Variogram Interpretation and OU Interpretation

91 views
Skip to first unread message

Alfons Patandung

unread,
Sep 25, 2022, 11:21:01 PM9/25/22
to ctmm R user group
Dear Chris 

We installed 10 VHF Telemetry by Lotek for maleo birds at 3 different nesting ground locations (2 inland/geothermal nesting ground and 1 beach nesting ground). All data is downloaded using Lotek Pinpoint Commander which is placed on the nesting ground. 4 individuals have returned.
1. Can you help us interpret the variograms of these 4 individuals (47177 : individuals from beach nesting ground). 
  2. I use ctmm.select for modeling and all the models in these 4 individuals show the OU model as the best model, does this have anything to do with nesting patterns?


Best 
Alfons 
47182.jpeg
47173.jpeg
47185.jpeg
47177.jpeg

Christen Fleming

unread,
Sep 26, 2022, 9:02:08 PM9/26/22
to ctmm R user group
Hi Alfons,

1-2 don't look resident. 2-3 either have a slight daily pattern or daily missingness in the data. 3 has a very impressive lunar pattern. 4 looks resident over ~weekly timescales, but is either continually shifting its range on a monthly scale or has a shift/dispersal in the data.

Getting an OU model rather than OUF indicates some coarseness in the data (relative to the movement behavior). So, for instance, you would have to compare diffusion rates instead of speeds if you wanted to compare movement activity.

Best,
Chris

Alfons Patandung

unread,
Sep 27, 2022, 12:01:27 AM9/27/22
to ctmm R user group
Thanks Chris 

What are the indications for numbers 1 and 2 don't look resident?

Best 
Alfons 

Christen Fleming

unread,
Sep 27, 2022, 11:35:03 AM9/27/22
to ctmm R user group
Hi Alfons,

In 1-2, the semi-variance keeps increasing with time lag, and doesn't clearly asymptote. It's possible with 1,2,4 that the behavior could look more resident on a weekly timescale if the data were segmented (around shifts/dispersals) or in a sliding window analysis.

Best,
Chris

Alfons Patandung

unread,
Sep 27, 2022, 11:42:02 AM9/27/22
to ctmm R user group
Thanks Chris 


Best 
AP

Reply all
Reply to author
Forward
0 new messages