New model that identifies Brook Trout streams

[raftman]

Not sure if this was posted already (if so, feel free to delete), but I came across this article today.

In sum, PSU has developed a model that can supposedly predict with high success which streams still hold native brook trout.

I would love to have access to this model...

New model identifies eastern stream sections holding wild brook trout

 

[Eccles]

raftman

I've only skimmed this myself (need to look carefully at some of the model assumptions before forming an opinion) but if you PM with an email I can send you a .pdf of this paper.

Eccles

 

[salmonoid]

TL;DR summary:

The most important determinant of Brook Trout occurrence probability was predicted water temperature, as would be expected.

And:

The study region and the factors determining Brook Trout population status are complex, and we recognize that our model does not account for many landscape attributes (e.g., impervious surfaces), stressors (e.g., acid mine drainage), biotic interactions (e.g., nonnative Brown Trout), or infrastructure (e.g., dams and culverts) that might have negative effects.

So, IMO, it seems this is more a model that predicts water temperature, and therefore could support brook trout; but there may very well be browns there now.

 

[laszlo]

Ok, here's where my minds at. To me the title suggests that when I opened this thread I would be viewing a very attractive female in a bikini pointing to a map.

 

[k-bob]

ez access here

http://www.coopunits.org/Oregon/People/Jefferson_Deweber/Publications/

 

[Chaz]

I knew about it, Tom Green mentioned it in a meeting I was at recently. At least I think it was him, but it may have been the T. U. program lead for the unassisted water.

 

[SBecker]

Remember when Tom Green was funny? Man, when he humped that dead moose.....my bum is on a moose. My bum is on a moose. Look at me, my bum is on a moose.

 

[poopdeck]

Penn state models have proven to be so effective in the past. Does it look like some type of sports equipment.

 

[troutbert]

I read through it briefly and it's an interesting idea.

But in PA at least I think there is no substitute for field exploration. As Mike discussed in another thread, there are often "surprises", i.e. brookies showing up where you might not expect them.

There are so many variables involved, I doubt that any computer model can really deal with them all.

Someone has to go out there on the streams and check. Some of the colleges are sending students out there to survey streams for wild trout. This is perfect, because it means a lot of people out there checking a large mileage of streams, plus it is an excellent learning experience for the students to get out in the field, rather than just doing "book learning."

 

[k-bob]

"I read through it briefly and it's an interesting idea.

But in PA at least I think there is no substitute for field exploration. As Mike discussed in another thread, there are often "surprises", i.e. brookies showing up where you might not expect them.

There are so many variables involved, I doubt that any computer model can really deal with them all."

I agree. Even if you have info on drainage area, geology, gradient, developed basin vs forest, etc., things like amd can still be out there in some patchy unpredicted way.

Even basic map info goes out of date. In counties like schuylkill, usgs topos show some old impoundments that are no longer there.

 

[k-bob]

right, a statistical model based on environmental data wont always be able to tell you where there are and are not trout...

in fact, even a survey-based list such the nat repro one cant do this. first, there are clearly many streams with trout not on the list (see university unassessed stream studies). but there are probably also streams on the list with no trout now. a board member and I once hit wyoming cty's tiny white brook, which is listed, and didnt see a ST. maybe we didnt find them or things changed after a big flood

 

[salmonoid]

I don't think anyone is claiming that this is a substitute for boots on the ground. However, boots on the ground are expensive, even if they are poor, starving college students and regardless, there are thousands of stream miles still to cover. A model can help guide where those boots on the ground might get the most bang for the research dollar. And it can help target remediation efforts to streams that theoretically could support wild trout, because the water is cold enough. Or it could help target remediation efforts to streams that are currently marginal temperature wise, but that might show water temperature improvements with some tender loving care.

Models can help surveyors target streams with a higher probability of having fish. But surveys are just a snapshot in time and a snapshot of a given stream section.

 

[k-bob]

"Models can help surveyors target streams with a higher probability of having fish."

right, and while I have no formal model, I have gone to fewer dry or fishless streams by looking at region, drainage area, gradient, geology, and recent satellite images of development/tree cover. if those things look ok, I'll give streams off the lists a shot.

 

[Eccles]

They have a great bunch of data (all the electrofishing surveys) with which to train the model (the sort of empirical dataset many modelers would give their eye teeth for) and I thought it was a good stab at the subject. The fact that the accuracy is variable just shows how inherently difficult predictions of these complex associations are.

And I think it is worth pointing out that most biological models are theoretical. They are not constructed to say something "is" like this but more to direct attention to the way something might be. As such models and empirical work (experiments, field studies) often run in parallel, each being refined and informed by the other. In fact models are often constructed to direct the initial design of complex experiments or field surveys whose findings then feed back to inform model accuracy.

I can also easily imagine this model being used to direct initial environmental impact surveys in the case of projected perturbations like housing, mining or industrial development and the like. It doesn't always have to be about telling fishermen where the fish are. Or aren't.

 

[pcray1231]

In a different field, but generally we use modeling to focus experiments, thus narrowing the scope and reducing cost. You still gotta get physical data. Then you run the data back through the model to focus the next experiment. Repeat until you have what you need.

You do eventually get to a stage where you are "beyond modelling capability". I.e. you've got the basic parameters down, but now you're tweaking on a level where too many unknown variables are at play. It's trial and error at that point. We don't know what's gonna happen, so lets do it and find out!

Translating this to the brookie model, I have no doubt that if we were starting from scratch and trying to catalog brookie streams, a model like this would be fantastic and lead us to focus our efforts on the most likely places.

But we're not starting from scratch. We're starting from a point where the vast majority of potential streams have been surveyed. We're likely already at the "beyond model capability" level. So I don't know how useful this is for PA.

On the other hand, I'd imagine the concept is useful SOMEWHERE, for instance in more remote states and regions where not so much prior work has already been completed.

 

[BrookieChaser]

poopdeck wrote:
Penn state models have proven to be so effective in the past. Does it look like some type of sports equipment.

Hahaha! We've all seen the priority sports take.

 

[Chaz]

Here's my formula, find a region you know has brook trout streams, find streams on the wild trout list, look for nearby streams not on the list, go fishing. If you fish a 1/2 mile of stream and don't catch any, it doesn't mean they are not there, just find another nearby stream. Come back if you don't catch any to check again, maybe the day just wasn't a good day to fish.

 

[raftman]

Here's my formula, find a region you know has brook trout streams, find streams on the wild trout list, look for nearby streams not on the list, go fishing. If you fish a 1/2 mile of stream and don't catch any, it doesn't mean they are not there, just find another nearby stream. Come back if you don't catch any to check again, maybe the day just wasn't a good day to fish.

Good formula... Have worked that one to success in the past as well.

 

[BrooksAndHooks]

I agree with Chaz. The PAFBC/PASDA has a list of the streams supporting wild trout, which is insanely extensive. Combine that with the Class A and B listings and the wilderness streams, and you can fish for a long time. Almost all of these streams will have brookies, and if you are catching browns, go upstream and you will start catching brookies most of the time. Get out and fish the streams that feed these, or that lie near by. I've discovered a ton of small streams that I'm sure feed brookies up the in mountains walking miles of wild trout waters. These are not listed, but I have fished some of them and pulled some nice brookies. Many times a stream that looks small where it flows in to the main water will have pools etc holding fish further up. I like to find water that I know holds brookies, or is near known water, and check out the water flow on low and high days, and check the elevation change etc to help me predict which streams to check out. I've found some duds that I thought would be good, but more often than not I find some nice little streams and a healthy population of fish. The best part about finding these small "blue liners" is that few other people fish them, so the fish are likely untouched (and aggressive) and all the more fun to catch as well. Not to mention the populations are likely thriving if there is little pressure.

I checked out the PSU data a few days ago, and I agree, it seems like it's a predictor of stream conditions that support the trout. That doesn't guarantee that there are fish there, but it can't hurt. I don't see the need for it, however, when there are already so many resources for streams if you look for them. Then again, I'm all for additional resources because I get half the fun of fishing from hiking/scouting out new streams. I don't like to hit the same stream too often so I'm always out checking out new stuff.

 

[raftman]

Combine that with the Class A and B listings and the wilderness streams, and you can fish for a long time.

Do you happen to know where I can find a list of Class B streams by any chance?


Nevermind.... found it