Determining Range Residency for Island-hopping Bird

[L D]

Hello,

I've been reviewing past posts on this topic and guidance materials, but still find myself uncertain about range residency variogram interpretation. I'm working with a migratory bird species. I'm looking specifically at the non-breeding season. So I'm only using non-breeding points for the home range analysis. 

Three of these seem okay (1-3), but I just wanted confirmation on them since this is my first time using these methods. Number 4 is the one I'm having the most trouble with. 

Numbers 1 & 2 remained in the same location throughout the non-breeding season. 

1) This bird only transmitted for two months from its non-breeding site.

2) This bird transmitted from one non-breeding site for 2.5 years.

3) Number 3 made a two week trip away from its main non-breeding site, before returning to the main site for the remainder of its transmission. I excluded this two week trip from the HR analysis, is this appropriate?

4) Number 4 is the one I am having the most trouble with. The variogram does not have an obvious asymptote. In addition, the bird was wintering between two neighboring islands. It alternated its time between the two islands for 1.5 years. It would remain on an island for a week up to 4 months at a time before switching islands. This does not seem like range resident behavior, but the two islands were close and it had a fairly regular pattern. In addition, it made four brief forays to distant islands. I left out these forays for these variograms. Would it be appropriate to continue with the analysis?

I know this is a common question in the group. Thank you to anyone who has the time to assist with another variogram question!

Liana

[Christen Fleming]

Hi Liana,

The first 3 definitely look more stable. For the 4th variogram, it will be more informative to compare it to the fitted model. If they don't match well, then you could cluster the datasets into the two islands and fit them separately.

Regarding the forays, if they are brief and rare, then they shouldn't make much of a difference and can be included as rare events.

Best,

Chris