I have 16S data from bacterial communities from 3 different oceans and within each ocean, they are separated by the morphology of the host they inhabit. So basically I want to test whether my communities are significantly different based on the two categorical variables (ocean and morphology). I see that I can use one way non parametric test with mothur (e.g. amova), but I wonder if there is any way of running a two way PERMANOVA.
I assume that I need to input a normalized OTU table (obtained with the sub.sample command) into the adonis function) for the adonis command. And also for any other function in the vegan package (such as metaMDS)?