Those practices can have effects on natural populations if the breeding strategies are not controlled very carefully. In British Columbia, for example, the fish that are used to produce the next generation in hatcheries are as much as possible selected from fish that have been born in the wild, so they're at least one generation wild fish. There is emerging evidence from the United States that rearing in hatchery conditions and other conditions can change the genetics of the fish population in one generation. These are very much emerging questions that we're trying to understand, and I'd say the jury is still out on that.
One of the complications is that there are both genetic and environmental influences on the characteristics of the fish. To try to pull those two influences apart to understand whether there are long-term effects is very difficult to do. Improving on research facilities that allow for natural streams under controlled conditions is required to answer those questions.