I agree with what most others have already said. Also, even if you are unsure of an observation, adding an ID will help experts of certain fields find it easier. For example, if I want to ID some fish I will search “ray-finned fishes” as a filter, instead of searching through everything
3 Likes
This!
I had been at the position @gcwarbler is describing for quite some time, even with taxa that cannot confidently be IDed to species in some regions at this time without certain traits that are usually not visible in photos. “Who the hell am I to determine what might be possible in the future?!”
But I also grew to come to the conclusion that using this DQA does not mean condeming those observations from being more accurately IDed if ever possible. It is actually a quite easy tweek of the filters to find those observations again and go fix them, if ever new information comes along. So I now also use it more frequently to clean the “needs ID”-pile.
However, I agree that one should be reasonably sure that more cannot be done AT THE PRESENT TIME to push an ID further and not hit that button too light-heartedly.
3 Likes
The bigger problem is leaving the ID as unknown - those tend to get ignored. I would say - just get it as close as you can as many will never get IDd to species from a photo anyway. If you know it’s a mushroom - put that as a minimum.
3 Likes
Thank you, everyone, for your thoughtful replies to my question. Very helpful.
This is a large part of what I use it for. As an example, we often get photos of spider webs, with no spiders in them. We can often make a good guess as to what family of spiders made the web, but there is no hope of making a specific ID in almost all cases. So IMO there is no reason to keep those as “needs ID” especially in a group that is woefully short on knowledgeable identifiers. The problem is that the “backlog” of needs-ID observations is (for several years now) growing at a rate that makes it impractical for anyone to sort through, so tagging things as unlikely-to-ever-get-a-specific-ID is one of the only ways to reduce the size of that pile.
I usually use one of a handful of copy/paste comments like “Unfortunately there is likely not enough detail in the photo(s) for a more specific ID” when using that DQA, so as to not seem impolite :)
5 Likes
@lynkos – I find that iNaturalist gives me lots of good opportunities to practice being wrong and being OK with it (or pretending to be, which is good enough). I even thank people for the correction sometimes, though I admit that sometimes gratitude is not exactly what I’m feeling.
5 Likes
Oh yes, I hear you 
. I just wish it wasn’t so hard for people to be kind. Be kind. Please!!!
2 Likes
I agree with the above, it is fine to put your best effort at an identification on your own posts, and preferably explain how you got to that name. I would be more cautious about giving an agreement to one of your observations or anyone else’s, because to my mind, an agreement does suggest a strong level of certainty. It can push the observation to research grade and all the consequences of that.
The other thing is, keep an eye on your notifications so that if there is a disagreement, you can quickly review whether you want to stick with your original id.
Yes! This is important and can correct any mistakes.
I think we are at the point now where we can speak of the world’s largest, most comprehensive field guide – a collaborative project by thousands of contributers from around the world. Like the Hitchhiker’s Guide to the Galaxy, it is only available in electronic form because the print edition would require a warehouse.
Say that you observed a nondescript, brown moth while on vacation in French Polynesia. You bring up the Field Guide to Organisms of the World and first you offer your photo of the moth to its computer vision. The computer vision replies, “We’re not confident enough to make a recommendation, but here are our top suggestions.” As it happens, there is only one suggestion: Rhesalides curvata. So, you thumb through the pages (as it were) until you come to the page about that species. There, you see a set of illustrations:
You can see that this is a very variable species, ranging from boldly patterned in black-and-white to almost plain brown. But as you study the illustrations, one thing stands out. There is a thin white line – like a curved “v” or a flared “u” – outlining the black, and also present in the same place on all the intermediate forms, including the plain brown ones. That’s probably a definitive character.
So then you look back at your moth.
The picture is blurry, and it is hard to make out much detail. But if you look carefully, you can make out that same white line in that same place, besides the same wing shape. You make a tentative identification on that basis, just as you would if you’d been working from a print-edition field guide.
This amazing publication is not actually titled Field Guide to Organisms of the World. It is titled… well, I think you’ve all figured it out.
Look what we have built! We – all of us together – have created the most comprehensive field guide ever published! A field guide that encompasses obscure Oceanic moths and well-known North American birds; African large mammals, and also African reef fishes; and new species constantly being added. Such a field guide is an invaluable identification resource.
7 Likes
I appreciate your comments. I think agreeing to reach RG is pretty common, although “actually having the knowledge to agree” might have different meaning to different persons. For example, if you suggest an ID that I didn’t make on upload, sometimes I will research the requirements for ID on Bugguide or in my books. Other times, I’ll just visually compare it to other photos I or others have taken—yeah, it looks just like it and there are no other species in the genus in my area. Done. Am I an expert—no. But if the info I obtain is enough to have suggested an ID, could it also be enough to confirm? Theoretically, even a confirmation to reach RG will also be corrected eventually, if wrong.
It’s mighty tempting when you have a new species for which there are only a few experts answering your queries, even after waiting months or years, to just click “Agree.” I try not to do that, but I see many prolific observers who have a friend who always confirms their IDs. Like clockwork. I’m not questioning the correctness of those IDs, but find it interesting how those kinds of working relationships help them increase their species stats.
2 Likes
I agree, it is amazing what has been done here!
I frequently search already identified to RG photos for patterns. Then I go up to the genus level and sometimes above it, searching the icon photos for other possible look-alikes. Quality icon photos are really important, so I’ve done quite a bit of photo curation with moths in the Eastern US, so that users can scroll down the list and only see sharp, representative photos. More to be done in that area, IMHO, as newer and often better photos become available all the time.
1 Like
I also photograph moths in my yard in Virginia. I have found that with moths it is best to use multiple sources. There are many good ones; BugGuide, FieldGuide.ai (which will also suggest an ID from a photo), Butterfliesandmoths.org, and BAMONA are all helpful. And North Carolina has a great site for moths Moths of North Carolina (ncparks.gov).
Usually, if I ID something that I’m not sure about or there’s not enough information in the photos to be sure, I will add a comment describing which other taxa it could be or what additional info is needed. That way others can gauge how much they trust my IDs. But I would rarely say I’m 100% certain since there’s a lot of variability and nuance in plant ID.
But plant and insect ID seems to be a lot less contentious than vertebrate ID.
1 Like
Two minor thoughts about this (I think it doesn’t matter if it is about own observations or somebody else’s):
- In order to identify some image as an X, one has not only to know X, but also all relevant non-X that might be similar (a potentially infinite space of possibilities). E.g. I only learned on Inat that there exists a subfamily Malachiinae (Malachite beetles), which look like Caraboidae but aren’t (they aren’t even closely related). I see a lot of insects in the “unknown” where the comment says “bee” in the observer’s language which are hoverflies (it depends on 2. here whether it would have been better if the observer had entered Anthophila into the ID field or not).
- How much one is willing to risk when proposing an ID depends on how rapidly one can react to disagreements (if one is not a specialist, that usually means withdrawal of one’s ID – I did this always with one exception where my ID turned out to be correct (Lepidoptera vs. Symphyta, I had counted the legless segments on the larva and someone else later IDd the caterpillar to species level)).
I’m rather hesitant with being too specific in order to keep my notifications manageable (if I miss a disgreement, the wrong ID sticks until it is overridden by enough opposing votes), that’s why I create lots of “fungi” (I still get them corrected into slime molds now and then).
My strategy is to try to look at all notifications and then unfollow those observations where my ID was refined (confirmation by the observer doesn’t count for that purpose) unless I’m interested in the final result.
PS: The API page https://jumear.github.io/stirfry/iNatAPIv1_identifications.html?user_id=XXX¤t=false&per_page=30&page=1 helps, where you insert your username for XXX.
It’s getting to the point where I make IDs of all my observations, using the technique I described earlier. Because if I don’t – there just aren’t enough identifiers; they’ll sit at a broad ID indefinitely.
3 Likes
Most of what I observe around home I know to species or at least genus, so I ID the observations to that level when I post them.
But I just spent three weeks on a new-to-me continent (Australia), so I knew virtually none of the species I observed. When I uploaded my Australia observations to iNat, I looked at what iNat suggested and usually went with that, if it made sense. If I had physical field guides for the species (birds, some plant species), I’d look up iNat’s suggestion to check on the ID. I usually posted plants at the genus level, because even the two plant field guides I own didn’t include all of the species in the small corner of the country that I got to see, so I didn’t know all the options within a genus.
Luckily for me there seem to be lots of Australian IDers and mostly they have confirmed or refined the initial ID I gave. However, plants I could only ID as Dicots when I uploaded them (not many) still sit at Dicots, but it’s only been 10 days since I uploaded the photos.
All of this is a long way of saying I think it’s always a good idea to add as fine an ID as you (the general you, not Jason in particular) can to your own observations, even if you’re not absolutely sure - but then pay close attention to your notifications. When an identifier corrects a mistake you made, you can check further yourself and rapidly learn what the species really is (or at least, what it’s called in the iNat world).
3 Likes
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.