Fall 2107 Cascade fish survey

[PercaMan]

Hey, there is a fishery survey report IDFG did on Cascade this last fall on there website?

https://idfg.idaho.gov/press/2017-survey-shows-perch-populations-remain-strong-lake-cascade

[steelerfaninboise]

I read this survey yesterday, link was posted on another site. I just want to say in no way do I want to give the impression that I am bashing fish and game by the statement I am about to make. No one can argue the fact that the perch fishery has made a tremendous come back because of their efforts, but when looking at the bar graph in this survey I am skeptical at best how they can form any accurate conclusions on age classes and percentages of perch based on such a small sample size. When you add the total number of fish up represented in the bar graph it comes out to 168 give or take a couple.

[dubob]

I used to evaluate radar performance for the USAF for about 15 years.  We did a lot of data collection and analysis and some of the very accurate performance predictions we made were based on a sample size of 3.  There is a confidence interval (the CI, also called margin of error) associated with the F&G data that isn't shown in that report.  The CI is probably less than 10.  There is also a confidence level (CL - also not shown) that tells them how sure they can be that the data is representative of the total population.  It is expressed as a percentage and represents how often the true percentage of the population lies within the confidence interval. Most researchers use the 95% confidence level which means you can be 95% certain that the number of perch in a size group will be accurate to within a plus or minus the CI of the number.  The wider the CI you are willing to accept, the more certain you can be that the whole population answers would be within that range.  I'd be very surprised to learn that they did not follow standard procedures for statistical analysis to determine the estimates given in that report.

[badger132]

I am just happy that they are managing this fishery. It is a once in a lifetime kind of experience to have a perch stuck in the hole because he is too big (Don't get that excited- I used to have a 4 inch auger, before I got a cordless drill and a Nils) I personally have caught more small perch this year than ever, and what I hope will happen is that if the fishery needs to change, like if they need to put size or bag limits on to maintain the unique size available in Cascade, that they will do it. I have a greater fear that the lake will return to what it was 30 years ago, chock full of 4-6 inch perch. I never saw one over 12 inches until this last reset of the lake, and the control of the squawfish population. There were a lot less fishermen then, and the gear used was primitive by comparison, so maybe we can fish them out if we are not careful. I leave the decision up to the professionals, just glad to hear they are watching things.

 

[ID_IceRookie]

SteelerFan:

First of all, congrats on your team securing a first-round bye.

You might consider contacting the author of the press release. I'm guessing the graph represents one site from their study, but I could be wrong. If they sampled with gill nets, they are lethal, so the may limit their use. They may have also had low success in their samples. The press release doesn't say anything about fish density or sample effort. I agree that it sounds more like an advertisement for the lake (and what IDFG has done) than a research publication.

[Idahogator]

PercaMan,
            I am fascinated in your looking to the future with your Title of this Topic and quantum leap of 90 years to 2107.

            It seems the main purpose of this/these studies is thinking/planning ahead and you have succeeded astoundingly well.

            Three cheers, Sir.       

[RB3]

I used to evaluate radar performance for the USAF for about 15 years.  We did a lot of data collection and analysis and some of the very accurate performance predictions we made were based on a sample size of 3.  There is a confidence interval (the CI, also called margin of error) associated with the F&G data that isn't shown in that report.  The CI is probably less than 10.  There is also a confidence level (CL - also not shown) that tells them how sure they can be that the data is representative of the total population.  It is expressed as a percentage and represents how often the true percentage of the population lies within the confidence interval. Most researchers use the 95% confidence level which means you can be 95% certain that the number of perch in a size group will be accurate to within a plus or minus the CI of the number.  The wider the CI you are willing to accept, the more certain you can be that the whole population answers would be within that range.  I'd be very surprised to learn that they did not follow standard procedures for statistical analysis to determine the estimates given in that report.

Nice post.  What was the standard deviation and p value??  I haven't seen many folks speak stat lingo like that since college!

[dubob]

RB3, I sent you a PM.

[Huge IceHole]

Had a chance to talk with the Biologist last week who wrote the piece and conducted the study....Paul Janseen.

Here are some points that I thought were interesting:
- These surveys are done on an annual basis...Started in 2012. 
- The methodology for the survey (number of nets used, length of survey, etc) was determined in 2012 by looking evaluating previous studies from bodies of waters with similar size and management objectives.  They may look at this in the future and decide to change.
- They run 30 total nets in different locations over a 7 day period.
- Nets are in place for about 22-24 hours.
- The holes vary in size...3/4" to 3".
- The total number of Perch caught in the study was 180.  This average is way down from previous years.  The average is around 400 with the highest ever being 739, 2014. 
-The week of the survey brought a cold front and fish activity/average number of fish of all species on the net sites was down.....Smallmouth, Trout, Suckers, Pikeminnow, etc.   
- I did not get the exact confidence intervals written down from the conversation, so a person wanting this precise information may want to call for themselves.  Historically they have a CI of around 10 for the different lengths.  F&G wants to monitor trends in lengths above and below 10".  This length represents a milestone from a management perceptive because these perch are no longer prey for the Pikeminnow.  For perch under 10 " the CI was actually better this year....CI of 4.  For Perch over 10" the CI was 6.  Even with less fish there was less distribution from the average between different net sites than in past years.   

In short, the conclusion I draw from this....The survey is not in depth enough to supply estimates of the total population in perch that I would feel comfortable enough to hang my hat on.  Limited by scope and size.  However, it does offer year over year data points on the trends of the population on Perch and Pikeminnow throughout the fishery as to effectively keep the predation in check.   
 

[IdahoSkies]

Thank you for asking important questions.  Any time your dealing with statistics the "how" of the question is usually more important than the underlying answer. 

[badger132]

I spoke to the fish biologist in charge of the lake a few years ago, and he told me that there was nothing special about the perch they planted, so I expect once the population increases to the point that it is in balance with the food again, the growth rates will be as they were before. That would mean that our current crop of large perch, which had all they could eat, will be replaced by larger and slower growing perch over time, until Cascade is like it was in the 80's- chock full of average sized perch. You can't get abnormally sized perch without something abnormal to cause them. I guess these big ones are the payback for those years in the middle where they were wiped out.