[R-sig-eco] Ordisurf in CCA is appropriate

[sunilmundra]

Hi everybody, 
I have some questions about Unconstrained ordination (GNMDS) and Constrained
Ordination (CCA) and use of ordisurf() in both method.
I am working on soil fungal sequence data (presence-absence matrix). I have
quite a few environmental variable to test with fungal community. Repeated
sampling was done for 9 weeks. Julian day was not related with fungal
community where as factorial time variable showed relationship with fungal
community. Which indicated that there is no direct relation ship with time,
but significantly related env variable (Moisture, NO3 and K) drive this
community. This is fine.
I run GNMDS analysis and it shows there is considerable variation due to
spatial heterogeneity. (Factor spatial variable shows approx 30% and when I
used UTM cordinate, Northing shows approx 10% variation, sampling distance
is within 1.5 x 1.5 km area, Can some one tell me which variable is more
appropriate for space, Factor variable or Cordinate). Therefore i need to
use appropriate method by which i can remove variation due to space and then
i want to test impact of env variable on community. 

Partial DCA can be good approach but i have no idea how to do it in R.
Suggestion would be helpful! Therefore in order to solve my purpose i used
CCA where i removed the variation due factorial space variable and and then
used Ordisurf to find the pattern. It was observed that soil water content
and NO3 content drive this community shift. 
The CCA ordination shows highly significant and perfectly ecologically
interpretable results. I would like to know is this OK to use ordisurf with
CCA or it is not appropriate and ordisurf should only be used with
unconstrained ordination.
Thanks a lot for your help! 
Best regards, 



--
View this message in context: http://r-sig-ecology.471788.n2.nabble.com/Ordisurf-in-CCA-is-appropriate-tp7579056.html
Sent from the r-sig-ecology mailing list archive at Nabble.com.