abline(v = 0, lty = "dotted")
abline(h = 0, lty = "dotted")
for (i in 1:length(grl)) {
points(sit.sc1[gr == i, ],
pch = (14 + i),
cex = 2,
col = i + 1)
}
text(sit.sc1, row.names(env), cex = 0.7, pos = 3)
# Add the dendrogram
ordicluster(p, env.w, col = "dark grey")
# Add legend interactively
legend(
locator(1),
paste("Cluster", c(1:length(grl))),
pch = 14 + c(1:length(grl)),
col = 1 + c(1:length(grl)),
pt.cex = 2
)
Hint See how the coding of the symbols and colours is automatically conditioned on the
number of groups: object grl has been set to contain numbers from 1 to the
number of groups.
5.3.3 PCA on Transformed Species Data
PCA being a linear method that preserves the Euclidean distance among sites, it is
not naturally adapted to the analysis of species abundance data. However,
transforming these after Legendre and Gallagher (2001) solves this problem
(Sect. 3.5).
5.3.3.1 Application to the Hellinger-Transformed Fish Data
This application is based on functions decostand() and rda(). The graphical
result is presented in Fig. 5.4. Readers are invited to experiment with other transformations, e.g. log-chord (see Sect. 3.3.1).
166
5 Unconstrained Ordination