aerodramus wrote:Hi there:
I'm working through (in RMark) nest survival analyses of a tropical bird species that has a nesting season of 365 days (i.e., year-round). Does coding the nocc as 365 seem appropriate? Has anyone worked with a year-round nesting species and coded their nocc as 365.
Any assistance is appreciated.
Thanks,
Nathan
365 occasions? Not a chance. You'll need to pool/bin in some fashion. Even if annual survival is low (say, for your basic dicky-bird, something like 0.5), then if you split the year up into 364 intervals, then you interval estimate would be
, which is so close to the boundary 1.0 as to be inestimable.
The other issue is -- think hard about the biology. Is daily survival really something that means anything for something other than (say) a fruit-fly? The time scale over which you estimate parameters should reflect intervals over which biologically interesting things happen.