Most likely, idk why tho a lot of translators often mistaken "treant" as "torrent".
Well, Google Translate gives it as "torrent", and the whole first page of Google results is BitTorrent and similar file sharing protocols. Jisho only lists "torrent (file); torrenting" and the location Trento (an Italian city). I can see why a translator might take a look at a string of katakana and assume that Google Translate or a dictionary knows what that word is, since the kana-romaji-loanword pipeline can be an absolute nightmare. Is the vowel in the first syllable there because that sound is actually there in the original word, or because
/tr/ isn't a permitted phoneme in Japanese and it had to be approximated? Now repeat this process for every kana and possibly adjacent pair of kana.
Also, not to point fingers at this group or any group in particular, but many JP-EN translations are done by people who are nonnative speakers of English, or at the very least may not be familiar with sometimes obscure English words that Japanese authors like to pull out of dictionaries. The Venn Diagram of "manga scanlators" and "people who really like DND/fantasy novels in English specifically" isn't necessarily separate, but its not a perfect overlap either.