Apart from doing this individually, since there is different regions involved, I do not think you ca reconcile the two. Unless you strictly use a taxonomic approach? I do not think you can do this in Mothur though.
As Alexandre mentioned, you could run everything through the phylotype pipeline and then make a comparison using mothur. But… I generally think mixing datasets - even if theyr’e the same region - is a bad idea since you’ll be combining different extraction methods, primer sets, etc and they aren’t an apples to apples comparison. Rather, I would run your analyses using each dataset separately and then compare the results like a traditional meta analysis.