Hello Dr. Souza,
I have not done a constrained NMDS as you ask, but I have used NMDS
in the program PC-Ord (developed by McCune) quite a bit. My standard
approach is to use 2 data matrices -- the 1st is the community data, and
you use this to create the ordination solution (i.e., the distribution
of points in space). The 2nd matrix is the environmental data, and you
use his to create overlays onto the community ordination (i.e., either
generate vectors or scale the size of ordination symbols to an
environmental value). In this way, you gain the benefit of the best
possible solution of the community data (i.e., the solution is not
constrained by the data of the 2nd matrix) and then you get to visualize
the relationship with the "controlling variables" of the 2nd matrix).
My apologies if you already knew this, as I know it isn't the
question you asked. But, maybe it was helpful....
Michael Cooperman
Research fellow of the National Research Council
Alexandre F. Souza wrote:
Hello All,
Do anyone has ever performed a constrained NMDS with community data? I
have read in McCune and Graces' Analysis of Ecological Communities that
the theory of such analysis already exists (i.e., associating vectors of
environmental variables to a NMDS ordination) but has seldon been used
by ecologists. It would be a better alternative to CCA, which has
stingent premises that ecological data does not match properly most of
times.
I suppose that this could be implemented in R, but I could not find it
in Vegan or other packages.
Thanks in advance,
Alexandre
Dr. Alexandre F. Souza
Programa de Pós-Graduação em Biologia: Diversidade e Manejo da Vida
Silvestre
Universidade do Vale do Rio dos Sinos (UNISINOS)
Av. UNISINOS 950 - C.P. 275, São Leopoldo 93022-000, RS - Brasil
Telefone: (051)3590-8477 ramal 1263
Skype: alexfadigas
afso...@unisinos.br
http://www.unisinos.br/laboratorios/lecopop