0

I have a data set with abundace of 4 species (S. spp, in bright colours here) across 72 sites and corresponding temperature, pH, altitude, moisture.

First, I performed a CCA where species abundance ~ temperature + pH + altitude + moisture. This graph captures the crux of the analysis:

enter image description here

Then, I wanted to explore how different the four species were. So I did a simple PCA with oordination plots as captured here:

enter image description here

On one hand, the first plot helps me interpret how the communities differ with respect to their environment and which of the abiotic variables most affects the various species. On the other, the second plot suggests that the communities are actually rather similar in their use of environment (temp, moisture etc).

I'm trying to understand how to reconcile what the two methods are teaching us. I would love nay input.

slicer
  • 603
  • 1
    These are biplots, a kind of overlay scatterplots. Since the theory and idea behind PCA and canonical correlation is different, the two pictures, while of the same kind, are not quite comparable. So try not to compare/reconcile them but rather to interpret both individually. – ttnphns Dec 22 '22 at 16:20
  • 1
    The ellipses probably add impression of the shapes of the clouds and sizes of their variation. – ttnphns Dec 22 '22 at 16:23

0 Answers0