Hi everyone,
I am working on a community analysis project. I am very happy with a graph
produced in package vegan using betadisper that is based off of a custom
community distance matrix (uni.dist) and a species abundance by site table
(shared). Betadisper produces a great PCoA plot of the communities
clustered by habitat type (convex hull around communities belonging to the
same habitat type, lines from community to the center of the cluster).
z <- with(enviro.variable.matrix, betadisper(as.dist(uni.dist),
enviro.variable.matrix$habitat))
plot(z)
What I would like to do now is to turn this into a biplot with the species
most important in determining community relatedness represented as arrows.
I can produce the PCoA plot without the convex hulls and centroids that
contains ALL of the species arrows (~200) using package ape.
uni.pcoa <- pcoa(uni.dist)
biplot(uni.pcoa, Y=shared, plot.axes=c(1,2))
Unfortunately, I cannot find any way to extract the loadings from the
biplot output. I've tried storing the biplot in a variable, but there does
not appear to be anything produced by the function except the plot (I
understand why uni.pcoa doesn't contain loadings). I've played with doing a
NMDS with the weighted averages of the species displayed as points instead,
but this does not communicate the same information that I was wanting. Any
suggestions would be appreciated!
Thank you,
Jo
[[alternative HTML version deleted]]
_______________________________________________
R-sig-ecology mailing list
R-sig-ecology@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
