The issue, for us, is that using an observation field for a taxonomic ID sidesteps all the amazing, sophisticated iNat infrastructure for crowdsourcing species IDs in an open, trackable, and dynamic way.
You’re right that we could definitely pull all the “interaction ->…” fields to the website with the API. But then we’d be more duplicative of GLOBI effort since that’s basically what they do (in addition to pulling records from the literature). Our niche is focusing more narrowly on interactions where there’s an observation for both partners, linked with URLs. So, basically, trading ease for (presumably) a bit higher data quality. One thing I have been doing is using those other observation fields (and predation-related projects) to find candidate observations to add to Who Eats Whom too.
Congratulations on this publication! I’m happy you invited me to be part of this project since this is such an interesting and important topic. I add observations as often as possible.
That’s basically what we do, right? The observation field we use is called “URL for partner observation”. And then by having that link to the partner observation we don’t really need a separate interaction field, since linking the two observations inherently links them as interacting (in the way our code works for the website). But maybe I’m not understanding!