Experiments and models to understand gene flow from transgenic fish in different environments.

Abstract

Transgenic fishes are nearing commercialization for aquaculture around the world. Farmed transgenic fish would likely escape from typical production facilities and interbreed with wild relatives. We tested methodologies for predicting the risk of gene flow from transgenic fish. We conducted the first study of gene flow in confined populations of transgenic animals. In two experiments several generations long, we released growth-enhanced transgenic (T) Japanese medaka (Oryzias latipes) into populations of wild-type (W) medaka in semi-natural environments. Transgene frequencies varied in the first experiment, but transgene frequencies all decreased in the second experiment. We measured six fitness traits in both genotypes, and found that T males were more fertile than W, but W males obtained more matings than T males. Next, we compared fitness traits of W and T medaka under four environments: (A) high food availability, predation absent; (B) high food availability, predation present; (C) low food availability, predation absent; and (D) low food availability, predation present. Overall, T females were more fecund than W, and fecundity was highest in Environment B. Offspring of TW and WT crosses had higher survival to sexual maturity than offspring of two W parents. Fish in Environment A reached sexual maturity sooner than fish in all other environments. W males had a mating advantage in Environments B and C. Finally, we observed gene flow in populations of T and W medaka in Environments A-D for 210 days. The final transgene frequency in Environment A was greater than in Environments C or D. We parameterized a demographic model with fitness trait values collected under the same environments, which predicted that transgene frequency in Environment A would be the highest, but also overestimated transgene frequency compared to observed results. Predicted transgene frequencies overlapped with observations in Environments B and C but not in the more extreme Environments A and D. Our results suggest that risk assessment of gene flow from T to W fish ought to consider the impact of limiting environmental factors on fitness components. Before using models to inform ecological risk assessments, predictions should be confirmed with data collected under relevant environmental conditions.