DSM over multiple regions

[Thomas F Johnson]

I have been monitoring a bird population with eight road transects (four in siteA, four in siteB), each about 15km long, and repeated about ten times on average. The aim being to  develop a distance function and dsm model, and then plotting this over a prediction grid. 

Within each site (A & B), the transects are in close proximity (c.5km away) from each other, but the sites (A & B) are hundreds of kilometers away from each other. What is the best way to handle this within a ds() and dsm() model? 

If both sites are included, the spatial smoothing s(easting, southing) is working over a large blank area between sites, is that okay? Would it be better to analyse the sites separately?

Tom

[Eric Rexstad]

Tom

It does not seem particularly prudent to try to make inference about bird density in a region midway between your sites, that is at least a hundred kilometers from any of your sampling effort.  Seems that would stretch your inference beyond the tolerance limits. Analysing the sites separately would be the more statistically sound approach to take.  That being said, you'll also want to ask whether relationships you find from along roadsides will remain relevant throughout the site to which you wish to make inference.

--
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/e0c44c01-a522-4979-a301-36d2b4f15621n%40googlegroups.com.

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

[Thomas F Johnson]

Thank you Eric. Much appreciated,

Tom