Skip to content

Guidance on adehabitat usage

1 message · Jefferson Ferreira Ferreira

Original
Jefferson Ferreira Ferreira
# 
Dear forum members

I had successful use adehabitatHR R package to estimate home ranges for
jaguar based on collar data. Now I'm trying to analyse relocations with
adehabitatLT package following the vignette (
http://cran.r-project.org/web/packages/adehabitatLT/vignettes/adehabitatLT.pdf).
But some doubts emerged and if some of you could clarify me I'll be pleased
to hear your suggestions.

My collar data have +/- 19% of missing data due to problems related to
receiving GPS signal from collar. So, I've filtered the data:

#re-subseting to exclude rows with no coordinates
coto_nonas2 <- subset(coto, coto$GPS_Latitude != 'NA')

(1) In your opinion, is this procedure correct?

Well, after this i've execute

#Placing the missing values in the trajectory
# The relocations have been collected daily, but there are many days during
which
# this relocation was not possible (storm, lack of field workers, etc.).
# We need to add missing values to define a regular trajectory.
# We have to define a reference date that will be used to check that each
date in
# the object of class ltraj is separated from this reference by an integer
multiple
# of the theoretical dt (here, one day), and place the missing values at
the times
# when relocations should theoretically have been collected.

refdata <- strptime('2013-02-13 00:00:00', '%Y-%m-%d %H:%M:%S')

# function setNA to place the missing values with theoretical collects
every 6 hours
coto_moves2 <- setNA(coto_moves, refdata, 6, units = 'hour')

# Even when filling the gaps with NAs, the trajectory is still not regular.
# The function sett0 can be used to 'round' the timing of the coordinates.
coto_moves3 <- sett0(coto_moves2, refdata, 6, units = "hour")


And from then, I filled with doubts.
My wawotest returned any correlation among the variables.

            dx           dy        dist
a   24.6905441  42.29083967  24.6541496
ea  -1.0000000  -1.00000000  -1.0000000
va 403.9675822 405.47166447 409.2888978
za   1.2782036   2.14988760   1.2680683
p    0.1005888   0.01578205   0.1023868


(2) should I discretize the the data in space? Or in time?

As say in other message in r-sig-geo forum, I'm a begginer and
unfortunately working alone. Could you please provide me some guidance with
this?
Any suggestions will be appreciated.
Follows attached a snippet of my data and my script if it can be useful.

Best regards