adehabitatHS - kselect analysis
Dear Roland,
while analyzing animal relocations using the kselect analysis in adehabitatHS (third-order habitat selection), I would greatly appreciate any help with the following two topics: 1) How to test for significance if the relocations are not independent? Since I am working with trajectories (time interval between relocations is one hour), I am afraid I cannot use rand.kselect for this purpose? Remark: To account for the dependence between successive relocations, I have used the BRB function to estimate the home ranges.
I do no know, but just a clue: maybe you should have a look at null model approaches. Have a look at the vignette of the package adehabitatLT, section 5. It may be possible to simulate trajectories at each step of the null model approach using some kind of movement model (which one is a complex question; there are no built-in functions to do that, and designing a sensible movement model is not an easy thing).
2) This topic deals with the visualization of the kselect results. Using plot.kselect, I get (among others) two diagrams showing a) the scores of the environmental variables (corresponding to x$l1 of the output) on the axes and b), the scores of the recentered marginality vectors on the axes (x$co). Just to be sure: Would it be mathematically correct to combine both scores into one single diagram, i.e. to create a kind of biplot?
Basically, the K-select analysis is a non-centered principal component analysis, so that biplots sensu Gabriel (1971) are indeed possible. Use the function scatter from the package ade4 to draw such biplots. Best regards, Cl?ment Calenge
Many thanks in advance! Roland Pape [[alternative HTML version deleted]]
_______________________________________________ R-sig-Geo mailing list R-sig-Geo at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Cl?ment CALENGE Cellule d'appui ? l'analyse de donn?es Direction des Etudes et de la Recherche Office national de la chasse et de la faune sauvage Saint Benoist - 78610 Auffargis tel. (33) 01.30.46.54.14