adonis2: strata with unbalanced nested designs?

[marinalaforgia]

Hi all,

I am analyzing differences between microbial communities in combinations of plant hosts and watering treatments where plant host subplots were nested in watering treatments plots, but our sampling design ended up being unbalanced. Because of the nestedness of this design, I thought it would be best to run the following code, but I am finding conflicting answers on whether or not this is best since our design is unbalanced:

adonis2(Dist ~ WaterTrt * PlantTrt, df, by = "margin", permutations = 9999, strata = df$Plot)

at the very least, I was thinking of including Plot as a covariate:

adonis2(Dist ~ WaterTrt * PlantTrt + Plot, df, by = "margin", permutations = 9999)

Can someone more familiar with permanova advise on the best way to analyze this type of nested, unbalanced data?

Cheers,
Marina