C-Hat dilemma! Which value to use?

[daniel.o31]

Hi all,

First time CMR analyst here.

I'm currently in the process of running GOF tests with my dataset. My dataset has roughly 4000 individuals (Atlantic puffins) over a 30 year period.

I ran GOF tests in U-CARE. Results from test CT and CL seem to say the data are not sufficient to resolve the tests. Results from test SR and SM are both significant, which was expected due to the nature of how the data was collected as well as the ecology of the species. Regardless, the value of the estimated C-hat from U-Care is 4.46. I then ran the tests over using the 'suppress first encounter' function to try and eliminate the transient birds. This helped a bit and gave a c-hat value of 3.10 (SR and SM still significant).

I then, following chapter 5, ran bootstraps in MARK. From 1000 simulations of my most general model (time dependent survival and recapture) I got a estimated c-hat value of 1.10 (12585.5180/11405.702).

Clearly these c-hat values are very different. I'm uncertain on how to proceed from here. Which estimated c-hat value do I apply to my eventual model set? I cant seem to think of a reason why applying the 1.10 value from the bootstraps would be justified given that U-care shows a lack of fit. Will I simply have to include trap dependence in my models and proceed with the 1.10 value? Trying my best to understand some of the math behind this problem based on the reading in chapter 5 but am having some difficulty. Any insight is much appreciated.

Thank you,
Daniel