Should curators have a critical point of view when changing iNat taxonomy?

I only note that from the POWO references you mentioned before that at least between 2011 and 2020 WCVP/POWO did not follow the Thiede & al. view.

Um, guys… Mexico is a Western country.

5 Likes

That’s the thing. People are acting like there is anything close to a current consensus. There isn’t. There isn’t even anything close to a median taxonomy we can revert to. It’s become mass chaos and that extends beyond iNat as well.

The advancement of DNA analysis has resulted in a lot of discovering of discrete cryptic taxonomic entities, but there’s no reason why they should be defined as species. That’s my whole point here. Not that taxonomists are wrong about their genetics or that entities aren’t distinguishable. They are if you have access to genetic sequencing equipment, they just shouldn’t be at species level, at least not on a citizen science website. That’s it. That’s the whole issue here.

Sure, or we could just do what i’ve been saying all along and put the cryptic entities at subspecies or form or something.

I honestly don’t think it is possible to create a usable taxonomy that is purely monophylletic and can also be used to engage in field research and is usable beyond genetics. Maybe that’s scientific heresy but that is what i am seeing here. it can never be perfect, in fact the way genetics is much more flexible and variable than we initially thought, the very concept of monophylly may be invalid. We certainly can’t fully anchor how we understand and classify life on that concept.

Maybe true from all sides of this, but the entirety of what now passes for Linnaean taxonomy is a small group of ‘westerners’ imposing a view from their subculture on anyone who wants to name any living thing beyond ‘green bushy thing’ or whatever.

Yes, and evolutionary convergence is a very very strong force to the point maybe it works on a level of the very structure of genetics too. What happens if we discover an earthlike planet that has life with an independent genesis that is nearly identical to Earth life?

5 Likes

I feel like which taxa you’re used to dealing with will impact your view on this a lot?

With birds I’m used to being able to get nearly everything to species, other than a very small number of species pairs or hard genera (e.g. Alder/Willow Flycatcher that won’t vocalize). Even with birds, doing field surveys for work with distant ducks there are a lot of individuals too distant to call beyond Greater/Lesser Scaup, Common/Red-breasted Merganser etc. Shorebirds are even tougher at distance.

With small insects I don’t ever expect species ID; I hope for genus and am content with family. Entomologists have been primarily using genitalia and other microscopic features to delineate species for ages, so field ID features had never been prioritized because the macroscopic features were more variable and subjective.

I feel like with plants it’s somewhere in between? Trees can almost always go to species (except willows maybe?) but Asteraceae are a mess and genus is often the best I can hope for.

4 Likes

Definitely agreed that it depends on taxa. Some groups have excellent consensus, some groups are a dumpster fire. For the dumpster fire groups, i really so think there is a lot of value in using sequencing to help figure out what species are legitimate and not. (Hellooooooo Inocybaceae)

But still, sequences should be accompanied by scientific due diligance and peer review

3 Likes

Oh yeah my impression is that fungi are similar to insects where microscopic spore analysis was being used to delineate species already, so field ID to species was often not practical even before genetics?

1 Like

For a lot of species, you are correct

1 Like

Ah, to be clear I was thinking of moving a taxon from one genus to another. In those cases, delaying a move doesn’t deny us a name to apply to an organism, the risk is only that we might use an outdated name for a while longer.

In cases of newly described species, the delay would come with a more substantial downside in lost/misidentified observations, so it would make sense to be less conservative in incorporating changes.

2 Likes

I think that it would be good to have a contribution by the iNat staff. In particular, in relation to the possibility to add to the curator guide a resume of some of the considerations expressed here in various posts.

2 Likes

Yeah, I don’t think “can be distinguished in field with hand lens” is really going to win acceptance as a species concept on or off iNat. I mostly do vascular plants, and it’s great that I can do that, for the most part (because it bypasses the issue of making collections!) but I’ve also dabbled in bryophytes and lichens, and there it’s just taken for granted that you’ll need to bring some material back with you if it needs a species-level ID.

I do think it might help to ease our guidlines a bit on creation of species complexes. I think the original idea was to avoid curators overloading the tree with a lot of amateur taxonomy, but I think it might be helpful.

Incidentally, my experience of taxon splitting making my life more difficult usually hasn’t involved either micromorphological traits or purely DNA-based species delineation. (Sharkey et al. have tried the latter for tropical braconid wasps and run into a lot of pushback.) In the examples I can think of (e.g., the splitting of Asarum canadense, Arisaema triphyllum, Caulophyllum thalictroides), we’ve gone from a circumscription that I could reliable diagnose vegetatively and for long periods of the year to ones that require actively flowering material to distinguish. Obviously there are plenty of vascular plants out there that are like that already, but it’s annoying even if justified. On the other hand, many of these subtle splits also seem to correlate with habitat or phenological differences that had previously been interpreted as polymorphism in a single species.

6 Likes

I completely agree.

Yes, those are good examples. Arisaema triphyllum provides a good illustration of the problem. POWO recently elevated the subspecies of Arisaema triphyllum to species rank, and iNat followed suit. As a result of this taxon change, the quality of iNat data is decreasing, and it will continue to decrease over time. This is because observers are now applying a species-level ID when a species complex would have been more appropriate. Since iNat does not treat species complex as a first-class taxon, it is routinely overlooked by observers.

I’ve given up trying to educate observers about the Arisaema triphyllum complex. It’s just too much work with too little reward.

There’s a one-to-one correspondence between a species complex with three species and a species with three subspecies. Theoretically, data users should be able map one to the other as needed, but the data are not the same. In the case of subspecies, the data are underspecified, while in the case of a species complex, the data are overspecified. The latter is the more dangerous of the two.

IMO, we should have intentionally deviated from POWO and left things as they were.

3 Likes

I’m curious what you mean by this?

1 Like

I am not sure what “first-class taxon” means here, but I think that there will always be an issue with fewer users choosing the species complex than the species, even when the species complex exists.

Many users don’t know what a species complex as a general thing is, so they are likely to be reluctant to choose it because they don’t recognize it.

I would guess that even for some users who do recognize the complex, they just want their observations to be IDed to species and will choose a species level ID if it is available.

In general, I support including species complexes where they are well-supported by research and allow for more specific IDs to be made, but I just don’t think that having them will be a huge solution. Once the complexes are available, IDers still need to use them, and it seems like this will almost always depend on a core of knowledgeable IDers doing those IDs.

I suppose that, if at some point so many IDs are changed to the complex, that the relevant species are no longer included as CV options, that might make a difference as it would reduce CV-based IDs. That said, if users ID as the genus instead of the species, that is almost more work for IDers! It’s much easier/quicker to find a specific species to bump back to a complex than it is to sort through a genus and pick out the observations that should be IDed to the complex with all the other observations of other species (or ones that should remain at genus).

3 Likes

I don’t think this realistically can happen in many cases (especially with commonly observed taxa), because both the CV and the iNat ID structure in general are so extremely biased towards getting species IDs. I’ve complained about this before here. The same issues occur beyond species complex level with all of these taxa.

2 Likes

Tangent question: how does e.g. GBIF deal with ‘Species Complexes’?
Whenever a species gets “promoted” to a so-called ‘Species Complex’, what happens to a former Species-level RG obs (and once exported as such to GBIF), which is now becoming a Complex-level RG obs?

It maps to the genus in GBIF

3 Likes

Thanks.
Yuck.

1 Like

I agree that this is unlikely to happen in many cases (and especially those with lots of observations). Perhaps more likely in species with a few hundred observations or fewer. I was just trying to address the potential outcomes of using complexes. This seems to be a more “conservative” approach (as opposed to creating and IDing to full species), so I think it’s good to discuss the costs/benefits.

I do think this is the best approach by GBIF to this (they aren’t interested in adding taxa that appear/disappear on all the sources that feed into them, and don’t have a level for species complex available I don’t think). However, because of this, I think that the species complex ID doesn’t add all that much value for research (since most end users of the data are using GBIF to access and thus won’t see it). Researchers could access the data via iNat of course, and I bet some active iNat users who are heavily invested in specific taxa do use IDs to complex (though I don’t know of any specific examples). I think another big benefit of complex IDs is that they often make other taxonomy curation tasks like future splits, atlasing, changes, etc. much easier.

2 Likes

Which raises the question of ecotypes. Some might consider ecotypes to be, in fact, cryptic species, whereas others might disagree.

Can confirm that it will suggest complexes if there are enough observations IDed to it.

When I search Hygrocybe conica, the first suggestion is the complex, not the species. Probably because its well known that there are several cryptic species hiding, and enough people have bumped H. conica species up to complex.

I’ve also seen this happen for Narceus americanus, and Allium tricoccum

1 Like