Are the ranges with “extant” in their name datasets that were created from just expert opinion or are they a combination of expert and observational data? If they are only expert opinion is there a way to generate a range based on expert and observational data?

Extant ranges are expert opinion, but our determination was that the expert's opinion of where the species are should include the observed data, so the expert range locations should include HUCs originally specified in the observed data. That said, the best way to be sure is definitely to include both data types. You can do this a few ways, depending on what your end goal is. If you just need a one-off layer that includes both of those data types, you can use the "Generate Layer from Query" tool, which will output a layer into the current map document. The query, in this case would be:

select distinct Zone_ID from Observations where Species_ID = "CMC01" and presence_type in (1,3,6,7,9)

Replace CMC01 with the PISCES species code for the fish you're looking for, which you can find in the Species table of PISCES.sqlite. I think you both have SQL experience, so I won't explain much about the rest of what's going on there, but let me know if you need anything else explained. For "presence_type in (1,3,6,7,9)" those correspond to these PISCES presence types, so you can modify that portion as appropriate.

If you wanted to make this a more permanent fixture that you can run for any number of species, you can make a custom map set with that query and a placeholder like ? in place of the species ID. You can find more information on that starting on page 5 of this tutorial1.