GOF testing with fixed parameters

[roodbergen]

Dear people,

I used Burnham's joint life and dead encounters models for the analysis of data on Black-tailed godwits (thank you Evan Cooch for your advice). However, I had to fix F=1, because both resightings and dead recoveries were only done in my study area. I would like to perform a GOF test on my best model, but somehow, this does not work. This probably has something to do with the fixed F, because the same model, but with F estimated instead of fixed at 1 does allow GOF-testing (both bootstrap and median c-hat, with pleasing results, chat <=1 and very small se).
The estimated F value is also 1, with a very small confidence interval, and the other parameters also remain unchanged, so I guess this does not really change my model (?).

Can I conclude that my model with fixed F=1 fits the data?
Or is there another way to get around this problem?

And can I conclude that Phi=S='true' survival, as F=1 (although most probably not biologically!) and that any migration from the area would be reflected in p?

Thank you very much for advice on this matter.

Maja Roodbergen