Multi-season multi-state model

[Loupeloupe]

Hello everyone,

I am trying to build a multi-season multi-state model with three states (unoccupied, low abundance, high abundance), but I met some problems with a simple model without covariate.

First, I am using the first parameterization, is Psi i(0-2) the probability of passing from the state 0 at time i to the state 2 at time (i+1) ? (which means having a colonization, isn’t it possible to have an extinction matrix and a colonization matrix as in the multi-season model?) Why some probabilities are missing (for instance, I don’t have Psi0(1-1), Psi i(1-0), Psi i(2-0)) ?
Is it possible to test if my case really is really Markovian ? How to test this hypothesis ?

Then, I would like to have a constant probability of detection and being equal to 1. How is it possible to enter a initial probability of detection = 1 ?

Finally, my first results gave me all my probabilities of extinction and colonization being equal to 0.5, which is completely impossible…How is it possible, did I miss something before running the model ?

Thank you very much for your answers.
Regards,
Lara MILLON