CCA: plot missing constrained variables?

Question

Im investigating how the abundance of a fish community along seasons of the year relates with environmental parameters of the water. So, both of my matrixes have four lines (Winter, Autumn, Spring a Summer) and on my species matrix the columns are my species and the values their mean abundance for each season; on my environmental data matrix, the columns are the water parameters (pH, salinity, temperature, oxygen and condutivity) and the values are their mean on each season (photos below).

When I plot my CCA result, it only plots 3 variables (out of 5)... Anyone knows why and how to solve it?

Here's the code Im using.

speE <- read.csv(choose.files(), row.names=1, sep=";")
envE <- read.csv(choose.files(), row.names=1, sep=";")

log.speE <-log(speE+1)
log.envE <-log(envE+1)

library(vegan)

cca1 <- cca(log.speE ~ ., data = log.envE)

plot(cca1, type="n")
text(cca1, dis="cn", col="dodgerblue3", cex=1.2)
text(cca1, col="chocolate2", bg="chocolate2", cex=1.1)
text(cca1, "species", col="black", cex=0.9)

summary(cca1)
cca1
anova(cca1, by = "axis")

****PHOTO EXAMPLES OF MY PROBLEMS****
[Environmental matrix by season] [1]: https://i.stack.imgur.com/zj0YR.png
[Species matrix by season] [2]: https://i.stack.imgur.com/unnZ0.png
[CCA that gone wrong missing environmental variables] [3]: https://i.stack.imgur.com/yu0yh.jpg

Jari Oksanen Jari Oksanen · Accepted Answer · 2020-11-21T20:12:04

I guess you were told that "Some constraints were aliased because they were collinear (redundant)"? If you did not see that information, please look again (it is there). For instance, if you have winter, autumn, spring and summer, you only have three independent variables: if it is not winter, autumn or spring, it must be summer. So summer will be aliased. Function alias(<myordination>) will tell you how this aliasing was done.

CCA: plot missing constrained variables?

1 Answers