While researching, I’ve come across species that are difficult or straight up impossible to ID without DNA. The most prominent for me being Bombus terrestris workers mixed with the B. lucorum complex, where terrestris is extremely similar to the near identical 3 species in the complex; the Eratigena atrica complex, where the the genitalia are needed to ID; and of course Taraxacum, where many minute features are needed for ID of alot of species. (however I’m not as knowledgeable when it comes to them.)
However, on the website, many species are bumped to research grade to the species, likely inaccurately. These species are common in Ireland and the UK (among others) and many people are IDing them as such species without the knowledge that they may be one of the others. I feel this is unrepresentative of the true numbers of the species themselves and I have thought about going through each observation in Ireland (at least) to bring them to a more accurate ID.
Is this worth it? Would I just backpedal the data?
With the bumblebees at least, I know the UK has it’s terrestris ssp. audax that is somewhat easier to differentiate from the lucorum complex, and the terrestris queens are confidently IDable. But it’s the ones that are like the complex bees that are the thorn in the side and have been clumped within one of the species, mostly terrestris and lucorum. For these individuals, the safest bet for them is to put them in the subgenus Bombus.
I don’t think its backpedaling data if you’re fixing potential inaccuracies. If there’s enough information to say that its more likely one lookalike species than another, that’s one thing, and I think species level IDs would be fine to leave at that point, but if there truly aren’t visual differences in the photo - leave it at genus.
red Russula mushrooms in the US are a big issue with this - Russulas are everywhere, they’re very common, there’s a bunch of different species and they all look the same and can basically only be IDed via microscopic examination of spores, DNA testing, or tasting them (and well, unless you’re really into mushrooms you probably don’t know that taste can be an important diagnostic.) If I ever want to do some mindless IDing I’ll just sit down and bump red russulas to genus with a note explaining why, and then mark it ‘good as it can be.’
unfortunately this will continue and get worse as iNat has really jumped onboard with the ‘splitter’ taxonomic framework where a lot of entities that can’t be defined in the field are elevated to species level, and there’s been a flurry of these lately. That being said it’s always been the case that some species can’t be distinguished in the field or from a photo, or can only be distinguished at certain times of year or via being dissected. I personally support creating species complexes when there are common cases of this issue, that ‘lump up’ to the highest level someone can reasonably identify in the field. While this is somewhat controversial there are others doing this too and they have been added for some taxa. Of course, it can be difficult to decide what is reasonably identifiable in the field, and creating too many species complexes could cause problems as well.
Explicitly disagreeing as you describe can be appropriate if you are confident there is no way the ID could have been known, but that can get sketchy because sometimes people do that when they can’t tell without realizing someone else might be able to tell. Also people use different taxonomic frameworks, as with dandelion where no one really has a workable taxonomy for that genus so everyone seems to do it differently, and forcing people to identify to microspecies level doesn’t sit well with me. So i tend to steer away from explicit disagreement unless I think the ID is wrong. But that’s kind of a controversial subject on iNat. I’d say if you plan to do a lot of explicit disagreement IDs, even justified ones, do be prepared for some pushback.
Ideally I would hope people only do it for species they are sure on. I generally avoid hitting ‘good as it can be’ for taxa that I’m less familiar with.
EDIT: Maybe because mushroom taxonomy - especially in the united states where so many look-alike species are misidentified as european species - is an absolute giant mess exacerbated by the ephemeral nature of mushrooms, I don’t know if I ever see mushroom experts (more expert than me) really bring up inats splitter policies? It seems more of a general acknowledging shrug at the difficulty with a push to collect and document modern specimens for herbariums so they can be properly examined and tested.
I don’t call myself an expert though, just a passionate enthusiast, so who knows.
well i don’t know mushrooms as well as plants and also have unpopular opinions on taxonomy (at least among taxonomists). Fungi are just so hard since all we see is the fruiting bodies most of the time. I often wish we had a good way to take fungal inventory of wetlands but because documenting only fruiting bodies is so biased, you’d really need soil genetics to do so and that’s expensive and inaccessible in most cases.
Yeah, outside of a few genera like Armillaria that have very distinctive rhizomorphs, its basically impossible to ID without the fruiting body (or as you said stuff like, soil genetics, and that seems like an almost sisyphean task)
Its also hampered IMHO by the United States completely reductive policies on Psilocybe, since that means if you are in a situation where you get stopped by authorities and have dried mushrooms on hand, you’re then left having to convince them that no, actually, they’re just for research, not drugs. Its not a huge worry for me but I can’t say the possibility hasn’t crossed my mind.
And with mushrooms, coming from an amateur, they are quite a pain. I’m sure with experience with IDing them, it will absolutely be easier, but right now, many look very similar, even with online ID guides, and I’m better off letting someone more experienced have a go at them. But I more or less know what features to photo for ID, so not all is lost. :)
I use iNat species distribution data frequently for building species distribution models and work closely with many other researchers who do the same. In my field at least (entomology) , where confident ID’s from iNat pictures are frequently difficult, there is a well known culture among researchers that research-grade observations of insects should be taken with a huge grain of salt. While I think iNat could potentially improve around the messaging and UX that always pushes ID’s to the species level at all opportunities, I think most researchers who use the data downstream are well aware of the shortcomings of the research-grade tag, especially for morphologically cryptic taxa.
For example, I recently worked on a project to model distributions of Thaumetopia moths - there are many Thaumetopia observations on iNat, but I wouldn’t ever be confident in the species if the picture was a blurry cellphone pic of a larvae
If it bothers you, and you can sort it out, go for it.
But do have a copypasta ready, and perhaps work in tandem with one or two others, so you can make visible progress together.
I absolutely and full get that there will be species that can’t be IDed from field marks alone, and I have come to accept that. I personally have been able to treat complexes, subgenus and genus IDs just as valuable as species IDs when the species can only be IDed with dissection and such.
I’ll probably stick to the taxa I’m sure about, like the bumblebees, when it comes to making these corrections, as I know what to look out for when confirming IDs, however, with taxa like the dandelions, I don’t know the full extent to the confirming features, and will include notes saying I’m not an expert and am mostly spreading awareness for the complexity through the genus IDs. I probably wouldn’t touch that genus unless it is one of my observations or it shows up on my dashboard.
I was thinking that researchers would be cautious when using research grade iNat data. Even when they do become research grade, someone may come along and prove you wrong, or you may prove yourself wrong on your earlier observations. I know I have.
The giant house spiders (E. atrica complex) is fairly interesting for two reasons; one, it was lumped into one species (“Tegenaria gigantica”) before being split again based on microscopy - so it is still recorded as a “senso lato” in many cases and most to all old records are treated as that “senso lato”. But it can often be told apart from other Eratigena here like Eratigena agrestis.
Second - a huge number of the species-level IDs on iNat are wrong. In the UK, E. saeva is probably dominant in the west of the country… on iNat, there are 3 here tagged as the species - none of these are research grade, but 2 provide microscopy. Meanwhile, E. atrica s.s has about 300 but is very localised and rare - most of these will be wrong!
I do wish the complexes were used more often. I think they’re very useful where there’s a “group” of species (the giant house spiders) that can readily be IDed and differentiated from the rest of the genus.
Extreme case - Dicranopalpus ramosus (“fork-palped harvestman”) is a complex as above, needing microscopy to tell apart from D. caudatus, recently split, etc, still recorded as an aggregate in the national recording schemes - both species present in the UK. They are distinctive and can very easily be narrowed down to that “complex”. Up until recently, those two were the only Dicranopalpus found here.
But now there is Dicranopalpus larvatus, first found 2020 and seemingly spreading very rapidly - is this an invasive species that needs to be tracked? I don’t think we can tell yet. Visually it’s fairly distinct but since you can’t tag a harvestman as the long-established D. ramosus ‘complex’, instead you have to go for the ‘vague’ genus level. Before 2020 that would mean that “D. ramosus group”, as there weren’t any others - but now, that’s ambiguous.
One thing that I constantly remind myself is that this is all a ‘work in progress.’ What we’re using now isn’t yet perfect – it’s a process towards progress! One easily identified species may be a group of cryptic species later. Or, some group of species/complexes may be resolved in the future at just some polymorphic single species! Taxonomy isn’t a stagnant science, in my opinion – it makes it fun!
One of my plant tax profs said, “If you’re comfortable with the names of plants, then you’re not up to date!” :)
It is worth checking out what the top identifiers of the taxa do. There are some serious experts working in the taxa you mention - study their approach, and if necessary start a flag on the taxa and tag the top people in for a discussion (that way the discussion gets recorded for posterity too!) Do check their profiles though - some will prefer you to contact them by direct message.
To be honest, i am cautious with any data. Most ‘professional’ data in my experience is collected in a frenzy of a short field season by low-pay field techs, interns, etc in adverse conditions without a lot of resources. Since the ‘field dirtbag life’ involves moving all the time and accepting any jobs you can, people tend to travel a lot and thus not be experts in any one place. And with the state of taxonomy a lot of species no one knows what to do with or people treat the same things differently. I don’t personally think that sort of data is ‘better’ than inaturalist data, just different. The upside of iNat is i can review all of the ‘research grade’ data myself and decide for myself whether to use it. Though i wish there were better tracking for this and there was a way for each of us to track things we reviewed and decided we did not have confidence in, without outright disagreeing with the ID. But for most past data from field techs in ecological surveys and such, i don’t have the great advantage of usually good quality photos, location, date, time, and the ability to comment and engage with others about them including the observer (unless they are gone). This is in my mind the main advantage of iNat over other ways of collecting field data, which mostly balances out the downsides. I can choose for myself if the data fits my purpose, and if i have questions i can often get answers.
Yes, I think it’s fine to do this, but as others have noted, be ready for some pushback and have some good documentation and patience on hand. Checking how other volume IDers/taxon experts approach these is also a great idea, as you don’t want to be working at cross purposes to someone else on lots of observations. You’re basically proposing an outreach project to educate folks about bumblebees, so, as long as you’re sure the IDs are correct and you’re happy to have discussions in comments, I don’t see an issue.
Yes, please do put disagreeing ID’s if something can’t be ID’d. As an observer, I would prefer my stuff be accurate. As someone who uses the database to learn more about what interests me, I really want it accurate.
I would also recommend writing a journal entry that you can link to explaining the issue with identifying those species that you can link to when asked. Quite frequently I will get tagged by someone in Europe on a Subgenus Bombus, and even though a couple of us will say those really can’t be species ID’d, the observer will still tag a bunch of others looking for that illusive species ID.
There may be some push-back and I wished there was a way to distinguish between disagreeing that something is ID’d correctly vs. disagreeing that something can be confidently ID’d to species. I would suggest to always put an explanation on any “disagreements” that are meant to bump things back to genus because it’s impossible to ID to species. The way the wording reads on these it comes across as “X disagrees this is species Y” suggesting that species Y can be ruled out. So in the bee example, if you bump things back to Bombus from Bombus terrestris someone might interpret that as “oh then it must be B. lucorum if it’s not terrestris.”