Dears,

I would like to do an AMOVA, but in my data set I have: a) one group with two populations and one group with six populations, b) some populations have 15/20 individuals and other 100/150

My questions are: 1°/ Is AMOVA robust or sensitive to such unbalanced data? 2°/ I could subsample the populations with 150 individuals and then, create populations with 20 individuals to deal with the issue “b)”, but how could I deal with the fact that my groups have very different numbers of populations (and thus, different numbers of individuals)?

I did not find any information about unbalanced data in AMOVAs, so your help will be greatly appreciated.

Thank a lot.