Federico Nave

Hi Carlos, if you want to estimate populations, yes.

I'm adding two papers that can be helpful.

Best regards,

Federico

Short answer is yes, but your inference will be limited to watering holes, i.e. you won't be able to make any statements about occupancy where there are no watering holes, and it is worth keeping in mind that what you are estimating is not abundance. There is also a problem of whether you selected your watering holes randomly, and whether you have repeat visits to the same site when the population can reasonably be considered 'closed', but answering those questions requires much more information from you.

Actually Occupancy-detection models were design with discrete patches of habitat in mind, but as Paul Glover-Kapfer mention your inferences will be about the usage (=occupation) of the watering holes. All other assumptions for these models (independence among sites, population closure between repeated visits, etc...) must be met.

If you are familiarized with R I would advice you to look for package Unmarked:

https://cran.r-project.org/web/packages/unmarked/index.html

If you prefer a "point'n'click" stand alone software, look for presence:

https://www.mbr-pwrc.usgs.gov/software/presence.html

There's also some Occupation models implemented in the Mark software but I'm not familiarized with them.

Federico Nave, you might be interested in this response to the Welsh et al. paper you shared, if you haven't seen it already: Article Ignoring Imperfect Detection in Biological Surveys Is Danger...

Carlos A. Gaitán and I have been communicating, and while theoretically he could use occupancy models, the number of waterholes sampled is very small so estimates of occupancy (from any method) are going to have large standard errors.

Great advice that people are giving here though!

Cheers

Darryl