Taxacom: replacing un-sequenceable types (was Re: Minimalist revision of Mesochorus)

[Jared Bernard]

Not only is biogeographical info not in DNA per se, but there are other
obvious risks of relying entirely on DNA. Clearly there is now a sentiment
that you can simply blast a barcode against a genetic repository to
identify something. Arguments about barcoding aside (and there are plenty),
a lot of stuff on genetic repositories is misidentified (e.g., Meiklejohn
et al. 2019, doi: 10.1371/ journal.pone.0217084; Fort et al. 2021, doi:
10.1111/1755-0998.13453; Cheng et al. 2023, doi:
10.3389/fevo.2023.1149839). So, someone who may not be an expert in
distinguishing species taxonomically has identified their species, and now
that name is forever associated with that genetic material. So when the
next user blasts their species' DNA, it comes back with an ID (or rather,
highest similarity), and that user probably doesn't question it. (Same
problem with the new AI tools.)

The task of the underappreciated field of taxonomy must therefore be to
distinguish species morphologically so that any following work, like
evolutionary relationships or ecology, won't be misapplied to the wrong
species. I always wonder, given the above emerging data on misidentified
material on genetic repositories, how many papers on ecology, conservation
biology, or evolution have misidentified the study subjects. This is the
problem with the widening disconnect between people studying biodiversity
and people who can identify the basic units of biodiversity.

And of course, this is aside from the fact that only an infinitesimal
amount of the world's biodiversity has any sequence in a repository,
incorrect or not (Kvist 2013, doi: 10.1016/j.ympev.2013.05.012).

If I'm restating an earlier comment, I apologize.