When identifying organisms (mostly birds, for myself) I sometimes come across observations where the photo itself is unidentifiable (due to distance or camera limitations), but the notes associated with the observation add some insight into the probably ID of the organism. Sometimes the photo is just a landscape photo, with the notes that the organism in question just flew off, and a brief description of the organism.
Should identifiers take these notes into account when identifying organisms? I know that drawings are accepted as evidence of an organism, but what about descriptions? Or links or video taken by the observer? Or links to other media taken by another person, which shows the organism in question better?
Bottom line is, should an observation ever be made research grade based on evidence not visible in the photo (or audio) uploaded to iNaturalist? This seems like one of the gray areas where people will have differing opinions, so Iām curious to see how most people treat this issue, and what the site staff would like people to do.
I think it is perfectly reasonable to use the description included. Is it possible to downvote the āincludes photoā criteria for Research Grade status? That could be a solution to the issue you see.
Iām happy to use comments about features that are hard to photo. They indicate that the observer knew what to look for and probably did see the feature. I used the descriptions mainly with plants where small details are needed but hard to photo.
However, I would probably draw the line at using only the description. I mean, a habitat photo without the bird but with the bird described? Thatās going too far to me, as far as iNaturalist records go. Verbal descriptions are accepted by some other venues and I think they should be posted there. One of the good things about iNaturalist observations is that they provide material that others can see or hear for themselves to judge the accuracy of the identification.
I do think thereās a gray area where the photo is suggestive but not quite enough by itself. (Too distant, bird against white sky, a little out of focus, some of the needed traits not showing, etc.) There, some description could tip me to agree rather than just move on, especially if I know the observer knows what to look for.
In my opinion, any evidence that can be used for that organism should. Itās up to the individual to weigh the reliability all information given. If you know something about the specimen, it doesnāt make sense to me to deliberately not use it to help with the ID just because itās not visible in the photo.
EDIT: Because it was also mention whether or not to trust third party sources, like media taken from another person, here is my thoughts. As long as you know that it is the same sighting or individual, it is fine to use that information. Like you can tell that all the other birds from both sightings are in roughly the same position, you know it was at the same place at the same time, and the better picture can provide evidence for the worse picture. Same thing with individuals from different sightings. If someone did not capture the diagnostic features that would identify the species, but it has a distinctive scar or something that would make the individual identifiable, as long as you are certain it is the same individual, you can use the other information as evidence. As stated previously, it is up to the identifier to assess whether this information is trustworthy, but if found so, there is no reason not to use it as evidence.
My issue with using an observers description is that I have seen descriptions that donāt match the observation. Subtle features that are subjective are often misinterpreted, and in extreme cases Iāve seen people describe size, shape and color in ways that were definitively incorrect based on the images uploaded. Even very experienced observers make mistakes.
I think that is grounds to take descriptions with a grain of salt, but not to throw them out entirely. Humans make lots of mistakes. I think it is a judgment call. But also, I just started using iNat this week, so maybe donāt take my advice.
If the feature is too subtle, then definitely take with a grain of salt if someone claims to have seen it. But if itās a more distinctive feature, they just failed to get it on camera, I feel like you can trust it. Itās really a case-by-case thing thatās up to the identifier what they can trust. If you believe the person did actually observe that feature beyond a reasonable doubt, thereās no reason to not use that as evidence.
It also kind of depends on whether it supports or contradicts the most likely ID. Like from a blurry photo of an orangish hummingbird, if someone mentions a seeing a male hummingbird with a mostly orange back, and they are far into the interior US, itās kind of safe to use that information as supporting evidence for a Rufous Hummingbird. But itās from Santa Cruz Island, then it becomes a bit harder to trust that same evidence. It might be true, but it is harder to accept. Like it doesnāt just support the ID, it supports the other evidence too. If you have a lot of evidence supporting each other, itās a good sign that itās the right ID. But if you have a lot of conflicting evidence, itās harder to know what to trust, so itās better to just rely on just what you can directly observe in the photos.
It depends. If the organism is visible in the photo, and I know that the field mark is reasonably obvious, then maybe. Problem is, a lot of the stuff I do is difficult for an expert to judge, Iām not inclined to take the word of someone Iāve never heard of that they judged the marks correctly.
Sometimes. I saw a blurry photo posted that was IDād to one species, but the visible parts of the photo gave a strong indication that it was another species. Thing is, this had a note that it was IDād by a pretty famous expert. I still sent it back to Subgenus due to what little was visible. Even experts make mistakes.
True. A big one to remember is while a lot of people do put effort into their ID on their observations (thank God for those people), many are still making an ID off of an online key they found an hour ago, and wonāt understand some of the nuance (Iāve had people point out how long the malar space is on something, missing just how small the difference is, it takes practice to use some of these marks). And then of course thereās the oh so fun argument: āI see these all the time, I know what they areā (the fish and game identifiers probably all collectively groaned).
These are where a lot of descriptions for IDās come from. So I take it on a case by case basis.
Some descriptive info may be very important for identification, such as size for small invertebratae. Or some context info like the substrate it was found on or if certain host plants of insects are present in the location.
Sometimes I do even ask for some descriptive info not visible in the photo, to get the ID.
I think itās fair to use the description as supporting, but not primary, evidence. For instance, iNat does allow medialess observations, but does not allow them to reach Research Grade. Uploading a photo that does not contain evidence of the organism should not be a loophole to allowing observations that arenāt supported by media that are evidence to reach Research Grade. I use a downvote in the Evidence of Organism DQA field for media that donāt contain evidence of the organism (like perches that used to contain a bird, etc.). If Iām not sure about where the organism is (eg, I might be missing it), I leave a comment asking the user where the organism is. Iāll wait a bit for a response before using the DQA. If observations that donāt have the focal organism in the media are not addressed with the DQA, the media could be inappropriately used in training the CV, etc.
In some observations where the focal organism is not present in the media, I also see users IDing other organisms that are present. For instance, lots of times the CV (often via Seek) will ID a tree or plant as an anole (āthinkingā an anole is perched there when it isnāt). Sometimes IDers will then ID the plant. I think this is fine too, especially in cases where the observer has been inactive for a long time. But itās possible that the observer really does want the observation to remain for the (unevidenced) focal organism, in which case it is best to use the DQA and leave the observation as Casual.
We try to take descriptive notes into consideration. For example, with an audio-only observation of a call that could be Red-shouldered Hawk or Blue Jay imitating it, a description from observer of the bird emitting the call may sway us.
Also, if the potential ID species is relatively common, we wonāt feel too worried about identifying it at only 90% certainly. For something rarer, the description may not be enough to sway us to ID.
Link to video observer took we will take into consideration and will leave comment with ID that says, ābased on video.ā Link to other observerās media with same time stamp we will also take as evidence.
Notes are invaluable.
E. g. when IDing grasses, the length of glumes, lemmas, spikelets can be distinguishing characteristics.
Glandular hairs may be easily visible with a 10x lens but not with the phone camera lens.
A short description of the biotope is useful.
If someone gives such a description, I would definitely take that into consideration in addition to the photo.
That is so true. My only observation of a white pelican was on a walk without phone or camera. So even though it will forever be a casual observation, it is a nice record for myself.
Am I using Notes wrong? I donāt typically put things that are visual, but things that would not be visible in photos, such as size in approximate mm or a relative comparison (āsimilar size to A. melliferaā), behavior (skittish, fast direct vs awkward slow flight, deliberately visiting each blossom on a plant, etc.), whether I saw one or more of the same, the plant it is on if not readily obvious. If it is an unknown bee and I hear buzz at the blossom I put that.
If it is a plant that has grown, I try to note how long since it appeared to reach this growth point, location within the garden (shady corridor, etc), insects on it.
Youāre making a good case for evaluating the supplemental information carefully which, of course, is always appropriate. That shouldnāt be an arguement to ignore or summarily discount observations (say, with poor photographic evidence) that offer such context.
We use notes on our obs as we choose. I remind myself of field marks I am learning. Catch details in photo, noted as - Usually yellow, but - for example.
You are not wrong! Not adding supplemental notes is perfectly OK. However, adding such notes is also OK. There are cases where it would be really helpful for identification and then if I did see the feature I try to remember to mention it, but often I donāt.
This is one that Iāve wondered about before, and in fact I just suggested an ID based on this today. A student and I both posted an observation of a fly that I was holding in my hand, but my photo came out clear and identifiable, while theirs was out of focus and not identifiable past genus. I suggested a species-level ID on their observation and linked to my observation of the same fly: https://www.inaturalist.org/observations/237642686
So technically, their observation is not identifiable to species-level based on what they posted, but the organism is identifiable to species-level based on someone elseās (my) picture. Iām not sure if there is any āofficialā policy on whether I should be adding an ID to their observation in a situation like this.
On a similar note, I sometimes see IDs on clearly unidentifiable observations by others who were present when the organism was observed, with comments like āthis was [Species X]; I was thereā, but without any links to additional media to support the ID. (Iām sure Iām guilty of doing this myself before.) This seems like maybe not a good practice, since the claimed ID canāt be verified by any media?
Iām curious what othersā thoughts are on these two scenarios.