Hello!
I think, there is no direct solution in Qiime2 to perform the task you described in your post.
But here are some workarounds you could try to apply:
Filter your ASV table to remove low abundant ASVs based on the relative abundances. Here is a link to the plugin. Not exactly what you asked for but I think it can work.
Collapse your table to certain taxonomy rank and filter it using the plugin above. It will do the trick but you will end up with collapsed table and I believe it is not what you asked for.
Download a csv file from taxabarplot visualization at desired taxonomy rank (you can find a link in upper left part of the visualization screen). It will give you a table with absolute counts of taxa, so you will need to manipulate it to obtain percent values (divide each value of taxa abundances in certain sample by the total count of all taxa in that given sample and multiply it by 100), find an average of each taxa accross all your samples and filter it to retain only taxa with abundances less than 1% and use that list to filter taxa from your table.
Note that in that way taxa, that are represented in certain group of samples in higher than 1% ratio can still be filtered out if it is underrepresented in other groups of samples.
