In many cases and for many taxa, subspecies identification is not widely discussed in the references available to amateur enthusiasts. This isn’t so much an issue of folks not identifying things to subspecies-level on iNat as it is the general emphasis on species-level taxa.
To my mind, the existing known range should stand until such a time as good evidence is presented to overturn it. The blurry image sounds like it isn’t good evidence.
I say this as someone who has extended the range of a species several times (and I continue to do so). I make sure I get a good, incontrovertible photo each time. I don’t try to say “Well, I’ve extended the range before, so clearly, the known range is still in flux, so this sketchy photo that kinda looks like it could be the species in question (rather than a similar species) should be good enough”.
Many iNat casual iNat users (and even some not-so-casual ones) don’t like their observations ID’d to the genus or complex level. They want them to be RG at the species level. They often submit photos that are very poor. If we don’t make use of known ranges, these observations would have to be left at the genus level (or maybe even higher). This will result in discouragement and possibly less participation.
I don’t see how this prevents us from re-considering known ranges if/when someone submits clear evidence of a species found outside the known range.
My point though is that we don’t use this reasoning for any other taxonomic level. Use the elephant example from mreith: should we stop at “Family Elephantidae” because you’re just supposed to know that Asian Elephants are the only species in Asia? Or with the moths, it seems like people agree just fine to ID the moths by location and not just leave them at genus. There are many benefits to having the ID be as specific as it reasonably can be.
I’m having trouble imaging a scenario where IDing to subspecies would obscure information. If anything, I would say that IDing to subspecies at all draws attention to the subspecies ID, and makes it more likely that an unusual subspecies find would be noticed.
As it is, most people do not know how subspecies are identified for most species. If identifying to subspecies was more normalized, people would have more data to work with regarding how to ID subspecies, so there would be more opportunity to catch unusual subspecies observations.
But I do want to reiterate that I’m referring to cases where people explicitly disagree with subspecies IDs even when they can’t give any reason why another subspecies could possibly be found in that location, when other subspecies are entirely on other continents. If we can’t ID by location in those cases, and there isn’t a visual difference to go off of, then the site really shouldn’t have those subspecies implemented at all: they’re just a trap that you’re never supposed to ID.
You make this statement
but don’t back it up with any examples. Yet you also state
when I gave multiple examples in my previous post.
There’s an important reason why
Because at those higher taxonomic levels, the observation will not reach RG on iNat and will remain in the Needs ID pool. A species ID moves the observation to RG and reduces the Needs ID pool.
Additionally, IDs for subspecies, more so than for other taxonomic levels, are more often based exclusively on location. This is the nature of how subspecies are often (but not exclusively) created/defined - for groups that have some minor level of variation which may be difficult or impossible to assess for most observers (requiring sequencing, microscopy, dissection, etc.). So while IDing by location alone is not limited to subspecies, it is more prevalent.
One of the points that I (and others) are making here is that this doesn’t work when species are not visually identifiable and have the potential to show up somewhere unexpected (e.g., birds). In such cases, identifying solely by location could obliterate data about the vagrancy of any unexpected subspecies. For some taxa, this might be notable/impactful.
You have raised some good points in this topic. Thanks.
If a species is split in a way that geographic “atlasing” will successfully separate the two newly described species entities, that seems like a much more efficient way to split the observations that individually having them all ID’d to subspecies. Individually taking all the observations in Location A to subspecies and then eventually raising the subspecies to species level seems like way more work than just doing a taxon swap that atlases all the observations in Location A to the correct new species name. If they’re going to be ID’d based on nothing but location anyway, may as well just automate it with an atlas and save time, no?
Where I could see a species split being hampered by lack of subspecies IDs is in a case like the North American Willets. If “Eastern” and “Western” Willet are ever split, it will be impossible to “atlas” the observations correctly, because all along the East Coast, the Willets in salt marshes in the summer are virtually all breeding Easterns while the ones on the outer beaches in the winter in those same counties are virtually all overwintering Westerns. This is a case where I’d argue that subspecies IDs add something, and will be helpful if a split occurs. Plus it’s worth considering whether vagrant Willets in oddball places are from the Eastern coastal population that’s supposed to go to the tropics in the winter or the inland Western population that’s supposed to winter partly along the temperate coast. A split here could potentially mean lots of observers and places get a “new species”, despite the majority of the IDs being largely based on geography/phenology/habitat rather than appearance of the organism itself. I see this as very different than “let’s label every Monarch in Canada as the migratory subspecies because Canada”, which I don’t see as a worthwhile way to spend time. I’m sure there are other more complex examples like the Willet out there, which are where I’d spend time if I were looking to do subspecies IDs.
Right, I agree that there are circumstances where you can’t be reasonably certain enough to ID subspecies, because subspecies ranges can overlap, and in those cases you need to leave the observation at species (or higher if necessary).
Since this has come up a couple times here, I’m concerned I may be being misinterpreted: I’m not suggesting people should only ID by location, when there are physical characteristics that can indicate a different ID. Of course we shouldn’t be using just the location to obliterate well-reasoned IDs based on other factors.
Keep in mind, the current default for 99% of IDs is that the observation doesn’t get ID’d to subspecies at all. Usually when someone IDs by location they are doing so because there is no other way to reach that level of identification, not because they are disagreeing with an ID made some other way.
Thankfully…
Obviously identifiers show distincts behaviors here, some will refrain from any IDing if there are no ways to ID based solely on the photo, a minority might even ID to the genus (which can be a bit frustrating if the original ID was at species level, and the location is in the range of this species). But most will agree with a species/subspecies ID in its known range, it seems.
On the other hand I get complaints on a regular basis because I’m obscuring location for all observations made on a private plot of land, and some users say that they can’t ID some insect species “without location”…
Just to add some more complexity to the question ;-)
Sounds like you did not obscure those observations, which should give the IDer usually enough information about the location to ID in most cased, but set the location to “private”, which gives the IDer no way of knowing even a world region of the observation.. I ID for example Pisaura spiders, a genus with many lookalike species, sometimes even cohabiting the same area. If I do not know the location I will need to push them back to genus..
Which brings me back to the original question. I basically agree with those saying that one has to find their own way of dealing with location based IDs.. for me it even varies with where and what I am IDing and how strong the current knowledge is. Pisaura is actually a great example - a rather well studied, often observed genus. The information on Pisaura mirabilis in central Europa is rather strong and there is a rather active community of arachnologists in certain parts of Europe. There are a couple of species around the Mediterranean area that cannot be reliably distinguished by photo alone, so we need to leave the ID at genus level. For example in Southern Italy there are several species, while in the North there seems to be only mirabilis. The body of literature on the genus in Italy is large enough for me to draw a rather clear line within the country even, up to which I feel comfortable to ID to species level. The picture is completely different when it comes to other countries or regions of Europe. E.g. in Ukraine it is known that a second species is present east of the river Dnepr, while there seem to be no record of this species on the western side.. but there are many records for Romania for example.. so I would assume the species will also occur west of the Dnepr within the Ukraine, filling this regional gap. I do ID accordingly conservative there.
Knowing some likely distribution borders like rivers, mounains etc. will also complete my interpretation of the known literature .. sometimes, when a species has been recorded from one side of a mountain line, but not the other, even though the locations are not that far apart, I might deem species ID still plausible enough until further evidence to the contrary is given. There is for example the case of Southern France (2 species, only genus ID) and northern Spain, for which only one species is known and I will ID it as such. the Pyrenees make it plausible enough for me, despite a distance of less then 100 km between two locations.. for which the second species is known versus it is not recorded.
For different reasons (like special knowledge about terrain or species behavior and such.. or maybe just varying degrees of boldness) different IDers will differ in their comfortability of IDing to species ( or actively disagreeingg) in such cases. .. which is fine. In the end community ID will decide and the absolute final desicion will be with the people using such data for their science.
What IDers decide here is not the end of the world either way 
I agree and would also note that a split with atlasing will likely be done anyways, unless all observations are IDed to subspecies. If there are some that are not, the atlasing will be done anyways. Atlasing will also be more consistent.
Information and prediction models can handle uncertainty, and prior probabilities are incredibly relevant in countless real world informational applications. Location contributes uncertain information, and should be included in models that include confidence estimates.
The problem with iNaturalist currently is that it lacks an actual uncertainty statistic associated with IDs and consensus IDs. This has been advocated for, but last time around we couldn’t get past people’s personal sensitivities about having confidence scores potentially assigned to their IDs 
lol I wrote ‘obscure’, not ‘private’, I’m not sure why you would say that…
Maybe I was not clear enough: the point is that some identifiers base their ID on the exact location even for animals, and this particular identifier is not happy with the wide area covered by obscured observations on iNat (most other platforms I know are using a 1 sq km grid by default, and protected species use a larger grid, or are hidden on some very local platforms)
I personally don’t think it’s a great idea to ID based on location when both ranges of similar looking species/subspecies are so close that they cover part of the same rectangle used by ‘obscure’.
And I find it even worse to complain to the observer about the fact that the location is obscured, btw!
I am so sorry, I did not want to be annoying here.
I honestly thought you might have confused the setting there, because I actually agree with you that it seems to be a weird IDing process if even an obscured location makes a difference for your suggestion.. it seems to be a very fine-grained exclusion process.
I am totally fine with obscured.. even with private actually, it´s the persons choice. But in the case of private without any further information just don´t expect too much from the IDers.
It depends. Some similar species might have overlapping ranges, but be very localized to certain habitats (or in some places, specific elevations). An ~27x27km rectangle is a pretty large area, and can contain a multitude of different habitats.
I ignore virtually all obscured observations. I don’t complain about them, I just filter them out. If I get tagged on an obscured observation, I point out that I don’t interact with obscured observations. Some of the responses I get do sound a bit like the observer feels I don’t have the right to choose which observations I will ID. So there’s enough unreasonableness to go around.
Wow. You have every right to use the obscure feature, and I have no idea why anyone would go out of their way to complain about it. That sounds like bullying. If someone isn’t comfortable IDing without knowing the exact habitat, they can just mark reviewed and move on. 
Unless you meant the complaints are coming from people you actually tag asking for help, in which case I think it’s reasonable for them to leave a comment acknowledging they saw the tag but aren’t able to ID.
I sometimes use iNat hoping to get an expert ID between two similar looking species.
Especially when I think something is out of its normal range - very important to get it right.
In this case, I add a comment like
“This ID is important for range research, no guesses please” or similar.
Agreed. Personally I ignore all “Private Location” obs when IDing insects, as nearly any insect species on any continent has a lookalike somewhere else in the world. I usually ID Obscured obs, but there are certainly times when the precise location (or precise date) would impact my suggestion. Insects can be hyper-specialized in microhabitats, and often a photo that I’d confidently take to species in one habitat I’d leave at genus in a habitat that could have other lookalikes. Not to mention that for aquatic species, the Obscuration could make it unclear if it’s from salt or fresh water, or in mountainous areas it could be unclear if it’s from an alpine rock slope or a hot sagebrush flat, etc.
That being said, I’d not just randomly comment about the obscuration if no one asked me. But if you obscure a location and then ask me for an ID that I can’t provide without habitat details, I’ve got to say something…
I find this article both entertaining and informative on the subject of identifying by location: ICYMI: Indeterminacy « ABA Blog
Thankfully, Thayer’s Gull and Iceland Gull have since been lumped. However we still have the three subspecies (and apparently a subspecies hybrid) that people can try to force individuals into, despite L. g. kumlieni probably just being a hybrid swarm between the other two subspecies.
There are lots of cases where assumptions like this need to be made. American/European Herring Gulls, Eastern/Western Wood-Pewees (and now Eastern/Western Warbling-Vireos), etc. Gulls are particularly frustrating since they’re so capable of vagrancy. When I see crows I assume they’re American Crow most of the time, but in certain places in my area I have less confidence because there are Fish Crows around. Fish Crow could occur away from those places but the chances go down dramatically. Generally outside of areas where a second species is remotely possible, you don’t suspect it unless you see a particularly extreme phenotype of the unusual species.
This starts to become an issue when the range limits or intergrade zones are unknown.
If the observation is within reasonable travel distance of another species, then I would consider that in the probability. A European Herring Gull is marginally more likely on the coast than inland North America. A previously undocumented introduced millipede species is more likely in an urban area with lots of potted plant imports, adjacent to another state which already has that species, than it is in a rural area several states away.