Interpreting Initial Occupancy Probability Covariates in ColExt Models

[Rory Macklin]

Hello all,

I am working with ColExt models and am wondering about the interpretive value of the coefficients in the initial occupancy probability submodel.

Say urban land cover proportion has a negative coefficient value in an initial occupancy submodel. The direct interpretation of this, in my mind, would be "sites with greater urban land cover are less likely to be occupied in primary period 1". Of course, the occupancy in primary period 1 is a product of colonization and extinction processes that occurred before primary period 1. Could one then suggest using the initial occupancy submodel coefficients as evidence, that "urban land cover has historically lowered colonization and/or increased extinction probability, leading to the observed lower initial occupancy probability observed in primary period 1"?

I imagine the counter argument would be that the lowered initial occupancy probability at sites with higher urban cover could be due to any range of things, and we can't test that because we don't have data for the primary periods prior to primary period 1, so we can't claim that it has anything to do with colonization/extinction probabilities. Perhaps there was a cull near urban areas that occurred just before we started surveying?

Generally speaking, I'm wondering whether the coefficient estimates in initial occupancy submodels provide ecological insight and what that might be, or whether they are just there to best inform the baseline from which estimates of occupancy in later primary periods (through estimated colonization and extinction probabilities) can be made.

Happy to discuss, thanks for your time!

Rory