One of the major concerns about transgenic plants and animals has always been that they could escape and that transgenes could then spread into wild populations with mostly unforeseeable consequences. ((Contribution by Michael Kubisch)) For most farmed animal species, cattle, goats or sheep for example, this is not much of a problem because there are no true wild populations with which escapees could hybridize. However, farmed fish, such as salmon or catfish, do have wild relatives, reproduce relatively fast and farmed fish do occasionally escape into the wild, even in large numbers. This has led to a number of estimates and models of what impact such transgenic escapees might have on resident fish populations or on their prey species.
A recent article tells a cautionary tale about the value of such predictions by demonstrating that advantages which transgenic animals have “down on the farmâ€, such as a faster growth rate if they carry extra copies of the growth hormone (GH) gene, may in fact be less obvious  in the wild. The article describes a study in which GH-transgenic and wild-caught coho salmon were compared in either a conventional hatchery or a simulated natural environment. Under hatchery conditions, in which fish were fed a commercially available diet, the transgenic salmon grew to nearly three times the size of their wild cousins. However, in the natural environment, in which fish were exclusively fed natural food items, transgenics had only a 20% weight advantage. When the salmon were introduced to prey species, in this case trout fry, the impact of transgenic animals on their prey was reflective of their environment and size and the impact of transgenics on prey was much reduced.
While this says relatively little about the actual impact of transgenic escapees on resident fish populations, it does show that accurate predictions may be much harder to come by than previously assumed.
On the subject of transgenics, you all might find this article interesting.
Titled “Darwinian Agriculture”, it’s actually about looking through the genetic history of plant species to find variations with desirable qualities for agriculture that have — for whatever reason — been selected against in the wild. Specifically, the article looks at aggressiveness of root-building, which takes up a significant portion of resources that could be useful elsewhere (for example, seed generation) in exchange for improving an individual plant’s competitiveness. In a farm setting, however, you would prefer that plants pursued cooperative strategies.
There seems to be an interesting upshot in that, like the salmon you discuss here, transgenic crops created on that theory aren’t really at risk of overtaking wild populations because they’re being selected for a quality that has traditionally been a disadvantage to wild plants.