temp emig and robust design

[storrence]

I have a couple of questions regarding what models to test using a robust design. I have 24 unconnected study sites (12 each of 2 years, 6 each year in 2 land use treatments), and need to estimate survival of metamorph spadefoots at each site. My goal is to compare survival between the 2 land use types. The number of encounter occasions are not similar among sites within years because the sites (playas) dry up at different rates.

First, I am 'playing around' with just 1 site for now until I learn MARK well enough to continue on with the rest of my data. I first tested for temporary emigration (by setting all gamma parameters to zero...they end up all being 1's when MARK runs the program). I found none. I have also run other models assuming different sort of emigration, but the change in the AICs between the NO emigration and these models is very large. Are there any other models assuming NO temporary emigration that I should run?

Second, can I compare the survival estimates from the best model for each study site using a ANOVA? What do I do if the best model for a few study sites is on WITH temp emigration (I'd be comparing survival estimates from sites where the best models are those with and without temp emig)?

Thanks in advance, storrence