Dear friends,
I have recently posted a message questioning how could one discount
the spatial variation in community species data (what has been
increasingly done through canonical variation partitioning, e.g. varpart
in Vegan) and still access the individual contribution of distinct
environmental variables to variation in species data.
I would like to tell the solution I found.
Step 1. Proceed a PCNM analysis of the spatial aspect of the dataset.
Step 2. Use the significant PCNM axis, taken as a group, as a covariate
in a PERMANCOVA (Anderson, 2001). This can be done in vegan whith the
code
adonis(sp ~ . + as.matrix(pcnm), data = env)
where sp means the rectangular matrix os species abundances, pcnm stands
for the significant pcnm variables, and env stands for the rectangular
environmental matrix.
This yields a ANOVA-like output table, where the effect of the
spatial structure is evaluated after removing the effect of the
environmental variables, but individual environmental variables are
still tested individually.
I believe this partly alleviates the need for canonical partitioning
several environmental matrices (i.e., subidiving the environmental
matrix in subsets).
Step 3. Proceed with a canonical variation partitioning of the data, to
evaluate the overall % contribution of pure spatial, pure environmental,
and spatially structured environmental matrices.
What do you think?
Best regards,
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
[email protected]
http://www.unisinos.br/laboratorios/lecopop
--
Esta mensagem foi verificada pelo sistema de antiv�rus e
acredita-se estar livre de perigo.
