Part of the problem is they have no clue as to how many fish returned during the peak because they count fish in only a small section of the river, they had no idea the area the counts were done in represented only about 1/3 of the run. This is what lead them to believe the numbers were either increasing or at least staying level when they were actually declining rapidly.

The area where they count in the river is the only area with any chums left, the rest of the river is now barren, this lead them to believe the run was doing OK when 2/3 of the run was gone. They didn't know this (or at least pay any attention to it) until I was looking through the data and brought it up a few years ago. The response was, "well the in river count method was established 30+ years ago and was poorly thought out, but that's how its done" .

If the beach counting of smolts was used to develop a method of predicting runs for the Snohomish system, then the method is flawed, and not in a small way. If 2/3 of the run is gone and you can't identify it in the beach counts then the method is worthless. Again, they only do counts on some beaches, and when the Bio's were at my house and were asked if they could identify which river system the smolts counted came from, they said no.

If you pick a beach to count that has ideal conditions to hold smolts, then even when there are very few in the salt they will congregate in that location, there may be millions elsewhere, or none, you have no idea. This is the same scenario as in the river. How can you base estimated returns on such incomplete and faulty data.