# Hellinger pre-transformation of the species data
spe.h <- decostand(spe, "hellinger")
(spe.h.pca <- rda(spe.h))
# Scree plot and broken stick model
screeplot(
spe.h.pca,
bstick = TRUE,
npcs = length(spe.h.pca$CA$eig)
)
# PCA biplots
spe.pca.sc1 <- scores(spe.h.pca, display = "species", scaling = 1)
spe.pca.sc2 <- scores(spe.h.pca, display = "species", scaling = 2)
par(mfrow = c(1, 2))
cleanplot.pca(spe.h.pca, scaling = 1, mar.percent = 0.06)
cleanplot.pca(spe.h.pca, scaling = 2, mar.percent = 0.06)
The species do not form clear groups like the environmental variables. However,
see how the species replace one another along the site sequence.
In the scaling 1 biplot, observe that 8 species contribute strongly to axes 1 and 2.
Are these species partly or completely the same as those identified as indicators of
the groups in Sect. 4.11?
For comparison, repeat the PCA on the original object spe without
transformation. Which ordination shows the gradient of species contributions
along the course of the river more clearly?
-0.5
0.0
0.5
-0.5
0.0
0.5
PCA biplot - Scaling 1
PCA 1
PCA 2
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Cogo
Satr
Phph
Babl
Thth
Teso
Chna
Pato
Lele
Sqce
Baba
Albi
Gogo
Eslu
Pefl
Rham
Legi
Scer
Cyca
Titi
Abbr
Icme
Gyce
Ruru
Blbj
Alal
Anan
-2
-1
0
1
2
3
-3
-2
-1
0
1
PCA biplot - Scaling 2
PCA 1
PCA 2
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Cogo
Satr
Phph
Babl
Thth
Teso
Chna
Pato
Lele
Sqce
Baba
Albi
Gogo
Eslu
Pefl
Rham
Legi
Scer
Cyca
Titi
Abbr
Icme
Gyce
Ruru
Blbj
Alal
Anan
Fig. 5.4 PCA biplots of the Hellinger-transformed fish species data
5.3 Principal Component Analysis (PCA)
167