due to the low skill level of the personnel. In addition, many respondents pointed out the risk for human error during net and fish handling operations such as handling of weights, changing of nets and at splitting, grading and stocking of the fish. The issue of poaching was also mentioned. This problem is a high risk factor for Gilthead seabream (Table 6.2.6 Chart 4.b) because they tend to swim near the water surface which makes them easy to catch. Biofouling, predators and fish behaviour are included as biological risk factors (Table 6.2.6 Charts 1-5.c). Biofouling represents several problems for the fish farmers as it obstructs water flow through the cage, and can also provoke undesirable biting behaviour among fish such as Atlantic cod (Table 6.2.6 Chart 2.c) and Gilthead seabream (Table 6.2.6 Chart 4.c). Net biting is not an issue for European seabass (Table 6.2.6 Chart 5.c) or Atlantic salmon. In Norway, removing biofouling from the nets with a high pressure cleaning disc is regarded as a high risk activity (Table 6.2.6 Chart 1.c). As for predators, the fin-fish farms in Ireland are ocassionally attacked by seals (Table 6.2.6 Chart 3.c), whereas Spanish respondents mentioned several species of wild fish that damage nets as they try to get inside cages, and then scare the farmed fish through the existing holes.

232

Table 6.2.6. Graphical representation of risk matrices.

www.preventescape.eu

233

Results

from net damage monitoring

Norway Three salmon sites and three cod sites were chosen to monitor net damage in Norway. The seawater production cycle in Norway is approximately 18 months for Atlantic salmon and 22 months for Atlantic cod, and is characterized by two periods. First, the fish are stocked into a smolt net, with small mesh size in the range 26 â&#x20AC;&#x201C; 40 mm. After approximately half a year, the smolt net is replaced by a net with a larger mesh size, in the range 44 â&#x20AC;&#x201C; 55 mm. Some of the fish farmers grade and sort fish during the net change period to achieve a more uniform biomass distribution in the cage. Others keep fish in one unit for the whole period to avoid stressing the stock. As already mentioned, the data on net damages was gathered at four stages: stocking; net change; intermediate inspection, and after harvest. Supplementary diver reports were collected where available. The quality of the information provided varied because the net servicer and diver companies often have their own protocols for producing service and repair reports. Figure 6.2.8 gives a summary of the areas of damage of nets monitored in Norway. Figure 6.2.8. Summary of areas of net damage - Norway.

Summary for Salmon cages: I: The area most exposed to environmental conditions is above the waterline and in the top 1 â&#x20AC;&#x201C; 3 m of water. The most probable causes of abrasion are; contact with other parts of the structure e.g. floaters, sinker-chains and ropes, exposure to environmental forces (waves, currents) and net handling operations. II: Significant damage was observed all around the body of the net, both on the vertical walls and in the conical bottom section. The most probable causes are environmental forces (currents) and net handling operations.

234

Summary for Cod cages: III: The area near the leadline and the central part of the bottom contains damage which is suspected to have been caused by cod biting.

Ireland There are approximately 35 current marine salmon farming licenses in Ireland. Organic Salmon production accounts for 83% of the total annual production. The growth cycle period is similar to that of Norway and Scotland. However, the conditions in which Irish salmon are grown differ greatly from the enclosed sea-lochs of Scotland and fjords of Norway. The majority of Irish fin-fish farms are located off the west coast, located in high energy, exposed and semiexposed areas. This has necessitated the development of strict Standard Operating Procedures (SOPs) for maintenance of all structures e.g. cages, moorings and nets. The introduction of SOPs has resulted in a significant reduction of escape incidents. In conjunction with SOPs, the development of correct weighting systems for pen netting has eliminated the use of predator nets. While the vast majority of companies employ SOPs for net maintenance, the results from this study indicate there is still some room for improvement. Figure 6.2.9 shows a summary of the areas of damage of nets monitored in Ireland. Figure 6.2.9. Summary of areas of net damage - Ireland.

I: The most exposed areas of the net are above the waterline, and the top 1 â&#x20AC;&#x201C; 3 m of net showed the greatest number of holes/abrasions. This most likely occurs due to contact with other parts of the cage structure e.g. floaters, chains and ropes. Environmental forces (wind, waves and currents) can be severe, resulting in net abrasion along and above the waterline. Net handling operations are also a contributing factor.

www.preventescape.eu

235

II: Damage was also observed near the leadline, along the net panels and close to the vertical lines in the junction between the net panels. The most probable cause for damage along the leadline was the individual weighting system. The damage along the walls could be attributed to predator attack (seals) and net handling operations.

Spain Spain has developed into one of the primary European aquaculture producers in terms of volume. In 2010, seabream and seabass were the two main marine finfish species being produced in terms of volume and value of production (APROMAR 2010). The total volume of production was approximately 43888 t consisting of 20360 t of seabass and 12495 t of seabream. Another important farmed species is the meagre (Argyrosomus regius); 3250 t was produced in sea-cages in 2010.

Figure 6.2.10. Summary of the areas of damage of net damage - Spain.

Summary

for

Gilthead

seabream and

European

seabass

I: As with the Norwegian and Irish nets, the most exposed area is above the waterline and the top water layer down to the 3 m depth. The most probable causes are contact with other parts of the structure (floaters, chains, ropes) and environmental forces (waves, currents). II: Damage was observed around the whole net enclosure. The most probable causes are environmental forces (currents), predators and biting. Damage to seabream nets is primarily caused by biting of the net by seabream themselves and predators trying to feed on the fish within the cage. III: The area near the leadline experiences extensive wear and tear due to current pressure and interactions with other parts of the construction, such as the sinker tube and sinker tube chain and rope.

236

Conclusions This study presents a qualitative evaluation of ‘escape-critical’ damages to nets from a variety of sea-cage designs, operations and fish species. Questionnaires and inspections showed that the causes of net damage are complex and that they vary among countries. Environmental conditions differ between countries; Spanish aquaculture facilities are situated in areas with high energy seas that experience several severe storms per year, whereas Norwegian farms are usually located within more sheltered locations. However, farms in the middle and northern areas of Norway may experience high current loads and ice on the nets and floaters. Irish conditions differ greatly from the enclosed sea-lochs and fjords of Norway and Scotland as the majority of farms are located off the west coast, in high-energy, exposed areas. The study covered several types of the sea-cage designs, although the dominant structure was a circular design with vertical net walls between 12 – 20 m deep and a conical bottom. Other cage designs were square or octagonal, with flat bottoms, circular nets, inclined walls and conical bottoms. A design that is custom-made, for one of the salmon producers in western Norway, has inclined walls in order to avoid contact of the netting against the chains attached from the floater down to the sinker tube. This new net pen design has been successful so far according to the manager in charge of the fish farm. In Norway, most of the net damage comes from wear and tear, especially near and above the waterline and near the bottom line, and cod biting behaviour. In Ireland, most of the damage occurs at, or above, the waterline. The farmers reported that leaving the nets loose on the cage during emptying resulted in damage, as wind causes the net to rub with the floaters and stanchions. They also reported predator attack, i.e. seals, as another main cause of small holes and abrasion. However, if the net walls are tensioned correctly, the attacks are minimised. Differing individual weighting systems were observed on the individual sites and it is thought that some improvements to weighting may reduce the damage due to predators. No escape or near-escape incidents occurred in these locations during the monitoring period. The Spanish situation differed greatly to Norway and Ireland. Over 95% of the holes and damages “routinely” encountered on the aquaculture nets corresponded to cages containing seabream. Their frequency increased during the spring-summer season and relate to fish biting habits. Fish biting was almost negligible in seabass facilities; any damage detected in these farms was likely related to the effects of the environmental loads and/or operational causes. The survey also showed that, for both fish species, net damages larger than 20 cm occur due to a number of causes such as: wear and tear due to friction between side/weight ropes and the net; environmental pressures and subsequent tearing of weak net mooring points; cannibalistic feeding habits on dead individuals lying on the bottom of the net, poor operational protocols and accumulative damage over time.

www.preventescape.eu

237

Recommendations • Producers should be encouraged to introduce Standard Operating Procedures (SOPs) for net maintenance. • Current SOPs should be reviewed. • Maintenance data should be kept for all nets. • Cage design should be improved based on the data showing where and how net damages occur. • The level of competence of the farm workers should be raised. • Carry out further investigations into updating the weight systems for nets. • The biting behaviour of seabream and cod should also be further studied.

238

References

cited

Bahr NJ (1997) System safety engineering and risk assessment: a practical approach. Philadelphia, U.S.A. Taylor & Francis, 251 pp. Gabiña G (2011) Damages on sea-cage nets. Internal report 23/03/2011, AZTI Technalia. ISO 14121-1:2007 Safety of machinery – Risk assessment – Part 1. Jensen Ø, Dempster T, Thorstad EB, Uglem I, Fredheim A (2010) Escapes of fishes from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquacult Environ Interact 1:71-83 Jensen Ø, Føre HM, Endresen PC (2011) "6.2.2 REPORT - Mechanical properties of net and ropes". Report on task 6.2.2 Prevent Escape, SINTEF Fisheries and aquaculture. Helle TA (2011) "Risk". Workshop presentation, 31 May 2011. (in Norwegian) Directorate of Fisheries. Moe H, Olsen A, Hopperstad OS, Jensen Ø, Fredheim A (2007) Tensile properties for netting materials used in aquaculture net cages. Aquacult. Eng. 37:252-265. Moe H, Gaarder R, Sunde LM, Borthen J, Olafsen K (2005) Escape-free nets for cod. SINTEF Fiskeri og havbruk Report SFH A 054041, Trondheim, Norway (In Norwegian). NS9415:2009. Marine fish farms. Requirements for site survey, risk analysis, design, dimension ring, production, installation and operation. Standard Norway.

www.preventescape.eu

239

6.3. Assessment

of escape critical loads and

damages from fish bite Cite this article as: Høy E, Volent Z, Moe H, Dempster T, Arechevala P, Sanchez-Jerez P (2013) Assessment of escape critical loads and damages from fish bite. In: PREVENT ESCAPE Project Compendium. Chapter 6.3. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

Authors: Erik Høy1, Zsolt Volent1, Heidi Moe1, Tim Dempster1, Pablo Arechevala2 & Pablo Sanchez Jerez2 1 2

SINTEF Fisheries & Aquaculture, Norway, University of Alicante, Spain

Introduction Within Norway, official statistics suggests that after the Norwegian technical standard (NS 9415) for sea-cage farms took full effect in 2006, the total number of escaped Atlantic salmon (Salmo salar) declined substantially, yet no similar decline has occurred for Atlantic cod (Gadus morhua; Jensen et al. 2010). This suggests that some causes of cod escape differ to those of salmonids (Moe et al. 2007a) and that different measures must be taken to reduce the amount of cod lost to the environment. Holes in the net play an important role in most escape episodes and are the cause of the greatest proportion of cod escapes by number (Jensen et al. 2010). Holes may form due to overloading of the sea-cage, abrasive contact with objects, from predator damage or through biting of net materials by the cultured species itself (Pemberton and Shaughnessy 1993, Moe et al. 2007b). Cod display different behaviour than salmon and they may actively escape through small holes in the netting (Hansen et al. 2008); this may lead to a ‘continuous leakage’, which is believed responsible for a considerable part of the total number of escapees (Moe et al. 2007a). In an assessment of commercial net-cage diver inspection logs across the Norwegian cod farming industry, Moe et al. (2008) estimated that small holes formed on average once per month per net cage. The small holes observed in cod cages may develop, or be accentuated as a result of biting of the nets if their initial appearance is due to other causes (Moe et al. 2007a, 2009). Cod spend a lot of time close to the net wall (Rillahan et al. 2010) and they seem to explore structures by biting or nibbling. The incidence of biting of the net wall may depend on factors such as feeding regimes, the distribution of feed, the level of biofouling and other factors (Moe et al. 2008). For species that display biting behaviour or spend a large proportion of their time exploring the net walls and bottom, a greater understanding of how the formation of holes can be prevented, through improved netting materials or operational procedures, is essential to prevent loss of farmed fish into the wild.

240

A procedure to test and compare the resistance of netting materials to repeated biting has been suggested (Moe et al. 2009). As data on the biting and pulling forces of cod were previously unavailable, this procedure assumed that cod pulled with a force corresponding to its weight in air, as occurs in other fish species (Steinberg 1963). Knowledge of both the frequency and force of the pull related to cod bites will improve the quality of such test results.

Objective We aimed to develop a method to measure lateral pulling forces exerted on nets by Atlantic cod during biting events. We then compared lateral pulling forces on two different netting materials from two different sizes of cod.

Materials

and methods

Experimental setup Trials were conducted in May 2010 over a 10-day experimental period in two tanks (2.5 m diameter, 1 m deep) with a continuous flow-through of seawater of 6-7째C and 70-80% oxygen saturation (Figure 6.3.1). Fish were held in the tanks for one month and fed rations of 1% of total biomass per day, following a schedule of feeding for two days, then non-feeding for two days. Periods of non-feeding were necessary in order to obtain an adequate number of biting events. The fish were farm-raised Atlantic cod obtained from Atlantic Codfarms AS and were of two different size distributions. At the start of the trial, one of the tanks were stocked with a group of about 300 cod of mean weight 178 g (hereafter referred to as 180 g), the other with 150 cod with mean weight of 609 g (hereafter referred to as 610 g). As tank volume was 4 m3, stocking densities were 13 kg m-3 for the 180 g fish and 23 kg m-3 for the 610 g fish. Figure 6.3.1. (A) Bite force sensor design, (B) example data indicating forces measured during a biting event, (C) set-up of the bite force sensors within the experimental tanks.

www.preventescape.eu

241

Bite force measurements A sensor for automatic logging of the lateral pulling force was built to obtain a large set of undisturbed, objectively collected bite events. The lateral pull measurement system consisted of a fixed, circular neodymium magnet (Sura Magnets N35, 5 x 5 mm) and a hall-effect sensor (Allegro A1321LUA-T) attached to the end of a rod of spring steel (3 mm Ø, 190 mm length; Figure 6.3.1). The hall-effect sensor measures magnetic field strength, and thus the position of the steel rod relative to the fixed magnet could be recorded. The sensor signal was recorded by a data acquisition device (LabJack U3-HV) and a logger system was developed and implemented in LabView code. Each sensor was then carefully calibrated with known forces applied with a 0 – 1000 g spring scale (Pesola Medio). A total of four sensors were used, two in each tank (Figure 6.3.1). The system was operating at 10 Hz and when a lateral pull exceeding a certain threshold (0.9 N) was registered, the system entered an “event log” state storing the last 3 s of history from a buffer and the next 3 s after triggering. As well as the high resolution event logs, the system recorded a continuous baseline log at a rate of 1 Hz to keep track of possible drift in the signal at zero loads. On one occasion, a video camera was deployed in the tank with the 610 g fish to monitor one of the nets. The video was later matched to the logged events so that registered biting events could be validated with video of the fish behaviour. Two different nylon netting panels where attached to the sensors to test if net type affected the pulling force from cod bite: a) 200 x 200 mm sheet of 20 mm bar length mesh size (mesh half-width), 1792 tex, black, nylon, hard laid, twisted line, knotted netting (hereafter referred to as twisted twine; TT): b) a 200 x 200 mm sheet of 30 mm bar length mesh size, 1688 tex white, nylon, medium laid, braided twine, knotless netting, (braided twine; BT).

Data processing and statistical analysis The 10 Hz event logs were analysed with respect to the power of each pull or tug at the net. The magnitude of a single pull on the net was defined by the value of the recorded spike. Some bite events were recorded as prolonged pulls which could span over several tenths of a second

242

and contain several peak values. These where then classified as continuous pull-events and characterized by their maximum peak value (see example in Figure 6.3.1b where 11 pulls are recorded and pulls 8 â&#x20AC;&#x201C; 11 represent a continuous pulling event). To test for differences in bite force exhibited by cod of different size, among the two net panels used and between singular and continuous pulls, we used independent Student's t-Tests to compare the mean of groups. As bite forces of the 180 g and 610 g cod differed, we tested whether bite forces of cod differed on the two net panels and between singular and continuous biting events separately for the 180 g and 610 g fish.

Results Observations of biting Video of the biting cod showed a behavioural pattern where cod actively targeted the net. Cod approached from below and first touched the net with the chin barbells before grabbing and biting into the net. To complete a biting event, cod commonly opened their mouth and expelled the net before swimming away. If the net became caught in the teeth, cod made vigorous, lateral head movements to free themselves from the net. Fibres in the twine that were caught in the teeth could break during the pull.

Bite pulling forces Over the 10-day experimental period, the bite force measurement sensors recorded 571 and 486 biting events for 180 g and 610 g fish, respectively. 36% and 42% of the recorded biting incidents had a measured bite force of â&#x2030;¤1.2 N (Figure 6.3.2a). The strongest 5% of all pulls averaged 3.7 N and 6.6 N for the 180 g and 610 g fish, respectively, with maximum recorded forces exerted of 5.9 N, and 9.4 N, respectively. When the average values for the 5% of the strongest pulls are compared to fish weight in air, this corresponds to 204% of the body weight in air for 180 g fish and 106% for the 610 g fish.

www.preventescape.eu

243

On average, 610 g cod made more forceful bites (mean ± S.E.; 2.6 N ± 0.05) than did 180 g cod (1.95 N ± 0.04; Fig. 6.3.2b). Net type affected bite force for both the 610 g and 180 g cod. The loosely laid, braided netting attracted 1.5 – 1.6 times more forceful bites on average from the 610 g cod (2.88 N ± 0.06) than did the hard laid, twisted twine netting (1.86 N ± 0.09; Figure 6.3.2c). Similarly, significant differences were detected in the bite forces exerted on different net types by 180 g cod (Figure 6.3.2c). Twisted twine netting (1.60 N ± 0.09) was subject to bite events with significantly less force exerted than towards the braided twine netting (2.14 N ± 0.05).

Fig. 6.3.2. (a) Frequency distribution of bite events by force for the 180 g and 610 g cod, (b) bite force by cod weight, (c) bite force exerted by 180 g and 610 g cod on the two different net panel types, and (d) bite force exerted by 180 g and 610 g during singular and continuous biting events. BT = braided twine net panel, TT = twisted twine net panel. S = bite event characterised by a single pull; C = bite event characterised by a continuous pulling at the net over several seconds (see example in Fig. 1). Data points for (b), (c) and (d) are mean±S.E.

244

Through analysis of the recorded bite sensor data, events were identified as continuous biting events or single bites to the net (see example in Figure 6.3.1b). For 610 g cod, continuous biting events were 1.9 times more forceful (4.54 N ± 0.38) compared to single bites (2.42 N ± 0.06) to the net (Figure 6.3.2d). For 180 g cod, a similar result occurred with continuous bites 1.4 times more forceful (2.64 N ± 0.14) than single bites (1.88 N ± 0.04) to the net (p< 0.001; Figure 6.3.2d).

Discussion Bite pull sensor method and relevance to sea-cage aquaculture The sensor we developed for the trials proved effective and robust by logging approximately 1000 bite events over the 10-day experiment period. While most bites involved relatively low pulling forces (< 2 N), strong bite pulling forces of up to 9.4 N were recorded. The breaking strength of a single fibre of nylon netting material is about 0.5 N (Moe et al. 2009). Thus, all of the recorded bite events were of a force sufficient to break multiple fibres. Sea-cage walls will vary in tightness, from extremely taut to loose in some sections when deformed in currents (Lader et al. 2008). In this trial, the items tested were loose hanging sections of netting. Thus the pulling forces due to biting measured in this study are relevant to actual farming conditions as cod bite damage is normally found in areas of loose netting, where there are loose ends between joins of different net panels or where structural, loadbearing ropes are sewn into the net (Moe et al. 2007a).

www.preventescape.eu

245

Differences in bite pull forces with cod size, net type and type of bite Larger cod (610 g) were responsible for more forceful pulls than the smaller cod (180 g), which corresponds with their larger total mass of muscle and larger fin-area. However, smaller cod pulled with a greater force relative to weight. The mechanism responsible for the greater pulling force relative to body weight of small cod remains unclear. The maximum load exerted towards the net by the fish will not solely be a product of the maximum pulling strength, but depend on the breaking strength of each individual fibre, the number of fibres caught in the teeth, and how easy it is to pull single fibres and undo the knitted or twisted construction of the thread. Different net material properties, such as the hardness of the knitted or twisted twine and different coatings, result in different degrees of damage from code bite events (Moe et al. 2009). Twines exposed to cod bite-events become frayed (Moe et al. 2009). In this trial, different net types influenced the recorded pulling force, depending on how readily the teeth of cod became entangled. The availability of single fibres was lower in the hard-laid twisted twine netting, whereas the braided twine net had a looser, Raschel knitted structure. Thus, it is unlikely the cod bit the braided net with more force per se, but greater entanglement of teeth in the fibres due to the structural properties of the net twine likely led to the greater measured pulling forces. We have not found any data which indicate that the colour of the nets influence on pull force. Colour might be a factor in motivation for interacting with the nets, but as this has been a technical study, the setup was not designed to evaluate behavioural aspects of cod biting. The measured pulling forces were higher for prolonged pulls, which is consistent with the mechanism of teeth becoming highly entangled in the net and the subsequent struggle to free the teeth and pull away.

Conclusion We have demonstrated that pulling forces applied to nets by cod during biting incidents can be measured empirically. With nearly 1000 analysed pulls recorded, our results shed new light on the mechanisms by which the biting behaviour of cod damages nets. Bite pulling forces measured for all recorded bites were sufficient to break several single fibres in standard nylon net twines. The structural properties of the net twine are an important factor in the mechanism of cod biting nets as load is applied directly to the fragile, single fibre structures when cod teeth become entangled in the netting. These results are relevant for the development of aquaculture nets and benchmarking their bite-resistance, when combined with additional information on the frequency and position of bites to nets. Several other species under industrial culture are known to bite nets (e.g. seabream; Sparus aurata) and similar methods could be applied to assess bite pulling forces exerted by these species.

246

References

cited

Hansen L, Dale T, Damsgård B, Uglem I, Aas K, Bjørn PA (2008) Escape-related behaviour of Atlantic cod, Gadus morhua L., in a simulated farm situation. Aquacult Res 40:26-34 Jensen Ø, Dempster T, Thorstad EB, Uglem I, Fredheim A (2010) Escapes of fishes from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquacult Environ Interact 1:71-83 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007a) Technological solutions and operational measures to prevent escapes of Atlantic cod (Gadus morhua) from sea cages. Aquacult Res 38:91-99 Moe H, Olsen A, Hopperstad OS, Jensen Ø, Fredheim A (2007b) Tensile properties for netting materials used in aquaculture net cages. Aquacult. Eng. 37:252-265 Moe H, Gaarder RH, Olsen A (2008) Codnet - Right Choice of Nets in Cod Aquaculture. SINTEF Report No. SFH80-A084016 (in Norwegian) Moe H, Gaarder RH, Olsen A, Hopperstad OS (2009) Resistance of aquaculture net cage materials to biting by Atlantic Cod (Gadus morhua). Aquacult. Eng. 40:126-13 Lader P, Dempster T, Fredheim A, Jensen Ø (2008) Current induced net deformations in fullscale sea-cages for Atlantic salmon (Salmo salar). Aquacult Eng 38:52-56 Pemberton D, Shaughnessy PD (1993) Interaction between seals and marine fish-farms in Tasmania, and management of the problem. Aquat Cons Mar Freshwat Ecosys 3:149-158 Rillahan C, Chambers MD, Howell HW, Watson WH (2010) The behavior of cod (Gadus morhua) in an offshore aquaculture net pen. Aquaculture 310:361-368 Steinberg R (1963) Monofilament Gillnets in Freshwater Experiment and Practice. Fishing Gear of the World 2. Fishing News Books Ltd., England, 111–115 pp Thorstad EB, Fleming IA, McGinnity P, Soto D, Wennevik V, Whoriskey F (2008) Incidence and impacts of escaped farmed Atlantic salmon Salmo salar in nature. NINA Special Report 36, 1-110 pp

www.preventescape.eu

247

6.4. Investigation

of the properties of materials

and technologies used in aquaculture Cite this article as: Jensen Ø, Endresen PC, Moe H, Mendiola D, Gabiña G, Mendia Huart L, Iturrioz JM (2013) Investigation of the properties of materials and technologies used in aquaculture. In: PREVENT ESCAPE Project Compendium. Chapter 6.4. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

Authors: Østen Jensen1, Per Christian Endresen1, Heidi Moe1, Diego Mendiola2, Gorka Gabiña2, Leire Mendia Huarte3 & Jose Miguel Iturrioz3 SINTEF Fisheries & Aquaculture, Norway AZTI-Tecnalia, Spain 3 INASMET-Tecnalia, Spain. 1 2

Introduction Two thirds of all fish escapes, in terms of both incidents and number of fish, occur through a hole in the net (Jensen et al. 2010). Holes form due to a number of reasons, the most common being: biting of the net by predators or the caged fish; abrasion; boat collisions; contact with flotsam, and cage handling procedures such as lifting. There are currently no written standards specifying the properties of the materials used in aquaculture, or explaining how the materials might degrade with time and cumulative wear and tear. This is particularly problematic for non-metallic materials such as polyethylene, polyamide and polypropylene; their functional properties are known to change with temperature conditions and over time.

Objective We aimed to determine the functional characteristics of commonly used aquaculture materials and establish objective methods for testing the materials used in sea-cage aquaculture nets.

248

Methods New and used net materials were collected from a number of net producers and fish farmers. Unfortunately, the history of the used net samples was generally not known; farmers do not keep fully detailed records about when they wash their nets (whether in-situ or onshore), whether or not the nets are coated, or details of any operations performed on the net during use that could influence the strength of the material. Some variation in the results is, therefore, to be expected. Strength reduction of new and used nets, with different mesh sizes and thread numbers, were compared. A number of factors can reduce the strength of the nets used in fish-farm cages, such as: onshore washing; coating; in-situ washing using high-pressure disks; creep during lifting operations; ultra-violet radiation, and abrasion. We tested how some of these factors would affect net strength. Nets may also be subject to extra loads from the weights or sinker-tube, which are used to maintain the shape and volume of the cage, plus any water currents. Continuous loads can introduce creep effects, especially on plastic materials such as nylon and Dyneema (http:// en.wikipedia.org/wiki/Creep_(deformation)). In addition, waves often introduce motions to the floating collar that propagate as dynamic tensions in the net. To investigate how this may affect the break-strength of the net, nylon and Dyneema samples were tested at increasing rates of deformation. The effect of onshore machine washing was investigated by washing net samples together with net cages that had been used to farm salmon for one production cycle. After washing the net panels 1 â&#x20AC;&#x201C; 5 times the mesh strength was tested. During summer and early autumn months the net cages, especially in mid- and southern parts of Norway are washed, using high pressure disks, as often as once per week to remove biofouling. Nylon and Dyneema net samples were washed using industry standard equipment to investigate the effect on mesh strength.

www.preventescape.eu

249

Testing procedure Numerous net materials were tested in a uni-axial tensile test machine (Figure 6.4.1). Two different test types were used to test the strength of the material; mesh strength tests (Figure 6.4.2) and uni-axial tensile tests to determine the tensile strength of the thread as well as the elasticity of the material (Figure 6.4.3). Tensile tests were conducted on 50 cm lengths of net. Pure nylon netting was cut while wet, whereas netting containing Dyneema fibres were cut while dry, using a butane powered cutting tool, to prevent filaments from slipping through the knots at the thread ends. Unless otherwise stated, samples were immersed in tap water at 20 – 23°C for at least 24 h prior to testing. Nylon and Dyneema specimens were subjected to a permanent load for 30 minutes. After seven days of relaxation the break-strength of the specimen was determined to investigate whether the 30 minutes creep-load had weakened the material. Quasi-static mesh-strength tests (ISO 1806:2002; ISO 2002) were performed on new and used material to serve as a reference for other tests. Equivalent test settings were adopted for the uni-axial tensile tests. Test methods were developed to investigate the influence from washing in-situ with high pressure cleaning disks (Figure 6.4.4), and load duration – both creep and dynamic tests. Two different nylon materials (N10 and N11), with (C) and without coating (U), as well as coated Dyneema (N12_C) net were subjected to one hour of controlled abrasion (Figure 6.4.5), and the strength of the material tested using uniaxial tensile tests.

Figure 6.4.1. Test machine used to determine net strength and elongation.

250

Figure 6.4.2. Test specimens for mesh strength testing.

Figure 6.4.3. Test specimen used in uni-axial tensile test.

Figure 6.4.4. High pressure disks used to clean net cages.

www.preventescape.eu

251

Figure 6.4.5. Equipment used to inflict controlled abrasion damage to net materials.

Results

and discussion

Net type Tests were performed on samples of the type of new materials that are supplied to the Spanish fish farming industry (Figure 6.4.6), and on used net samples from European seabass (Dicentrarchus labrax) and gilthead seabream (Sparus aurata) cages (Figure 6.4.7). A 30% decrease in the break-load and a 15 â&#x20AC;&#x201C; 40% decrease in the length of elongation at rupture were found in nets that had been used for one production cycle (Figure 6.4.6). The Dyneema nets were found to have much higher break-load and stiffness compared to the nylon nets. The difference between nets from different producers was relatively small. The mesh strength of nets used to farm Atlantic cod (Gadus morhua) and Atlantic salmon (Salmo salar) were determined (Figures 6.4.8, 6.4.9 and FigureÂ 6.4.10). There was no obvious correlation between production year or the number of wash cycles and degradation in the strength of the net. Figure 6.4.6. Mesh strength and elongation at break - virgin and used Spanish net materials.

252

Figure 6.4.7. Mesh strength and elongation at rupture - new nets from different producers.

Figure 6.4.8. Mesh strength for nets of varying age.

www.preventescape.eu

253

Figure 6.4.9. Changes in net strength due to washing of the net.

Figure 6.4.10. Force vs strain curve for different net materials.

254

Creep loads Tests showed that a 30 minute creep-load had a positive effect on the break-load for the Dyneema material (N6; Figure 6.4.11). For the C90 case, (i.e. where the creep-load is equal to 90% of the break-load), the break-strength increased by close to 20%. An opposite effect was observed for nylon nets (N7 and N8) where the break-strength was reduced by 5 – 10%. Figure 6.4.11. Stress at break divided by stress at break for virgin material.

Mesh

strength and deformation rates

The relative mesh strength, normalized against the mesh strength reported in ISO 1806:2002 (ISO 2002), of one of the two nylon nets tested, increased by 10% as the rate of deformation increased (Figure 6.4.12). However, no significant change was detected for the second nylon sample and the Dyneema mesh strength was reduced by up to 10%. The highest rates of deformation tested gave a total load time of 1.2 second for the Dyneema net, 1.6 seconds for one of the nylon nets, and 2.7 seconds for the other. These values are within the range of

www.preventescape.eu

255

loads expected due to wave action. Snap loads will have even shorter load times. These results indicate that short periods of extra load bearing, such as from waves, should not be a problem for nylon nets, but suggest that for Dyneema nets these loads may be problematic, particularly if snap loads with even higher rates of deformation occur.

Figure 6.4.12. The effect of rate of deformation on relative mesh strength.

Washing Washing of nets led to a distinct drop in strength after only one wash (Figure 6.4.13). A reduction in strength of 20% was observed after the net panels had been washed four times. According to the Norwegian technical standard NS9415 (Standard Norge 2009) net cages must be discarded if the remaining strength is reduced by 35%. Net washing therefore accounts for more than half of the allowable strength reduction. In comparison, there was no reduction in strength even after 40 wash cycles of in-situ high pressure cleaning (Figure 6.4.14). However, escape incidents have occurred during high pressure cleaning, especially in connection with the use of cranes, but these incidents have occurred due to the incorrect use of equipment by their operators.

256

Figure 6.4.13. How the washing of the net effect the mesh strength.

Figure 6.4.14. Effect of high pressure claning on the mesh strength.

www.preventescape.eu

257

Abrasion Following abrasion testing, the reduction in strength was significant for all of the tested nets (Figure 6.4.15). The coating had a positive effect on the abrasion resistance of nylon nets. Even though the tensile strength of new Dyneema net is higher than that of nylon, the relative reduction following abrasion was far larger than for the nylon nets. This result may reflect the fewer filaments in the Dyneema thread.

Figure 6.4.15. The effect on abrasion on tensile strength.

Recommendations • Used net materials showed a significant reduction in strength, even after just one production cycle. However, no clear relationship was found between the duration of net use and the subsequent reduction in strength. • High pressure cleaning had no effect on mesh strength. • Cleaning of nets in washing machines lowered the strength of the nets by up to 20% after 4 wash cycles. • Abrasion significantly reduced the strength of the nets. Coating limited the reduction in strength, but could not eliminate it.

258

References

cited

Jensen Ă&#x2DC;, Dempster T, Thorstad E, Uglem I, Fredheim A (2010) Escapes of fish from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquaculture Environment Interactions 1:71-83 The International Organization for Standardization, 2002. ISO 1806. Fishing nets -Determination of mesh breaking force of netting. Standard Norge (2009) Norwegian standard NS 9415.E:2009 Marine fish farmsâ&#x20AC;&#x201D;requirements for site survey, risk analyses, design, dimensioning, production, installation and operation. Standard Norge, Lysaker

www.preventescape.eu

259

6.5. Sea-load

exposure

Cite this article as: Jensen Ø, Lader P, Kristiansen D, Mendiola D, Gabiña G, Sanz V, Rico A (2013) Sea-load exposure. In: PREVENT ESCAPE Project Compendium. Chapter 6.5. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Østen Jensen1, Pål Lader1, David Kristiansen1, Diego Mendiola2, Gorka Gabiña2, Veronica Sanz2 & Antonio Rico2 1 2

SINTEF Fisheries & Aquaculture, Norway AZTI-Tecnalia, Spain

Introduction Escapees can have negative ecological and genetic effects on populations of wild fish, and the present level of escapees is regarded as too high and a problem for the sustainability of seacage aquaculture (Naylor et al. 2005). The escapes problem is largely caused by technical and operational failures of fish farming equipment. Atlantic salmon (Salmo Salar), rainbow trout (Oncorhynchus Mykiss) and European seabass (Dicentrarchus labrax) primarily escape after structural failures of containment equipment, whereas Atlantic cod (Gadus morhua) and gilthead seabream (Sparus aurata), escape due to holes in the net made by the farmed fish as well as predators (Jensen et al. 2010). Structural failures is by far the dominating cause of escapes and may be generated by severe environmental forcing in strong winds, waves and currents, which may occur in combination with how the components have been installed or operated (Jensen 2006; Jensen et al. 2010).

260

Objective The objective in this task was to study the dynamic behaviour of typical nets and cages used in European aquaculture. The emphasis was to develop knowledge on how extreme weather loads lead to loss of structural integrity. The model tests focused on net deformation and how holes in the net form due to contact between net cage and weight system whereas the numerical simulations was motivated by a need to better understand how multiple cages in a system behave when subjected to waves and current.

Methods Model tests were performed in a tow tank at the United States Naval Academy to investigate at which combinations of waves and current the net deforms and come in contact with the sinker tube chain. The wave considered in this study is a severe swell with a full scale period of 12.7 s and a wave height of 5.6 m. A model in scale 1:40 was used to represent a fish cage with a circumference of 120 meters and net depth of 40 m. The cage model was composed of a model net, a flexible floating collar of the circular plastic type and a weight system. Two different net design and four different weighting systems were tested (Figure 6.5.1, Figure 6.5.2). The model was moored to a towing carriage and subjected to regular waves in the small wave tank. Multiple tow tests were conducted in a single run by starting the towing carriage at the lowest velocity, and by stepwise increasing the speed until all the conditions were conducted. Tension forces in the mooring lines were measured during the towing at a sampling rate of 10Â Hz. The model was filmed by an underwater camera placed at the side of the model pointing perpendicular to the towing direction so that the geometry of both the cage and the weight system was visible. A modified floating collar design with purpose to reduce the probability of contact between support chains and net was also tested. The modified collar had a larger diameter of the outer ring compared to the conventional floating collar. This caused the separation between the support chains and the net to be increased. Detailed numerical simulations of a typical Spanish fish farm were performed to investigate the structural integrity of the floating collar and the mooring grid. Wave and current parameters were chosen to represent condition found at the Canary Island, Mediterranean Sea and Atlantic Ocean. Detailed models were developed to represent typical mooring grid (Figure 6.5.3) and net cage designs. For more details on the numerical models see.

www.preventescape.eu

261

Figure 6.5.1. Sliding and fixed connection between weight system and lower end of net cage.

Figure 6.5.2. Net and weight system test setup.

262

Figure 6.5.3. Model used for mooring analysis.

www.preventescape.eu

263

Results In Figure 6.5.4 and Figure 6.5.5 the net deformation for the different combinations of net design and weight system when subjected to current and waves are presented. Both figures show clearly how the deformation increases with increasing current. This follows the same pattern of deformation with increasing velocity as reported in previous studies (Lader and Enerhaug 2005), the front of the cage deforms more than the rear and the bottom rises due to the front and rear deformation. For the case where the model was subjected to waves, the two pictures shown were taken when the model had the largest horizontal displacement and deformation, approximately when the wave crest (bottom picture) and wave through (top picture) passed the centre of the cage. In the figures the contact between the net and the weight/sinker tube chain is indicated by a thick black line. This contact was found by visual

Figure 6.5.4. Deformation and contact for a cylindrical net in current and waves.

264

Figure 6.5.5. Deformation and contact for a conical net in current and waves.

inspection of the images and is thus somewhat approximate, but gives nevertheless a good impression of the amount of abrasion in each condition. The sinker tube chain on the front side goes into slack, and this represent a potential for snap loads which can cause dangerous high loads in the chain (Lader and Fredheim 2006). The experiments showed that the sinker tube performs better than individual weights. Both for a cylindrical net in current only and for conical net in current and waves, contact occurred at a lower current velocity when using individual weights compared to using a sinker tube (see Figure 6.5.4 and Figure 6.5.5). In addition it was observed that contact occurred at a higher velocity and over a smaller area when using the conical net compared to the conventional cylindrical net. This was independent of weighting system (sinker tube or individual weights, sliding or fixed) and true for both current only and combination of waves and current. No difference was observed, between the sliding and fixed condition used to attach the net to the sinker tube or individual weights. This final observation may, however, be influenced by how the sliding connection was modelled in the experiments. If the friction between the sliding connection and the modelled sinker tube chain were too large, compared to the full scale cage, the overall behaviour of the system may have been different for the model compared to a full scale cage. Additional results are presented by Lader et al. (in preperation).

www.preventescape.eu

265

The modified floating collar with increased diameter of the outer ring was tested using the conical shaped net. Deformations of the net at the downstream side of the cage lead to contact between the net and the support chain for the two largest towing speeds, as with the conventional floating collar. However, the model tests suggests that the modified cage design will reduce the probability for contact at operating conditions of full scale cages as the velocity where contact first occurred was significantly higher, see "Figure 6.5.6 Net deformation and contact with weight system when using modified cage with larger outer ring". Hence, contact between the net and the support chains should be expected for typical full scale net cages at least for a dimensioning current speed of 1 m s-1. Similar results from the 2009 tests with the normal floating collar design and the same net model, showed that contact between the support chain and the net occured for current velocities equal to 0.5 m s-1 and above. Hence, the modified collar design yields a significant improvement of the contact problem relative to the normal collar design.

Figure 6.5.6. Net deformation and contact with weight system when using modified cage with larger outer ring.

Based on the experiments it can be concluded that independent of net and weight system design, contact will occur between the sinker tube chain and the net even at moderate current levels with the conventional net, floating collar and weight system design. This observation is confirmed by recent incidents in Norway; more fish have escaped due to hole in the net (Figure 6.5.7). Contact between the net and the sinker tube chain has occurred at multiple instances (Figure 6.5.8) and abrasion damages on the side of the net has been detected during net inspections (Figure 6.5.9). From the pictures it can be seen that the biofouling has been removed by the chain rubbing against the net. The strength of the net was tested in the laboratory, and it was confirmed that the contact had introduced a significant reduction of the strength of the material (Chapter 6.2.2, this compendium). In addition to visual observations of net deformation, force acting on the net cage were measured. The dominating hydrodynamic forces acting on the structure due to the waves can be divided into viscous drag forces on the net, and Froude-Kriloff, diffraction, radiation and viscous drag forces on the floating collar. Mean peak to peak force amplitudes for the conical shaped net are obtained from time intervals of about 10 wave periods from the measured time-series. Obtained values are compared with estimated theoretical values in Figure 6.5.10. There are several wave periods where cancellation of the wave-induced horizontal forces occurs. For the model tested, effects of force cancellation are found to be most pronounced for and where, according to theory, the total horizontal force has a local minimum. There is a local maximum of the wave-induced horizontal forces when. Correspondingly for the full scale cage, cancellation of wave-induced forces occur for the wave periods and, while the

266

Figure 6.6.7 Contact between net cage and weight system at a commercial farm - hole in the net.

Figure 6.5.8. Contact between net cage and weight system at a commercial farm.

Figure 6.5.9. Contact between net cage and weight system at a commercial farm - biofouling removed.

www.preventescape.eu

267

wave period corresponds to the local maximum. The wave induced drag on the sinker tube is negligible due to the exponential decay of the particle velocity with depth. The effect of net deformation is neglected when computing the wave induced drag force on the net cage. Current-induced drag forces on the fish cage model were investigated for the two net design with different weight configurations. The conical net experiences significantly lower drag forces compared to the cylindrical net, primarily due to a reduced exposed area, see Figure 6.5.11. The drag coefficient on net panels shows a strong dependency on the Reynolds number (Rn) when Rn is small, typically Rn < 100 (Fridman 1986). This effect of Rn on the induced drag is observed for the lowest test velocities in Figure 6.5.12 as an increase of the non-dimensional drag force with decreasing current velocity. As the deformations of the net increases with increasing current velocities, the projected area of the net in the towing direction is reduced. The result is a reduced drag force compared to that of an undeformed net for the same current condition, as illustrated by the solid and stippled lines in Figure 6.5.13 obtained from theory. In Norway, all main components of a fish farm (e.g. floating collar, net cage, mooring system and feed barge) are required by legislation to be product certified according to the technical standard NS9415. In the standard the significant wave height and corresponding peak period are calculated based on wind velocity and measured effective fetch lengths. It is not required that the designer consider how the calculated dimensioning wave lengths correspond with the geometry of the cage or system of cages. As Figure 6.5.10 shows, combinations of cage geometry and wave conditions can give cancellation of force, which in a design and dimensioning process could be critical as other combinations of wave length and height which at first glance appear less critical can give significantly higher forces on the system. It was found that there are many wave periods where cancellation of wave induced forces on the model occur. These cancellation wave periods are within the range of dimensioning wave periods commonly used for testing of fish farm structures and hence are important to be aware of. The simulations showed that when current and wave direction have the same angle of attack, three locations (Canary Island, Mediterranean Sea and Atlantic Ocean) have common lines which collapse, with the exception of the Mediterranean Sea, which have fewer number of lines that break. In fact, failures occur in the rope section, not in chain lines. In general, the environmental conditions are worse at the Atlantic Ocean location, causing the mooring line

268

loads to be larger than at the Canary Island and Mediterranean Sea locations. According to the simulations, the cages will deform prior to the bridle lines reaching its breaking load. The integrity of the cage is only guaranteed when the supported stress is below a third of the breaking load from grid to cage lines. It is necessary to remark that the weakest cage components are the brackets due to the manufacturing process, where the material properties are weakened. These brackets are the origin of cage collapse due to the appearance of cracks. This is consistent with findings in the Spain and Norway industries (Figure 6.5.13).

Figure 6.5.10 Drag forces during towing tests on different net and weight system designs.

www.preventescape.eu

269

Figure 6.5.11. Wave induced drag forces on net cage.

Figure 6.5.12. Drag coefficient as a function of Froude number.

270

Figure 6.5.13. Cracks in bracket.

Recommendations • Equipment used at a site should be designed and dimensioned, using validated methods, to ensure that the equipment is suitable to withstand the environmental conditions at the site. • It should be validated through analysis or by other means that equipment used at a site (such as floating collar, net, weighting system and mooring) fit together without potential of the individual parts damaging other parts. • Effort should be put into designing new solutions for net cages, floating collars and/or weight system since systems used today often experience abrasion between net and weight system even at moderate current velocities. • Mooring grids should be designed to withstand the expected environmental conditions at a site, the environmental conditions should be determined using appropriate methods and the mooring grid should be designed in such a manner that failure of one mooring line should not lead to a total loss of integrity of the entire farm.

www.preventescape.eu

271

References

cited

Fridman A L (1986) Calculations for Fishing Gear Designs. Fishing News Books Limited. Jensen Ø (2006) Assessment of technical requirements for floating fish farms—based on escape incidents January 2006. Rep no SFH80 A066056. SINTEF, Trondheim (in Norwegian) Jensen Ø, Dempster T, Thorstad E, Uglem I, Fredheim A (2010) Escapes of fish from Norwegian sea-cage aquaculture: causes, consequences and prevention. Aquaculture Environment Interactions 1:71-83 Lader, P. and B. Enerhaug (2005). Experimental Investigation of Forces and Geometry of a Net Cage in Uniform Flow. IEEE Journal of Ocean Engineering 30(1). Lader, P. and A. Fredheim (2006). Dynamic properties of a flexible net sheet in waves and current - A numerical approach. Aquacultural Engineering 35(3): 228-238. Lader P, Kristiansen D, Jensen Ø, Fredriksson D., (In preperation) Experimental study on the interaction between the net and the weight system for a gravity type fish farm.

272

6.6. Recapture of escaped the Mediterranean Sea

seabass and seabream in

Cite this article as: Bayle-Sempere JT, Sanchez-Jerez P, Fernandez Jover D, Arechavala Lopez P (2013) Recapture of escaped seabass and seabream in the Mediterranean Sea. In: PREVENT ESCAPE Project Compendium. Chapter 6.6. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

authors: Just T. Bayle-Sempere, Pablo Sanchez-Jerez, Damian Fernandez-Jover & Pablo ArechavalaLopez 1

Department of Marine Sciences and Applied Biology, University of Alicante, Spain.

Introduction The problem of escaped seabream and seabass Farmed fishes may escape into the wild due to technical or operational failures of farm facilities. Consequently, escapees could mix with local stocks, leading to negative genetic and ecological consequences through interbreeding, predation, competition for food or habitat and the transmission of pathogens to native populations (Prevent Escape Compendium Chapter 1). Within the Mediterranean Sea ecosystem, however, the knowledge concerning these potential harmful effects of escapees is still sparse, despite large numbers of seabass and seabream escapes (Prevent Escape Compendium Chapter 2; Figure 6.6.1 and 6.6.2). Results from the Prevent Escape project have identified that escaped seabass (Dicentrarchus labrax) and seabream (Sparus aurata) are not only able to swim away from farm facilities to nearby farms (Figure 6.6.2), local fishing grounds and coastal habitats (Figure 6.6.3, 6.6.4), but they can also exploit natural resources, such as food and habitat (Prevent Escape Compendium Chapters 4.3 and 4.4; Arechavala-Lopez et al. 2011, 2012). Moreover, if the number of escapees is high, they may bias estimates of wild populations if not accounted for. Therefore, it is essential to identify the extent of farmed escapees in native stocks and recapturing them should be a primary goal to avoid negative consequences. Wild seabass and seabream are commonly fished in the Mediterranean Sea, withstanding a considerable commercial and recreational fishing effort (e.g. in the Spanish Mediterranean Sea, about 3000 commercial fishing vessels target these species among others; MARM 2008). Consequently, using the capacity within commercial fisheries is a serious option to consider for the recapture of escaped seabass and seabream, and may complement other mechanisms of mitigation such as raising sterile individuals or ensuring low numbers of escapes by better management practices. Recapture techniques could also be implemented in regulations to reduce the dissemination of escapees, as they are in certain US states and in Scotland (Washington State Legislature 2012; Scottish Government 2012).

www.preventescape.eu

273

Figure 6.6.1. A school of escaped seabass (Dicentrarchus labrax) photographed in a shallow water coastal habitat on the Canary Islands after an escape event. Photo: Carlos Sangil

Figure 6.6.2. A school of escaped seabream (Sparus aurata) photographed beneath a fish farm on the Canary Islands after an escape event. The bottom of the cage is visible above the fish. Photo: Arturo Boyra

274

Figure 6.6.3. Seabream (Sparus aurata) in a natural sea grass habitat in the Mediterranean Sea. Photo: Pablo Sanchez-Jerez

Figure 6.6.4. Adult seabass (Dicentrarchus labrax) in a natural coastal habitat in the Mediterranean Sea. Photo: Maite Vazquez-Luis

www.preventescape.eu

275

Objective We aimed to develop and test suitable methods for recapturing escaped seabream and seabass.

Recapture capabilities of specific fishing gears Data on acoustic and external tagging and short-term behavioural observation of seabream and seabass escapees (Prevent Escape Compendium Chapters 4.3 and 4.4) provided valuable information for targeting their recapture after an escape event (Table 6.6.1). The results revealed that seabream catches of artisanal fisheries are composed, to a large extent, by escaped seabream, sometimes up to 75% of the total seabream catches in the reproductive season (early winter). The most successful fishing gear for recapturing seabream was gill nets, as evidenced by the data obtained in several acoustic tagging experiments where the fate of individual escaped seabream and seabass were followed through time (Arechavala-Lopez et al. 2011, 2012). Nets were most effective during the dark, when seabream moved close to the bottom and became vulnerable to capture by these nets. Recreational fishing, though occasionally successfully (Figure 6.6.5), had lower success catching escaped seabream than commercial fisheries. Consequently, after a massive escape event, local artisanal fisheries should be alerted to increase pressure immediately after the incident. Fish farmers may consider having one or more emergency gill nets prepared to be deployed as soon as possible around the farm to recapture escaped seabream. After escape, seabass suffered high mortality and their movement patterns and susceptibility to be fished differed to seabream. Mortalities were numerous within the first days and survivors tended to move towards the coastline. Recreational fishermen, mainly using fishing rods, were more successful capturing seabass than seabream. Therefore, we recommend that efforts to inform recreational fishers at local marinas, fishing shops or fishing clubs should be undertaken after large escape events. Moreover, the use of beach-moored barrier nets (e.g. something like the moruna, used along the Spanish Mediterranean coasts to catch seabass; GarcĂa Alcaraz et al. 2005) may be the most useful method in the case of a massive seabass escape incident, given the behavioural tendency of seabass to approach close to the shore after an escape. Similar gear types have proved partially successful in recapturing salmon (Chittenden et al. 2011). However, the efficiency of this fishing gear has not been tested for seabass and therefore deserves more research. During the Prevent Escape project, baited fish traps were also tested; these had marginal success in recapturing escaped seabass, seabream and meagre.

276

Figure 6.6.5. A professional catch of seabream (Sparus aurata) on the Spanish Mediterranean coastline after an escape event. Only 1 out of 9 fish was determined as a wild individual by analysis of scales. Photo: David Izquierdo-Gomez

Fishing gear type Species

Gill netting

Fishing rods

Moored beach gears

Traps

Seabream

+++/f

+/c

+/f

Seabass

+/c

+++/c

+/f

(+++): very effective; (+): less effective; f: deployed close to the fish farm and widespread some km around; c: used along the coastline

Table 6.6.1. Summary of gear types tested to recapture seabream and seabass escapees and their effectiveness.

Discussion While various fishing gears had greater or lower rates of success in recapturing escapees, all recapture attempts in the Prevent Escape Project proved only partially effective in recapturing escaped seabream and seabass. Despite relatively intensive fishing efforts, usually through simultaneous use of several fishing methods, we never recaptured more than 10% of escapees in any single simulated release. This low recapture rate is likely due to several reasons, including: 1) poor survival of the escapees after release, possibly due to the abundant wild fish predators

www.preventescape.eu

277

around fish farms; 2) rapid dispersal away from the point of escape, possibly due to escapees responding to reduce predation risk, which reduced the capture efficiency of fishing gears; and 3) under-reporting of recaptures by commercial and recreational fishers. Even if survival is poor and under-reporting of recaptures occurs, significant numbers of escapees are likely to enter coastal environments and begin to interact with their wild counterparts. Our results emphasize that the focus in preventing escapes should first and foremost be placed upon improving farming regulations, component technologies and operational routines so that fish do not escape in the first place, rather than relying on trying to recapture fish after they have escaped. While only partially effective, recapture fisheries may be a realistic option in specific circumstances to reduce the impact of escape events (Uglem et al. 2010, Skilbrei and JĂ¸rgensen 2010), although consideration must also be given to the likelihood that by-catch of nontarget species will be high before such measures are implemented. The most effective gears were gillnets for seabream and fishing rods for seabass. These results are supported by the fact that farm-aggregated wild fish are actively exploited by artisanal gill net fisheries in the Mediterranean (Akyol & Ertoluk 2010) and escape incidents are often followed by increased catches in these local artisanal fisheries, recording high recapture rates. Trawling appears to be ineffective in recapturing salmon escapees (Skilbrei & Jorgensen 2010) despite fish being present along the towing tracks. However, trawling may work better for seabream, as they are a demersal species, but this remains untested. Predation of escapees by wild fish appears to be important in the early stages after an escape event in rapidly reducing the number of escapees that survive (Arechavala-Lopez et al. 2011, 2012). Therefore, measures to ensure that populations of piscivorous predators are maintained around fish farms will assist in reducing the survival of escapees. Many Mediterranean seabream and seabass farms (e.g. Spanish Mediterranean farms) have a no fishing requirement within the leasehold area. This regulation will likely assist in mitigating escapes. We conclude that after a major escape event, a significant proportion of escaped seabream and seabass could be recaptured if fishing effort is implemented, both in the immediate vicinity of the farm and up to several kilometres away in likely natural habitats where the escapees may disperse to. The fishing efforts should last for several weeks to reduce the number of escaped seabream and seabass that survive.

278

Recommendations • As all methods of recapture are only partially effective and have implications for other species (e.g. by-catch), the primary focus of authorities should be upon implementing measures to prevent escape (Prevent Escape Compendium Chapter 7), with re-capture methodologies a secondary consideration. • Predation of seabream and seabass escapees by wild fish is important in the early stages after an escape event in rapidly reducing the number of escapees that survive. Therefore, spatial protection from fishing to ensure predator populations occur around fish farms will assist in reducing the survival of escapees and mitigate their effects. • Gillnets are the most suitable gear to recapture seabream. After a large-scale escape event, they should be deployed over several kilometres around the fish farm for several weeks. • Fish farmers should consider having one or more emergency gill nets prepared to be deployed as soon as possible around the farm to recapture escaped seabream. • Fishing rods and beach moored gear are more effective in catching escaped seabass near the coastline. Coastal fishers should be made rapidly aware of large-scale escape incidents.

www.preventescape.eu

279

References

cited

Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Bayle-Sempere JT, Sanchez-Jerez P (2011) Immediate post-escape behaviour of farmed seabass (Dicentrarchus labrax L.) in the Mediterranean Sea. J Appl Ichthyol 27: 1375–1378 Arechavala-Lopez P, Uglem I, Fernandez-Jover D, Bayle-Sempere JT, Sanchez-Jerez P (2012) Post-escape dispersion of farmed seabream (Sparus aurata L.) and recaptures by local fisheries in the Western Mediterranean Sea. Fisheries Research 121–122:126-135 Akyol O, Ertosluk O (2010) Fishing near sea-cage farms along the coast of the Turkish Aegean Sea. J Appl Ichthyol 26:11–15 Chittenden CM, Rikardsen AH, Skilbrei OT, Davidsen JG, Halttunen E, Skardhamar J, McKinley RS (2011) An effective method for the recapture of escaped farmed salmon. Aquacult Environ Interact 1:215-224 García Alcázar A, Chereguini O, Porta JM, Porta J, Álvarez Herrero MC, Abellán Martínez E (2005) Creación y caracterización de un banco de esperma de lubina Dicentrarchus labrax (L., 1758) de una población mediterránea. Bol Inst Esp Oceanogr 21:213-217 MARM (2008). Estadísticas pesqueras. Secretaría General Técnica. Subdirección General de Estadística. MInisterio de Medio Ambiente y Medio Rural y Marina, Madrid. 152 pp. Scottish Government (2012) What to do in the event of an escape of fish from a fish farm. http://www.scotland.gov.uk/Resource/0038/00387515.pdf Skilbrei OT, Jørgensen T (2010). Recapture of cultured salmon following a large-scale escape experiment. Aquacult Environ Interact 1:107–115 Uglem I, Bjørn PA, Mitamura H, Nilsen R (2010) Spatiotemporal distribution of coastal and oceanic Atlantic cod Gadus morhua sub-groups after escape from a farm. Aquacult Environ Interact 1:11–19 Tlusty MF, Andrew J, Baldwin K, Bradley TM (2008). Acoustic conditioning for recall/recapture of escaped Atlantic salmon and rainbow trout. Aquaculture 274:57–64 Washington State Legislature (2012) Marine finfish aquaculture - Escape reporting and recapture plan required. http://apps.leg.wa.gov/wac/default.aspx?cite=220-76-120

280

6.7. Recapture of escaped juvenile morhua) in Northern Norway

cod

(Gadus

Cite this article as: Serra-Llinares RM, Nilsen R, Bjørn PA, Noble C, Uglem I (2013) Recapture of escaped juvenile cod (Gadus morhua) in Northern Norway. In: PREVENT ESCAPE Project Compendium. Chapter 6.7. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

Authors: Rosa Maria Serra-Llinares1, Rune Nilsen1, Pål Arne Bjørn1, Chris Noble1 & Ingebrigt Uglem2 1 2

Nofima Marine, 9291 Tromsø, Norway. Norwegian Institute of Nature Research, Tungasletta 2, 7485 Trondheim, Norway.

Introduction Atlantic cod (Gadus morhua L.) is an important species for the Norwegian fish farming industry. Due to its highly exploratory behavior, cod seem more prone to escape than other cultured species (Moe et al. 2007). Between 2005 and 2010, on average 1.3% of the live stock of cod kept in the sea escaped annually, adding up to more than 1 million individuals over those years (Norwegian Directorate of Fisheries 2010). Escapes from fish farms are regarded as an environmental problem, and interactions between farmed and wild cod are likely to occur. Effective and practical routines to recapture escapees are urgently needed to reduce the ecological impact of cod escapes, should they occur. In addition, live recapture of escapees could reduce economic losses associated with escape events. The recapture rates of adult cod escapees through local commercial fisheries can be high, even with a small recapture effort (up to 52 % according to Uglem et al. 2008). We aimed to evaluate the efficiency of a recapture program aimed at juvenile cod escapees. For this purpose, a wide range of recapture methods were tested both immediately after a simulated escape and over the longer-term in a surveillance fishery. We simulates a series of escape incidents of cod in Nordfjord (67°07'N, 14°17'E), Gildeskål, Norway (Figure 6.7.1) in 2010 and 2011. During the study period, two cod farming locations were operating in the area, as well as a local, semi-commercial fishing fleet of 5 to 6 boats. Three escapes of approximately 1000 externally tagged juvenile cod each were simulated during the study (Table 6.7.1). The fish were tagged with external T-bar tags before being released from one of the farming locations. Each tag had the words “DUSØR UTBETALES” (“REWARD”) and the necessary contact information printed on it to promote the reporting of recaptures from the local community. Intensive recapture efforts were implemented immediately after the second and third releases. Live recapture methods were tested in the immediate vicinity of the farm during the first week post-escape. Recapture through local fisheries was encouraged through a reward program that continued for over six months after the last escape was simulated.

www.preventescape.eu

281

ilsen Rune 6 mar, 15:21 liminado: Espacio

Group

Total length (cm)

Weight (g)

Release date

Mean

1033

24.8

1.8

171.74

50.86

16 Sept 2010

874

28.9

2.5

331.25

10 Oct 2010

870

36.37

3.22

691.37

198.93

20 March 2011

Table 6.7.1. Overview of the three deliberate escapes of externally tagged Atlantic cod. Figure 6.7.1.Study area and the location of study farms.

Figure 6.4.1.Study 0 area and the location of study farms

Live recapture methods included: • six cod pots in three different sizes (1.8, 8 and 14 m3) distributed evenly around the farm (Figure 6.7.2); • two fyke nets fixed directly onto the net wall of the commercial net pens (Figure 6.7.3); • a net pen provided with a gate that could be opened and closed from the surface (“smart pen”); and • a large crane-operated dip net (Figure 6.7.4).

282

Figure 6.7.2. Fishing pots used for recapture of Atlantic cod (Gadus morhua) farm escapees.

Figure 6.7.4. The large crane operated dip net in use.

Figure 6.7.3. Fyke net used for recapture of Atlantic cod (Gadus morhua) farm escapees.

www.preventescape.eu

283

The idea behind the last two methods was to attract the escaped fish into the traps by offering commercial fish feed into the sea. In addition, two local semi-professional fishing vessels were hired to attempt to recapture juvenile escapees with gill-nets. The mesh size of the nets was chosen in accordance with the size of the fish in each release group. The nets were deployed 2 km north and south of the release site at depths ranging from 10 to 45 m, perpendicular to the coast in a line from shallow to deep waters. The gill-net fishing lasted for six days after each simulated escape. Posters with information about the recapture reward were distributed in strategic points around the area of study (Fig. 6.7.5).

Figure 6.7.5. The poster made for the reward program.

284

There was no recapture of tagged juvenile cod with any of the live recapture methods tested in the vicinity of the fish farm. However, large amounts of saithe (Pollachius virens) and cod (both wild and farmed from previous escape events) were caught. Large aggregations of piscivorous fish have been previously reported around fish farms all around the world (Boyra et al. 2004, Dempster et al. 2002, 2009, Uglem et al. 2009). Wild fish attracted to farms also tend to be large adults (Dempster et al. 2002). The lack of recapture of tagged juveniles in the immediate vicinity of the farm might have been a consequence of the predator-avoidance behavior of the escapees or that escapees were eaten by the large wild fish in the vicinity of farms after the escape. In agreement with this hypothesis, we detected high predation pressure of the tagged fish. More than 100 tags, most of which belonged to fish from the first release group, were recovered by examining the stomach of large cod and saithe caught close to the net pens. A total of 40 tagged escapees were recaptured in the gill-nets. However, the by-catch outnumbered the recapture: more than 500 individuals from non-targeted species were recorded, with crabs (Cancer pagurus), saithe and cod (wild and farmed from previous escape events) the most abundant species (Figure 6.7.6). The by-catch also included tusk (Brosme brosme), haddock (Melanogrammus aeglefinus), pollack (Pollachius pollachius) and mackerel (Scomber scombrus), among other species. Only eight tagged escapees were recaptured through the reward program by local fishermen. All these recaptures occurred within 2 km from the release farm, in a time span from six days to six months post release. There were no reported recaptures in the gill-nets used by the semicommercial fishermen operating in the fjord, probably due to a mismatch between the size of the standard cod nets used in the fjord and the small size of the tagged fish (<40 cm total length), which keeps the juvenile escapees out of the fishing window of the local fisheries. In summary, of a total of almost 3000 juvenile cod that were released for this study, only 48 (1.7%) were recaptured, another 110 (3.9%) escapees were found in the stomach of predators and the remaining 94.3% were unaccounted for one year after the first escape event was performed. In conclusion, the present study indicates that small juvenile farmed cod escapees might suffer a high predation pressure as soon as they enter the wild. Those that are able to escape from the predators aggregated around the farm will rapidly seek protection in shallow waters close to the coast. Recapture efforts should therefore not concentrate around the farm, but in shallow coastal waters. However, the small mesh size of the gill-nets needed for the recapture of juveniles seems to be significant drawback to the efficiency of this recapture method; the high by-catch rates and the resulting workload make this method non-operational at large scale. The potential benefits of a recapture program should be considered in relation to the potential for depleting local wild fish stocks and the costs of recapture. With the suite of recapture methods evaluated in this study, the recapture rates achieved were too low to defend the efforts devoted to recapture. This emphasizes the need to prevent escape incidents by improving farming technology and operational routines rather than relying on trying to recapture fish after they have escaped. In other words, the battle to avoid possible negative ecological effects due to escape of juvenile farmed Atlantic cod should take place before the cod escape from the sea-cages.

www.preventescape.eu

285

Figure 6.7.6. Total catch in gill-nets after one week of recapture effort in both October 2010 and March 2011.

286

References

cited

Boyra A, Sanchez-Jerez P, Tuya F, Espino F, Haroun R (2004) Attraction of wild coastal fishes to Atlantic subtropical cage fish farms, Gran Canaria, Canary Islands. Environ Bio Fish 70(4):393–401 Dempster T, Sanchez-Jerez P, Bayle-Sempere JT, Giménez-Casalduero F, Valle C (2002) Attraction of wild fish to sea-cage fish farms in the south-western Mediterranean Sea: spatial and shortterm temporal variability. Mar Ecol Prog Ser 242:237–252 Dempster T, Uglem I, Sanchez-Jerez P, Fernandez-Jover D, Bayle-Sempere J, Nilsen R, Bjørn PA (2009) Coastal salmon farms attract large and persistent aggregations of wild fish: an ecosystem effect. Mar Ecol Prog Ser 385:1–14 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007) Technological solutions and operational measures to prevent escapes of Atlantic cod (Gadus morhua) from sea-cages. Aquac Res 38:91–99 Norwegian Directorate of Fisheries, 2010. http://www.fiskeridir.no/statistikk/akvakultur. Uglem I, Bjørn PA, Dale T, Kerwath S, Økland F, Nilsen R, Aas K, Fleming I, McKinley RS (2008) Movements and spatiotemporal distribution of escaped farmed and local wild Atlantic cod (Gadus morhua L.). Aquac Res 39:158–170 Uglem I, Dempster T, Bjørn PA, Sanchez-Jerez P, Økland F (2009) High connectivity of salmon farms revealed by aggregation, residence and repeated movements of wild fish among farms. Mar Ecol Prog Ser 384:251–260

www.preventescape.eu

287

6.8. ‘Escape

through spawning’: solutions to

reduce the escapes of viable fish eggs from sea-cages Cite this article as: Dempster T, Uglem I, Sanchez-Jerez P, Somarakis S (2013) ‘Escape through spawning’: solutions to reduce the escapes of viable fish eggs from sea-cages. In: PREVENT ESCAPE Project Compendium. Chapter 6.8. Commission of the European Communities, 7th Research Framework Program. www.preventescape.eu ISBN: 978-82-14-05565-8

Authors: Tim Dempster1, Ingebrigt Uglem2, Pablo Sanchez-Jerez3, Stelios Somarakis4 SINTEF Fisheries & Aquaculture, Norway Norwegian Institute of Nature Research, Norway 3 University of Alicante, Spain 4 Hellenic Centre of Marine Research, Greece 1 2

Introduction Fish farming in sea-cages is increasing worldwide; both the volumes produced and the numbers of fish species cultured are increasing. During the last decade, culture of species that may reproduce within sea-cages has become more common. Examples of such species within European aquaculture are Atlantic cod (Gadus morhua; Jørstad et al. 2008) and seabream (Sparus aurata; Dimitriou et al. 2007). Knowledge of the extent and ecological effects of reproduction of farmed fish within sea-cages is, however, sparse. In the culture of Atlantic cod, some fish mature during the first year of culture, while a majority of farmed cod are believed to mature during the second culture year. This means that almost the entire culture stock in any particular farm has the potential to spawn in the sea-cages before they are slaughtered. Results from the Prevent Escape project (Prevent Escape Compendium Chapter 5.3, Uglem et al. 2012) indicate that spawning in sea-cages is likely to be widespread in the industry and that eggs and larvae (Figure 6.8.1) are likely to survive and spread throughout coastal waters. Therefore, there is considerable potential for larvae from escaped cod eggs to experience favourable conditions for survival and recruitment to coastal cod stocks. This may cause significant ecological and genetic effects in wild populations in the future.

288

Figure 6.8.1. Atlantic Cod eggs ready to hatch and a newly hatched larva. Photo: Tora Bardal, IBI, NTNU.

In the Mediterranean region, information about spawning by fish kept in sea-cages prior to the Prevent Escape project was sparse. Results from the Prevent Escape project (Prevent Escape Compendium Chapter 5.2) indicate that spawning in sea-cages is likely to be widespread. While eggs and larvae produced by farmed seabream were of poor quality compared to wild fish, large numbers may still survive and flow out from farms into natural environments. They may therefore recruit to wild seabream populations. As for cod, this may cause significant ecological and genetic effects in wild populations in the future. The emerging species of interest to sea-cage aquaculture in the Mediterranean, the meagre (Argyrosomus regius) did not mature in sea-cages and does not appear to mature (Prevent Escape Compendium Chapter 5.4). It appears unlikely that egg escapes of this species could occur, unless production sizes alter dramatically, or the size at first maturity changes.

Overarching

general principle to avoid egg escape

From the information generated in the Prevent Escape project and existing literature on the production of other species, it is possible to derive a general principle: pelagic spawning species, where both sexes may be present in the same cage at the same time are at risk of escape through spawning if the weight at maturity is smaller than the harvest weight (Figure 6.8.2). For new species entering culture, this principle will give the capacity to predict whether they will be susceptible to escape through spawning. Data on the average size or weight at maturity from wild populations cannot be substituted for data on size or weight of maturity in cultured populations. This is evident for both cod and seabream, where the weight or size at maturity is lower than for wild populations. Some capacity to predict whether cultured individuals of a species will have a different size at maturity than wild populations may be available from wild populations if there is evidence for spatial or temporal differences in size at maturity which would indicate that this reproductive trait was relatively plastic and thus able to change.

www.preventescape.eu

289

Fig. 6.8.2. Weight at maturity verses harvest weight for pelagic broadcast spawning species where weight at maturity has been documented in aquaculture settings. Boxes indicate the ranges for different species. Boxes that fall below the black dashed line are at risk of creating â&#x20AC;&#x2DC;escape through spawningâ&#x20AC;&#x2122;.

Specific

recommendations for

Atlantic

cod

Recent developments within the cod farming industry have involved increased growth rates in the sea-cage phase. Slaughter of farmed fish before spawning during the second season in the sea is now common. Hence, a simple, realistic and profitable action to drastically reduce the risk of ecological effects as a result of egg escape is to make slaughter before the second spawning season mandatory. Technical solutions aimed at preventing spawned eggs to enter the sea have also been suggested. Such solutions could, for instance, involve the use of closed sea cages and mechanical filters to remove eggs. However, closed cages are still not an economically viable solution due to high technological and operational costs. Furthermore, development of technology for closed cage systems is still in an early stage. Photoperiod manipulation will most likely reduce significantly possible effects of egg escape during the first season through reduced survival of eggs and larvae in the wild. In addition, the use of photoperiod manipulation to culture cod in submerged cages is promising (KĂ¸rsoen et al. unpublished data). Photoperiod manipulation may be more effective in submerged cages as the typically high natural surface light dissipates with depth and allows the artificial lights to provide more constant light conditions. Recent results indicate that while this technique does not stop maturation altogether, it may delay maturation such that when spawning occurs it does so outside of the main spawning season. Presumably, this will mean that eggs and larvae will then be subject to environmental conditions that are less suited to their survival. Other options to reduce the amounts of eggs escaping are hybridization, sterilization and polyploidy. Recent research has shown that production of triploid Atlantic cod may practically eliminate the risk of egg escape, as gamete production by triploid females is delayed and dramatically lowered compared to diploid females and that any larvae produced by triploid

290

fish were non-viable (e.g. Feindel et al. 2010, 2011). This is promising as it indicates that triploid females will not mature before harvest, and growth rates will increase through reduced investment in gonad production. However, problems such as initially higher mortality, greater fingerling costs, maturation of triploid males and consumer acceptance need to be solved before production of triploid fish is taken up by industry (Triantafyllidis et al. 2007; Feindel et al. 2010).

Key Recommendations • Make slaughter before the second spawning season in the sea mandatory • Introduce photoperiod manipulation • Investigate the combination of culture in submerged cages and photoperiod manipulation to enhance the suppression of sexual maturation • Investigate triploidy as a viable option to reduce the ability of farmed cod to spawn and reduce the viability of any spawned eggs and larvae

Specific

recommendations for gilthead seabream

There are two specific cases in which the probability of producing more eggs, or that larvae may recruit to the wild populations, are increased for seabream: (a) the sex ratio in cages is balanced (close to 1:1) (b) farms occur where seabream are able to close their life cycle, e.g. close to lagoons. The use of a device such as a curtain-like egg collector as constructed as used in the Prevent Escape project could prevent the dispersion of eggs away from the cages. However, there are significant technical challenges in using such devices and their effectiveness remains to be tested at full-scale. Further, restricting water flow through the upper part of a cage can compromise oxygen levels for the cultured fish and compromise their welfare (Oppedal et al. 2011). Future studies that look into filtration as a method to remove spawned eggs must therefore also take the welfare of cultured fish into account.

www.preventescape.eu

291

Other options to reduce the amounts of eggs escaping would be the use of hybridization, sterilization and polyploidy. However, problems such as initially higher mortality, greater fingerling costs, poorer growth and consumer acceptance need to be solved first. Available information on seabream ecology is sparse, but indicates that it is probably an estuary-dependent species. Thus, a precautionary mitigation measure to reduce the likelihood of negative ecological interactions arising from escape through spawning would be to restrict the culture of large seabream (of sizes beyond that of sex reversal) in areas close to known nursery grounds of the species, such as lagoons.

Key Recommendation â&#x20AC;˘ Restrict the culture of large seabream (of sizes beyond that of sex reversal) in areas close to known important nursery grounds of the species, such as coastal lagoons

292

References

cited

Benetti DD, O’Hanlon B, Rivera JA, Welch AW, Maxey C Orhun MR (2010) Growth rates of cobia (Rachycentron canadum) cultured in open ocean submerged cages in the Caribbean. Aquaculture 302: 195-201 Feindel NJ, Benfey TJ, Trippel EA (2010) Competitive spawning success and fertility of triploid male Atlantic cod Gadus morhua. Aquacult Environ Interact 1:47-55 Feindel N, Benfey T, Trippel E (2011) Gonadal development of triploid Atlantic Cod (Gadus morhua). J Fish Biol 78: 1900-1912 Jørstad KE, van der Meeren T, Paulsen OI, Thomsen T, Thorsen A, Svåsand T (2008) "Escapes" of eggs from farmed cod spawning in net pens: recruitment to wild stocks. Rev Fish Sci 16: 1-11 Oppedal F, Dempster T, Stien L (2011) Environmental drivers of Atlantic salmon behaviour in sea-cages: a review. Aquaculture 311: 1-18 Uglem I, Knutsen Ø, Kjesbu OS, Hansen ØJ, Mork J, Bjørn PA, Varne R, Nilsen R, Ellingsen I, Dempster T (2012) Extent and ecological importance of escape through spawning in sea-cages for Atlantic cod (Gadus morhua L.) Aquacult Environ Interact (in press)

www.preventescape.eu

293

294

seeks to reduce the number of fish that escape from European aquaculture through research to improve fish farming techniques and technologies.

PREVENT ESCAPE is financially supported by the Commission of the European Communities, under the 7th Research Framework Program.

7. Recommendations and guidelines of fish farms, management and

for the design operation of

equipment Cite this article as: Fredheim A (2013) Recommendations and guidelines for the design of fish farms, management and operation of equipment. In: PREVENT ESCAPE Project Compendium. Chapter 7. Commission of the European Communities, 7th Research Framework Program. www. preventescape.eu ISBN: 978-82-14-05565-8

authors: Arne Fredheim1, Ă&#x2DC;sten Jensen1 and Tim Dempster1 1

SINTEF Fisheries & Aquaculture, Norway

www.preventescape.eu

295

The Prevent Escape project has estimated that more than 9 million fish have escaped from European fish farms between 2007 and 2009, with a range of ecological and genetic consequences possible. In addition, the direct cost, at the point of first-sale, was estimated to be as high as â&#x201A;Ź47.5 million per year in Europe. The environmental and economic costs of escape provide clear justification for a greater focus on development and implementation of methods to prevent escapes in European countries where fish farming is a significant industry in coastal waters. Here, we present the case for technical standards and other measures to prevent escapes, summarise the technologies currently used by the sea-cage farming industry, and outline a range of implementable measures that can be taken by various stakeholders to better prevent escapes.

The

need for technical standards for sea-cage fish farms and

current status across

Europe

Once implemented, technical standards for the design, management and operation of fish farms are a demonstrated, powerful, industry-wide tool to prevent escapes (Jensen et al. 2010). Technical standards used to regulate industry performance should be subject to frequent updates and improvements. The technical standards also need to keep pace with changes in technology, the types of physical environments in which farms are located and

296

practical aspects related to use of equipment of ever-increasing size. For technical standards to be applied and enforced consistently across the industry, legislative support is required. For example, a recent spate of escapes in Scotland has led to calls for a technological overhaul of the industry; to support such an overhaul, legislation will be required to ensure the standards. Technical standards, codes of practice, legislation and legislative support, and the reporting of escapes vary between countries across Europe. Norway is the only European country with legislation (NYTEK), certification bodies (accredited by Norwegian Accreditation) and a technical standard (NS9415.E:2009) specifying the requirements for fish farms in terms of their component technologies, design and engineering. The introduction of the NS9415 technical standard is believed to have significantly reduced the volume of escapes to the current level; ~0.1-0.3% of the approximately half a billion salmon held in sea-cages in Norway each year (Jensen et al. 2010). Although other countries have voluntary codes of conduct and best practice guides for marine fish farm management, mooring procedures and the manufacture of net cages in situ (Table 7.1), these fall short of the regulatory power of technical standards backed by legislation. A movement towards development of technical standards in other major aquaculture producing countries appears to be gaining momentum. For example, in Scotland, a proposal for a technical standard (SARF073 2012) is in progress. Furthermore, an international technical standard is under development (ISO 2012), and the various results from the Prevent Escape project have proved valuable input.

www.preventescape.eu

297

Country/ region

Norway

Existing industry codes

• NS 9415

• Code of conduct

Scotland

Ireland

Canada

• The Code of practice for the prevention of stock escapes of Irish farmed salmonids

Legislation and certification System.

The Aquaculture & Fisheries (Scotland) Act 2007 came into force in August 2007 and makes relevant legal powers and provisions in relation to fish farms for containment and fish farm escapes. The Code of Conduct is mandatory for all members of the Scottish Salmon Producers Organisation (SSPO).

Industry voluntary

• Code of containment for the culture of salmonids in Newfoundland and Labrador Industry voluntary • Code of containment for culture of Atlantic salmon in marine net pens in New Brunswick

• FEAP Code of Conduct http://www. aquamedia.info/ consensus/

Voluntary

International

• NASCO Guidelines on Containment of Farm Salmon; CNL(01)53 (http://www.nasco.int/pdf/ ag reements/williamsburg.pdf)

Industry voluntary

International

• ISO Standard (ISO)

Industry voluntary

Europe

Chile

298

• Draft technical standard published (SARF0732012)

Level of enforcement

None

Reporting of escapes

Required to report all incidents, number of escaped fish, how handled and what happened.

Other escape reducing actions/efforts

Comments

Directorate of fisheries do inspections and follow-up on escape incidents. Official Escape Commission with experts who surveyed escape incidents existed from 20062011.

Legislation concerning the reporting of fish farm escapes has been in place since 2002 and requires that Marine Scotland must be notified of an escape or suspected escape immediately by phone and in writing by completing an Initial Notification Form.

Marine Scotland established â&#x20AC;&#x2DC;The Improved Containment Working Groupâ&#x20AC;&#x2122; in 2009 to make proposals and a strategic Framework for Scottish Aquaculture to reduce escapement.

A Final Notification Form is required within 28 days of the Initial Form. Escapes must be reported to the Department of Communications, Marine and Natural Resources (DCMNR). An escape incident must be reported by phone within 24 hours, and in writing within 1 week. Atlantic Salmon Watch (Fisheries and Oceans Canada) must be informed. Farmers will, in the event of escapes, take immediate action, co-operate and inform the respective authorities to ensure that the appropriate actions are taken.

Focus on management. Technical requirements for net cage design and manufacturing.

Farmers will seek to minimize the potential risks that are presented by farmed fish escapes to wild fisheries.

No mention of standards for technical equipment or counting of escaped fish.

Under development based on the Norwegian standard NS 9415 and the Scottish draft technical standard Initial work started to develop a technical standard for fish farming equipment.

Table 7.1 Status of reporting requirements, legislation, technical standards and codes of practice/ conduct in major finfish producing countries.

www.preventescape.eu

299

Technical

equipment

Most floating fin-fish farms are based on the same general principles and component technologies; a floating collar, a net cage and a mooring system (Fredheim & Langan 2009). The floating collar acts as an attachment point for the net cage, helps maintain its shape, distributes forces to the mooring system, and provides a working platform for daily operations. The floating collar is the main structural component of a floating fish farm, integrating the other parts. The two most common set-ups are circular plastic fish farms, made of high density polyethylene (HDPE) pipes, and interconnected hinged, steel farms (hereafter known as steel fish farms) (Figure 7.1).

Figure 7.1. A circular high density polyethylene pipe (HDPE) fish farm (left) and a typical interconnected hinged steel fish farm with attached feed blowers and a storage barge next to it (right) (Photos by SINTEF Fisheries and Aquaculture).

The circular plastic fish farm is constructed by welding the pipes together into preferred lengths. The complete pipe length is then forced into a circle and the two free ends are welded together. These fish farms systems are delivered as single, double or triple rim systems. The rims are connected together using clamps, made either of HDPE or steel. A HDPE fish farm will normally be moored with a grid mooring system of ropes, connector plates and floats. The steel fish farm is made up of steel bridges, and their flotation devices, connected with hinges. Each bridge is normally 12 m and they are connected in various configurations to make up the fish farm. The flotation devices are generally made of expanded polyester, covered with a protective material, and attached underneath the steel bridges. The hinges allow for rotation around one axis in the horizontal plane, but not around the vertical axis. Steel fish

300

farms are relatively stable, with large areas for walking and equipment. The flotation can be controlled and this allows for heavy auxiliary equipment such as feed blowers, and the equipment required to handle net cages and fork lifts on board. Net cages serve as the containment system to keep the fish in place and, in traditional fish farms, are suspended inside the floating collar. The netting material (most commonly polyamide) is attached to a frame of ropes that distributes the resultant forces. Weights are attached to the netting, to maintain the correct shape and volume (see Fridman 1992 for more information about materials and Moe et al. 2007a for net cages in particular). Modern high strength materials such as â&#x20AC;&#x153;ultra-high molecular weight polyethyleneâ&#x20AC;? (UHMWPE) are used in some new fish farm and net cage designs. These materials have different market brand names like Dyneema and Spectra (Moe et al. 2005). Fish farm mooring systems primarily consist of ropes, floats and anchors. Several smaller components; shackles, connection plates, chains, rings etc., are used in the mooring system to connect together the primary parts. The purpose of a mooring system is to secure the fish farm in the desired position. Mooring requirements are determined by the size and characteristics of the fish farm and conditions like bottom topography and weather conditions at the specific site. There are two main methods for mooring a fish farm; either by independent lines attached directly to the floating collar, or, by a grid mooring system to which one or several floating collars are connected (Figures 7.2 and 7.3).

Figure 7.2. Illustration of a HDPE collar fish farm with a grid mooring system (left) and of an steel fish farm with moorings (right) (Illustrations by SINTEF Fisheries and Aquaculture).

www.preventescape.eu

301

Figure 7.3 Illustration of a grid mooring system for HDPE collar fish farms (Illustration by Aqualine AS).

Mitigating

and preventing escapes

A wide range of measures are available to better prevent fish escapes, many of which have been identified during the Prevent Escape project, from other relevant work, and general experience dealing with fish escapes in Norway over the past decade. The biggest contributors to both the number of fish escaping and the number of escapeincidents are mooring system failures and hole formation after failures in the net cage. The main mechanisms to prevent mooring system failure are: 1) detailed site surveys; 2) rigorous implementation of technical standards for equipment; and 3) proper requirements for mooring system analysis. To reduce the incidence of hole formation in net cages, measures include: 1) implementation of technical standards; and 2) implementation of operational measures to prevent fish from biting the net and predators attacking the net cage. The operation and regulation of the fish farming industry encompasses a number of diverse stakeholders, from fish farming companies to equipment manufacturers, and local, regional and national governments. Recommendations relevant to all of these stakeholders are presented here, but, not all of the recommendations are directed to all of the stakeholders (see Table 7.2 for an overview).

302

Principles for sea-cage design Risk is a product of probability and consequence; some measures that reduce consequence might increase probability. Further, is it not possible to have zero probability, so it is always necessary to define what the acceptable consequences of an event occurring are in relation to the likelihood of occurrence. Risk awareness can be applied at several levels; the probability for a failure and consequence for that failure can be assessed at the individual incident level (sea-cage), or on a more aggregated level (farm site or industry as a whole). Failure of components, operational errors and human mistakes cannot be completely prevented, so it is important to design equipment in such a way that failures of individual components, or human error, do not lead to total equipment breakdown. For example, a mooring component failure should not lead to the failure of the whole mooring system or farm, and furthermore, should not lead to fish escaping. This should be the key design principle for fish farms and is the basis used here to formulate the design and operational guidelines outlined in this report.

General

recommendations

1) Mandatory reporting of all escape events, including: • The number of fish that escaped, their age, size and health status. • A description of the sea-cage technology involved, including the age of the equipment, manufacturer and whether or not the equipment is designed or certificated according to any technical standards. • Description of site and relevant environmental data including water depth, maximum current, wind and waves. Estimate of current, wind and wave conditions at time of escape if relevant to escape cause. • Categorization of the operational circumstances. • An estimated cause of escape.

www.preventescape.eu

303

A specific form for reporting should be developed for each country. Existing forms exist in Scotland and Norway and could be used as guidelines. 2) Evaluation. Introduction of a defined procedure to collate, analyse and learn from the data collected during mandatory reporting. The information should be provided to equipment suppliers and fish farmers so that equipment and operational improvements can be made. This could be a government service, a joint industry effort, or a combination of both. 3) Risk evaluation and identification. Risk analyses should be completed for the planning, design, production, delivery, installation and operation of the fish farm. Each risk analysis should be broken down into the probability and consequences of an incident occurring, and be based on approved procedures (e.g. according to requirements laid down in the ISO manual). The analyses should be clearly documented to facilitate review. 4) Operational procedures. Some fish farming operations are likely to pose a higher risk of an escape event occurring if they are done incorrectly, for example; correct anchoring and mooring of the net cage, connecting net cages to floaters, and correctly weighting the net cages in currents. These critical operations should be identified through a risk evaluation (see above) and risk reduction strategies employed. Copies of operational procedures should be readily available at all sites. Relevant personnel should be aware of the procedures and receive appropriate training before commencing duties. 5) Training. A large percentage of escape incidents are due to operational errors. Fish farming is a complicated, multi-disciplinary activity, and expertise in several different areas is required. Until now, education has focused mainly on the biology of fish farming, with less attention paid to the technical aspects. The industry should consider mandatory training of all relevant personnel concerning the environmental consequences of fish escapes, as well as the relevant operations and technical measures required to reduce escape incidents. A training course could be a joint industry effort. 6) Technical standards. Detailed technical standards should be introduced. There are relevant standards, such as NS 9415 (NS 9415.E:2009, 2009) and the preliminary Scottish standard (SARF 073, 2012), and also plans to develop an ISO standard for aquaculture technology similar to these. There are also general purpose standards for the design and dimensioning of constructions, such as the EN Eurocodes (EUROCODE), which are relevant to fish farm. The same constructional design principles apply, independent of construction type, and our recommendation is to use the Eurocodes standards as the basis for the design of floating fish farms. The structure that is unique to fish farms is the net cage. It has distinctive properties, both in terms of construction and with respect to its behaviour and handling of loads subject to currents and waves. There are several relevant design standards and codes of conduct for the use of net cages (Table 1) and methods and procedures for calculation of behaviour and loads (LĂ¸land 1992; Lader et al. 2008; Moe et al. 2010; Kristiansen and Faltinsen, 2012). The latter is also described in NS 9415 and the preliminary Scottish standard.

304

7) Hand books/User manuals. The manufacturers of fish farm equipment should supply a detailed handbook describing the product, including: • the design capacity relative to current, wind and wave levels; • inspection and maintenance plans; • integration and installation with other main components. These handbooks should be available at all fish farms and all personnel should be familiar with their content. 8) Certification system. A certification system should be introduced for equipment approvals, based on technical standards and requirements for construction design: • Certification of products and relevant service providers for procedures such as anchor handling and site survey, the main components of the fish farm, such as the mooring system, net cage and floating collar, and the feeding barge and relevant auxiliary equipment such as the feed pump. • Equipment integration should be documented, for example attaching the net cage to the floating collar. • Service providers should to be certified according to specific standards to ensure competence and safety in procedures and the reduction of fish escape incidents. • Site approval – checking that the components and equipment used in all fish farms meets technical standards and is installed according to approved procedures and instructions. Site approval should be sought prior to fish being introduced to the farm. The comprehensive system that exists in Norway may not be relevant or feasible for other countries. Rather than focusing on such an extensive system, which might lead to opposition from other countries, prioritising those elements that have the greatest impact on the reduction of fish escapes would be prudent. For instance the introduction of site specific farm certificates incorporating: • A site survey stating the environmental conditions and bottom topography. • A mooring analysis. • Control of the interconnection and coupling of the main components according to the relevant handbook.

www.preventescape.eu

305

Specific recommendations - fish farm: 1) Site surveys. Currents, wind and wave characteristics should be determined for all of the sites to enable the design and selection of suitable equipment for the site-specific weather conditions. The net cage and floating collar need to withstand variable environmental loads, but will not necessary be designed specifically for each site, rather they could be categorized into classes suitable for conditions up to a maximum current and wave force. Mooring systems, on the other hand, need to be specifically dimensioned for each site, due to differences in farm layout, water depth and bottom topography. 2) Weight system. The weight systems need to be designed and installed so that it does not introduce unnecessary loads onto the netting material or make direct contact with the netting, causing abrasion. The weight system is an integrated part of the complete fish farm and its interactions with other components of the farm must be considered both by the manufacturer of the floating collar and of the net cage.

Specific recommendations - farming equipment: 1) Design. Products and components should be designed according to a technical standard, with an independent body to enforce the standard. Manufacturers should provide data supporting the strength and capability of their product and a third party should then control of the presented data. 2) Risk based design. Farms should be designed so that an individual failure will not lead to larger failures and fish escapes.

306

Specific recommendations - mooring system: 1) Dimensioning. The design should be based on the bottom topography and environmental loads at each site. The procedure should follow accepted standards and protocols. A proper numerical mooring analysis should be performed. 2) Redundancy. The design should allow for breaks in individual mooring lines, without the loss of integrity of the mooring system. 3) Connector plates. Connector plates should be handled with care and the design and use allow for damage, fracture or rupture of a connector plate to occur without the loss of integrity of the mooring system. 4) Anchors. The holding power of all anchors should be tested and documented after installation. Accepted procedures should be followed. 5) Mooring system components. Only certified steel mooring components should be used. Shackles need to have reliable safety pins installed. Care should be taken when steel components of different quality are attached to each other to avoid galvanic corrosion. Mooring chains in the surface zone are at risk of corrosion. In addition, metal components in the surface part of the mooring system may damage the netting and their use should be minimised. 6) Ropes and knots. Ropes used in the mooring system are subject to UV degradation and knots generally reduce the strength of a rope by 50%.

www.preventescape.eu

307

Specific recommendations - net cage systems: 1) Fish biting. Fish biting is a major cause of escape incidents, but certain preventative measures are possible (Moe et al. 2007b; Høy et al. 2012) and these may also reduce the effect of predator attacks: • Stronger net cage constructions. • Avoid minor damage to nets from handling and abrasion. • Inspect the net cage frequently for holes. • Ensure net repairs are of a good quality. • Keep the cage nets clean - biofouling may attract the fish. • Make sure that cage nets are taut. • Check that the mesh width is suitable for the fish size. • Make sure that the fish are fully fed at all times. • Frequently grade fish for size. • Provide a stimulating cage environment to distract the fish from biting. 2) Abrasion and wear. The formation of holes, due to wear and tear, and net abrasion, due to contact with other components such as the weight system, are a major problem. Contact between vertical chains or other hanging components of the fish farm and the net cage should be prevented. A well-designed weight system that remains taut at all times is highly beneficial. The weight system and net cage need should be designed/selected together. The manufacturers of the net cage and floating collar should provide detailed instructions on how to attach, inspect and use the weight system in order to avoid contact with the netting. 3) Predator attacks. It is difficult to prevent predator attacks completely; however, it is possible to reduce the number of attacks and the effect of the attacks: 1) remove dead fish daily, or even more frequently at sites with high predation risk; and 2) introduce a double net around the dead fish removal system to avoid wear. A well-designed weighting system which keeps the sea-cage net taut will also reduce the capacity of predators to make holes.

308

Recommendation

Government Fish farmer (legislation)

Equipment producer

Relevant towards cause of escapement

Mandatory reporting

All

Evaluation

All

Risk evaluation and identification

All

Operational procedures

All

Training

All

Certification system

All

Technical standards

All

Hand books/User manuals

Operational failure and human errors

Site surveys

Structural failure, Mooring failure

Weight systems

Hole in net

Design farm system

Structural failure

Operational failure, Structural failure, Mooring failure and human errors

Mooring failure

Risk based design

Dimensioning

Redundancy mooring

Mooring failure

Connector plates

Mooring failure

Anchors

Mooring failure

Mooring system components

Mooring failure

Ropes and knots

Mooring failure

Fish biting

Hole in net

Abrasion and wear

Predator attacks

Research and development

Hole in net Hole in net

All

Table 7.2 Overview of recommendations and their relevance to different stakeholders and major cause of escapement.

www.preventescape.eu

309

Recommendations for further research and development As technological development in the industry is ongoing, a focus on improving the quality of fish farm equipment is constantly required. Much relevant knowledge and competence has been developed through Prevent Escape and other national and international research projects specifically related to escapes. Significant future gains may also be made by drawing upon and integrating developments and discoveries in other industrial sectors and scientific disciplines. Net cage failures are the most common causes of escape, both with respect to the number of incidents and the number of fish escaping. Thus, improvements to net design, and/or netting material would be of great value in reducing escapes. A large number of new materials and net cages have been proposed, but there are challenges in updating systems in terms of cost and difficulty in replacing single components within operational systems. The netting materials developed for other industries have very different properties to those currently used in fish farms, therefore, the use of new net cage materials would involve a major change in fish farming operations, accessory equipment and staff training. If these new materials are to be properly integrated, research and development on their properties and correct use will be required.

310

References

cited

Bjorn PA, Finstad B, Kristoffersen R (2007) Differences in risks and consequences of salmon louse, Lepeophtheirus salmonis (Kroyer), infestation on sympatric populations of Atlantic salmon, brown trout, and Arctic charr within northern fjords. ICES J Mar Sci 64: 386-393 EUROCODE. The EN Eurocodes. European Standard Organization. (http://eurocodes.jrc. ec.europa.eu/home.php) Fredheim A, Langan R (2009) Advances in technology for off-shore and open ocean aquaculture. In: “New technologies in aquaculture”. Woodhead Publishing Limited and CRC Press LLC. Fredheim A, Jensen Ø, Dempster T (2010) Norway: Escapes of fish from aquaculture. In: Advancing the Aquaculture Agenda OECD report. ISBN: 9789264088726 DOI:10.1787/9789264088726-16-en Fridman AL (1992) Calculations for fishing gear design. Fishing News Books Ltd. Henriksen K, Jensen Ø, Høy E (2012) Effects and consequences of a requirement of double net cage for fish farming of cod. SINTEF Report A22620 (In Norwegian). Høy E, Volent Z, Moe-Fore H, Dempster T(2012) Loads applied to aquaculture nets by the biting behaviour of Atlantic cod (Gadus morhua). Aquacult Eng 47:60-63 ISO (2012) International Standards Organization ISO/TC 234/WG 3 on aquaculture technology. Jensen Ø (2006) Assessment of technical requirements for floating fish farms – based on escape incidents January 2006. SINTEF report no. SFH80 A066056 ISBN 82-14-03953-8 (In Norwegian) Jensen Ø, Dempster T, Thorstad EB, Uglem I, Fredheim A (2010) Escapes of fish from Norwegian sea-cage aquaculture: causes, consequences, prevention. Aquacult Environ Interact 1:71-83 Kristiansen T, Faltinsen OM (2012) Current loads on aquaculture net cages. J Fluid Struct (in press) http://dx.doi.org/10.1016/j.jfluidstructs.2012.04.001 Lader P, Dempster T, Fredheim A, Jensen Ø (2008) Current induced net deformations in fullscale sea-cages for Atlantic salmon (Salmo salar). Aquacult Eng 38: 52-56 Løland G (1991) Current forces on and flow through fish farms. PhD thesis, Dept. of Marine Hydrodynamics, The Norwegian Institute of Technology.

www.preventescape.eu

311

Moe H, Gaarder R, Sunde LM, Borthen J, Olafsen K (2005) Escape-free nets for cod. SINTEF report no. SFH80 A054041, Trondheim, Norway (In Norwegian) Moe H, Olsen A, Hopperstad OS, Jensen Ø, Fredheim A (2007a) Tensile properties for netting materials used in aquaculture net cages. Aquacult Eng 37(3):252-265 Moe H, Dempster T, Sunde LM, Winther U, Fredheim A (2007b) Technological solutions and operational measures to prevent escapes of Atlantic Cod (Gadus morhua) from sea-cages. Aquacult Res 38: 90-99 Moe H, Fredheim A, Hopperstad OS (2010) Structural analysis of aquaculture net cages in current. J Fluids Struct 26: 503-516 Naylor R, Hindar K, Fleming IA, Goldburg R, Williams S, Volpe J, Whoriskey F, Eagle J, Kelso D, Mangel M (2005) Fugitive salmon: assessing the risks of escaped fish from net-pen aquaculture. BioScience 55:427-437 NS 9415.E:2009 (2009) Marine fish farms - Requirements for site survey, risk analyses, design, dimensioning, production, installation and operation. Standard Norway. (http://www.standard. no/en/Sok-og-kjop/produktkatalogen/Produktpresentasjon/?ProductID=426625) NYTEK. Regulations about requirements for technical standards for floating aquaculture insulations. FOR 2011-08-16 nr 849 (http://www.lovdata.no/cgi-wift/ldles?doc=/sf/sf/sf20110816-0849.html) Prevent Escape (2012a) Final work package report on the extent, technical and operational causes of cost of escape events. Prevent Escape (2012b) Final work package report on tasks 6.1 “Biological, design and operational issues” and 6.2 “Hydrodynamic loads and structural response”. Prevent Escape (2012c) Deliverable 3.1 Internal report on the population variation and individual variation in inherent behavioural escape motivation. SARF 073 (2012) A report presenting proposals for a Scottish Technical Standard for Containment at Marine and Freshwater Finfish Farms. Scottish Aquaculture Research Forum. ISBN: 978-1-907266-45-4. (http://www.sarf.org.uk/reports/) Taylor M, Kelly R (2010) Assessment of Protocols and Development of Best Practice Contingency Guidance to Improve Stock Containment at Cage and Land-based Sites Volume 1: Report. pp 74. ISBN: 978-1-907266-30-0. (http://www.sarf.org.uk/reports/)

312

Taylor M, Kelly R (2010) Assessment of Protocols and Development of Best Practice Contingency Guidance to Improve Stock Containment at Cage and Land-based Sites Volume 2: Supporting Information. pp 177. ISBN: 978-1-907266-31-7. (http://www.sarf.org.uk/reports/) Thorstad EB, Fleming IA, McGinnity P, Soto D, Wennevik V, Whoriskey F (2008) Incidence and impacts of escaped farmed Atlantic salmon Salmo salar in nature. Report from the Technical Working Group on Escapes of the Salmon Aquaculture Dialogue.

www.preventescape.eu

313