Great question! The basic answer is that OTU clustering happens first, then taxonomy assignment happens second. During the initial OTU clustering process, all OTU representative sequences must be >3% different from each other (or if you used dada2 or deblur, they might be only a few base pairs different from each other!) After clustering OTUs, the best possible taxonomy is assigned to each OTU, and the best match might happen to be to the same microbe in the taxonomy database. So while the OTUs should all be a little different, it’s possible for the taxonomy to be the same.
Does that help answer your question?
Colin
P.S. It looks like this data set was made with Qiime 1. Qiime 2 offers newer (and better!) ways of clustering OTUs and assigning taxonomy, so you could consider using Qiime 2 to improve your results.
