Predictive occupancy modeling with non-extrapolatable detection covariates

[Olivia Meyers]

Hi, 

I am trying to produce a predictive single season model of a snail species' occurrence across a designated national park in Canada. The goal is to extrapolate the model across the national park using environmental GIS layers as input. My challenge is that while my occupancy covariates are spatial layers of the park-- the detection covariates I am fitting (eg. date of sampling, precipitation) cannot be extrapolated. 

My question is, do I fit and select my top model with both detection and occupancy covariates and then remove the detection covariates from the top model equation when I extrapolate/predict? Or should I just perform a regular, non-heirarchical logistic regression?

Thank you so much for your insight! I really appreciate the help. I am new to modeling. 

All the best,

Olivia 

MSc Student 

Department of Vet Med at University of Calgary 

[Marc Kery]

Dear Olivia,

the beauty of these types of hierarchical models is that they keep completely separate the occupancy and the detection process. So, to predict occupancy, you can ignore detection probability, because true occupancy does not depend on it.

Best regards  --- Marc

---

From: unma...@googlegroups.com <unma...@googlegroups.com> on behalf of Olivia Meyers <omme...@gmail.com>
Sent: Thursday, March 28, 2024 19:17
To: unmarked <unma...@googlegroups.com>
Subject: [unmarked] Predictive occupancy modeling with non-extrapolatable detection covariates

--
*** Three hierarchical modeling email lists ***
(1) unmarked (this list): for questions specific to the R package unmarked
(2) SCR: for design and Bayesian or non-bayesian analysis of spatial capture-recapture
(3) HMecology: for everything else, especially material covered in the books by Royle & Dorazio (2008), Kéry & Schaub (2012), Kéry & Royle (2016, 2021) and Schaub & Kéry (2022)
---
You received this message because you are subscribed to the Google Groups "unmarked" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unmarked+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/unmarked/034a288d-b81a-4c10-9121-36f59a4a1e03n%40googlegroups.com.

[Olivia Meyers]

Hi Marc,

Thank you for the insight!

All the best,

Olivia