RE: Graphs

From: Sylvain DolÚdec (sylvain@biomserv.univ-lyon1.fr)
Date: Fri Dec 12 2003 - 11:07:00 MET


  Dear all,

  I did Co-inertia analysis for Table 1=SamplesxEnvir variables and Table
2=SamplesxSpp.

  Could you please tell me how to plot samples and species abundance along
the F1 co-inertia axis defined by the environmental variables?
  I saw that in DolÚdec S and Chessel D 1994, Coinertia analysis: an
alternative method for studying species-environment relationships.
Freshwatre Biology 31.

  I guess you refer to Fig. 9 (?) Just use the Curves module with bar option
your --.iil1 file as x-axis and your faunistic table as Y (species were
previously selected). Then you rework the graphs under any drawing
software... But you can get a better result with the graphical routines of R
(eg plot and type="h") including a direct selection of species that you want
to draw, etc.

  I have already used module Scatters-Values + (file --.iil1) + (Spp as G
values).
  Could you please confirm this is really the graphic representation of the
species abundances at samples positioned acording to the co-inertia F1/F2
axes defined by the environmental variables?
  I think so... --.iil1 contains the sample scores computed from your
environmental table. But this is not done in the paper you mention above.

  I would also appreciate to know how to use ADE4 for ploting samples at
axisXX=(CoInertia axis F1 for species) and axisYY=(Coinertia F1 axis for
environmental variables).
  The idea (according to the same authors) is to plot samples at these axis
so that to show correlation between the environmental ordination of samples
and the species ordination of species.
  Trying to do that, I did extract 1st column of --i.il1 and 1st column
of --.iil2 to plot samples at other graph software.
  Could you please tell me if it is correct?
  Anyway I would prefer doing that at ADE4...is it possible?
  I suppose you refer to Fig.8a of the paper (?). You can do this with the
Curves module and the label option of ADE4 ("a nice software") if you
prefer. Put --.iil2 as X-axis and --.iil1 as Y-axis. Note that this is a
reverse choice compared to the paper you mention.

  Best regards.
  Sara Mouro

  Best regards as well
  Sylvain



This archive was generated by hypermail 2b30 : Tue Sep 07 2004 - 13:45:26 MEST