Revisit rarefaction, NMDS, dbRDA & PCoA

[jibarozzo]

• Are we rarefying the date before NMDS?
• Is NMDS the right approach?
• Revisit Hellinger transformations and dbRDAs, are they adequate?
• Relative abundance transfromation before Bray-Curtis or after?

I think relative abundance before Bray-Curtis is leads to values at zero or close to zero which then metaMDS doesn't handle well.