If the ID (here, for plants) is based on inflorescence characters, NOT range, do you still see a reason why the observer would want this?
1 Like
My response as an identifier: if the subspecies is identifiable from the documentation provided, I think identifiers should indicate the subspecies, regardless of what the observer “wants.”
My response as someone who uploads observations: Given the new way iNat handles subspecies identifications with regards to “Research Grade” status coupled with the huge number of observations misidentified at subspecies level (but correct to species), I understand not wanting someone to add subspecies on this platform.
7 Likes
One of my least favourite parts about subspecies is how debates over species vs subspecies labels will not allow the species to become research grade – even though both species and subspecies internally agree with the species label.
I would love if iNat could have a way to treat subspecies IDs separately, ergo if everyone suggests the same species it should allow data to be useable at species level even if subspecies is debated.
9 Likes
A typical situation from this observer is that an identifier has already confirmed the species, making this RG. If I then add the subspecies (and another identifier confirms it), what is the downside? The complaint was that the observations were being “turned into” subspecies when this subspecies is the primary one in the area.
Assuming they understand how the system works, it sounds like maybe a personal preference/stance. Like personally I get annoyed when people identify my rock pigeons as feral pigeons because it seems so unnecessary. It’s not logical on my part, but we’re not wholly rational creatures. I suspect this is similar, but you’d have to ask them directly. Everything we say would be speculation.
7 Likes
Can you give an example of what you’re thinking of? In my experience, subspecies IDs don’t affect Research Grade status once it’s RG at species level.
If the ID started at subspecies level, then it’ll stay at Needs ID until someone confirms or denies the subspecies. I’d agree that this is awkward but it’s really because there’s no signifier in the UI for a better option: you shouldn’t push the observation to Research Grade at subspecies if the subspecies isn’t confirmed, but if the observer wants their observation at subspecies then it wouldn’t be good to have the species-level IDs overwrite the subspecies.
1 Like
I understand it to be a preference, yes. I’m just trying to understand whether there are practical implications that I haven’t known about. I didn’t get an answer to my comment on the observation about my reasons for adding the subspecies.
1 Like
Sorry. I worked on lots of observations and can’t pull this one up, especially as I removed my ID’s since they were causing such a problem. If/when I run into others I’ll try to remember to send the links to you.
1 Like
The single subspecies ID without disagreement only prevents RG if it’s first. If the first ID was treated like the others, there wouldn’t have been the issue that they recently tried to fix.
1 Like
Perhaps we need an Opt Out of ssp choice - but only for each obs. (Then mine can be Needs ID at ssp regardless of who yells FIRST!)
I think this kind of exemplifies something I’ve tried to explain on a number of occasions when I’m confronted about subspecies ids. People are often comfortable with specie level ids but suddenly get squirmy or even passively aggressive when it comes to subspecies. For example, I have no problem with identifying Northern American Kestrels (ssp. sparverius) without adequate morphological features because it’s the northernmost subspecies, so in most of its distribution, it is going to be the only option… ever. The only time I would reconsider such ids is when you start getting into the southeast US, where paulus breeds but sparverius winters. Nevertheless, I will still get messages stating you can’t identify by distribution despite the improbability of another option. Yet, they would be perfectly okay with their Black-billed Magpie ids, when there is no morphological distinction between them and Eurasian Magpies. I see a theme that distribution is often shut down as a means of id until it involves a species-level taxa.
7 Likes
Some differences between the magpies are described here, although they seem pretty subtle and subjective. But for sure essentially everyone identifying them is going by range, same with American and European Herring Gulls and many other species. I think opposition to geography-based subspecies ID comes from a combination of discomfort with geography-based ID in general and discomfort with subspecies taxonomy together. Since species are the basic unit people are forced to be more pragmatic more often. More discussion of this here.
5 Likes
Yes, it’s completely subjective. There’s a reason why there’s no ABA record of Eurasian, and it’s because even if an observer says what they perceive to be a short-tailed, shorter-winged, etc magpie, you need to convince a rare records committee that your report is legit. I just can’t see it happening without a DNA sample.
2 Likes
Please keep in mind that us botanists treat infraspecific ranks completely different to zoologists. I constantly see arguments from zoologists saying that subspecies are essentially irrelevant, and this is absolutely not the case in botany. (I’m not implying that you said that, just using it as an example of the difference).
5 Likes
There are definitely relevant subspecies in zoology; some species look quite different in other parts of their range and they deserve their own category (although I expect most of the obvious examples of these will eventually be split into their own species…). But at least with North American birds I get the impression there was an excessive amount of new subspecies being described at some point in time and no one has cared enough to clean them up. eBird often lumps them into more identifiable informal “subspecies groups” which iNat doesn’t support currently.
The only plant subspecies I know is European Reed (Phragmites australis australis) which makes sense to me compared to the American subspecies. But sedges, asters, sowthistles etc. are hard enough to ID to species that it confuses me when someone randomly decides to take one to subspecies…
3 Likes
Yes, that is precisely what happened. In the 1890’s, the last few North America species were described (primarily grassland species in the northern Great Plains), and the rising generation of biologists basically got left with nothing. As such, there is a rather consistent trend of subspecies naming from 1900-1930. The American Ornithological Union was responsible for the reasonable curation of subspecies taxonomy, but when they acquired Central America and the Caribbean as part of the taxonomy in 1957, they decided to forego subspecies and focus on species-level decisions. And that in turn raised a generation of scientists that could care less about subspecies, and hence why we have a current taxonomy that’s 60+ years out-of-date.
3 Likes
In my experience, some folks don’t want the “common name” that’s displayed as the community taxon for the species to be replaced with the Latin name for the subspecies (since many subspecies don’t have a common name) when they’re reviewing their own observations.
Which is why I’m hopeful this feature request will someday be implemented:
https://forum.inaturalist.org/t/display-common-name-for-species-when-infraspecies-e-g-subspecies-lacks-one/59485
3 Likes
Speaking as someone who generally likes subspecies and tries to learn the scientific names anyway, I still 100% agree with this. The way the site displays subspecies definitely needs some kind of update if subspecies taxonomy is going to be used.
2 Likes
Agreed. I’ve argued elsewhere on the forum that if range were universally discounted when identifying, an unacceptably high percentage of insect observations on iNat could never be ID’d to species level. Many if not most insects have virtually-identical doppelgängers somewhere in the world, and “this is Canada, not Norway” is the only way to make an ID on them without DNA or dissection. If someone’s standard is “range can’t be used to make an ID”, then there are millions of insect observations they can start bumping back to genus or family level.
7 Likes
In my experience using range to ID is only an issue when the location precision bubble (what is called ‘error’ even though it’s really precision) is so large as to encompass several lookalikes. This mostly happens when oberservers use broad named locations (e.g., Panama) in which case the ‘error’ bubble contains the entirety of that named location. If accurate location data are provided to within a few km then I find using location to ID is not just helpful, but a must (especially when trying to ID to ssp.)
3 Likes