Detection Only

[brink2010]

I am only interested in detection rates. Is it acceptable to run a model without occupancy? Do I have to include psi(.)?

[darryl]

Do you have any units were you never detected the species so are potentially unoccupied? Detection probability at unoccupied units is 0, so if you ignore that then your estimated detection rate may be biased low.
Darryl

[brink2010]

I have only one transect at my field site and there is very little variation in habitat structure. We found salamanders throughout the entire site and therefore would not benefit from calculating occupancy. Can we just run models for detection?

[darryl]

If you're going ignore occupancy, which is the equivalent to modelling psi=1, then you can easily get the detection probabilty by dividing the number of detections by the number of surveys, or just use standard logistic regression if you have some covariates of interest.
Darryl

[jhines]

Darryl is correct about using the logistic regression, or the simple math to calculate site detection probabilities. This should be equivalent to running an occupancy model with psi fixed to 1. So, there isn't anything wrong with doing it as an occupancy analysis, if that's easier for you.