Pradel models... (and misidentification)

[cebert]

Hello,
I wanted to use Pradel models fo analysis of a non-invasive genetic capture-recapture data set with the aim to test if there is violation of closure in our data... In order to do this, it would be necessary to be able to set survival to 1 and recruitment to 0. Now I would like to ask if this is possible using MARK and if so, how it is done (I did not manage to find out until now...)
I have another question concerning closed capture models with misidentification: Is it possible for me to define alpha (the possibility of being identified correctly) myself? Because we defined a misidentification rate (1-alpha) via tests and would like to use this experimentally defined rate in the analysis. If we let alpha be an estimated parameter, we always get really strange results with an exorbitant misidentification rate...
Thanks in advance for help!!
cebert

[Eric Janney]

Cebert,

It is easy to fix parameters in MARK. After you "click" the model run icon, the "Setup Numerical Estimation Run" page will open. There is a "fix parameter" button you can click that will allow you to fix parameters. I'm not sure if this is the best way to test for demographic closure? You might want to read Stanley & Burnham (1999). "A closure test for time-specific capture-recapture data". They present a goodness-of-fit test for testing the closed Mt model against the open Jolly-Seber model. I'm not sure about your second question, you should read Lukacs and Burnham's work on the topic.

Good luck,

Eric