Spatial correlation model for avian point count clusters

[northernbio]

I am analysing some older avian point count data collected in spatial clusters over a single season. Each cluster has 6 or more points, and points are spaced about 250 m apart. The same bird cannot be heard between points. For each cluster the points are visited sequentially in a single moring. Location of the first point is more or less random, but the location of the other 5 points are systematically located based on the first plot location, in a rectangular pattern. Each cluster occupies a single habitat type, so the site is defined as some radius around the 6 points. I've sketched the point pattern below.

. . .
. . .

It seems to me that the model that corrects for spatial correlation would be the appropriate one to use, but I suppose one could also argue that the simple occupancy model, where each point is treated as a replicate observation, is also valid. Also, for the spatial model, it is not clear to me whether the option to treat the intial thta0 to equal the rest of the thta estimates should be selected.

Any thoughts on this?

Rob

[darryl]

You could try the spatial correlation model, although it assumes a linear rather than 2D detection process so you'll have to keep that in mind with your interpretation.

If your initial point is randomly located, then you should NOT set the initial theta equal to thta0.

Cheers
Darryl