Carcassman,

Good question about how this can occur in two generations but not the first. As I am not a geneticist that has knowledge in this area, I won’t even attempt to try to explain how it would occur. I suspect that you know some good people in this area though, so perhaps they can. The purpose of my response was to try to make clear what these studies actually showed. The fact that some of the results were not statistically significant is important. You know that. Like I said in my first response, the authors got to choose how they approached this question completely. They chose the tests they would use as well as the level of confidence needed to draw any real conclusions. When you then get non-significant results you should report that clearly, not muddle up the results. Back to your first question though.

Since the Araki and Blouin work, there have been a number of papers looking at reduced fitness in hatchery fish spawning in the wild. Some have shown loss of fitness, others have not. What I find more interesting is that they have actually been able to identify specific traits that are having an effect on productivity. Some of these might shed some light on your question.

After the Hood River papers, Araki, Berejikian, Ford, and Blouin used a simple quantitative genetics model to evaluate whether domestication selection can explain rapid fitness declines (BTW, I would consider 2 generations in the hatchery to be very rapid domestication). What they concluded was that if selection acts on a single trait, rapid effects can be explained only when selection for that trait is very strong both in the hatchery and in the natural environment. If selection works on multiple traits throughout the life cycle, rapid fitness declines are plausible. With that in mind, there are also several studies (numerous authors and locations) doing some interesting multivariate analysis, trying to identify the traits that may be responsible for changes in fitness. To date, here are a few of the traits that have been identified as being associated with reduced reproductive success: Origin (natural v. hatchery), sex (M or F), location of spawning, fish length, return date, number of same sex competitors, behavior of offspring and spawning adults. I’m sure this is not a complete list of the factors identified, but I didn’t take the time to look up additional results. I certainly am not surprised with most of these on the list, and I doubt that you are either. The point being that if all of these factors have a bearing on RRS, and selection acts on all of them, I would guess that there is any number of potential negative correlated traits that could be expressed very rapidly between the 1st and 2nd generation. Again, excuse my ignorance on genetics, but that is the best I can do. I suspect it would be easy for someone with knowledge in this field to explain the exact mechanism(s). I cannot.

Just for fun, though, let’s just look at one trait that I will throw out for consideration. BTW this particular one has bugged me since I looked at the hatchery programs on the Hood River and considered the results of the Araki papers. This study combined the results of spawners from both the summer steelhead and winter steelhead programs to improve the power of the tests (yes, that tricky little issue about statistics again). My recollection of the details of these programs is that the pond that released the summer SH was located well up the Hood River in the middle of really good SH spawning habitat, but the location of the pond that released winter steelhead was located relatively low in the river not in what you would call good SH spawning habitat.

So here is an hypothesis. In the first generation, using natural fish collected at the old dam, you collected fish that would have spawned throughout the watershed. Even though you released them from at least one pond located in the lower part of the river, they generally did a good job of dispersing themselves upon return (as the fish that you collected were used to doing just that). In the second generation, you now use fish that were only accustomed to the river based on their hatchery experience. The summer SH would probably distribute themselves around the rearing pond in good SH habitat, but the winter SH, or at least some of the winter SH would distribute themselves in the spawning area nearer their rearing pond in less productive spawning habitat as they now have no cues to home to any other part of the river. Could that 2nd generation then be less productive in the wild based on only this one trait (spawning location) than the 1st generation? I suspect so.

BTW, the scenario described above is very much like what happens in the integrated spring Chinook program on the Wenatchee River. A disproportionate number of adults return to spawn near the rearing pond, in marginal habitat. Combine this with (and I’m just making this up) differences in spawning behavior and competition with more adults returning to this area to spawn (not making this up, as it is also a factor on the Wenatchee), and you now start having some additive (multiplicative?) changes in reproductive success from using fish collected from the natural environment versus fish having gone through just one generation in the hatchery.

What I find more interesting about these current studies is that they lead you to think about the specific biological factors at play and how you might be able to address them. That interests me much more than the debates around the “absolutes” that you often hear on this forum arguing the intrinsic good or bad of wild and hatchery fish.

Finally, I’m not going to respond to your point about even Chambers Creek (or Skamania) fish occasionally being able to develop self-sustaining populations. I’m not sure what you are getting at. It is clear though, even using the point estimate for RRS (85% of wild fish), that at least the first generation hatchery fish in the Hood River studies have productivity an order of magnitude greater than the non-local hatchery populations.