SEARCH

SEARCH BY CITATION

Keywords:

  • competition;
  • ecology;
  • genetic modification;
  • invasion;
  • risk assessment;
  • transgenic;
  • trout

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  1. The first genetically modified (GM) fish intended for human consumption has recently stimulated significant scientific discussion and regulatory scrutiny regarding food safety and environmental risks. Currently, no experiments with transgenic fish have been performed in nature, yet such data are needed to facilitate predictions of ecological consequences should engineered fish escape to the natural environment.
  2. To address this limitation, we conducted experiments under natural conditions but within a contained environment to assess the impact of invasion of growth-enhanced GM coho salmon Oncorhynchus kisutch (Walbaum) on survival and growth of three naturally cohabitating fishes: Chinook salmon Otshawytscha (Walbaum), steelhead trout O. mykiss (Walbaum) and conspecific wild-type coho salmon.
  3. We found that the impact of stream-reared GM coho salmon on invaded specimens was similar to the impact of non-GM coho salmon. However, GM fish significantly reduced survival and growth of the invaded populations if they were first allowed to grow larger under hatchery conditions before being released.
  4. Synthesis and applications. Our results show that the ecological impact of fish genetically modified (GM) for rapid growth on closely related fish species may not be high in stream environments, unless these fish are first reared under culture conditions where they are able to realize their genetic growth potential. As such, first generation escapes of GM fish into the natural environment should be a main concern in the short term, whereas later generations, which are more similar to naturally occurring genotypes, are expected to have significantly weaker effects but which could persist for longer periods.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Non-native, invasive organisms cause large ecological and economical costs each year. Ecological costs can occur as extinction of endemic species, reduced genetic variation in native species, changing habitats, reduced ecosystem services and increased pathogen exposure, while economic costs can be observed as reduced production in aquaculture and fisheries and impacts on international trade (Vilà et al. 2009; Lovell, Stone & Fernandez 2006). Genetically modified (GM) animals share many similarities with invasive species (Jeschke, Keesing & Ostfeld 2013). Some GM animals are intended to be used in nature, whereas others are intended to be kept strictly in contained production facilities, but there is always a risk of escape (Alphey et al. 2002; Forabosco et al. 2013; Thresher et al. 2013). Among the latter are several species of fish that have been genetically modified to enhance growth rate to meet the increasing demand for aquaculture products world-wide. These fish will, in most cases, be farmed in contained aquaculture settings, but still pose a potential threat to native biota due to the risk of escape (Naylor et al. 2005).

Fish can respond strongly to growth hormone (GH) overexpression by transgenesis due to the indeterminate nature of their growth physiology (Nam et al. 2001; Devlin, Biagi & Yesaki 2004). Associated with this transgenesis is a plethora of pleiotropic effects on the morphology, physiology, foraging and risk-taking behaviours and life-history traits of the animal (Devlin, Sundström & Muir 2006; Pennington & Kapuscinski 2011). Some GM fish lines have recently been subjected to extensive assessment and debate as they proceed through regulatory evaluation for use commercially (Fox 2010; Smith et al. 2010; Van Eenennaam & Muir 2011). Although potential ecological consequences of fast-growing GM fish have been discussed extensively for decades (Tiedje et al. 1989; Snow et al. 2005; Kapuscinski et al. 2007), very little empirical work has been done using GM fish under the conditions likely to be encountered by the fish once in nature (Devlin, Sundström & Muir 2006; Devlin et al. 2007).

To date, studies with GH transgenic fish conducted under conditions mimicking nature have focused on intraspecific predation (Sundström et al. 2007b), competition during early developmental stages (Dunham et al. 1999; Sundström, Lõhmus & Devlin 2005; Moreau et al. 2011), competition during breeding (Bessey et al. 2004; Fitzpatrick et al. 2011; Moreau, Conway & Fleming 2011) and multi-generation competition (Pennington & Kapuscinski 2011). Along the same lines, modelling exercises have focused not on ecological consequences, but rather on fitness comparisons of GM vs. non-GM conspecifics following invasion by GM fish (Muir & Howard 1999; Hedrick 2001; Aikio, Valosaari & Kaitala 2008; Ahrens & Devlin 2011). The only direct assessment of the impact of GM fish on other species has been through quantification of predation effects (Sundström et al. 2009) where GM salmon reared in the hatchery consumed more prey than non-GM reared in the hatchery, but compensatory growth responses in prey resulted in final biomass among prey being similar across predator treatments.

It is very likely that the insertion of a transgene into a wild strain that modifies physiological and behavioural processes will in the majority of cases cause a reduction in fitness from phenotypes that have been under natural selection conditions, although this may not always be the case [e.g. insertion of disease resistance genes (Mao et al. 2004)]. An observation emerging from the previous work is that the genetic modification can cause plastic phenotypes in transgenic fish to respond differently than non-GM fish to experimental and rearing conditions (genotype by environmental influences), complicating the use of data from culture conditions for prediction of how fish from natural conditions will behave in all cases (Sundström et al. 2007b). Because growth-enhanced transgenic and artificially selected fast-growing fish can develop more extreme phenotypes (e.g. body size, feeding behaviour) relative to wild type in the hatchery compared with nature, most studies on hatchery-reared fish are typically expected to overestimate risks compared with when such fish are studied under natural conditions. However, it is also possible that there may be conditions in the culture environment for which the relative performance of GM fish to wild type is underestimated as compared to natural conditions. Reduced phenotypic effects of GH transgenes under natural conditions may actually allow transgenes to persist and cause long-term population effects, compared with more extreme hatchery-reared GM fish which are likely to experience strong selection and elimination after entry to nature.

To address the shortage of studies examining consequences of GM fish on other ecosystem members beyond predation effects, we here report on four experiments focusing on competitive interactions with two ecologically similar species, Chinook salmon and steelhead trout, as well as conspecific coho salmon. We simulated invasions by GM coho salmon on these wild-type populations (we use the term wild type for fish that were invaded) and measured growth and survival following invasion. To control for the increased density arising from the presence of young GM coho salmon, we compared the performance of the invaded wild-type fish with populations invaded by an equal number of invading non-GM coho salmon (these invaders were also of wild genotype but are referred to as non-GM to distinguish them from the invaded fish).

We developed contained naturalized stream environments, closely mimicking natural stream systems for young salmon and resulting in growth rates similar to that found in nature (Sundström et al. 2007b). In the hatchery environment, the transgenic strain used in this study have the capacity to grow 2–3 times that of the unmodified non-GM genotype, resulting in fish that reach 2500 g within 2 years instead of the normal four (Devlin, Biagi & Yesaki 2004). This rapid growth is achieved mainly by an increased food intake and partly by increased food conversion efficiency (Raven, Devlin & Higgs 2006). Previous knowledge of the relative survival and growth of growth-enhanced GM fish under various conditions was used to design the experiments: susceptibility to predation (Abrahams & Sutterlin 1999; Sundström, Lõhmus & Devlin 2005; Duan et al. 2010), effects of rearing conditions (Sundström et al. 2007b) and habitat complexity (Devlin et al. 2004; Sundström & Devlin 2011).

To sustain their more rapid growth and greater food consumption, we expected the effects of GM coho salmon, relative to effects by non-GM conspecifics, to reduce growth and possibly survival of their cohabitants more than an equal number of invading non-GM coho. However, in the presence of predators, we expected the effects of GM fish on the invaded fish to be less owing to increased predation mortality on GM fish compared with non-GM fish. We also expected effects on the cohabitant from the GM fish to be greater in less complex habitats and when GM fish experienced a hatchery-reared period before invasion.

Material and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Experimental Animals and Facilities

Research was conducted in a contained facility at Fisheries and Oceans Canada's Centre for Aquaculture and Environmental Research in West Vancouver, Canada, designed to prevent escape of fish to the natural environment. All experiments were conducted under Canadian Council for Animal Care approved Animal Use Protocols.

Genetically modified (transgenic) genotypes of coho salmon were 5–7th generation offspring of the same strain (M77) containing a hemizygous copy of the OnMTGH1 GH gene construct (Devlin et al. 1994; Devlin, Biagi & Yesaki 2004). The transgenic strain was derived from the Chehalis River, located in south-western British Columbia. To maintain a wild genetic background and avoid effects of domestication, GM fish populations were produced each generation by crossing males homozygous for the transgene insert with multiple (>10) wild females derived from the Chehalis River, and wild-type non-GM coho were produced as half-sibs from the same wild females, utilizing wild males as sires. Chinook salmon and steelhead trout were all produced by crossing multiple wild-caught parents and experimental specimens were randomly collected from a common pool of crosses.

All fertilized eggs were incubated under similar conditions in 10 °C flow-through well water until the first-feeding stage. When fish were reared in the hatchery, they were fed formulated stage-appropriate salmon feeds (Skretting Inc. Vancouver, BC, Canada) to satiation five times per day. Stream-reared fish were reared under conditions as described below under the heading Experimental Stream Tanks. When sampled, tanks were emptied, and fish captured from the tanks, counted, weighed and length measured. All ages of fish are specified in time post-first feeding.

Experimental Stream Tanks

The experimental setup consisted of 12 stream tanks [water volume 2000 l; 5 × 1 × 0·4 m; except in the experiment on Chinook parr which was carried out in 12 smaller tanks (water volume 450 l; 2·5 × 0·6 × 0·3 m)] that were landscaped with large gravel, numerous large rocks, bushy material and two large logs to provide refuge for fish and for predators when present. In all experiments, flow-through creek water followed natural variations in water parameters (e.g. pH, temperature and turbidity). Additional water movement was created by a pump (2 m3 h−1) in one corner of each tank.

During rearing and experiments in the stream tanks, fish were stochastically fed roughly 2–3% wet weight per body weight, 1–3 times per day. Food consisted of frozen or live natural feed (e.g. artemia, mysis, blood worms, tubifex, crickets, fruit flies etc.) items on average 3 days of 4, and natural prey was continuously entering the facility through the creek water (larvae of stoneflies, caddisflies, mayflies). This feeding regime is known to make non-GM fish grow at the same rate observed in nature (Sundström et al. 2007b).

Steelhead Trout Invaded at First-Feeding Stage

Fifty first-feeding steelhead trout fry (28·9 ± 1 mm) were introduced into each of the 12 stream tanks together with 50 non-GM (44·4 ± 7·0 mm; in six tanks) and 50 GM (45·9 ± 5·8 mm; in the other six tanks) coho salmon fry that had been reared in the hatchery for 1 month. After 10 days, two wild-type coho salmon presmolts (21·6 g ± 2·6 SD; 12·4 mm ± 0·61) were introduced into every other tank as predators. This resulted in a 2 × 2 factorial design with predator present/absent and invading genotype (non-GM vs. GM) as the two factors. Fish were sampled after 27 days.

Chinook Salmon Fry Invaded at First-Feeding Stage

Fifty first-feeding Chinook salmon (37·4 ± 1·3 mm) were introduced with 50 first-feeding non-GM (31·0 ± 1·5) or GM coho salmon (33·3 ± 1·5 mm) in each stream tank. After 1 week, half of the tanks received one cutthroat (161 mm ± 9·7 SD) and one rainbow trout (335 mm ± 11·2) predator. This was a 2 × 2 factorial design with predator present/absent and invading genotype (non-GM vs. GM) as the two factors. Fish were sampled after 166 days.

Chinook Salmon Parr Invaded by Hatchery- and Stream-Reared Fish

Chinook salmon parr reared in stream tanks (49·4 ± 2·7 mm; from the non-predator treatment in experiment on Chinook fry) for 6 months were transferred to each of 12 smaller stream tanks with five fish in each tank. These stream tanks were set up similarly to the large stream tanks but with smaller size structure, a water inflow of 0·5 m3 h−1 and without the additional pump. After 6 days, each of these tanks received five non-GM or GM coho salmon reared either in stream tanks (non-GM: 42·5 ± 3·7 mm; GM: 45·6 ± 3·7 mm) or in the hatchery (non-GM: 73·1 ± 6·1 mm; GM: 124 ± 8·7 mm). Hence, each tank contained five Chinook and five coho salmon in a 2 × 2 design with invader genotype (non-GM vs. GM) and invader rearing background (stream tank vs. hatchery) as the two factors. Fish were sampled after 62 days.

Coho Salmon Invaded

In half of 12 large stream tanks, large structures were removed leaving only gravel on the bottom. These tanks were ‘simple habitats’, whereas those with structure remaining were ‘complex habitats’. Next, each tank received 15 wild-type coho reared in the hatchery for 1 month (42·7 ± 4·2 mm). After another month, each tank received 15 non-GM (59·5 ± 3·4 mm) or GM (60·9 ± 3·9 mm) invaders reared in the hatchery. Hence, this experiment was a 2 × 2 factorial design with habitat complexity (simple vs. complex) and invading genotype (non-GM vs. GM) as the two factors. An adipose fin clip was used to distinguish between wild-type (invaded) and invading coho salmon, with the clipped group alternating across tanks. No effect of clipping was detected among replicates. Fish were sampled after 231 days.

Statistical analyses

Invaded fish and invading fish were analysed separately using each tank as the independent sample unit. We used R version 2.15.1 (R Development Core Team 2009), and the cut-off for considered being significant was = 0·05.

We analysed survival with generalized linear models (GzLM), and growth performance (specific growth rate (SGR) calculated as G = 100 × (ln(tank average final size) – ln(tank average initial size))/t, where t is the number of days between measurement of initial and final sizes providing 12 independent replicates) with general linear models (GLM).

The GzLM on survival was fitted using a binomial or quasibinomial (if data were overdispersed) distribution. After fitting the full factorial model, the model selection procedure followed that of Crawley (2007) using analysis of deviance by first removing non-significant interactions terms followed by removal of non-significant factors. If the interaction contributed significantly to the model, detailed analysis was performed by testing only nearby groups (e.g. wild fish in presence of predators was tested against wild fish in the absence of predators excluding all GM fish). Results are reported as the model deviance (D).

General linear model on growth is reported as the anova output using type III sums of squares from the car package in R (http://cran.r-project.org/web/packages/car/index.html). When the interaction term was significant, the two-way design was collapsed into the four combinations and a one-way anova was run followed by Tukey's HSD to discern differences between groups.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In three of four invaded populations, survival was the same regardless of whether the invaders were GM or non-GM fish (Fig. 1a). A difference in survival was found when Chinook salmon parr were invaded by stream- and hatchery-reared coho salmon invaders (Genotype × rearing background: D = −5·3, = 0·021). When invaders were reared under semi-natural conditions prior to invasion, the survival of the invaded populations was still similar; 93% when invaded by GM salmon and 100% when invaded by non-GM salmon. However, when the invading GM coho salmon had been reared under hatchery conditions for 6 months prior to invasion, they did reduce survival (to 20%) of the invaded population, whereas non-GM invaders had no effect on survival of Chinook parr (100% survived). In the two experiments (with steelhead and Chinook fry) including predators (Fig. 1a), survival was significantly reduced in the invaded population in the presence of predators, but their survival did not depend on whether the invaders were GM or non-GM coho salmon (steelhead: D = −328, < 0·001, Chinook: D = −209, < 0·001). Habitat complexity had no effect on survival of the invaded wild-type coho population after 231 days (Fig. 1a), with survival being very high overall (>97%).

image

Figure 1. Survival (a) and specific growth in length (SGRL; (b) of invaded wild-type fish of three salmonid focal species (steelhead trout, Chinook salmon and coho salmon) in four different experiments after invasion by wild-type (non-GM) coho salmon or growth-enhanced genetically modified (GM) coho salmon. Terms on x-axis describe experimental treatments [presence/absence of predators; rearing background (hatchery- or stream-like); and structural complexity of the invaded habitat (only gravel on the bottom or the addition of larger rocks and logs)]. The performance of the invading fish is shown in Fig. 2. Due to differences in fish starting size among experiments, growth performance should only be compared within experiments. Vertical bars show standard error of mean (obscured by the point when small). Note that because there was only one survivor in the predator treatment for Chinook fry, no SGRL was calculated.

Download figure to PowerPoint

Specific growth rates of the invaded populations matched that of survival (Fig. 1b). Growth of invaded steelhead trout and first-feeding Chinook salmon did not depend on the genotype (GM vs. non-GM) of the invaders after 27 days (invader genotype effects on length growth: F = 0·51, = 0·53) and 166 days (F = 0·71, = 0·46), respectively. However, as for survival, the growth of the older, invaded, Chinook salmon parr was reduced in the presence of hatchery-reared, but not stream-reared, GM coho salmon relative to growth of Chinook invaded by non-GM coho regardless of rearing background (Genotype × Rearing background: F = 10·4, = 0·015). When wild-type coho salmon were invaded, their growth was faster in the presence of GM coho salmon in the complex habitat, while non-GM invaders and the GM invaders in simple habitats had similar effects on growth of the invaded coho (Fig. 1a).

Survival of the invading GM and non-GM coho salmon generally either did not differ or was higher in the GM fish across experiments (Fig. 2a). Presence of predators reduced survival of both genotypes similarly when invading first-feeding steelhead trout and Chinook salmon; however, in the absence of predators, GM salmon had better survival than the non-GM invader when invading the Chinook salmon fry (Genotype × Predator: D = −6·4, = 0·003). When invading Chinook parr, there was no statistical difference (Genotype: D = −0·18, = 0·67), but both types of GM fish had 100% survival, whereas the non-GM had 93% survival if stream-reared and 73% survival if hatchery-reared. Overall, survival of both GM and non-GM coho was high when invading conspecific wild-type coho, survival rate was 80% or higher in all replicate habitats and showed a significant interaction between genotype and habitat complexity (D = −7·6, = 0·006), but with no significant pairwise test (Fig. 2b).

image

Figure 2. Survival (a) and specific growth in length (SGRL; (b) of the invading non-GM and GM coho salmon after invading three wild-type salmonid species (steelhead trout, Chinook salmon and coho salmon) in four different experiments. Terms on x-axis describe experimental treatments [presence/absence of predators; rearing background (hatchery- or stream-like); and structural complexity of the invaded habitat (only gravel on the bottom or the addition of larger rocks and logs)]. The performance of the invaded fish is shown in Fig. 1. Due to differences in the initial size of fish among experiments, growth performance should only be compared within experiments. Vertical bars show standard error of mean (obscured by the point when small). Note that because there was only two non-GM and no GM survivors in the predator treatment for Chinook fry, no SGRL was calculated.

Download figure to PowerPoint

Despite their much greater growth potential, GM coho salmon invaders did not consistently outgrow their non-GM comparators (Fig. 2b). To the contrary, non-GM coho invaders grew faster than GM coho when invading both steelhead trout (F = 11·7, = 0·011) and Chinook salmon fry (F = 8·9, = 0·041). Invasion of Chinook parr presented a more complex picture partly because invading fish were of very different sizes and growth is size-dependent (Fig. 3). Nevertheless, SGR and absolute growth rates (in mm and g) were congruent: GM invaders grew more than non-GM when both had been reared under stream conditions prior to invading, but the reverse was true when both were hatchery-reared (genotype × rearing background: F = 46·3, < 0·001). When invading conspecific coho salmon, GM coho grew larger in the simple habitat compared with the complex habitat in comparison with the non-GM in the two habitats (genotype × habitat complexity: F = 150, < 0·001; Fig. 2b).

image

Figure 3. Example phenotypes of the coho salmon invading Chinook salmon parr before and after the invasion. Left column shows the genetically modified (GM) and the right column the non-GM (wild-type) invaders. The two top rows show fish reared in a contained stream habitat, and the two bottom rows show fish reared in a hatchery environment. Phenotypes are only representatives and were scaled to the average length of the group they represent.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Our results show that an invasion by genetically modified (GM) growth-enhanced coho salmon that were reared in naturalistic stream environments had, for the life stages examined, no or only small effects on invaded resident populations of related species from the same genus Oncorhynchus compared with the corresponding invasion of non-GM coho salmon. In contrast, a significant effect of invader genotype was observed when GM fish were first reared in the hatchery and obtained a much larger body size than GM fish reared in a semi-natural environment (Table 1). Hatchery-reared GM salmon significantly reduced both survival and growth of the invaded wild-type target populations, possibly also acting as predators due to their large size. However, the large body size is unlikely to be the only determinant of effects as the invading non-GM coho reared in the hatchery were also much larger than the invading stream-reared non-GM coho salmon, but their effects on the invaded Chinook salmon did not differ from their smaller stream-reared comparators.

Table 1. Summary of survival and growth effects observed for the invaded wild-type population after invasion by genetically modified (GM) and wild genotype (non-GM) coho salmon
Species invadedEnvironmental factorsSurvivalaGrowtha
  1. a

    Rearing history of invading fish in the hatchery (h) or in semi-natural streams (s), studied in a complex (c) or simple environment (s). E.g. GMh < non-GMh means that invasion by genetically modified (GM) coho salmon reared in the hatchery (h) reduced growth of the invaded species (Chinook) compared to the growth of the invaded species when it was invaded by wild genotype (non-GM) fish reared in the hatchery (h).

  2. b

    Predation reduced survival but did not depend on genotype of the invaders.

  3. c

    Growth was higher in the invaded population in the presence of GMs compared to the presence of any non-GM.

SteelheadPredation riskbGM = non-GMGM = non-GM
ChinookPredation riskbGM = non-GMGM = non-GM
ChinookRearing historyGMh < GMs, non-GMh, non-GMsGMh < GMs, non-GMh, non-GMs
CohoHabitat complexityGM = non-GMGMsc < non-GMs = non-GMc ≤ GMc

These data together show that cultured GM salmon have a predator/competitive phenotype distinct from other groups examined, and suggest that escapes or releases of older GM coho salmon from culture facilities may pose the greatest concern to sympatric species. Thus, important factors to consider for determining the impact of escaped growth-enhanced GM fish would be (i) age at escape as very young fish may develop into more wild-like phenotypes (Sundström et al. 2007b), whereas older GM fish appears to retain their hatchery phenotype even after being in a stream-like environment for several months (Sundström et al. 2009; Sundström, Lõhmus & Devlin 2010), (ii) number and frequency of escapes as repeated ‘trials’ increase the chance that at least some escaped individuals will be successful (McGinnity et al. 2003; Kolar & Lodge 2001), (iii) longevity and dispersal ability of escaped individual (Sundström et al. 2007a; Uglem et al. 2008; Solem et al. 2013) as this determine the temporal and spatial extent of impact. The first course of action would therefore be to prevent or minimize escape events from the culture facilities through physical containment, focusing on older fish (Devlin & Donaldson 1992; Mair, Nam & Solar 2007).

Longevity and dispersal ability of GM fish would influence their impact and would be affected by geographical containment by rearing fish in an environment that does not afford long-term survival. Fish that are by design poorly adapted to the receiving environment may also provide an approach for reducing their fitness. In the present experiments, fish were bred in a way that introduced the transgene into an otherwise wild genetic background. However, GM fish may in some cases be transgenic for a single trait but also experience artificial selection in the hatchery environment in order to further modify production traits, leading to domestication effects as occurs for many non-transgenic aquaculture species (Duarte, Marbá & Holmer 2007; Teletchea & Fontaine 2012). Studies show how domestication reduces fitness following escape from culture conditions (Jonsson & Jonsson 2006; Araki, Cooper & Blouin 2009; Milot et al. 2013), presumably arising because a rapid growth phenotype (selected for under hatchery conditions) is often costly in nature (Biro et al. 2004; Dmitriew 2011; Saikkonen, Kekäläinen & Piironen 2011). Notwithstanding, rapid growth and domestication may not always reduce fitness – indeed, enhancement of growth in wild-type brown trout in nature by treatment with exogenous GH did not cause a reduction in survival (Johnsson & Björnsson 2001). Hence, the severity of escaped hatchery-reared GM fish is difficult to assess and is likely to depend on characteristics of the receiving environment (Levin, Zabel & Williams 2001).

In the present study, habitat complexity influenced the experimental outcome over a longer time period (230 days), with invasion by GM coho salmon fry causing reduction in growth of the invaded coho population in simple, but not in complex environments (compared with effects of non-GM invaders). These effects, albeit relatively weak, agree with GH having a greater effect under hatchery conditions than under natural conditions (Johnsson et al. 2000). Growth hormone stimulates appetite and can increase aggression in fish (Johnsson & Björnsson 1994), while a complex habitat can reduce the advantage of aggressive phenotypes (Höjesjö, Johnsson & Bohlin 2004) which may explain why GM fish had a less of an impact in the complex compared with simple habitats. However, it is important to note that these relationships may not be linear across environments, such that other effects may be observed at intermediate complexity compared with little or much complexity, or indeed at further extremes (Crowder & Cooper 1982).

Mortality and growth rates of the invaded populations were less influenced by invader genotypes than by invader rearing background or experimental conditions, but the performance of the invaders differed somewhat depending on whether they were GM or wild type. In nature, coho salmon can be sympatric with Chinook salmon (Taylor 1991) and steelhead trout (Hartman 1965) and when so each species occupies different microhabitats. Hence, survival and growth of GM and wild-type coho salmon can differ without necessarily affecting sympatric fish species. Alternatively, effects on survival could be due to size differences, trout being smaller and Chinook fry initially being larger than invaders, making them relatively easier or more difficult for predators to capture compared with the coho invaders.

Our results show that growth-enhanced GM fish have, at least during early rearing in stream conditions, limited detectable impact on other closely related ecosystem members under naturalized laboratory conditions except if first given the opportunity to grow large in the hatchery. This suggests that a main short-term ecological concern with fish genetically modified to overexpress GH comes from such individuals that have first been able to grow at a rate that is probably not achievable in most natural environments. Currently, long-term multigenerational data are not available for potential impacts of fish that have lived their entire life history in nature (but see Pennington & Kapuscinski (2011) for an additional laboratory study under naturalized conditions), and thus, caution should be applied to extrapolate the current data to a full risk assessment scenario. Nevertheless, for both short-term and long-term scenarios, containment strategies are critical and should include preventing the escape of larger and older fish from culture conditions for which impacts have been demonstrated in the present data. Our data also show that, to reveal complex influences on phenotype and consequences of GM salmon, it is important to develop contained experimental facilities that mimic natural conditions to as great an extent as possible for risk assessment research, despite the difficulty of achieving full nature simulation. Clearly, impacts of GM salmon are dependent on the specific conditions under which they are assessed, as would be expected for an organism whose life history and phenotype plastically respond to complex interactions with multitude of environmental variables.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank two reviewers for their helpful comments on the manuscript and views on risk assessment. The work was carried out with financial support from the Canadian Regulatory System for Biotechnology (RHD). LFS was funded by a postdoctoral grant from the Swedish Research Council FORMAS, an assistant professorship from the Swedish research council Vetenskapsrådet, and a Marie Curie Outgoing International Fellowship under contract MOIF-CT-2005-8141 from the European Community's Sixth Framework Programme. The present work does not necessarily reflect the community's views and in no way anticipates its future policy in this area. We have no conflict of interest to declare.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Abrahams, M.V. & Sutterlin, A. (1999) The foraging and antipredator behaviour of growth-enhanced transgenic Atlantic salmon. Animal Behaviour, 58, 933942.
  • Ahrens, R.N.M. & Devlin, R.H. (2011) Standing genetic variation and compensatory evolution in transgenic organisms: a growth-enhanced salmon simulation. Transgenic Research, 20, 583597.
  • Aikio, S., Valosaari, K.-R. & Kaitala, V. (2008) Mating preference in the invasion of growth enhanced fish. Oikos, 117, 406414.
  • Alphey, L., Beard, C.B., Billingsley, P., Coetzee, M., Crisanti, A., Curtis, C. et al. (2002) Malaria control with genetically manipulated insect vectors. Science, 298, 119121.
  • Araki, H., Cooper, B. & Blouin, M.S. (2009) Carry-over effect of captive breeding reduces reproductive fitness of wild-born descendants in the wild. Biology Letters, 5, 621624.
  • Bessey, C., Devlin, R.H., Liley, N.R. & Biagi, C.A. (2004) Reproductive performance of growth-enhanced transgenic coho salmon (Oncorhynchus kisutch). Transactions of the American Fisheries Society, 133, 12051220.
  • Biro, P.A., Abrahams, M.V., Post, J.R. & Parkinson, E.A. (2004) Predators select against high growth rates and risk-taking behaviour in domestic trout populations. Proceedings of the Royal Society B: Biological Sciences, 271, 22332237.
  • Crawley, M.J. (2007) The R Book. John Wiley & Sons Ltd, West Sussex.
  • Crowder, L.B. & Cooper, W.E. (1982) Habitat structural complexity and the interaction between bluegills and their prey. Ecology, 63, 18021813.
  • Devlin, R.H., Biagi, C.A. & Yesaki, T.Y. (2004) Growth, viability and genetic characteristics of GH transgenic coho salmon strains. Aquaculture, 236, 607632.
  • Devlin, R.H. & Donaldson, E.M. (1992) Containment of genetically altered fish with emphasis on salmonids. Transgenic Fish (eds C.L. Hew & G.L. Fletcher), pp. 229265. World Scientific Publishers, Singapore.
  • Devlin, R.H., Sundström, L.F. & Muir, W.M. (2006) Interface of biotechnology and ecology for environmental risk assessments of transgenic fish. Trends in Biotechnology, 24, 8997.
  • Devlin, R.H., Yesaki, T.Y., Biagi, C.A., Donaldson, E.M., Swanson, P. & Chan, W.K. (1994) Extraordinary salmon growth. Nature, 371, 209210.
  • Devlin, R.H., D'Andrade, M., Uh, M. & Biagi, C.A. (2004) Population effects of growth hormone transgenic coho salmon depend on food availability and genotype by environment interactions. Proceedings of the National Academy of Sciences of the United States of America, 101, 93039308.
  • Devlin, R.H., Sundström, L.F., Johnsson, J.I., Fleming, I.A., Hayes, K.R., Ojwang, W.O., Bambaradenyia, C. & Zakaria-Ismail, M. (2007) Assessing ecological effects of transgenic fish prior to entry into nature. Environmental Risk Assessment of Genetically Modified Organisms. Vol 3. Methodologies for Transgenic Fish (eds A.R. Kapuscinski, K.R. Hayes, S. Li & G. Dana), pp. 151187. CAB International, Oxfordshire.
  • Dmitriew, C.M. (2011) The evolution of growth trajectories: what limits growth rate? Biological Reviews, 86, 97116.
  • Duan, M., Zhang, T., Hu, W., Guan, B., Wang, Y., Li, Z. & Zhu, Z. (2010) Increased mortality of growth-enhanced transgenic common carp (Cyprinus carpio L.) under short-term predation risk. Journal of Applied Ichthyology, 26, 908912.
  • Duarte, C., Marbá, N. & Holmer, M. (2007) Rapid domestication of marine species. Science, 316, 382.
  • Dunham, R.A., Chitmanat, C., Nichols, A., Argue, B., Powers, D.A. & Chen, T.T. (1999) Predator avoidance of transgenic channel catfish containing salmonid growth hormone genes. Marine Biotechnology, 1, 545551.
  • Fitzpatrick, J.L., Akbarashandiz, H., Sakhrani, D., Biagi, C.A., Pitcher, T.E. & Devlin, R.H. (2011) Cultured growth hormone transgenic salmon are reproductively out-competed by wild-reared salmon in semi-natural mating arenas. Aquaculture, 312, 185191.
  • Forabosco, F., Lõhmus, M., Rydhmer, L. & Sundström, L.F. (2013) Genetically modified farm animals and fish in agriculture: a review. Livestock Science, 153, 19.
  • Fox, J.L. (2010) Transgenic salmon inches toward finish line. Nature Biotechnology, 28, 11411142.
  • Hartman, G.F. (1965) The role of behavior in the ecology and interaction of underyearling coho salmon (Oncorhynchus kisutch) and steelhead trout (Salmo gairdneri). Journal of the Fisheries Research Board of Canada, 22, 10351081.
  • Hedrick, P.W. (2001) Invasion of transgenes from salmon or other genetically modified organisms into natural populations. Canadian Journal of Fisheries and Aquatic Sciences, 58, 841844.
  • Höjesjö, J., Johnsson, J.I. & Bohlin, T. (2004) Habitat complexity reduces the growth of aggressive and dominant brown trout (Salmo trutta) relative to subordinates. Behavioral Ecology and Sociobiology, 56, 286289.
  • Jeschke, J.M., Keesing, F. & Ostfeld, R.S. (2013) Novel organisms: comparing invasive species, GMOs, and emerging pathogens. AMBIO: A Journal of the Human Environment, 42, 541548.
  • Johnsson, J.I. & Björnsson, B.T. (1994) Growth-hormone increases growth rate, appetite and dominance in juvenile rainbow trout, Oncorhynchus mykiss. Animal Behaviour, 48, 177186.
  • Johnsson, J.I. & Björnsson, B.T. (2001) Growth-enhanced fish can be competitive in the wild. Functional Ecology, 15, 654659.
  • Johnsson, J.I., Jönsson, E., Petersson, E., Järvi, T. & Björnsson, B.T. (2000) Fitness-related effects of growth investment in brown trout under field and hatchery conditions. Journal of Fish Biology, 57, 326336.
  • Jonsson, B. & Jonsson, N. (2006) Cultured Atlantic salmon in nature: a review of their ecology and interaction with wild fish. ICES Journal of Marine Science, 63, 11621181.
  • Kapuscinski, A.R., Hayes, K.R., Li, S. & Dana, G. (2007) Environmental Risk Assessment of Genetically Modified Organisms. Vol 3. Methodologies for Transgenic Fish. CAB International, Oxfordshire.
  • Kolar, C.S. & Lodge, D.M. (2001) Progress in invasion biology: predicting invaders. Trends in Ecology & Evolution, 16, 199204.
  • Levin, P.S., Zabel, R.W. & Williams, J.G. (2001) The road to extinction is paved with good intentions: negative association of fish hatcheries with threatened salmon. Proceedings of the Royal Society B: Biological Sciences, 268, 11531158.
  • Lovell, S., Stone, S. & Fernandez, L. (2006) The economic impacts of aquatic invasive species: a review of the literature. Agricultural and Resource Economics Review, 35, 195208.
  • Mair, G.C., Nam, Y.K. & Solar, I.I. (2007) Risk management: reducing risk through confinement of transgenic fish. Environmental Risk Assessment of Genetically Modified Organisms. Vol 3. Methodologies for Transgenic Fish (eds A.R. Kapuscinski, K.R. Hayes, S. Li & G. Dana), pp. 209238. CAB International, Oxfordshire.
  • Mao, W., Wang, Y., Wang, W., Wu, B., Feng, J. & Zhu, Z. (2004) Enhanced resistance to Aeromonas hydrophila infection and enhanced phagocytic activities in human lactoferrin-transgenic grass carp (Ctenopharyngodon idellus). Aquaculture, 242, 93103.
  • McGinnity, P., Prodöhl, P., Ferguson, A., Hynes, R., Ó Maoiléidigh, N., Baker, N. et al. (2003) Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farmed salmon. Proceedings of the Royal Society B: Biological Sciences, 270, 24432450.
  • Milot, E., Perrier, C., Papillon, L., Dodson, J.J. & Bernatchez, L. (2013) Reduced fitness of Atlantic salmon released in the wild after one generation of captive breeding. Evolutionary Applications, 6, 472485.
  • Moreau, D.T.R., Conway, C. & Fleming, I.A. (2011) Reproductive performance of alternative male phenotypes of growth hormone transgenic Atlantic salmon (Salmo salar). Evolutionary Applications, 4, 736748.
  • Moreau, D.T.R., Fleming, I.A., Fletcher, G.L. & Brown, J.A. (2011) Growth hormone transgenesis does not influence territorial dominance or growth and survival of first-feeding Atlantic salmon Salmo salar in food-limited stream microcosms. Journal of Fish Biology, 78, 726740.
  • Muir, W.M. & Howard, R.D. (1999) Possible ecological risks of transgenic organism release when transgenes affect mating success: sexual selection and the trojan gene hypothesis. Proceedings of the National Academy of Sciences of the United States of America, 96, 1385313856.
  • Nam, Y.K., Noh, J.K., Cho, Y.S., Cho, H.J., Cho, K.N., Kim, C.G. & Kim, D.S. (2001) Dramatically accelerated growth and extraordinary gigantism of transgenic mud loach Misgurnus mizolepis. Transgenic Research, 10, 353362.
  • Naylor, R., Hindar, K., Fleming, I.A., Goldburg, R., Williams, S., Volpe, J. et al. (2005) Fugitive salmon: assessing the risks of escaped fish from net-pen aquaculture. BioScience, 55, 427437.
  • Pennington, K.M. & Kapuscinski, A.R. (2011) Predation and food limitation influence fitness traits of growth-enhanced transgenic and wild-type fish. Transactions of the American Fisheries Society, 140, 221234.
  • R Development Core Team (2009) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, http://www.R-project.org.
  • Raven, P.A., Devlin, R.H. & Higgs, D.A. (2006) Influence of dietary digestible energy content on growth, protein and energy utilization and body composition of growth hormone transgenic and non-transgenic coho salmon (Oncorhynchus kisutch). Aquaculture, 254, 730747.
  • Saikkonen, A., Kekäläinen, J. & Piironen, J. (2011) Rapid growth of Atlantic salmon juveniles in captivity may indicate poor performance in nature. Biological Conservation, 144, 23202327.
  • Smith, M.D., Asche, F., Guttormsen, A.G. & Wiener, J.B. (2010) Genetically modified salmon and full impact assessment. Science, 330, 10521053.
  • Snow, A.A., Andow, D.A., Gepts, P., Hallerman, E.M., Power, A., Tiedje, J.M. & Wolfenbargerh, L.L. (2005) Genetically engineered organisms and the environment: current status and recommendations. Ecological Applications, 15, 377404.
  • Solem, Ø., Hedger, R., Urke, H., Kristensen, T., Økland, F., Ulvan, E. & Uglem, I. (2013) Movements and dispersal of farmed Atlantic salmon following a simulated-escape event. Environmental Biology of Fishes, 96, 927939.
  • Sundström, L.F. & Devlin, R.H. (2011) Increased intrinsic growth rate is advantageous even under ecologically stressful conditions in coho salmon (Oncorhynchus kisutch). Evolutionary Ecology, 25, 447460.
  • Sundström, L.F., Lõhmus, M. & Devlin, R.H. (2005) Selection on increased intrinsic growth rates in coho salmon Oncorhynchus kisutch. Evolution, 59, 15601569.
  • Sundström, L.F., Lõhmus, M. & Devlin, R.H. (2010) Migration and growth potential of coho salmon smolts: implications for ecological impacts from growth-enhanced fish. Ecological Applications, 20, 13721383.
  • Sundström, L.F., Lõhmus, M., Johnsson, J.I. & Devlin, R.H. (2007a) Dispersal potential is affected by growth-hormone transgenesis in coho salmon (Oncorhynchus kisutch). Ethology, 113, 403410.
  • Sundström, L.F., Lõhmus, M., Tymchuk, W.E. & Devlin, R.H. (2007b) Gene-environment interactions influence ecological consequences of transgenic animals. Proceedings of the National Academy of Sciences of the United States of America, 104, 38893894.
  • Sundström, L.F., Tymchuk, W.E., Lõhmus, M. & Devlin, R.H. (2009) Sustained predation effects of hatchery-reared transgenic coho salmon Oncorhynchus kisutch in semi-natural environments. Journal of Applied Ecology, 46, 762769.
  • Taylor, E.B. (1991) Behavioural interaction and habitat use in juvenile Chinook Oncorhynchus tshawytscha and coho Oncorhynchus kisutch salmon. Animal Behaviour, 42, 729744.
  • Teletchea, F. & Fontaine, P. (2012) Levels of domestication in fish: implications for the sustainable future of aquaculture. Fish and Fisheries, doi: 10.1111/faf.12006.
  • Thresher, R.E., Hayes, K., Bax, N.J., Teem, J., Benfey, T.J. & Gould, F. (2013) Genetic control of invasive fish: technological options and its role in integrated pest management. Biological Invasions, doi: 10.1007/s10530-013-0477-0.
  • Tiedje, J.M., Colwell, R.K., Grossman, Y.L., Hodson, R.E., Lenski, R.E., Mack, R.N. & Regal, P.J. (1989) The planned introduction of genetically engineered organisms: ecological considerations and recommendations. Ecology, 70, 298315.
  • Uglem, I., Bjorn, P.A., Dale, T., Kerwath, S., Okland, F., Nilsen, R., Aas, K., Fleming, I. & McKinley, R.S. (2008) Movements and spatiotemporal distribution of escaped farmed and local wild Atlantic cod (Gadus morhua L.). Aquaculture Research, 39, 158170.
  • Van Eenennaam, A.L. & Muir, W.M. (2011) Transgenic salmon: a final leap to the grocery shelf? Nature Biotechnology, 29, 706710.
  • Vilà, M., Basnou, C., Pyšek, P., Josefsson, M., Genovesi, P., Gollasch, S. et al. (2009) How well do we understand the impacts of alien species on ecosystem services? A pan-European, cross-taxa assessment. Frontiers in Ecology and the Environment, 8, 135144.