I’ve seen other posts about similar issues, but not exactly what I am seeing here.
For the moth subfamily Cerurinae [https://www.inaturalist.org/taxa/1185181-Cerurinae] (which I think was recently added to the taxonomy of the moth family Notodontidae… not by me) I am seeing discrepancies between “total observations” and the actual total number of observations listed in the taxonomy section. There are over 3,000 observations of Cerurinae, but the “total observations” that you see in a more standard view/search is just 785. It almost seems like some daughter taxa are not showing up in a search for “Cerurinae” despite them belonging to that subfamily (which I verified by following back the parent taxon from species to genus to subfamily).
One possible explanation is that some of the taxa that belong to Cerurinae are still showing up when I search for the tribe that they used to belong to (Dicranurini). But to me this does not make sense. When I search for Dicranurini I see some species of Cerurinae (the species Furcula occidentalis is an example), but when I investigate the taxonomy of the Cerurinae species showing up in a Dicranurini search I see that it is still assigned to the parent taxon (a genus, Furcula) whose parent taxon is Cerurinae. It seems like certain species of Cerurinae are showing up in either Cerurinae OR Dicranurini, thereby splitting the totals of Cerurinae, but I do not know how this is possible.
Note that Dicranurini does NOT have a subfamily-level parent taxon (it should, but doesn’t), Dicranurini is simply curated as a daughter of family Notodontidae. Could this be what’s causing these issues?
This type of issue has seemed to crop up periodically when taxa get moved - lower taxa somehow retain deprecated parantage. I think I had fixed this once by just moving a taxa back and forth again (though this is obviously not practical nor ideal). I think there might be some periodic reindexing too that will eventually resolve it.
That makes sense, I’ve actually seen the number of Cerurinae observations tick up slowly over time (more than would be expected by simply adding new observations). So perhaps reindexing will eventually resolve it.
