Question Author
Thanks for replying, divebuddy. I nearly bumped it after no replies in 5 hours but remembered that this knocks it off the 'unanswered' tab and it really would have got ignored.
i) Yes, hunted to extinction is one possible fate. We came perilously close to losing Cod, not so long ago.
ii) Dying after the eggs are fertilised is part of the normal life cycle. It is all well understood and makes sense in that, having done their one task in life, dying immediately stops them from hoovering up any of the oxygen or food in the stream and also ensures the hatching fry do not get eaten.
What's gone wrong here is that they're dying way before they even reach the spawning grounds.
iii) //OK, but what do you contend at other times.//
Its corollary, which is that, if an environmental niche persists for eons then there is no pressure to evolve and the species occupying that niche will endure with few changes. Crocodile, alligator, coelocanth, sharks, dragonflies, stromatolites and so on. Few visible differences from their fossil equivalents. If a design is a good fit to its environment, then variants (deviants, mutants, if you insist) are more likely to be a substandard fit than to be an improvement.
In the salmon's case, they need to have a heat tolerant variant, to keep that river's population going just for this crisis year and could revert to the norm in generations to come.
I will assume that the species uses other rivers too and will not be utterly wiped out, it is just a substantial chunk of population at stake, being that it is a big river with many tributaries. It could be recolonised after a catastrophe if some fish have adventurous tendencies and do not conform to the "return to river they spawned in" cliché. Finding a depopulated major river would have a major payoff for the lucky few who get to breed there first.
It's early a.m. and I need sleep. Will return to this tomorrow.