Hi @Masanobu_Hiraoka,
Thanks for sharing the data and commands — this allowed me to track down the error.
It looks like the issue is not due to the tree-building pipeline, but rather a known bug related to the use of dereplicate-sequences
without further clustering.
This issue — and possible workarounds — is also described here:
Using cluster-features-de-novo
should get you moving, even if you cluster at 100% similarity.
This is an unrelated issue. I do not see anything in the QIIME 2 part of your pipeline that would cause this issue (since you are just dereplicating the input sequences, no filtering involved), so I think you are right to look to the qiime1 forum to solve this issue... if you rule out qiime1 as the issue, please open a new topic to get help for this issue. Thanks!
Good luck! I hope my workaround proves suitable to you.