Hi.
I'm trying to decide the 'best' way to analyze my data....
Basically, I am interested in knowing if habitat type (binomial) of a
population and/or distance between populations (ibd) is effecting genetic
distances between populations. My genetic distances are pair-wise Fsts (an
nxn matrix), and I have a vector of 0's and 1's for my habitat type
descriptor. For looking at ibd, I have the raw lat/long data and a matrix
of pair-wide geographic distances. I wanted to use adonis, but am unclear
on how or if I could use my geo distance data. I know that for things like
RDA you can use pcnm's of the distances, but since that only takes the
positive eigenvalue-related eigenvectors and adonis takes all eigenvectors
(I believe, at least in vegan), that seems, well, wrong. I'm thinking that
just using the lat and long values is not what I'm looking for (not worried
about a clinal effect). I saw that Pilot et al. 2006 used DISTLM (aprox.
the pay money version of adonis?) to look at the significance of their
environmental effects separately from a partial Mantel test to look at for
significant IBD. I'd rather look at both simultaneously because that would
make more sense, and I'd think should be possible. Any tips?
Thanks,
Ginnie
[[alternative HTML version deleted]]
_______________________________________________
R-sig-ecology mailing list
R-sig-ecology@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
