Smalma-
Thanks for the response, maybe not though, because I lost a bit of sleep thinking about the ESA, salmon, and evolutionary significant units. In ESA terms an ESU is a distinct population and it seems that for a population to be distinct it must have a high degree of reproductive isolation. Good examples are the desert pupfish living a few hot springs and rainbow trout in the great basin. Anadromous salmon on the other hand have low reproductive isolation. That is one of the features that insures that they utilize all available suitable habitats and also make it so difficult to determine what is an ESU.
It was certainly an error to say that there are no distinct genetic or phenotype differences among naturally spawning populations of chinook in Puget sound. That a Nooksack chinook stock seem to be more related to the Fraser group does make sense but are they a distinct population in Puget Sound. Without hatcheries geographic isolation might lead to a distinct population difference between chinook stocks in the far south sound and the north but the question is where would the boundary between the populations be.
Your example of the phenotype differences of the Snoqualmie fall chinook is interesting. If late spawning does not occur in the other rivers in Puget Sound then any strays entering the Snoqualmie would not likely be fall spawners. In that sense the late spawners would be reproductively isolated. 20 to 30% of the returning fish exhibiting this phenotype could be caused by a random gene expression and might not be enough to consider them a distinct population. If a selection process favored the late spawners and they became the normal spawning population then they might be considered a distinct ESU.
The question of what is a distinct population of salmon in the ESA context is not easy to determine. Genetics can't necessarily identify distinct populations, it only takes a few strays to insure gene flow between populations. It is likely that genetics can't differentiate the late spawning Snoqualmie fall chinook phenotype from the south Sound/Green River falls. Any competent scale reader could easily distinguish between returning fish of the two groups.