Skip to content

partial mantel tests "Ecodist"

4 messages · Sarah Goslee, Penner, Johannes, ddepew at sciborg.uwaterloo.ca

Original
ddepew at sciborg.uwaterloo.ca
# 
Hi all,
I'm searching for a little clarification on partial mantel tests  
(ecodist package)

I've a distance matrix (x,y), and several others containing  
environmental/chemical variables.
Based on the help file, and the package instructions I've managed to  
implement the tests as;

var1 ~ env1 + space

to partial out the effect of space and test the relationship between  
the variable of interest vs variable env1.

My questions are as follows;

1) can "raw" data be used to construct the dissimilarity matricies? or  
should they be standardized? different variables have different  
measurment scales, my inclination is to standardize, but I don't know  
if this will dampen relationships between variables.

2) If env1 and another variable are correlated, is the appropriate test
         var1 ~ env1 + env2 + space?,
or
  var1 ~ env1 + space and then var1 ~ env2 + space?

3) interpretation... Does the value of "r" (i.e. + or -) imply spatial  
overlap (+) or spatial exclusion (-)?

Any assistance would be greatly appreciated!

Thanks,
Sarah Goslee
# 
Hi David,
On Fri, May 8, 2009 at 10:27 AM, <ddepew at sciborg.uwaterloo.ca> wrote:

            

      
      
      
    

        
I'd standardize, especially if you're using Euclidean distances. The Goslee
and Urban JSS paper on the ecodist package goes into more detail (as
do some of the references cited therein).

            

      
      
      
    

        
Test for what? The first one partials out both env2 and space from the
relationship of var1 ~ env1, a very different thing than the second
example.

            

      
      
      
    

        
A negative value for r is usually uninformative (unless you've used
particular data transformations or something otherwise unusual). The
Mantel test question is generally: do differences in X correspond to
differences in Y, so the test you want is whether r > 0. Again, see the
JSS paper discusses this further.

Sarah

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Penner, Johannes
    

    

      
      #
    

  


  

      
Dear all,

I am trying to calculate barriers with the monmonier algorithm
(adegenet).

mon1 <- monmonier(mycoordinates, mydistancamatrix, network$cn, ...)

The network beforehand looked alright. However, I always get the error:
"cn is not a nb object". I am not really sure what this means, probably
that the network is not recognised properly. So far I was not able to
detect my fault, though I already installed the newest version of R and
updated all packages (which was suggested in help files I have found
during the search). Any help would be highly appreciated!

Thanks in advance!

Best regards
Johannes

  
  


  


      

    

    

      
      

        


  

        

      ddepew at sciborg.uwaterloo.ca
    

    

      
      #
    

  


  

      
Thanks Sarah,
Looking again at #2, I see your point.
As for the standardization, I didn't see it mentioned in the JSS  
paper, but I'll have another look.
Assuming a significant r is returned, I guess I would need to look at  
the raw data to infer the type of relationship (+ or -).