I flagged the fact that a snake was showing up in the “similar species” list for Painted Turtle. The curator informed me that:
‘Similar species’ is an automatic list of other species that Painted Turtles have been mistakenly reported as. Oddly this means, that on at least 2 separate times, users have identified a turtle as this snake species.
I think it would be good for curators to “tag” species for inclusion or omission from similar species - in addition to the automatic algorithm that exists. I think it goes against the purpose of the site to show new users a snake as part of a similar species list.
I’ve generally found the automatically generated suggestions to be surprisingly good. I imagine what happened in this case is that there were observations of a turtle that only showed it poking it’s head out of the water. In that case, I could understand how it could be misidentified as a snake.
I also agree regarding adding species. Often, we get cryptic species complexes of insects where most species have no suggestions for observations, and the lack of these species being listed under “similar species” often creates false positives as they more often get identified upward as “this is out of range for Species A, but this complex also including Species B through Species K can’t be separated apart from microscopy.” So they likely will never get legitimate suggestions.
I disagree on this one, espeically with regards to excluding things. For the purposes of iNaturalist and users, I think knowing what mistakes are being made as is currently shown is more valuable than what mistakes someone thinks could be made.
I agree with Chris… I am often recommending to people to reconsider IDs, and if they do, to LEAVE their first ID present as a withdrawn ID rather than deleting it, so that it adds to the “similar taxa” data. In other situations, I recommend to DELETE the errant ID if it is a situation of just clicking the wrong one in a list etc, ie there is no similarity involved.
Twice ppl have mis-ID a locust for an ant. I’m guessing it must have been grasshopper hatchlings? It would be useful if one could click on the tiny ‘2’ next to the ant and have pop up a clickable list of the obs where the mis-IDs happened. If anyone else agrees I’ll make it a feature request.
The link for the screenshot is: https://www.inaturalist.org/taxa/119211-Dictyophorus-spumans
(you might have to filter location to Southern Africa to see the ant)
I don’t think similar species need curating at present, but I suspect that, in time, the list of mis-IDs will grow to the point where the similar species page will be so full of far-fetched mis-IDs that it becomes hard to use.
