Pop abundance with specified recruitment

questions concerning analysis/theory using program MARK

Re: Pop abundance with specified recruitment

Postby Fish_Boy » Wed Nov 28, 2018 2:44 pm

Any suggestions on which of the approaches here is most 'correct'? The apparent survival estimates for the unfixed N produces comparable estimates to wild fish, but the confidence intervals would indicate there is an issue.

A CJS provides the same estimates and wide confidence intervals. While fixing N in POPAN to the numbers stocked, produces lower estimates, but good confidence intervals.
Fish_Boy
 
Posts: 57
Joined: Fri Oct 28, 2005 2:12 pm
Location: Winnipeg

Re: Pop abundance with specified recruitment

Postby jlaake » Thu Nov 29, 2018 1:25 pm

A few things here.

1) This post was started in reference to estimating abundance. If you are only interested in survival then use CJS.

2) The reason for your large CI on survival is probably due to your small p and possibly small n which I don't know. That is not going to go away. Not sure why fixing N would have had any effect and if it did I would be very suspicious of that model given that it didn't match the CJS results.

3) It is unclear to me that you followed my original suggestion which was to add a 1 at the beginning of the capture history for released fish and a 0 for wild fish. Then pent=1 and p=1 for released fish and pent=0 and p=0 for hatchery fish for that first occasion/interval. Is that what you did? If it was then I would have expected that N for released fish would have matched the number released because you are saying that you "caught" all released fish that first occasion and all of them were released that first occasion. For wild fish their should be an estimate of N.
jlaake
 
Posts: 1031
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: Pop abundance with specified recruitment

Postby Fish_Boy » Thu Nov 29, 2018 1:55 pm

jlaake wrote:A few things here.

3) It is unclear to me that you followed my original suggestion which was to add a 1 at the beginning of the capture history for released fish and a 0 for wild fish. Then pent=1 and p=1 for released fish and pent=0 and p=0 for hatchery fish for that first occasion/interval. Is that what you did? If it was then I would have expected that N for released fish would have matched the number released because you are saying that you "caught" all released fish that first occasion and all of them were released that first occasion. For wild fish their should be an estimate of N.


I limited the encounter history to stocked fish only, so I added a 1 at the beginning for fish released, pent = 1 for first period and zero thereafter, p = 1 for first period and solved for thereafter.

As follows each individual is known, the 2014 cohort is stocked out, then monitored annually since. The group attributes are two locations (n1 = 424, and n2 = 415)as follows:

/* 900067000055036 */ 10100 1 0;
/* 900067000055037 */ 10000 1 0;
/* 900067000055038 */ 10000 1 0;
/* 900067000055039 */ 10000 1 0;
/* 900067000055040 */ 10000 1 0;
/* 900067000055041 */ 10000 1 0;
/* 900067000055042 */ 10000 0 1;
/* 900067000055043 */ 10000 0 1;
/* 900067000055044 */ 10001 1 0;
/* 900067000055045 */ 10000 1 0;
/* 900067000055046 */ 10000 1 0;
Fish_Boy
 
Posts: 57
Joined: Fri Oct 28, 2005 2:12 pm
Location: Winnipeg

Re: Pop abundance with specified recruitment

Postby jlaake » Thu Nov 29, 2018 2:18 pm

If you are limiting to stocked fish only why wouldn't you stay with CJS which is for releases. If you are limiting to releases there is no reason to be using POPAN. I think you are missing the point of this original post which was for a mix of released and wild fish and the focus was abundance estimation.To be honest without doing some simulation, I'm not certain my idea would work. Notice that I prompted for others to chip in but no one did so. Regardless, I think for your case, you should be using CJS.
jlaake
 
Posts: 1031
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: Pop abundance with specified recruitment

Postby Fish_Boy » Thu Nov 29, 2018 2:44 pm

I have returned to the CJS as the stocking intervals are every two years, while the monitoring is every year. I had constructed an encounter with both stocked and wild Lake Sturgeon. I had concerns over the intervals differences, but your suggestion provides a potential solution to that based on pent = 1 at stocking.

At any rate, your suggestion does converge to comparable survival and recapture values comparing CJS and POPAN, so it works as it should.
Fish_Boy
 
Posts: 57
Joined: Fri Oct 28, 2005 2:12 pm
Location: Winnipeg

Previous

Return to analysis help

Who is online

Users browsing this forum: No registered users and 1 guest