Over the last few days, I collected some minor “stats” when changing things from “unknown” to coarse clades (usually as broad as “spiders”, “anura” etc. as I rarely know better):
People love frogs, beetles, spiders, snakes and butterflies - many of the things I changed were IDd (or at least specified further) in minutes
Nobody loves weeds, so labelling them monocots, dicots etc. doesn’t seem to help (I didn’t change any from somebody else, I have enough thereof in my on observations). Fungi and Hymenoptera also are refined rarely (maybe because there are so many of them and my label still is almost as broad as “unknown”). OTOH, things i labelled “insects” sometimes get improved to “ants”.
Some people seem to first uplod lots of pics as “unknown” and later (on the same day) add the exact species, probably to save time or online data volume when outside of their Wifi. To avoid unnecessary labor, I’ll look only for unknonws that are a few days old from now on.
It does help… a little. There are people who are looking for non-vascular plants, or for monocots… so any preliminary sorting will eventually be appreciated.
also I do love weeds. their resilience and ability to carve out a new niche for themselves is cool.
edited to add:
also, it will notify the people who made earlier IDs, and I find that they often take notice and refine the ID further!
Please continue to refine fungi. The issue is less that people arent IDing there just arent that many fungi IDers, theres tons of fungi that just look vaguely blobby and are hard to ID without specific pictures, and a lot of folks just… dont know what to photograph to get a species level ID.
From an outsiders’ perspective, but one who knows some mycologists, fungus taxonomy just seems to be in so much flux that it’s pretty daunting to even an attempt a fine or semi-fine ID unless it’s something like Amanita muscaria. They sure are cool, though.
Well. That too. But even then with the stuff in flux you can usually get it down to at least like, family. Usually.
Edit: like if I’m going through top level fungi obs i usually try to knock even the obscure ones down below lichens so its a little easier to filter that out.
Edit 2: i know theres a couple different DNA projects across various states for mushrooms too, Ohio for sure has one going, so that may help in the future refine some of the more obscure ones.
Non-vascular plants are essentially mosses (which I think don’t exist cladistically anymore, I haven’t seen them here anyway) and Red+Green Algae. I didn’t find them in the unknowns (maybe there are some in the microscopic images, but I don’t recognize them anyway). Since Brown Algae aren’t plants but look like them, I keep my hands off all of those to avoid errors.
PS: I just found another reason for me to not change unknowns into “weeds”: 90% of unknowns are dicots anyway, so if anyone wants to look at plants, just selecting unknowns is a place to go.
Unrelated question: is there a way to mark somebody’s observation so I get messaged if it gets an ID? There are weird beings out in the world and I’d love to know what they are, e.g. https://www.inaturalist.org/observations/158390598
I might add a second ID as “unknown” but that seems silly.
Fun detail: for https://www.inaturalist.org/observations/159140536, the auto-ID recommended Cochineal. The picture is some Opuntia with no insects whatsoever, but it seems that more pictures were uploaded that looked like this with the parasite being the interesting item. This reminds me of AI-misidentification of X-ray images where the prospects of patients had to be predicted by an AI, but that just found out that bad cases were treated in another hospital with another X-ray-machine than the easy cases and the AI detected just the machine type.
Are the fuzzy white bits around the thorns not a symptom of some kind of infestation? The fuzzy bits are almost certainly what the AI thinks is evidence of presence for Cochineal; if I scroll through and pick a random nearby Opuntia observation without those fuzzy bits, it does not suggest Cochineal.
I think I found the champion of the lazy observer contest: A tree labelled with a sign describing the species in latin and the local language and the familiy of the species, observation labelled as “unknown”.
I am reasonably confident that is what the CV is seeing; I have no idea whether or not it is right. I have for sure seen the behavior you are describing though. For example probably like 90% of acer glabrum (rocky mountain maple) observations are infested with at least a small patch of maple erineum mite (aceria calaceris), so the CV always suggests maple erineum mite for rocky mountain maple observations, even on the handful that happen to have no trace of it (like this one https://www.inaturalist.org/observations/166146288), because it is hard for it to generalize the difference when so much of its training data contains both.
Checking only the box “benötigt ID” in the section “Qualitätsstufe” should limit the results to observations that are not captive. Is this not happening for you?
Observations are only “needs ID” if they are not casual (Hobbyqualität, i.e., captive or missing data) and not already “research grade”.
I just found a phenomenon in the ID module that might be relevant for everybody who does IDs: If one identifies things, they fall out of the filter criteria and items from the next page shift down into the current one. This means that if one is done on one page and goes to the next page, the first few items are missing because the went internally to the page on has just left. Pressing Ctrl-r before going to the next page avoids that, the downshifted items are now visible near the end.
