Journal of Fish Biology (2014) 85, 1972–1991 doi:10.1111/jfb.12544, available online at wileyonlinelibrary.com
Silver spoons in the rough: can environmental enrichment improve survival of hatchery Atlantic salmon Salmo salar in the wild? L. J. Roberts*, J. Taylor†, P. J. Gough†, D. W. Forman* and C. Garcia de Leaniz*‡ *Swansea University, Centre for Sustainable Aquatic Research, Department of Biosciences, Swansea SA2 8PP, U.K. and †Natural Resources Wales, Cynrig Fish Culture Unit, Llanfrynach, Powys LD3 7AQ, U.K.
This study tested the ‘silver spoon’ hypothesis which posits that individuals that develop under favourable conditions should enjoy a fitness advantage later in life because they are more likely to recognize and settle in high-quality habitats. Atlantic salmon Salmo salar of two age classes (0+ and 1+ years) were reared in environmentally enriched or standard hatchery tanks for a short period (c. 10 weeks), were then released into a natural river and sampled on repeated occasions to test for silver-spoon effects. Compared with controls, enriched fish had a 6⋅4% higher recapture rate and settled in higher velocity habitats when they were stocked as 0+ year fry, but not when they were stocked as 1+ year parr. The opportunity for selection was generally higher for environmentally enriched fish than for controls, and also higher for 0+ than for 1+ year fish. Selection favoured individuals with high condition factor, extensive fat reserves and longer than average pectoral fins in both age classes but favoured a small body size in 1+ year and a large body size in 0+ year releases. Stomach analysis showed that enriched fish ate more, and adapted quicker to natural prey than controls. These results provide support for silver-spoon effects in fish and indicate that enrichment can improve post-release performance in conservation programmes, but seemingly only if fish are not kept in captivity for too long. © 2014 The Fisheries Society of the British Isles
Key words: domestication; feralization; fitness; hatcheries; salmonids; selection.
INTRODUCTION Transitions into novel environments often pose important survival risks to animals that rely on familiarity with local conditions to find food, seek shelter or escape from predators (Dahlgren & Eggleston, 2000; Cimprich et al., 2005). Such transitions are particularly problematic for hatchery fish released into the wild because hatcheries impose very different selective pressures from those encountered under natural conditions (Fleming & Einum, 1997; Hard et al., 2000; Araki et al., 2008; Burns et al., 2009; Mayer et al., 2011). Hatchery fish often perform poorly in the wild (Jonsson et al., 2003; Saloniemi et al., 2004; Kallio-Nyberg et al., 2006; Serrano et al., 2009) as ‡Author to whom correspondence should be addressed. Tel.: +44 (0) 1792 29 53 83; email: [email protected]
1972 © 2014 The Fisheries Society of the British Isles
E N V I R O N M E N TA L E N R I C H M E N T I N S A L M O S A L A R
1973
they tend to settle in unfavourable habitats (Munakata et al., 2000), lack the necessary antipredatory skills (Alvarez & Nicieza, 2003; Seiler & Keeley, 2007; Sparrevohn & Stottrup, 2007; Roberts & Garcia de Leaniz, 2011; Roberts et al., 2011) and are unable to forage efficiently (Orlov et al., 2006; Larsson et al., 2011). This often results in high mortalities, particularly during the first few days post-release (Olla et al., 1998; Brown & Day, 2002). Environmental enrichment (i.e. the addition of structurally complex features during captive rearing) could promote behavioural flexibility and enhance the cognitive capacity of animals to adapt to novel situations (Kempermann et al., 2002; Kerridge, 2005; Kihslinger & Nevitt, 2006; Näslund et al., 2012). For example, behavioural flexibility can be enhanced in captive-bred Atlantic cod Gadus morhua, L. 1758 through the addition of spatial and foraging cues (Braithwaite & Salvanes, 2005; Salvanes & Braithwaite, 2005). Esocids reared with artificial vegetation show reduced startle responses and preference for vegetated habitats (Einfalt et al., 2013). In salmonids, enrichment and predator conditioning have been shown to improve crypsis (Donnelly & Dill, 1984; Maynard et al., 2005), foraging ability (Brown et al., 2003a, b; Rodewald et al., 2011), neural plasticity (Salvanes et al., 2013), antipredatory behaviours (Berejikian et al., 2003; Brown, 2003; Vilhunen, 2006; Roberts et al., 2011) and use of shelters, which may reduce cortisol levels (Näslund et al., 2013), at least under laboratory or semi-natural conditions. More generally, development under favourable conditions (i.e. those most likely to be preferred in the wild) could increase fitness later in life through ‘silver spoon’ effects because individuals may be better able to recognize and disperse into high-quality territories (Stamps, 2006). Yet, despite these potential benefits, there is conflicting evidence regarding the effects of enrichment on post-release survival of fish in the wild. For example, Tatara et al. (2009) found no difference in recapture rates of steelhead trout Oncorhynchus mykiss (Walbaum 1792) that had been reared in conventional or enriched environments. Similarly, no difference in recapture rates was found between juvenile Atlantic salmon Salmo salar L. 1758 reared in conventional or in structured tanks (Brockmark et al., 2007), or between control fish and fish conditioned to the threat of predation (Hawkins et al., 2008). In contrast, predator exposure and shelter conditioning increased survival of seabream Diplodus sargus (L. 1758) released in the wild (D’Anna et al., 2012), and enrichment also had a positive effect in the migration speed and post-release survival of hatchery-reared S. salar smolts (Hyvärinen & Rodewald, 2013). One reason why enrichment may not necessarily translate into better post-release survival may rest in the timing of enrichment in relation to the duration of rearing (Näslund et al., 2012). Captive fishes may become increasingly domesticated with duration of captive rearing (Huntingford, 2004) and the innate capacity to respond to predators may be suppressed with increasing time spent in the hatchery environment (Huntingford et al., 2006), particularly if development of antipredatory behaviour takes place early during ontogeny (Hawkins et al., 2008; Larsson et al., 2011). Furthermore, captive rearing may select for phenotypes with a reduced capacity for learning and reduced ability to retain learned information (Brown et al., 2013). Therefore, the benefits of enrichment may be limited to an early age (when fish are least domesticated and are most receptive to predators; Hawkins et al., 2008), or may also depend on timing of release and presence of other fishes, because prior residency may confer resident fish a competitive advantage (Huntingford & Garcia de Leaniz, 1997).
© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1972–1991
1974
L . J . R O B E RT S E T A L.
To address these issues, the rearing environment of juvenile S. salar was enriched at two different stages during ontogeny (0+ and 1+ years) but for the same length of time. The post-release performance of enriched fish was then compared with that of controls reared under standard hatchery conditions while maintaining rearing densities constant. Experimental fish were stocked upstream of a barrier impassable to spawning migrants and devoid of conspecifics to avoid confounding prior-residency effects (Huntingford & Garcia de Leaniz, 1997), and several fitness-related traits were recorded before and after stocking to better understand the targets of selection. The hypothesis was that fish reared under enriched conditions would perform better than controls owing to silver-spoon effects (Stamps, 2006). The expectation was that enrichment would have the strongest effect on the youngest, least domesticated individuals (Näslund et al., 2012).
MATERIALS AND METHODS The study was carried out during 2007–2008 using two age classes (0+ year fry and 1+ year parr) of juvenile S. salar reared at the Natural Resources Wales Cynrig Fish Culture Unit and derived in both cases from the stripping of broodstock caught in the River Taff (South Wales, U.K.).
E N V I R O N M E N TA L E N R I C H M E N T A N D P R E D AT O R CONDITIONING Two different cohorts were used for this study, 1+ year parr (mean ± s.e. body mass 29⋅30 ± 0⋅58 g) were enriched in 2007 and 0+ year fry (mean ± s.e. body mass 0⋅80 ± 0⋅02 g) were enriched in 2008 following a similar procedure in both years. Fish used during both years were reared under standard husbandry conditions (c. 4 months for 0+ year fry, and 1 year and 4 months for 1+ year parr). Two identical fibreglass circular tanks (5 m diameter × 1 m deep for 1+ year parr, and 3 m diameter × 1 m deep for 0+ year fry) each containing the same number of fish were selected for enrichment and two were kept under standard hatchery conditions to serve as controls. Loading densities were 4⋅51 g l−1 for 1+ year parr (3022 fish per tank, flow 120 l min−1 ) and 0⋅56 g l−1 for 0+ year fry (5000 fish per tank, flow 70 l min−1 ). To enrich the tanks, Lawson cypress Chamaecyparis lawsonia branches (2007) or pieces of cargo netting (2008) were anchored at the bottom of the tanks, camouflage netting was used to provide 30–40% overhead cover and natural prey (frozen bloodworms Chironomus sp. for 1+ year parr, and live Artemia sp. and natural invertebrates for 0+ year fry) were delivered via four submerged plastic pipes (100 cm × 5 cm) in each tank 10 times a day. The addition of natural prey amounted to
Silver spoons in the rough: can environmental enrichment improve survival of hatchery Atlantic salmon Salmo salar in the wild? L. J. Roberts*, J. Taylor†, P. J. Gough†, D. W. Forman* and C. Garcia de Leaniz*‡ *Swansea University, Centre for Sustainable Aquatic Research, Department of Biosciences, Swansea SA2 8PP, U.K. and †Natural Resources Wales, Cynrig Fish Culture Unit, Llanfrynach, Powys LD3 7AQ, U.K.
This study tested the ‘silver spoon’ hypothesis which posits that individuals that develop under favourable conditions should enjoy a fitness advantage later in life because they are more likely to recognize and settle in high-quality habitats. Atlantic salmon Salmo salar of two age classes (0+ and 1+ years) were reared in environmentally enriched or standard hatchery tanks for a short period (c. 10 weeks), were then released into a natural river and sampled on repeated occasions to test for silver-spoon effects. Compared with controls, enriched fish had a 6⋅4% higher recapture rate and settled in higher velocity habitats when they were stocked as 0+ year fry, but not when they were stocked as 1+ year parr. The opportunity for selection was generally higher for environmentally enriched fish than for controls, and also higher for 0+ than for 1+ year fish. Selection favoured individuals with high condition factor, extensive fat reserves and longer than average pectoral fins in both age classes but favoured a small body size in 1+ year and a large body size in 0+ year releases. Stomach analysis showed that enriched fish ate more, and adapted quicker to natural prey than controls. These results provide support for silver-spoon effects in fish and indicate that enrichment can improve post-release performance in conservation programmes, but seemingly only if fish are not kept in captivity for too long. © 2014 The Fisheries Society of the British Isles
Key words: domestication; feralization; fitness; hatcheries; salmonids; selection.
INTRODUCTION Transitions into novel environments often pose important survival risks to animals that rely on familiarity with local conditions to find food, seek shelter or escape from predators (Dahlgren & Eggleston, 2000; Cimprich et al., 2005). Such transitions are particularly problematic for hatchery fish released into the wild because hatcheries impose very different selective pressures from those encountered under natural conditions (Fleming & Einum, 1997; Hard et al., 2000; Araki et al., 2008; Burns et al., 2009; Mayer et al., 2011). Hatchery fish often perform poorly in the wild (Jonsson et al., 2003; Saloniemi et al., 2004; Kallio-Nyberg et al., 2006; Serrano et al., 2009) as ‡Author to whom correspondence should be addressed. Tel.: +44 (0) 1792 29 53 83; email: [email protected]
1972 © 2014 The Fisheries Society of the British Isles
E N V I R O N M E N TA L E N R I C H M E N T I N S A L M O S A L A R
1973
they tend to settle in unfavourable habitats (Munakata et al., 2000), lack the necessary antipredatory skills (Alvarez & Nicieza, 2003; Seiler & Keeley, 2007; Sparrevohn & Stottrup, 2007; Roberts & Garcia de Leaniz, 2011; Roberts et al., 2011) and are unable to forage efficiently (Orlov et al., 2006; Larsson et al., 2011). This often results in high mortalities, particularly during the first few days post-release (Olla et al., 1998; Brown & Day, 2002). Environmental enrichment (i.e. the addition of structurally complex features during captive rearing) could promote behavioural flexibility and enhance the cognitive capacity of animals to adapt to novel situations (Kempermann et al., 2002; Kerridge, 2005; Kihslinger & Nevitt, 2006; Näslund et al., 2012). For example, behavioural flexibility can be enhanced in captive-bred Atlantic cod Gadus morhua, L. 1758 through the addition of spatial and foraging cues (Braithwaite & Salvanes, 2005; Salvanes & Braithwaite, 2005). Esocids reared with artificial vegetation show reduced startle responses and preference for vegetated habitats (Einfalt et al., 2013). In salmonids, enrichment and predator conditioning have been shown to improve crypsis (Donnelly & Dill, 1984; Maynard et al., 2005), foraging ability (Brown et al., 2003a, b; Rodewald et al., 2011), neural plasticity (Salvanes et al., 2013), antipredatory behaviours (Berejikian et al., 2003; Brown, 2003; Vilhunen, 2006; Roberts et al., 2011) and use of shelters, which may reduce cortisol levels (Näslund et al., 2013), at least under laboratory or semi-natural conditions. More generally, development under favourable conditions (i.e. those most likely to be preferred in the wild) could increase fitness later in life through ‘silver spoon’ effects because individuals may be better able to recognize and disperse into high-quality territories (Stamps, 2006). Yet, despite these potential benefits, there is conflicting evidence regarding the effects of enrichment on post-release survival of fish in the wild. For example, Tatara et al. (2009) found no difference in recapture rates of steelhead trout Oncorhynchus mykiss (Walbaum 1792) that had been reared in conventional or enriched environments. Similarly, no difference in recapture rates was found between juvenile Atlantic salmon Salmo salar L. 1758 reared in conventional or in structured tanks (Brockmark et al., 2007), or between control fish and fish conditioned to the threat of predation (Hawkins et al., 2008). In contrast, predator exposure and shelter conditioning increased survival of seabream Diplodus sargus (L. 1758) released in the wild (D’Anna et al., 2012), and enrichment also had a positive effect in the migration speed and post-release survival of hatchery-reared S. salar smolts (Hyvärinen & Rodewald, 2013). One reason why enrichment may not necessarily translate into better post-release survival may rest in the timing of enrichment in relation to the duration of rearing (Näslund et al., 2012). Captive fishes may become increasingly domesticated with duration of captive rearing (Huntingford, 2004) and the innate capacity to respond to predators may be suppressed with increasing time spent in the hatchery environment (Huntingford et al., 2006), particularly if development of antipredatory behaviour takes place early during ontogeny (Hawkins et al., 2008; Larsson et al., 2011). Furthermore, captive rearing may select for phenotypes with a reduced capacity for learning and reduced ability to retain learned information (Brown et al., 2013). Therefore, the benefits of enrichment may be limited to an early age (when fish are least domesticated and are most receptive to predators; Hawkins et al., 2008), or may also depend on timing of release and presence of other fishes, because prior residency may confer resident fish a competitive advantage (Huntingford & Garcia de Leaniz, 1997).
© 2014 The Fisheries Society of the British Isles, Journal of Fish Biology 2014, 85, 1972–1991
1974
L . J . R O B E RT S E T A L.
To address these issues, the rearing environment of juvenile S. salar was enriched at two different stages during ontogeny (0+ and 1+ years) but for the same length of time. The post-release performance of enriched fish was then compared with that of controls reared under standard hatchery conditions while maintaining rearing densities constant. Experimental fish were stocked upstream of a barrier impassable to spawning migrants and devoid of conspecifics to avoid confounding prior-residency effects (Huntingford & Garcia de Leaniz, 1997), and several fitness-related traits were recorded before and after stocking to better understand the targets of selection. The hypothesis was that fish reared under enriched conditions would perform better than controls owing to silver-spoon effects (Stamps, 2006). The expectation was that enrichment would have the strongest effect on the youngest, least domesticated individuals (Näslund et al., 2012).
MATERIALS AND METHODS The study was carried out during 2007–2008 using two age classes (0+ year fry and 1+ year parr) of juvenile S. salar reared at the Natural Resources Wales Cynrig Fish Culture Unit and derived in both cases from the stripping of broodstock caught in the River Taff (South Wales, U.K.).
E N V I R O N M E N TA L E N R I C H M E N T A N D P R E D AT O R CONDITIONING Two different cohorts were used for this study, 1+ year parr (mean ± s.e. body mass 29⋅30 ± 0⋅58 g) were enriched in 2007 and 0+ year fry (mean ± s.e. body mass 0⋅80 ± 0⋅02 g) were enriched in 2008 following a similar procedure in both years. Fish used during both years were reared under standard husbandry conditions (c. 4 months for 0+ year fry, and 1 year and 4 months for 1+ year parr). Two identical fibreglass circular tanks (5 m diameter × 1 m deep for 1+ year parr, and 3 m diameter × 1 m deep for 0+ year fry) each containing the same number of fish were selected for enrichment and two were kept under standard hatchery conditions to serve as controls. Loading densities were 4⋅51 g l−1 for 1+ year parr (3022 fish per tank, flow 120 l min−1 ) and 0⋅56 g l−1 for 0+ year fry (5000 fish per tank, flow 70 l min−1 ). To enrich the tanks, Lawson cypress Chamaecyparis lawsonia branches (2007) or pieces of cargo netting (2008) were anchored at the bottom of the tanks, camouflage netting was used to provide 30–40% overhead cover and natural prey (frozen bloodworms Chironomus sp. for 1+ year parr, and live Artemia sp. and natural invertebrates for 0+ year fry) were delivered via four submerged plastic pipes (100 cm × 5 cm) in each tank 10 times a day. The addition of natural prey amounted to
