Canonical Analysis on Principal Coordinates

Canonical Analysis on Principal Coordinates (CAP) is related to Principal Coordinate Analysis and allows the exploration of complex, non-linear relationships between two sets of variables, typically species data and environmental factors.

Description

Canonical Analysis on Principal Coordinates (CAP) is a multivariate statistical technique commonly used in ecological and environmental studies to explore relationships between two sets of variables, typically species data and environmental factors. This method is related to PCOA working directly with dissimilarity matrices derived from the data.

Example

wetlands <- readRDS("wetlands.rds")

library(vegan)

cap_ord <- capscale(wetlands$cross ~ groundwater + tmean + tseason + prsum +
    prseason,
        data = wetlands$env, dist = "bray")
cap_ord

Call: capscale(formula = wetlands$cross ~ groundwater + tmean + tseason
+ prsum + prseason, data = wetlands$env, distance = "bray")

               Inertia Proportion Rank
Total         143.2120     1.0000     
Constrained    26.8696     0.1876    5
Unconstrained 130.7534     0.9130  169
Imaginary     -14.4110    -0.1006  154
Inertia is squared Bray distance 
Species scores projected from '$' 'wetlands' 'cross' 

Eigenvalues for constrained axes:
  CAP1   CAP2   CAP3   CAP4   CAP5 
10.738  6.546  6.321  2.611  0.654 

Eigenvalues for unconstrained axes:
  MDS1   MDS2   MDS3   MDS4   MDS5   MDS6   MDS7   MDS8 
14.944 12.391  7.744  7.536  6.443  6.014  5.100  4.508 
(Showing 8 of 169 unconstrained eigenvalues)

plot(cap_ord)