Skip to content
Prev 5940 / 7420 Next

interpreting ecological distance approaches (Bray Curtis after various data transformation)

Dear Tim,

You are right: Bray-Curtis distance will be non-zero if two communities 
differ in size (sum of abundances), even if the relative abundances is 
the same. If you have number of individuals data, rarefying is the best 
solution. If you cannot apply it (e.g. because only cover data are 
available), you can calculate distance from relative abundance, and yes, 
this case BC is equivalent to Manhattan. Note that using relative 
abundances don't remove fully the effect of different sampling effort, 
because rare species could missing from the smaller sample.

I don't recommend calculating BC-distance from Hellinger-transformed 
data, because sum of transformed abundances are meaningless.

Best regards,

Zoltan

2019. 04. 02. 17:15 keltez?ssel, Tim Richter-Heitmann ?rta: