Heterogeneity in detection probabilities

[Aimie Baribeau]

Hi!

I have a quick question. I have heterogeneity in my detection history for my sites. I have 132 sites with 14 days of survey (using camera traps). For each day, I have a 0 if no deer was detected and a 1 if a deer was detected throughout the day (24 hours). But, when I look at my detection history, some sites have mostly a detection everyday and some as no detection at all. This, I heard, can cause bias in my occupancy estimation. For now, my models estimate a mean detection probability for all my sites but I would like to know if there is a way to account for heterogeneity in detection in my models in R using unmarkedFrameOccu.

Thanks,

Aimie

[Kery Marc]

Dear Aimie,

you could use the so-called Royle-Nichols or heterogeneity occupancy model. See also the interesting paper by Dorazio on the relationship between three types of species distribution models and also the two papers by Royle.

Dorazio, R.M., 2007. On the choice of statistical models for estimating occurrence and extinction from animal surveys. Ecology 88, 2773-2782.

Royle, J.A., 2006. Site occupancy model with heterogeneous detection probabilities. Biometrics 62, 97-102.

Royle, J.A., Nichols, J.D., 2003. Estimating abundance from repeated presence-absence data or point counts. Ecology 84, 777-790.

You can view the RN model as a method to estimate abundance from detection/nondetection (a. erroneously k.a. presence/absence) data or as a method to estimate occupancy in the presence of heterogeneity in detection probability that is caused by among-site variation in abundance.

Best regards -- Marc

---

From: unma...@googlegroups.com [unma...@googlegroups.com] on behalf of Aimie Baribeau [lapi...@gmail.com]
Sent: Monday, January 11, 2021 15:43
To: unmarked
Subject: [unmarked] Heterogeneity in detection probabilities

--
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/d994459b-93b0-4598-a8b4-ea33775bca89n%40googlegroups.com.

[Benedikt Schmidt]

Dear Aimie,

 

There is probably a reason why there is detection heterogeneity. The Royle-Nichols model is a good model if variation in abundance causes detection heterogeneity. You may have site covariates which may explain detection heterogeneity. Alternatively, you might fit a model with a site random effect. You can't do this in 'unmarked' (I think) but you can fit some models in 'ubms' (https://kenkellner.com/blog/ubms-vignette.html) or JAGS.

 

Best wishes,
Beni

 

 

 

Gesendet: Montag, 11. Januar 2021 um 16:05 Uhr
Von: "Kery Marc" <marc...@vogelwarte.ch>
An: "unma...@googlegroups.com" <unma...@googlegroups.com>
Betreff: RE: [unmarked] Heterogeneity in detection probabilities

To view this discussion on the web visit https://groups.google.com/d/msgid/unmarked/F801018864E940468323F405DF95D2DD7628ED4C%40MailNT.vogelwarte.ch.

[Aimie Baribeau]

Thanks for your answers.

I’ll look into that!

Aimie

To view this discussion on the web visit https://groups.google.com/d/msgid/unmarked/trinity-54b79bfd-4eff-4632-9d50-504b6dacad09-1610384015624%403c-app-gmx-bap05.