Abstract
There is persistent commercial interest in the use of growth modified fishes for shortening production cycles and increasing overall food production, but there is concern over the potential impact that transgenic fishes might have if ever released into nature. To explore the ecological consequences of transgenic fish, we performed two experiments in which the early growth and survival of growth-hormone transgenic rainbow trout (Oncorhynchus mykiss) were assessed in naturalized stream mesocosms that either contained predators or were predator-free. We paid special attention to the survival bottleneck that occurs during the early life-history of salmonids, and conducted experiments at two age classes (first-feeding fry and 60 days post-first-feeding) that lie on either side of the bottleneck. In the late summer, the first-feeding transgenic trout could not match the growth potential of their wild-type siblings when reared in a hydrodynamically complex and oligotrophic environment, irrespective of predation pressure. Furthermore, overall survival of transgenic fry was lower than in wild-type (transgenic = 30% without predators, 8% with predators; wild-type = 81% without predators, 31% with predators). In the experiment with 60-day old fry, we explored the effects of the transgene in different genetic backgrounds (wild versus domesticated). We found no difference in overwinter survival but significantly higher growth by transgenic trout, irrespective of genetic background. We conclude that the high mortality of GH-transgenic trout during first-feeding reflects an inability to sustain the basic metabolic requirements necessary for life in complex, stream environments. However, when older, GH-transgenic fish display a competitive advantage over wild-type fry, and show greater growth and equal survival as wild-type. These results demonstrate how developmental age and time of year can influence the response of genotypes to environmental conditions. We therefore urge caution when extrapolating the results of GH-transgenesis risk assessment studies across multiple life-history or developmental stages.
Highlights
There is persistent interest in the production of transgenic organisms for use in basic and applied bio-medical research, plant and animal agriculture, and aquaculture
At the beginning of this experiment, there was no genotypic difference in the relationship between initial mass and length of fry; body mass did not differ between genotypes, despite a significant co-varying effect of body length (ANCOVA full model F3,39 = 63.84, P
To place our results in a larger context, our study suggests that if the present strain of GHtransgenic trout ever found its way into nature, its success would likely be constrained by the interacting effects of low resource availability and high predation risk
Summary
There is persistent interest in the production of transgenic organisms for use in basic and applied bio-medical research, plant and animal agriculture, and aquaculture. Despite an appreciation for the potential ecological and evolutionary impacts that transgenic organisms might exact if ever released into nature (either intentionally or unintentionally), the pace at which empirical risk-assessment data have been generated has lagged behind advances in the technology [1]. This is complicated by the inherent difficulty in making ecologically realistic riskassessments, with robust estimations of fitness parameters [2,3]. Riskassessment research must depend on the use of contained laboratory mesocosm experiments that mirror, to the best extent possible, the physical and biological complexities found in nature. It is hoped that from such experiments fitness parameters might be estimated, allowing the ecological and evolutionary effects of transgenesis to be extrapolated to wild settings [1,2,3,4]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
