Re: [unmarked] time-of-detection

154 views
Skip to first unread message

Richard Chandler

unread,
Aug 1, 2012, 9:38:43 AM8/1/12
to unma...@googlegroups.com
Hi Tracy,

If I'm not mistaken, time-of-detection models are normally used for estimating abundance, not occupancy. You can fit the abundance models in unmarked using the multinomPois() or gmultmix() functions. Each has a help page, and there is this vignette: http://cran.r-project.org/web/packages/unmarked/vignettes/cap-recap.pdf

Richard

_____________________________________
Richard Chandler, post-doc
USGS Patuxent Wildlife Research Center
301-497-5696



From: Tracy <trap...@gmail.com>
To: unma...@googlegroups.com
Date: 08/01/2012 09:11 AM
Subject: [unmarked] time-of-detection
Sent by: unma...@googlegroups.com





Hello all,  
I have single-visit point count data with spatial replicates, and so the replicates lack independence.  I am trying to use a time-of-detection model for occupancy to get around this problem, but I am not sure how to set this up.  I was using 0/1 notation up until the first detection, and then missing values (NA) after that, but this does not seem to work.  Any ideas on the best way to set up this model?  Thanks for any help!
 
Tracy

Tracy

unread,
Aug 2, 2012, 3:34:24 PM8/2/12
to unma...@googlegroups.com
Hi Richard,
 
Thanks for the reply.  In looking into this more, I do not have the information to use a time-of-detection method, but would like to use a modification of the removal method (i.e. using the time to the first detection of the species, and no other information about the site after the first detection, as described in MacKenzie and Royle 2005).  I have count data, but I don't think using the counts for this type of method is correct?  And so if I am only using the time to the first detection, can this sort of data be used to model abundance using the multinomPois() function?  I assume it would need to be modified a bit from the typical "removal" sampling method.  Alternatively, can I use the occu function for occupancy, entering the data as a matrix of 0s and1s, with missing values after a species is detected at a site?  Thanks for your help!
 
Tracy
Reply all
Reply to author
Forward
0 new messages