# Extraction of the richness difference matrix
fish.rich <- as.matrix(fish.pod.j$rich)
# Plot of the richness difference with respect to site 30
fish.rich.30 <- fish.rich[29, ][-29]
site.names <- seq(1, 29)[-8]
plot(site.names,
fish.rich.30,
type = "n",
xaxp = c(1, 29, 28),
main = "Doubs fish data: richness difference with respect
to site 30",
xlab = "Site number",
ylab = "Richness difference"
)
lines(site.names, fish.rich.30, pch = 24, col="red")
points(
site.names,
fish.rich.30,
pch = 24,
cex = 1.5,
col = "white",
bg = "red"
)
text(3, 0.85, "Upstream", cex = 1, col = "red")
text(27, 0.85, "Downstream", cex = 1, col = "red")
Hint Observe how the richness difference between sites 1 to 29 and site 30 has been
obtained: the richness difference object of class “dist” is first transformed into a
square matrix (fish.rich) and its last row is extracted, excluding the last
value (richness difference of site 30 with itself; remember that site 8, where no fish
had been captured, has been removed from the data set).
Figure 8.6 shows that upstream from site 30 there is little richness difference up to
site 17, with the notable exception of sites 23, 24 and 25. In the mid twentieth
century, these three sites were heavily polluted (urban pollution). Since more species
are found upstream from these sites, one can deduce that the species pool has been
depleted in these three sites due to major anthropic perturbation. Upstream from site
17, richness difference steadily increases to site 9, then drops at sites 6, 5 and 4, then
increases again up to site 1, where it is the highest. For some reasons (probably
pollution downstream of the city of Pontarlier at the time of sampling in the 1960’s),
species richness is quite low in sites 7–13 when compared to the other sites in the
upper course of the river.
We will now plot together Podani’s Jaccard-based replacement (Repl J ) and
richness difference (RichDiff J ) indices as well as the Jaccard dissimilarity. Contrary
to Legendre’s example, however, we will plot the values between neighbouring
sites, not with respect to site 30 (Fig. 8.7), to emphasize local (pairwise) features.
8.4 Beta Diversity
397
fish.rich <- as.matrix(fish.pod.j$rich)
# Plot of the richness difference with respect to site 30
fish.rich.30 <- fish.rich[29, ][-29]
site.names <- seq(1, 29)[-8]
plot(site.names,
fish.rich.30,
type = "n",
xaxp = c(1, 29, 28),
main = "Doubs fish data: richness difference with respect
to site 30",
xlab = "Site number",
ylab = "Richness difference"
)
lines(site.names, fish.rich.30, pch = 24, col="red")
points(
site.names,
fish.rich.30,
pch = 24,
cex = 1.5,
col = "white",
bg = "red"
)
text(3, 0.85, "Upstream", cex = 1, col = "red")
text(27, 0.85, "Downstream", cex = 1, col = "red")
Hint Observe how the richness difference between sites 1 to 29 and site 30 has been
obtained: the richness difference object of class “dist” is first transformed into a
square matrix (fish.rich) and its last row is extracted, excluding the last
value (richness difference of site 30 with itself; remember that site 8, where no fish
had been captured, has been removed from the data set).
Figure 8.6 shows that upstream from site 30 there is little richness difference up to
site 17, with the notable exception of sites 23, 24 and 25. In the mid twentieth
century, these three sites were heavily polluted (urban pollution). Since more species
are found upstream from these sites, one can deduce that the species pool has been
depleted in these three sites due to major anthropic perturbation. Upstream from site
17, richness difference steadily increases to site 9, then drops at sites 6, 5 and 4, then
increases again up to site 1, where it is the highest. For some reasons (probably
pollution downstream of the city of Pontarlier at the time of sampling in the 1960’s),
species richness is quite low in sites 7–13 when compared to the other sites in the
upper course of the river.
We will now plot together Podani’s Jaccard-based replacement (Repl J ) and
richness difference (RichDiff J ) indices as well as the Jaccard dissimilarity. Contrary
to Legendre’s example, however, we will plot the values between neighbouring
sites, not with respect to site 30 (Fig. 8.7), to emphasize local (pairwise) features.
8.4 Beta Diversity
397