Playing with timeintervals

[Luc te Marvelde]

Dear all

In the first three weeks, my chicks are hatching. We started minotoring from the 5th week... so.. the first 4 weeks, we got no resightings. When i run a multistrata model (in which i seperate the first 4 weeks from the other weeks) i still get movements between strata (for me these are different woodlots)for the first 4 weeks, and that while i got no single resighting in that period. MARK should estimate NO movement at all in these 4 weeks.

Maybe this has something to do with my timeintervals. My data is based on weekly resightings. I also would like to estimate weekly survival/movements (multistratamodel). My weekly survivalrates are very high (around 0.98) and since MARK has trouble estimating values near 0 or 1, i was told that one can lower the estimates by decreasing timeintervals. So, in stead of using timeintervals of 1, i use timeintervals of 0.125 (so i estimate a 8 weekly survival rate for every week).

Does these small timeintervals explain why MARK estimates movements in the first 4 weeks? or if so, can you explain what is realy happening in the calculations? Does it uses information from the previous and following weeks to culculate a specific survivalrate? or is it just something like 'estimate^8' ?

Thanks in advance,

Luc te Marvelde
University of Groningen (The Netherlands)