I am analysing movement data from 18 satellite-tagged individuals from an Old World vulture species. Upon inspecting the data I found there were 3 sampling regimes used:
(i) hourly, 2am to 7pm,
(ii) hourly, 3am to 8pm, and
(iii) every 3 hrs, continually, day and night.
I wonder how this will affect my calculations of monthly home range size (using 95% mcps and kernels). For the birds where fixes were not taken at night, will the importance of their nocturnal roost sites be underestimated in the home range estimates? Do I need to do some data cleaning before I start estimating home range size? And if so, how can I account for these 3 different sampling regimes in my calculations of monthly home range size?
Also, would cross validated be a better smoothing method than Href for these birds (which I expect to move mainly between roost/nest site and the same feeding site each day).
And is there any point in including mcps? The recent literature seems to imply that they are generally included only because other researchers include them, so 'for comparison with other studies', but they don't seem to be the best way of estimating home range size in a large bird which will have a few sites it spends most of its time at (roost/nest site and daily feeding site, for those birds that feed at vulture restaurants).
Lastly, would the package 'adehabitatHR' in R be the best way to get monthly home range sizes? Or would folks recommend T-LoCoH?
Thanks.
Hi,
you know my feelings about T-LoCoH (I've seen the procedure and outputs) and I think MCPs should be considered absolete and to a lesser degree KDEs. I think your best option would be "iii" but, you might want to use a "coarser" time resolution/stamp e.g. every 6 hour fixes. I also think that including night fixes are important and day fixes ONLY can give a skewed perception of space use in this bird. After all, roosting sites are important.
Regards,
I expect your (iii) to produce the smallest home range size, and at only 5 readings during daylight hours, then not very useful. While (i) and (ii) should be identical, my questions about Hooded (?) Vultures are: when do they go to roost, when do they fly off in the morning, and do they ever change position during the night, perhaps especially on moonlit nights? Readings at 7/8 pm and 6 am will answer that. Readings during the day at one-hourly intervals of a bird that can glide at >80 kph are not enough, but too bad. I'd have thought that MCPs are ok, because we all understand them, but certainly a vulture (as would a seabird) needs a smarter approach.
Thanks for your comments Prof Mundy, yes these data are for Hooded Vultures which Dr Keith Bildstein and his collaborators have collected over the last few years. That's a good idea to look at times the birds go to roost and their movements (or lack thereof) at night thank you, it will confirm info from camera traps at nests and behavioural observations at vulture restaurants. And thanks Rion, don't worry, I will try and use the T-LoCoH method if I have time after I'm done with kdes, it will be interesting to see the differences between the 2 methods.
I am hoping to compare home range size between birds from 3 populations, so I'll need to use the data from all the birds, which means including the 3 different sampling techniques (or 2 really, since Prof Mundy is right that (i) and (ii) shouldn't be too different). All I could think of to account for the differences in sampling is to calculate monthly home range for each bird and using 'sampling method' as a fixed factor in a linear mixed model, but I'm not sure if that's enough.
I think mcps provide limited information on the biology of any individual and with good quality data derived from trackers like this it is important to focus on extracting value from the derived points (regardless of how they may differ). Like how do birds use the space over which they move, seasonal and daily patterns in selecting and avoiding certain habitats, and how can we use this in conserving these birds. Good luck with this very interesting (and challenging) project.
- Badges
- Science topic
- Similar topics
- Medicine
- Disease
- Infectious Diseases