You really should read some of the literature that's out there. There is a compelling amount of information indicating that significant decreases in survival occurs when hatchery fish spawn in the wild or if there are hatchery/wild interactions--even if the hatchery fish are just 1 generation removed from the wild. These progeny just don't survive to adulthood to the same degree as wild fish progeny. So simply being more selective or increasing the number of fish you spawn doesn't help the situation.

With that said, I don't blame hatcheries for decreases in wild production. Like you say, the function of hatcheries is to provide fishing opportunity, both in the total numbers of fish and year to year consistency of runs. The latter is very important because wild populations can fluctuate by an order of magnitude because of natural environmental conditions; its the nature of the anadromous beast. And according to some of the data that Bill McMillian presented last night, you can have just as many good years with lots of fish as poor years with very few, and it's alll completely natural. It is definitively not the case where the poor year is the exception.

In this setting, I submit that you CAN'T manage for harvest at current fishing levels without hatcheries. The natural fluctuations of wild fish is such that during the good years there may be an adequate surplus for harvest, but during the poor years, which can be half the time, sooner or later you're going to fish a run into extinction or remnance.

So without hatcheries, you're left with either 100% C&R, or an ultimately failing attempt to manage wild populations for harvest. With a no hatchery scenario, managing wild steelhead for C&R would be the only viable way to go.

Throw in economics now. In Canada, when BC went to 100% C&R in the mid-80s, there was a substantial decline in license sales over the next 3-4 years. Only after BC began hatchery brood stock programs did angling begin to increase to historical levels. Without hatcheries, and 100% C&R, I think the same thing would happen here. There would probably be declines in local businesses, except perhaps on trophy rivers, and lower total state revenues from license sales and taxes.

Now throw in politics. Do you think there wouldn't be pressure to manage wild stocks for harvest if run sizes suddenly increased by an order of magnitude? You betcha. How quickly people would forget a previous poor year and think, whatever went wrong has been corrected. I don't think people will ever realize that things aren't wrong, during some years, but naturally, cyclically low. And the tribes will always be harvesting, and harvesting at MSY, which doesn't work.

Where am I going here? I believe hatcheries are necessary to keep fishing opportunities at there present levels. However, hatchery policy as it sits now will cause further declines in wild production. I think the solution lies in the retooling of existing hatchery policy such that a goal of minimal interference to wild stocks be achieved. I think Bill presented substantial evidence that temporal isolation of spawning periods (i.e., breeding for early returners) doesn't effectively isolate hatchery and wild fish. But unlike Bill, I believe that there is a threshold of hatchery interaction that will minimally affect wild stocks, and that this threshold can be reached with changes in hatchery practices.

Practices such as discharging smolts in hatchery feeder streams where excess fish can be collected. Thorough hatchery imprinting. Limiting the trucking of early returning adults . No smolt discharges in mainstems. Marking fish such that straying rates can be determined and monitored. Relocating/eliminating facilities proximal to rebuilding runs. Elimination of hatchery plants in small streams, etc.

Some of the above ideas are probably prudent, others maybe not, others would need to be studied. But the most important consideration is to produce a strong policy driven objective to keep hatchery and wild fish separate. That presently is not the case. Those that think that the temporal separation of hatchery runs is adequately protecting wild runs are either ignoring a body of data, misinterpreting those data, or are satisfied with the present level of wild production.