Smalma,

You're right about the WDFW always being wrong. There are so many perspectives and opinions out there that no one decision will ever make more than a handful of folks totally happy with it.

I'll clarify my above post.

First, the WSC has not developed an opinion on what the numbers that justify a targeted cnr season should be. I'd guess that many factors in our mission statement would control that.

While I think that WDFW is using the best information it has available, I think we can agree that run prediction models are fairly accurate at best, at least for steelhead. While 80% of target escapement might not be tenable at current technology, it might be very tenable if there was better confidence in the run predictions.

Most of my personal knocks on management are not based on poor policies as much as policies are based on poor data. I don't mean to say that WDFW chooses to use poor data, but that it may not have access to anything better, whether it be due to cost or technology.

The spawner-recruit models work great in a vacuum, but it seems to me that they aren't so great in the real world. Many variables either aren't accounted for, or are accounted for by a general protection buffer that may or may not have any correlation to the variables they are buffering agains.

MSH/MSY is based on those s/r models, and besides being a shaky model so far as conservation goes, it's further weakened by being based on them.

The last possibility for correction in both of those models would be in-season assessments, but there are only a couple of ways I know of to do that. Those are sonar counting (very expensive, but seems to work OK so far in Alaska) and weirs/traps where fish are actually counted and passed on. The latter would likely be very accurate, but would introduce other problems.

My assessment:

We take faulty numbers, put them into faulty models, toss in a general protection buffer, then start bonking fish.

When the season's over, we go out and count fish again to see how we did. Sometimes we did above average (so far as escapement goes compared to what we thought), which is great. Sometimes we hit it pretty close, which is OK, too. Sometimes we over-estimated the runs, which is not Ok at all.

Our management scheme tells us that if we are over-escaped that somehow we failed to harvest enough. I think that such an idea is funky enough when applied to a commercial commodity such as most salmon, but is extremely funky applied to a statutorily mandated recreational fish.

If we hit it close to target, then we did a heck of a job this year. Can't have those darn fish spawning on top of each other and dying without our help.

If we come in under escaped, then either marine conditions were bad, the river conditions were such that the tribes had greater than anticipated success, or the sporties had really good luck due to nice weather. Or one of many other unmeasured variables took effect.

My conclusion:

If our models are not very accurate, and the safeguards in them only work in pretty much average conditions with average runs, which by definition almost never happen, let's develop new ways to gather data and better models to put that data in.

While we're getting that done, let's stop directly harvesting under the old models.

The idea, for me, is to balance angling opportunity and recreational payoffs to all the businesses that rely upon steelhead fishing for all or part of their income with conservation. The more fish we have, the more days we get to fish and the more fish we get to catch. The more we fish, the more money we spend.

I would immediately cease any wild fish harvest during the hatchery season, say November through February. I'd even consider extending the hatchery harvest season everywhere so that the hatchery fish are more effectively removed from the system. That would have to depend, of course, on the health of the wild fish that would be incidentally caught.

Then I'd have catch and release seasons. Where the balance point comes between opportunity and conservation hasn't really been defined yet to my satisfaction. Perhaps it would be when a river is at or exceeding 100% escapement. However, what model are we going to use to measure that?

Wherever it comes down, I'd want to have high confidence in my prediction models and some sort of in-season assessment that has at least a moderate level of accuracy.

Barring that (which is where we're at now), I'd feel better about fishing over an 80% escaped run if I knew that the run hadn't been subjected to a harvest season already. It's pretty clear that we won't be able to accurately assess the run until after we've bonked on it, if at all. At least if we're cnr'g the wild fish, we can be confident that somewhere around 95%-98% of those we release will still be out there spawning.

I'd prefer to have every river's escapement and capacity measured with great accuracy, I'd like to have models that are fool proof, and plug in indisputable numbers. Then I'd prefer that all the rivers come out to be around 100% escapement and that we all can fish until April each year on each river. I bet all the gas stations, mini marts, resaurants, hotels, and guides that depend on steelhead would like those things, too.

Without those things, I'd stop having kill seasons on wild fish, I'd put my management money into creating better science, and I'd promote cnr as a great way to fish a lot, spend a lot, and do it year after year.

Whew. "I got blisters on my fingers."

Fish on...

Todd.