To play devil's advocate:

Someone once defined insanity as doing the same thing, over and over again, and expecting different results....

We model our salmon fisheries to kill every salmon, wild or hatchery, above established escapement goals. While we don't always do as well predicting how many fish will be available, we do a pretty fair job of harvesting down to the last paper fish. I think it's safe to say that in the modern era, escapements well above the floor are viewed as management failures. Evidently, our managers aren't failing very often in that regard.

It seems like more and more science is pointing the finger at habitat as the primary limiting factor in salmon abundance. It would be pretty arrogant (and misguided) for someone like me to say the science is fatally flawed, so I'll stop short of that, but I think it may be incomplete.

I recall from high school chemistry that any good experiment contains a control sample. If the hypothesis states that harvest is not a leading factor, shouldn't the observations include what happens when harvest (and I don't mean just in-river sport harvest) is removed from the equation? I realize why that experiment has not been conducted, and for the same reason, it's not likely that it will ever happen. I guess my point is that politically-tainted science, while it may yield results that help fisheries managers sleep at night, probably fails (often by design) to reveal the whole truth.