4th assumption of Cormack-Jolly-Seber

[susan.hurley@gmail.com]

4th assumption: all samples are instantaneous, relative to the interval between occasion (i) and (i+1) (Cooch and White 2004)

Hello. I'm studying Snowy Plover chick survival over the pre-fledging period in Northern California. Due to our survey methods and constraints of getting to sites on an interval, my data is set up in 4 day intervals with the recapture periods lasting 4 days and the survival interval lasting about 24 hours. Clearly I am horribly violating assumption 4, but does this render my results useless? Is there any problem resulting from this violation that I can look for in the statistics?

Thank you for any help!!!

[ganghis]

Susan,

No, it doesn't necessarily jeaporadize your inferences. Where biases creep in is when a) the capture hazard rate (as in survival models) varies between samples or when b) the mortality hazard rate varies between samples. You probably have some idea of the first... For instance, if plover captures occur primarily at the beginning of one 4 day sample period, and then occur primarily at the end of the following 4 day sample period, you are likely to have an underestimate of survival... assuming you are defining you survival interval between midpoints of sampling periods. The mortality hazard varying a bit is not usually as big of a deal. What I'd suggest is to make a histogram of the times of first capture for individuals in each sampling period. If the histograms look the same between capture periods, you're in good shape. The best published source on this is

Smith, D.R. and Anderson, D.R. (1987). Effects of lengthy ringing periods on estimators of annual survival. Acta Ornithologica 23, 69-75.


Cheers, Paul