fourth.aravo <- fourthcorner(
tabR = aravo$env,
tabL = aravo$spe,
tabQ = aravo$traits,
modeltype = 6,
p.adjust.method.G = "none",
p.adjust.method.D = "none",
nrepet = 49999)
# Correction for multiple testing, here using FDR
fourth.aravo.adj <- p.adjust.4thcorner(
fourth.aravo,
p.adjust.method.G = "fdr",
p.adjust.method.D = "fdr",
p.adjust.D = "global")
# Plot
plot(fourth.aravo.adj, alpha = 0.05, stat = "D2")
This representation allows a detailed interpretation. For instance, SLA (specific
leaf area) and N_mass (mass-based leaf nitrogen content) are positively associated
with Snow (mean snow melt date) and Form.5 (concave microtopography), features
that can also be observed in Fig. 6.22c, d. This shows that these traits are likely to
favour species that tolerate a longer period of snow cover: a higher nitrogen content,
partly due to nitrogen storage in snowpacks and partly to the protective effect of
snow on soil temperature and water content (Choler 2005), warrants larger reserves,
and a larger leaf area allows a larger rate of photosynthesis once the plant is
eventually exposed to the sun. Conversely, these two traits are negatively associated
with PhysD (physical disturbance due to cryoturbation), which tends to occur more
often in areas without snow and therefore more exposed to large temperature
oscillations.
The complementarity between RLQ and fourth-corner analyses can further be
exploited by representing traits and environmental variables on a biplot (inherited
from the RLQ analysis), and adding blue lines for negative and red lines for positive
associations (inherited from the fourth-corner tests) (Fig. 6.24):
# Biplot combining RLQ and fourth-corner results
plot(fourth.aravo.adj,
x.rlq = rlq.aravo,
alpha = 0.05,
stat = "D2",
type = "biplot"
)
In this biplot, the positive relationships between the traits SLA and N_mass and
the environmental characteristics Snow and Form.5, discussed above, show up
clearly as a tight group of associations that take place in concave-up sites where
snow takes time to melt. Of course, many other relationships can be identified in this
graph.
6.11 Relating Species Traits and Environment
295
tabR = aravo$env,
tabL = aravo$spe,
tabQ = aravo$traits,
modeltype = 6,
p.adjust.method.G = "none",
p.adjust.method.D = "none",
nrepet = 49999)
# Correction for multiple testing, here using FDR
fourth.aravo.adj <- p.adjust.4thcorner(
fourth.aravo,
p.adjust.method.G = "fdr",
p.adjust.method.D = "fdr",
p.adjust.D = "global")
# Plot
plot(fourth.aravo.adj, alpha = 0.05, stat = "D2")
This representation allows a detailed interpretation. For instance, SLA (specific
leaf area) and N_mass (mass-based leaf nitrogen content) are positively associated
with Snow (mean snow melt date) and Form.5 (concave microtopography), features
that can also be observed in Fig. 6.22c, d. This shows that these traits are likely to
favour species that tolerate a longer period of snow cover: a higher nitrogen content,
partly due to nitrogen storage in snowpacks and partly to the protective effect of
snow on soil temperature and water content (Choler 2005), warrants larger reserves,
and a larger leaf area allows a larger rate of photosynthesis once the plant is
eventually exposed to the sun. Conversely, these two traits are negatively associated
with PhysD (physical disturbance due to cryoturbation), which tends to occur more
often in areas without snow and therefore more exposed to large temperature
oscillations.
The complementarity between RLQ and fourth-corner analyses can further be
exploited by representing traits and environmental variables on a biplot (inherited
from the RLQ analysis), and adding blue lines for negative and red lines for positive
associations (inherited from the fourth-corner tests) (Fig. 6.24):
# Biplot combining RLQ and fourth-corner results
plot(fourth.aravo.adj,
x.rlq = rlq.aravo,
alpha = 0.05,
stat = "D2",
type = "biplot"
)
In this biplot, the positive relationships between the traits SLA and N_mass and
the environmental characteristics Snow and Form.5, discussed above, show up
clearly as a tight group of associations that take place in concave-up sites where
snow takes time to melt. Of course, many other relationships can be identified in this
graph.
6.11 Relating Species Traits and Environment
295