31

ECOM II

7

Canonical Correspondence Analysis

CCA seeks an ordination of the sites (samples) in terms of their biological components

which is optimised in terms of the influence of environmental factors.

What is CCA

?

Environmental variables

Selecting Environmental variables

Linear combinations of environmental variables

Dummy Environmental variables

Transforming environmental variables

Circular environmental variables

Biological variables

The normal situation is for the biological data to comprise a sample (column) by

species (row) array. The species data would comprise quantitative counts of the

number of individuals in each sample. (See 

Creating and editing a data set

).

Presentation of Results   Your output

CCA produces an ordination of the species and sites. It also produces biplots that

show the inferred ranking of the species along an environmental variable. The output

from ECOM is divided into a number of different sections of a tabbed notebook. These

are described below.

Eigenvalues

Species Scores

Sample Scores

Ordination Plot

Rank Plot

Correlation of Env. Vars

Binary Environments

Biplot Scores

Canonical Coef

7.1

What is CCA?

Canonical correspondence analysis (CCA; ter Braak 1986, 1994) is an ordination

method in which the ordination of the biological (main) matrix by correspondence

analysis or reciprocal averaging is constrained by a multiple regression on the

variables included in the environmental matrix.

In ecological terms, the ordination of sites and species is constrained by their

relationships to environmental variables. If the environmental variables included are

major determinants of community structure and abundance of species changes along

these environmental gradients then this technique aids interpretation of community

structure and identification of the features that mould it.

The alternative approach is to use ordination techniques such as Correspondence

Analysis or Principal Components Analysis which ordinate the community (sites 

species) data alone and then undertake an auxiliary analysis to identify the

Copyright 2004  PISCES Conservation Ltd