If the samples are all distributed within the same region, then you cant add region as a factor in your analysis. the region would be included as an upper level in a hierarchal AMOVA, so you can understand how molecular variation is partitioned within populations, among populations within region, and among regions. You could, however, add other information at this level, say if your populations were from two different looking ecotypes or from two distinct habitats.

Agree with Sean's comments. You can use DARWIN,as it has an option where you can see how the two populations relate but could have another level for regions.

what kind i of analysis are you interested in? What is your research question?

I want to use AMOVA to show that there is a difference between my two population and that the location factor is not associated with genetic structure... all the isolates of the 2 populations are collected from 3 regions

Genealex is only taking the data when each population comes from different regions

Anyone can help please