Multi state Barker model with state uncertainty

questions concerning analysis/theory using program MARK

Multi state Barker model with state uncertainty

Postby LisenkadeVries » Thu May 31, 2018 7:49 am

The subject of a multi-state Barker joint live dead model was discussed on this forum a while ago (“Resightings in multi-state models?” in 2011 and “Resightings in multi-state models?”in 2017), but since then not a lot has been said about this specific model.

At the moment, I am working with a dataset with ringing data, resightings and recoveries of barnacle geese. These geese are ringed or recaptured in both summer and winter and can be resighted and recovered all year round. After reading the manual and the literature, a barker joint live dead model seems to be the best fit here.

The geese are caught in the summer in three different breeding areas. In winter, there is only one sampling area. Because we are interested in the transition between the different breeding areas, I would like to extend the barker model to a multi-state barker model. Because we are not able to directly assign geese ringed in the winter to a specific breeding area, state uncertainty must be taken into account.

Since this model was last discussed on this forum, it has been implemented into Mark.

Because the explanation of this model is not yet included in the manual, I have read the help file supplied by MARK and any articles on multi-state models with state uncertainty I could found. The explanation in the help file is fairly concise and there are a number of parameters that I do not understand completely.
It concerns the parameters rho, rho’, b and pi.

If I understand correctly, delta would say something about the probability of giving the animal the correct state?

It is a very complicated model, so I am a bit anxious about the analyses. Any help would be greatly appreciated.

Thanks!
LisenkadeVries
 
Posts: 4
Joined: Tue May 29, 2018 4:39 am

Re: Multi state barker model with state uncertainty

Postby Guillaume Souchay » Wed Jul 25, 2018 5:07 am

Hi,

I would be also interested in knowing a little bit more on all these parameters. By analogy to the Barker joint live and dead model, I would guess that rho and rho' are equivalent to F and F', but I'm not sure of this...
Also, when I try to implement this model with CH like "A0A0B0B0A1", I have an error message regarding the presence of '1' in the CH.

So, if anybody has some news about the implementation of the MS Barker model, I would be glad too.

Thanks in advance,

Guillaume
Guillaume Souchay
 
Posts: 64
Joined: Wed Feb 27, 2008 5:20 am
Location: Nantes, FRANCE

Re: Multi state barker model with state uncertainty

Postby gwhite » Wed Jul 25, 2018 10:16 am

Check the help file that comes with MARK, specifically the Multi-state Barker Model page that details data types 176 and 179. For both data types, dead encounters where state is known with certainty are the lower-case letter of the state codes (required to be upper case). For data type 176, uncertain states are coded as U for live encounters and u for dead encounters. For data type 179, dead encounters with uncertain state are specified when you set up the analysis. The only time a "1" code would be used is if you have specified "1" as the code for uncertain events.

Gary
gwhite
 
Posts: 329
Joined: Fri May 16, 2003 9:05 am

Re: Multi state barker model with state uncertainty

Postby Guillaume Souchay » Wed Jul 25, 2018 11:37 am

Hi Gary,

Thanks a lot for the tip, the help does contain all the needed information ... Sorry to have missed that.
The model is know able to run. Despite the use of lower case for dead recoveries, MARK shows the following warning
* * WARNING * * Data type declared MSBarker, but only live found.


The model is however able to properly estimate r, the probability of detecting a dead marked animal.

I will take another look at the data and how properly to constraint the PIM to ensure I didn't miss something else.

Thanks again.

Guillaume
Guillaume Souchay
 
Posts: 64
Joined: Wed Feb 27, 2008 5:20 am
Location: Nantes, FRANCE


Return to analysis help

Who is online

Users browsing this forum: No registered users and 9 guests