Is it recommendable to use a three way PERMANOVA analysis for an unbalanced design?

I have a data set that consist of observations in four ponds and one stream. I have three factors: locality (each lake and the stream), habitat (permanent pond, temporal pond, stream) and...

31 May 2015 6,998 3 View