point counts as covariate?

[Virginia]

Hi all,

If abundance induces changes in detection probability, can I add point counts (relative abundance) as a covariate?
I counted frogs in each of 4 visits (at 50 sites), and I was able to identify "possible" re-sights, so not all counts are from different individuals.
I haven't been able to run Royle Biometrics models, maybe because I have missing observations? (I don't get a warning message or result)

Please, if anyone has any suggestions I'll be most grateful.

Virginia

[darryl]

Using the counts as a covariate is double-dipping in a sense as the same information is being used to determine detection so you're always going to have counts>0 when detection=1. Unless your counts are independent of the detection data. Other option is the use the recently developed multi-state models and categorize your counts in to none/low/high (for example)

Darryl