Dataset with many non-detections and few detections

[Hannah Madden]

Good morning,

Question re: R package 'Distance'.

I have been surveying a species that appears to be in severe decline and only have 34 detections out of >300 transects. Given the limitations of this dataset, I want to ensure that Distance in R is suitable for such a small sample size. 

Any advice would be greatly appreciated,

Hannah Madden

[Eric Rexstad]

Hannah

This question arises frequently; how many detections are enough.  There is not a simple answer to this.  If the data are "nice", the detection process might be reasonably represented by a smaller number of detections.  However, if the data are not well behaved, e.g. half of the detections are at perpendicular distances <5m, then even a large number of detections may not lead to reasonable results.

Look at the histogram of the perpendicular detections and ask yourself whether they accurately depict the detection process; this is easiest to do if you were the collector of the data (or can speak with the data collectors).

--
You received this message because you are subscribed to the Google Groups "distance-sampling" group.
To unsubscribe from this group and stop receiving emails from it, send an email to distance-sampl...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/a9fcf8fd-0143-4136-b040-8b546c185407n%40googlegroups.com.

-- 
Eric Rexstad
Centre for Ecological and Environmental Modelling
University of St Andrews
St Andrews is a charity registered in Scotland SC013532

[Hannah Madden]

Hi Eric, thanks for your reply. About 20% of the detections are <5m. I am aware that with the desktop version of Distance with very small samples sizes you could use availability as a multiplier in the density estimator but am not sure if that's is still necessary with the R package (and if so, how). 

Hannah

[Eric Rexstad]

Hannah

Not quite certain what you mean about availability in association with small sample sizes.  However, it is possible to use multipliers in the R package when, for example, converting density of cues into density of individuals.  See an example of this here:

https://examples.distancesampling.org/Distance-cues/cuecounts-distill.html

Multipliers are described in the section "Introducing a new function `dht2`"

To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/31c20951-3786-46b2-b46c-038ef202f305n%40googlegroups.com.